USER GUIDE

Version 1.3.2
22. Jul 2021

U-HE - HECKMANN AUDIO GMBH - BERLIN

 Table of Contents
Introduction 3
Installation ................................................................3
Resources ................................................................3
The Satin Story ........................................................4
Features and Specifications.....................................5
User Interface ..........................................................6
Quick Start Guide.....................................................8

Preset Browser 9
Overview ..................................................................9
Directory Panel ......................................................10
Presets Panel .........................................................13
Preset Info Panel ...................................................15
Installing Soundsets ...............................................16
Preset Tagging .......................................................17
Search Functions ...................................................18

Upper Panel 20

Lower Panel 22
Studio Mode ...........................................................22
Delay Mode ............................................................24
Flange Mode ..........................................................26

Service Panel 28
Tape Machine Primer .............................................28
Tape Parameters....................................................30
Repro Head Parameters ........................................31
Analyzer .................................................................33
Circuit Parameters .................................................33

Configuration 35
MIDI Remote Control .............................................35
Preferences Page ..................................................37

Tips and Tricks 39

Frequency Response Plots 42

NKS 55

Glossary 56

2
 INTRODUCTION

Introduction
Installation
Go to the Satin page, grab the appropriate installer for your computer, unZIP and run it. Satin is
fully functional in demo mode except for a mild crackling which disappears after authorization.
By default, Satin uses the following directories:
Win presets (Local) C:\Users\*YOU*\Documents\u-he\Satin.data\Presets\Satin\
presets (User) C:\Users\*YOU*\Documents\u-he\Satin.data\UserPresets\Satin\
preferences C:\Users\*YOU*\Documents\u-he\Satin.data\Support\ (*.txt files)
If a previous installation into the VstPlugins folder caused no problems, you can reinstall there.
Mac presets (Local) Macintosh HD/Library/Audio/Presets/u-he/Satin/
presets (User) Macintosh HD/Users/*YOU*/Library/Audio/Presets/u-he/Satin/
preferences Macintosh HD/Users/*YOU*/Library/App…Support/u-he/com.u-he.satin…
resources Macintosh HD/Library/Application Support/u-he/Satin/
To uninstall, delete the plugin files, then Satin.data (Windows) / the two Satin folders (Mac). Please
read all relevant texts included with the installer, especially license.txt.

Resources
For frequency response curves, go to satin plots
For support, news and downloads, go to u-he.com
For lively discussions, go to our KVR forum
For online friendship, visit us on facebook
For tutorials and more, go to our youtube channel
For extra presets, go to patchlib

u-he team 2021 (Q3)

Urs Heckmann (boss, concepts)
Jayney Klimek (office management)
Howard Scarr (user guides, presets, necessary grump)
Rob Clifton-Harvey (IT admin, backend development)
Sebastian Greger (GUI design, 3D stuff)
Jan Storm (framework, more code)
Alexandre Bique (all things Linux)
Oddvar Manlig (business development)
Viktor Weimer (support, presets, the voice)
Thomas Binek (QA, bug-hunting, presets)
Henna Gramentz (office supervision, support)
Frank Hoffmann (framework, new browser)
Alf Klimek (tagging, rock-stardom, studio)
Sebastian Hübert (media, synthwave)
David Schornsheim (more code)
Stephan Eckes (yet more code)
Luca Christakopoulos (communication design)

Special thanks to Riccardo Pasini for Studer® A827 tests, Torsten Bader for lending us his vintage
NR units, Clemens Heppner for his genius while at u-he, beta-testers and friends who contributed
presets… and to Brian “you can call me Brian” Rzycki for maintaining the original Patchlib.

TOC intro presets studio delay flange service config tricks plots nks gloss 3
 INTRODUCTION

The Satin Story

The quality of recent audio applications would have been unthinkable just a few decades ago. Now
we have crystal-clear sound, miles of headroom, finest resolution. Perfection, the Holy Grail...
But do we really want perfection? Legendary audio engineers enthuse about the special sound of
vinyl and magnetic tape. Some even say the truly great albums only came alive because of the
imperfections (in both performance and technology), and that absolute perfection would be boring.
Perhaps that should be taken with a pinch of salt, but one thing is certain: our ears are thoroughly
analogue, human hearing follows rules similar to those in mechanical and electromagnetic
recording processes. The most important lesson here is that nature itself is non-linear.
Tape recording and playback involves a whole host of non-linearities. Although only a few of them
are immediately discernible, they all contribute in one way or another to the final sound: like e.g.
guitar strings and bodies they interact. In real life, things can get complicated!

non-linearities typical of tape machines

• magnetic tape saturation
• smoothing / blurring of transients
• compression of high frequencies
• ‘head-bump’ effect
• asperity ‘3D’ effect
• ‘glue’ effect with multiple tracks

When we first started outlining the

concept, we had little idea just how
much research and functional testing
would be involved.
Satin’s specifications also grew far
beyond what we originally planned.
While most of the ‘modeling’ plug-ins
emulate a single piece of hardware,
we were aiming for a true all-rounder
with a highly flexible character... the tape machine equivalent of Diva, our analogue modeling
synthesizer.

We decided to model each element of a real tape machine as a separate unit within a modular
architecture, as this would allow us to shape the sound in more ways than the original. The core
parameters accurately emulate all the important electrical and electromagnetic processes, but
Satin’s alternative modes and ‘service’ parameters make it a more flexible processor than any real
world tape machine could ever be. Satin isn’t a one-trick pony, it’s a tape effects construction kit.

So, apart from delivering authentic tape machine effects, Satin is capable of much more ‘left-field”
tape sound processing (including tape flanging and multi-tap delay).

TOC intro presets studio delay flange service config tricks plots nks gloss 4
 INTRODUCTION

Features and Specifications

• Internal sample rate: 352-384kHz (depends on project sample rate, 8 x oversampling 44.1kHz)
• Bias oscillator frequency: 118 to 128kHz
• Bias level: variable, +/- 5 dB
• Tape speed: 1.875 to 30 inches per second, continuously variable
• Tape types: vintage, modern
• Total harmonic distortion (at -20dB Input, 30 ips, bias +2dB, AES30-EQ)
vintage 2.4%
modern 1.2%
• Wow & Flutter, maximum (DIN 45507/IEC 60386)
vintage 0.55% at 1.875 ips to 0.28% at 30 ips
modern 0.52% at 1.875 ips to 0.19% at 30 ips
• Record / repro EQ standards (individually selectable)
IEC/CCIR 7.5 ips, IEC/CCIR 15 ips
NAB
AES 30 ips
• Compander types (standard noise reduction systems in Studio mode)
A-Type, A-Type mod
B-Type
uhx Type I or II (improved since v1.1.0)
• Frequency response (+/-2 dB, -12dB input, Modern tape, pre-emphasis 50%)
1.875 ips 35Hz to 9kHz (IEC 7.5)
3.75 ips 25Hz to 15kHz (IEC 7.5)
7.5 ips 27Hz to 18kHz (IEC 7.5)
15 ips 35Hz to 20kHz (IEC 15)
30 ips 40Hz to 22kHz (AES 30)
• Pre-emphasis (18kHz)
1.875 ips 0 to 56dB
3.75 ips 0 to 48dB
7.5 ips 0 to 40dB
15 ips 0 to 33dB
30 ips 0 to 27dB
• Maximum delay times (in Delay mode)
1.875 ips 4.267s
3.75 ips 2.133s
7.5 ips 1.067s
15 ips 533ms
30 ips 267ms
• Input headroom 0 to +18db
• Gap width (repro head) 1 to 5µm
• Head bump 0 to 8dB, typically 4dB
• Crosstalk (at 1kHz) -80 to -20db
• Static tape noise -100 to -40dB
• Asperity / mod noise -100 to -50dB
• Azimuth range +/- 2 arc-minutes
• 0 VU reference level -24 to 0dBFS (AES-17 standard)
• System noise -120db

TOC intro presets studio delay flange service config tricks plots nks gloss 5
 INTRODUCTION

User Interface

GUI appearance
The size of Satin’s user interface can be adjusted to suit your personal preference / monitor size –
simply right-click anywhere in the background of Satin’s window (if the MidiLearn menu pops up
instead, try clicking a bit further away from the knobs). Select a GUI size between 70% (854 x 504
pixels) and 200% (2440 x 1440 pixels) in 10% steps.
Note that the setting here is temporary. You can set a default GUI size in the global Preferences.

knobs
coarse control..... Click+hold with the lefthand mouse button, then drag up or down
fine control.......... For steps of only 0.01, hold down either SHIFT key before moving the knob
scroll wheel.........Hover over a knob and roll the wheel (fine adjustment via SHIFT)
reset to default....Double-clicking a knob reverts to a default value (e.g. tape speed = 15 ips)

switches

Although they look similar, Satin actually has three different kinds of switch. Some slide left to right
(like the one in this image), others are on/off types that are only highlighted when switched on. The
Studio panel’s Group switches are mutually exclusive ‘radio buttons’.

parameter locking

Clicking with the righthand mouse button on any knob or switch in Satin opens a tiny context menu
containing only Lock. A locked parameter can still be adjusted, but the value won’t change when
you switch presets, and it won’t affect (or react to) other instances of Satin assigned to the same
Group. To unlock a control again, right-click and untick ‘Locked’.

the control bar

The bar at the top of Satin’s window is almost self-explanatory – see the next page, then go here.

TOC intro presets studio delay flange service config tricks plots nks gloss 6
 INTRODUCTION

the main panels

The rest of Satin’s GUI is divided vertically into three panels:

UPPER PANEL

LOWER PANEL

SERVICE PANEL

Upper Panel
Together with the control bar, this area always remains visible because all the controls apply in any
of Satin’s operation modes (see lower panel). For more about the upper panel, go here.

Lower Panel
The Mode selector in the middle switches the functionality and appearance of the lower panel...
Studio: Use this mode whenever you want to add some tape machine ‘sheen’. Satin lets you group
multiple instances together. For more about Studio mode and Groups, go here.
Delay: This mode emulates a monster tape machine with four stereo repro heads set up for delays.
For more about Delay mode, go here.
Flange: Emulates classic tape flanging, a dramatic analogue ‘whoosh’ effect that requires skill and
practice. Satin can do this automatically. For more about Flanger mode, go here.
Service Panel
In typical studio sessions of the past, a well-calibrated multitrack tape recorder was essential for
professional work – assuming you wanted the highest audio fidelity possible. A technician would
sometimes have to adjust the bias (pre-magnetization) to suit the kind of tape being used, the aim
being to optimize the recording system for a flat frequency response with minimum distortion.
Most of the parameters in Satin’s service panel practically reconfigure or even replace the virtual
‘hardware’: For instance, changing the head gap width would mean replacing expensive parts. For
more about the service panel, go here.

TOC intro presets studio delay flange service config tricks plots nks gloss 7
 INTRODUCTION

Quick Start Guide

To use Satin purely as a ‘tape sound’ processor i.e. to add some tape-typical warmth and cohesion
(enough flowery words for now), all you really need to do is experiment with the two large gain
controls at the top: the Input and Output knobs...

Here goes: Insert an instance of Satin into an audio channel of your DAW / host application (a
default preset will be loaded) and play around with the Input control. Watch the VU meter, but
concentrate more on the sound i.e. what Satin is doing to your audio material.
You can compare with the untreated signal by clicking on the Bypass button. But beware! Judging
which settings actually sound better means keeping the perceived volume as constant as possible
– volume can be a significant factor here, as louder almost always sounds ‘better’.
Check whether Makeup is switched on. If so, the output level is automatically reduced when the
input is increased (note that the Output knob doesn’t move automatically, so you can always ‘fine-
tune’ the output level).

A few questions you should be asking yourself while adjusting the input and output levels:

• What’s happening to the transients in the signal when I give it more ‘oomph’?
• How are the high frequencies in particular being affected by different settings?
• At what point does the signal start to distort? In which frequency range?
• What is happening to the (perceived) stereo width at different gain settings?

Depending on certain properties of the audio material, these effects can be quite subtle, especially
if you are not very familiar with the ‘tape sound’. Tip: You could train your ears by processing very
different audio material e.g. drums, vocals, bass. After some practice, you should start to develop a
feel for the above processes and how they interact.
Patience helps, and the cumulative effect of applying Satin to several (or even all) tracks in a song
will make everything appear more obvious (see groups).

the next few chapters include a description of each and every element in Satin’s user interface...

TOC intro presets studio delay flange service config tricks plots nks gloss 8
 PRESETS

Preset Browser
Overview
You can load any preset in the current folder by clicking on the data display, or step through
presets by clicking on the arrow symbols either side of the data display. Of course Satin also
includes a preset browser – click on the PRESETS button to the right of the Data Display:

Folders are on the left, presets in the centre and information about the current preset on the right.
If you can’t see any presets at all in the central area, click on Local. If you can’t see the PRESET
INFO panel, click on the button in the top right and select Show Preset Info.
The Local root directory contains a representative selection of presets copied from the subfolders.
After loading a preset (by clicking on its name) you can step through all the others using your
computer’s cursor keys.

That’s all you really need to know for now!

For users who want to dig deeper, however, Satin’s browser offers many interesting features,
including a powerful search engine. For details, please read the rest of this chapter.

Default preset
When Satin starts it checks whether the Local root contains a preset called default, which is then
loaded instead of the standard one. Note: Any preset called ‘default’ will not appear in the browser.

TOC intro presets studio delay flange service config tricks plots nks gloss 9
 PRESETS

Directory Panel
If you don’t see this panel on the left of the browser, click on the DIRECTORY tab.

Local
The factory presets are sorted into 4 main folders: Delay Mode, Flange Mode, Studio Mode and
Templates (just 4 “vanilla” presets – read the Preset Info about each one)
We recommend that you do not add or remove any presets here, but save all your own creations
as well as presets from third parties to ‘User’ folder.
MIDI Programs
‘Local’ also contains a special folder called ‘MIDI Programs’ which is normally empty. When the first
instance of Satin starts, up to 128 presets from that folder are loaded into memory, to be selected
via MIDI Program Change messages. Certain DAWs automatically route MIDI into effect plug-ins
while others achieve the same via more complicated means. You might have to spend some time
looking through the appropriate documentation for information on how to do this!
As the presets in MIDI Programs are accessed in alphabetical order, it makes sense to prefix each
name with an index ‘000 rest-of-name’ to ‘127 rest-of-name’ or similar. IMPORTANT: Unlike regular
presets, MIDI Programs can’t be added, removed or renamed on the fly. Changes will only take
effect after the host software is restarted.
The MIDI Programs folder can contain up to 127 sub-folders of 128 presets, switched via MIDI
‘Bank Select’ messages (CC#0) preceding the Program Change message. The MIDI Programs
folder itself is bank 0, sub-folders are addressed in alphabetical order starting with bank 1.

TOC intro presets studio delay flange service config tricks plots nks gloss 10
 PRESETS

When Satin receives a program change, it will display the bank and program numbers to the left of
the preset name e.g. “0:0” for the first preset in the first bank. In certain hosts, however, the first
bank / preset is designated “1” instead of the correct “0”.
To avoid another possible source of confusion, please make sure that there are no junked presets
in the MIDI Programs folder. All files there are adressed, even if they are hidden.

User
The best address for your own creations as well as presets from other sources. You can either
select User immediately before saving the preset, or set a global preference which ensures that it
will always be saved to this folder – see the preference Save Presets To.
Tip: It’s worth finding out where the User folder actually resides on your computer. Right-click on
the User folder and select Open in Finder / Explorer.

Smart Folders
The other folders do not contain files, but display the results of querying a database of presets.
The content is therefore dynamic i.e. it will change whenever the underlying data changes.

You can drag & drop any smart folder content onto e.g. ‘User’ or the desktop (see External
Drag & Drop) to create folders containing real copies of those presets!

Search History
Click on this folder to display the results of past searches (maximum 10). Whenever you need to
make the results of a search more permanent, right-click and select save Search... The entry will
be moved to the Saved Searches folder – see below. To remove all searches from the list, right-
click on the Search History folder and select clear.
Saved Searches
This folder contains searches that have been saved via right click from Search History. To remove
individual saved searches, right-click on the search and select delete. Tip: Entries dragged from
Saved Searches and dropped onto real folders within Local or User will create a folder containing
copies of all found presets!
Banks
These smart folders reference metadata about preset origin – the version of the factory or third
party library with which the preset was installed. See Preset Info a few pages down. Banks are
predefined for factory presets.
You can even create your own custom banks: Drag & drop one or more presets onto the ‘Bank’
folder then enter a suitable name into the dialogue box.
To remove Bank attributes from selected presets, either drag & drop them onto the ‘no Bank’ folder
you will see at the bottom of the Bank list, or right-click on the Bank and select Remove Presets
from Bank. Empty Banks will disappear.
Favourites
8 smart folders, one for each Favourite colour. See Presets context menu a few pages down. Only
one Favourite colour/number can be set per preset. Presets dropped onto one of the Favourites
folders will be marked as such. Favourite status can be removed from all presets of one particular
colour by right-clicking on the ‘Favourite’ folder and selecting Remove All Favourite (n) Marks.

TOC intro presets studio delay flange service config tricks plots nks gloss 11
 PRESETS

Junk
A smart folder pointing to all junked presets. See Presets context menu on the next page. Presets
dropped here will disappear from the rest of the browser unless made visible (see Show Junk in
the Presets context menu). Like Favourites, Junk can be exported/imported (as Junk.uhe-fav). See
External drag & drop.
Tags
Smart folders for each Category, Application and Character tag. Presets dropped onto these
folders will adopt the corresponding tag. Presets dropped onto the Untagged folder will have all
Category, Application and Character tags removed.
Author
Smart folders for each Author. Tip: Instead of signing each of your creations, you could sign just
one of them, then select them all and drag them onto Author/(You)/. As the process cannot be
undone, please use this feature with caution!

Directory Context Menu

Right-clicking on any folder within Local or User will open this menu:

Refresh
Create New…
Rename…
Open in Finder *

Move to Trash *
On Open Expand to
Show Folder Icons
Refresh
Update the contents of the browser. Depending on your operating system, this may be necessary
after you have moved, added, removed or renamed any folders or presets using Explorer / Finder.
Create New…
Insert an empty subdirectory.

Rename…
Edit the folder name.

Open in Finder / Explorer

Opens a system window for the clicked folder. After adding, removing or renaming preset files or
folders outside of Satin’s own browser, remember to refresh the directory (see Refresh above).
On Open Expand to
These options determine how deeply the browser will open subdirectories whenever the GUI is
opened again or the refresh function is called. The first option (none) collapses all folders, while
the final option (all levels) reveals all nested folders.
Show Folder Icons
Deselect this option if you find the folder icons distracting.

TOC intro presets studio delay flange service config tricks plots nks gloss 12
 PRESETS

Presets Panel
The central, unlabelled area of the browser is where you click to load presets...

Presets context menu

Right-click to open a menu containing functions that can be applied to individual presets.

Mark as Favourite 1
Mark as Favourite 2
Mark as Favourite 3
Mark as Favourite 4
Mark as Favourite 5
Mark as Favourite 6
Mark as Favourite 7
Mark as Favourite 8
Mark as Junk

Show Junk
Select All
Deselect
Rename…
Copy to User Folder *
Show in Finder *
Convert to h2p *

Move to Trash *
Mark as Favourite
Choose one of 8 ‘favourite’ marks. The selected entry will be replaced with unmark as favourite.

Mark as Junk
Instead of deleting presets, you can mark them as ‘junk’ so that they disappear from the browser…

Show Junk
Activate this option to display junked files (see above) marked with a STOP symbol instead.

Select All, Deselect

See Multiple selection below.

Rename…
You can change the names of presets using this function. Note that only the most recently selected
preset can be renamed i.e. you can’t rename multiple files at once.
Copy to User Folder / Duplicate
The entry here depends on the status of the preference Save Presets To and whether the source
presets are in the Local or User folder. Selected presets are copied with an index appended to the
name (like the ‘Auto Versioning’ preference).
Show in Finder / Explorer
Opens a system window for the right-clicked file. In smart folders only, holding down an option key
(Mac) or ctrl key (Windows) replaces this entry with Show in Browser, which shows the currently
selected file in its original location within Satin’s browser.

TOC intro presets studio delay flange service config tricks plots nks gloss 13
 PRESETS

Convert to native / h2p / h2p extended / nksfx

Converts the selected preset(s) to the format previously selected via right-click on the SAVE
button. The nksfx option is only available in the VST2 version – see the NKS chapter.
Move to Trash / Recycle Bin
This function moves all selected preset(s) to the system trash. Caution: This also works on files in
smart folders i.e. the originals will land in the system trash.

RESTORE
While in the browser you can audition presets for as long as you like without losing track of the one
that was previously loaded: Clicking the Restore button gets you back to where you started.

Multiple selection
A block of adjacent presets can be selected via shift+click, and individual presets can be added to
the selection via cmd+click (Mac) / alt+click (Windows). Presets can be moved to a different folder
via drag & drop. To deselect, either click on a deselected preset or pick Deselect from the context
menu.

Internal drag & drop

You can drag and drop single or multiple files from the preset panel onto any folders in the
directory panel. Files dragged onto regular folders will be moved unless you hold an option key
(macOS) or ctrl (Windows), in which case they will be copied instead. Files dropped onto smart
folders will adopt the attribute of that folder: For instance, you can set the Author or Favourite
status of several presets at once.

External drag & drop

To manage your preset library externally you can drag presets and folders between Satin’s
browser and your desktop (or any system window). On the Mac most Finder operations will
automatically update the browser. Updating might not be immediate when using multiple formats
or multiple host applications, but all it usually takes is a click on the GUI or in the directory tree
(sets the focus to the clicked instance of Satin). On Windows systems, a Refresh (see Directory
context menu) will be required before changes to the contents of the browser appear.
Another little helper: If you drag a Satin preset from e.g. your desktop and drop it onto the Data
Display, that preset will be loaded (but not automatically saved).
Exporting smart folders
Drag any smart folder onto the desktop to create a new folder containing those presets. Drag an
entry from your Search History, or the Category ‘Saturation’, Favourites or one of the Authors.
Exporting favourite / junk status
You can export Favourite status, all at once or individually: Shift+click and drag the ‘Favourites’
folder onto the desktop to create a file called Favourites.uhe-fav. Similar for sub-folders: If you
shift+click and drag e.g. ‘Favourites 5’, this will create a file called Favourite 5.uhe-fav.
The same method works for Junk status, creating a file called Junk.uhe-fav. Such files can be
imported into Satin’s browser on a different computer (for instance), via drag & drop onto (or
anywhere within) the Favourites folder, or to the Junk folder.
Note that importing .uhe-fav files from another computer will only work 100% correctly if all preset
names and locations are identical on both computers!

TOC intro presets studio delay flange service config tricks plots nks gloss 14
 PRESETS

Preset Info Panel

The panel to the right displays information about the selected preset:

If you can’t see this panel, click on the button in the top right corner and tick Show Preset Info.
Below the preset name you should see its path (from /Local or /User), the Bank and the Author
(which also appear as Smart Folders).
DESCRIPTION and USAGE text is entered immediately before saving a preset.
CATEGORIES, APPLICATION and CHARACTER are the tags for the current preset. You can
remove or add tags directly – see ‘Tagging via PRESET INFO’ a few pages down.
If you prefer to see less information, hide the tags only or the entire PRESET INFO panel. The
FLAG button opens the Tagging window.

TOC intro presets studio delay flange service config tricks plots nks gloss 15
 PRESETS

Installing Soundsets
Any soundsets we distribute ourselves will (eventually) be available in .uhe-soundset format. Third
parties are also encouraged to use this package format for their own commercial soundsets (for
details please contact our support team).
Standard Method
To install, drag & drop the .uhe-soundset file into Satin – anywhere will work. The soundset should
appear in the ‘User’ folder. If a soundset with the same name already exists in that location, any
modified files will be backed up and the location of the backup file will be displayed.
Alternative Method
Soundsets in .uhe-soundset format can also be installed by clicking on the u-he badge, selecting
Install Soundset... from the menu and navigating to the .uhe-soundset file. This option is
especially useful for Linux, as the browser version for that platform does not support drag & drop.
Regular Folders
Folders containing Satin presets can be manually copied or moved into the ‘User’ folder. You
might have to refresh the browser (see Directory context menu) before they appear there. A
refresh is generally necessary in Windows but not in macOS.
Note: As .uhe-soundset files are basically ZIP-compressed folders, you can rename them i.e.
replace the long file extension with ’zip’, then extract the presets and documentation.

TOC intro presets studio delay flange service config tricks plots nks gloss 16
 PRESETS

Preset Tagging
“Tags” are elements of metadata, information that you can add to presets so that they can be found
according to certain attributes.
IMPORTANT: Clicking on SAVE isn’t required, as tags are updated automatically. The obvious
advantage is that presets don’t have to be saved every time you edit tags. The main drawback is
that you should only edit tags after saving your preset. If you edit tags while in the process of
creating a new version of something, you would also be changing the tags in the original preset!

The Tagging window

Right-click on the [SAVE] button and select Tag this preset:

Category describes a preset by process type, Application offers technical classifications, and
Character tags are pairs of opposites from which you can choose only one.

Tagging via PRESET INFO

In the PRESET INFO panel, right-click on CATEGORIES, APPLICATION or CHARACTER and
select or deselect tags from the context menu. Tip: If you right- click on an existing tag, the first
option in the menu becomes Remove Tag...
The function Create Search from Tags searches for presets with ALL the same tags.

Tagging via the Tags smart folder

You can also tag presets by dragging & dropping them onto one of the ‘Tags’ smart folders. To
remove all tags, drag them onto the ‘[no Tags]’ smart folder.

TOC intro presets studio delay flange service config tricks plots nks gloss 17
 PRESETS

Search Functions

Search by Tags
Click on the TAGS tab to open this view. The buttons here let you set up search criteria according
to existing tags with just a few mouseclicks:

There are four sets of buttons. The first three correspond to the tags in the tagging window (see
the previous page), and the bottom row lets you find any presets tagged as Favourites. Clicking on
the [^] icon to the right of each heading hides the options for that set of tags.

Practically…
Click on the DIRECTORY tab, right-click on the ‘Search History’ smart folder and select Clear.
Double-click on ‘Local’ to restrict the scope of the search to that particular folder (presets in the
‘User’ folder will not appear in the search results now). The selected path appears immediately
below the Search field. To exit the “restriction” folder again, click on the ‘up’ [^] symbol to the left.
Click on the TAGS tab and select the Categories [Chorus] and [Tone]. Presets tagged with either
of those categories will appear in the presets panel. Click on the DIRECTORY tab again: The text
“#Modulation:Chorus #Filter:Tone” appears in the Search field and in ‘Search History’ (which also
shows the number of presets found in the search). If you are feeling adventurous, try editing the
contents of the Search field now – the results are updated accordingly.
Unlike selecting several Category tags, which will expand the scope of the search, selecting
Application, Character or Favourites tags will restrict the scope i.e. you will get fewer hits.

TOC intro presets studio delay flange service config tricks plots nks gloss 18
 PRESETS

Search by Text
The Search field lets you find presets according to a string of text. Here’s an easy example: If you
remember that the preset you’re looking for has the word “clock” in its name or description, enter
“clock” into the Search field and hit [Return]…
The search routine normally looks into the preset name, author, the DESCRIPTION and USAGE
(see the PRESET INFO panel). Searches are not case-sensitive, and quotes are not required
unless you need to include spaces between words.
To restrict the search to a particular search path, double click the folder. This path will appear
immediately below the Search field. The [^] button to the left moves the Search path up one level,
while the [X] button to the right sets the search path to the default (all Satin presets). Alternatively,
you can navigate directly to any higher level by right-clicking on the path.
Try it: Enter three or four letters then hit [Return]. For instance, star will find all files containing the
text string star (e.g. mustard or starters). Entering "star wars" (with the quotes) would find e.g.
Battlestar Warship, if such a preset existed.
Scope
You can limit the scope of the search to just the preset name or specific parts of PRESET INFO by
using name (preset name), author, desc (description) or use (usage) followed by a colon. For
instance, author:the finds all presets by sound designers whose author names contain ‘the’.
Similarly, desc:space will find all presets with the word ‘space’ in the description.
Logic
The following logical operators can be used between text strings, but not between tags.
AND requires that presets contain both words. It can be written explicitly, but is not necessary. For
example, star AND wars (or simply star wars) will find presets that contain both ‘star’ and ‘wars’.
OR means that presets can contain just one of the words… or both. For example, star OR wars will
find presets that contain ‘star’ as well as presets that contain ‘wars’.
NOT excludes presets containing the specified word. To find all presets that contain ‘star’ but don’t
contain ‘wars’, enter star NOT wars.

Including Tags
Tags can be entered into the Search field if preceded by a ‘#’.
For example, name:int #Delay:Tape will find presets with "int" as part of the name that are tagged
with the Tape Delay category.
In the current version of the browser, tags must appear after any text. For technical reasons, they
take the form #type:category (the type is invisible in the TAGS panel).

TOC intro presets studio delay flange service config tricks plots nks gloss 19
 UPPER PANEL

Upper Panel
This part of Satin’s window contains all the basic switching and tape machine parameters:

Bypass
Activating this button disables all processing. Note that grouped instances are bypassed together!

Data Display
The display at the top shows the name of the selected preset or, while a parameter is being edited,
its current value. Clicking on the arrows to the left and right of the data display steps forwards and
backwards through presets. Clicking directly on the data display opens a list of all presets in the
current directory (so you don’t always have to open the browser to select presets).
Recently added feature: If you drag a preset from outside the plugin and drop it onto Satin’s data
display, that preset will be loaded (but not automatically saved anywhere).

Presets / Save
See the chapter describing Satin’s preset browser.

Undo / Redo
The undo and redo buttons let you step back and forwards again through the edit ‘history’.
Although the number of steps in the buffer is limited to 10, you can even undo a change of preset –
so switching presets before saving doesn’t mean losing your work.

u-he Badge
Click to open a dropdown menu containing links to this user guide, to the documentation folder, to
our website, to our user support forum at KVR as well as to our address in various social networks.

Makeup
When Makeup is switched on, the output level is automatically reduced when the Input knob is
turned up (the output level is effectively linked to the Input knob). Note: the Output knob doesn’t
magically turn while you’re adjusting the Input, you can still use it to fine-tune the output level.
Caution: Turning the input volume way down while Makeup is on can seriously increase any hiss
and asperity you have dialed in (see tape parameters). Also, in Delay mode, quickly turning the
input volume down can cause the delayed signal to get very loud for a short while.

TOC intro presets studio delay flange service config tricks plots nks gloss 20
 UPPER PANEL

Tape
This switch affects the type of emulated tape. When driven into saturation, the ‘vintage’ tape tends
to distort sooner. It loses treble, and the low-to-mid range is stronger. The thinner modern tape has
a more even response, less distortion and doesn’t lose as much treble.

Metering
This switches what the VU meter displays: either the input or the output levels.

RMS
This feature affects how the VU meter displays output levels. When RMS is disengaged, the VU
meter follows its traditional ‘ballistics’, which means about 300ms integration time – a good
compromise for monitoring fluctuations in the human voice as well as most musical instruments.
Engaging RMS causes the VU meters to react more steadily to the energy content of the signal,
and integrating over a longer period matches perceived ‘loudness’ better.

0VU Ref
Having a standard reference while monitoring average (RMS) levels within a project makes a lot of
sense. For example, if your project is at -18 dBFS, it would be best to set -18.00 here. The VU
needles will be displaced by -18dB so that e.g. -20dB will appear as -2dB.

Soft Clip
When the button is engaged, output levels approaching 0dBFS are gently limited using a very
smooth transfer function similar to a Class-AB tube or FET-driven amplifier stage that gradually
changes from clean to clipped. Disengage the button If you want the occasional peaks to be
handled elsewhere within your mixing environment (overshoots are only visualized, not treated).

Input / Output
These knobs adjust input and output levels. They are so large because they are the most
important controls in Satin! Use them wisely, young Skywalker...

VU meter
Volume Unit (VU) meters are often seen ‘decorating’ audio equipment, from the most expensive
professional units where they actually make sense, to the cheapest consumer devices where they
are included more for aesthetic reasons than for precise monitoring.
Originally developed around 1940 for telephone companies, the VU standard was soon adopted by
the broadcast industry. VU meters have a relatively slow response, which is perfect for measuring
average levels. Traditionally, users were expected to aim for 0 VU whenever possible.
Note: Satin’s VU meter complies to the AES-17 standard i.e. it is calibrated with sine waves (as
opposed to square waves) and therefore displays a 0dBFS sine wave as 0 VU – a non-AES-17
meter would show about 3dB less.
Peak Meter: The central bar is a stereo peak meter with a significantly faster response than the
VU needles. Use the peak meter to monitor the levels of faster transients e.g. drums.

MIDI Activity
This indicator flashes whenever Satin is receiving MIDI data.

TOC intro presets studio delay flange service config tricks plots nks gloss 21
 LOWER PANEL

Lower Panel
Mode
Switches between Satin’s three basic operation modes: Studio, Delay or Flange.

Studio Mode
This is Satin’s default, the mode you should use for all your ‘normal’ tape processing tasks:

Speed ips
How quickly, in inches per second, the virtual tape is drawn past the heads. Faster tape speed
usually means better overall audio fidelity (especially with high frequencies) and less noise. Also,
the lower frequency limit moves up slightly. How these effects arise is described in the chapter
about Satin’s service panel. Note: The most common tape speeds are 7.5, 15 (default) and 30 ips.

Pre-Emphasis
Pre-emphasis has nothing at all to do with the EQ, it mainly addresses the ‘gap loss’ phenomenon.
The gap of the repro head determines its high frequency response – the smaller the gap, the better
the response. Wavelengths smaller than the gap width more or less fall into the gap, the worst
case being when the wavelength is exactly the same as the gap width, resulting in zero current.
This gap loss frequency is proportional to tape speed: about 8kHz at the regular 15 ips (16kHz at
30 ips and 4kHz at 7.5 ips). See the response plots showing rolloff without pre-emphasis.

No Group
NOT assigning an instance of Satin to any group is the default state, and pressing the momentary
button No Group reverts to this default. See Group 1-8 below.

Bypass Tape
Press this button to take the tape section out of the main signal path. The compander and record /
repro circuits remain active. The “off” state of this button is ignored in Delay and Flange modes.

Group 1-8
Studio mode lets you assign the current instance to one of eight groups. Each group acts as a
container for all instances so they behave as a single entity while editing. This comes in handy

TOC intro presets studio delay flange service config tricks plots nks gloss 22
 LOWER PANEL

when using Satin as a multi-track processor (for a practical example, see Multitrack Coherence).
Groups can be named by double-clicking on the fields.
Important: Using groups in simultaneously opened projects in your host application(s) can lead to
unexpected results – the groups of one project will be overwritten by those defined in the next. Also
please note that Satin cannot be grouped if the processes are sandboxed. If you find that Satin’s
grouping feature does not work as described, please refer to the documentation for your DAW.

Compander
No tape machine plugin should be without a noise reduction encoder/decoder! Use these for
coloration effects or e.g. decoding old cassettes which have been recorded with noise reduction.
Satin includes 5 types – all models of well-known hardware, despite the “obscured” names...

A-Type
First implemented in early professional video recorders, this type also became the noise reduction
standard for multitrack tape and, to a lesser extent, optical movie soundtracks. A-Type processes
four different frequency bands, with the two higher bands overlapping so that typical tape hiss
frequencies are companded more strongly. The A-Type typically provides about 12 dB of noise
reduction (A-weighted). Click here for a related article in Wikipedia.
A-Type Mod
This type mimics the 'Cat-22' modification that was popular among users of the original A-type
hardware from the early 1970s. All four bands were realized on a single card, and the signal that
was added to (or subtracted from) the main path was mixed via four resistors. By simply cutting or
desoldering the resistors of bands 1 and 2, only the treble bands remained active, resulting in the
very 'airy' sound heard on numerous hit records. See the dolby trick.
A-Type Mod works especially well with vocals, acoustic guitars or anything that would benefit from
a very 'up-front' and bright top end without sounding harsh or shrill.
B-Type
This was added to countless consumer products, for instance pre-recorded compact cassette
tapes. B-Type is a single-band system that only processes high frequencies. With its relatively mild
compansion, B-Type typically provides about 9 dB of noise reduction (A-weighted).
uhx Type I
This 2:1 compander was meant for professional systems using tape with a signal-to-noise ratio of
at least 60 dB and a flat (+/-3 dB) frequency response from 30 Hz to 15 kHz.
uhx Type II
This type was meant for cheaper, more noisy consumer media with a more restricted frequency
response. Type II rolls off high and low frequencies in the control signal path (the sound isn’t
affected) to desensitize the system to frequency response errors.
Note: Both uhx types were improved for Satin version 1.3.1. If either is selected, the Legacy Mode
button will appear – if you prefer, you can switch this off before resaving.
Same as Enc
Sets the decoder option to the same as the encoder.

Mix
This control should normally be set to maximum, but you can set lower values to soften the effect,
especially when using the encoder circuit as an enhancer (see dolby trick). Note: This ‘noise
reduction’ can only reduce on-tape noise, not noise generated before the plug-in. This also applies
to ‘system noise’ generated by the analogue-modeled circuitry.
TOC intro presets studio delay flange service config tricks plots nks gloss 23
 LOWER PANEL

Delay Mode
With so much ‘tape goodness’ already built into Satin, it made sense to include a tape delay mode:

Delay mode panel with two repro heads

Delay mode panel with four repro heads

Repro Heads
The sliding switch at the top left selects whether you will be presented with 2 or 4 repro heads for
the delays. Selecting ‘2’ here saves a few CPU cycles and memory.

Speed ips
How quickly, in inches per second, the virtual tape is drawn past the heads. Faster tape speed
usually means better overall audio fidelity (especially with high frequencies) and less noise. Also,
the lower frequency limit moves up slightly. How these effects arise is described in the chapter
about Satin’s service panel. Note: The most common tape speeds are 7.5, 15 (default) and 30 ips.
Note: Unless Tempo Sync (see below) is active, delay times will depend on the current tape speed:
the faster the tape, the shorter the delays. While Tempo Sync is on, Speed ips only affects the
colour of the sound (of course this would be impossible using a real tape machine).

Tempo Sync
In Tempo Sync mode, delay times are relative to song tempo. Distances between the repro and
recording heads are set in musical note lengths relative to a tape speed of 15 ips). See Distance.

TOC intro presets studio delay flange service config tricks plots nks gloss 24
 LOWER PANEL

Routing
With the typical Multi-Mono, each stereo channel (L & R) feeds the output back into its own input.
Selecting Cross instead feeds each output to the other input. Ping-Pong sums the stereo signal to
mono first, then feeds this sum to either the left or right input. Note: if the Balance (see below) is
set to the extreme opposite stereo channel from the Ping-Pong, you won’t hear any delay.

Distance
These sliders set the distance (in inches) between each repro head and its record head. The delay
times depend on the current tape speed except in Tempo Sync mode...
When Tempo Sync (see previous page) is active, clicking anywhere on the track of the slider then
dragging left or right will snap to 16th-note values. Note that the centre position corresponds to
4/16 i.e. a quarter note. When the slider is clicked and dragged in the normal manner, you can set
arbitrary values and/or fine tune them by holding down SHIFT beforehand.

Mod Rate / Mod Amount

The distance i.e. length of each delay can be independently modulated by a sine wave LFO. Use
this feature whenever you need extra movement or more ‘organic’ textures. Careful editing of the
modulation parameters here can even result in stunning, reverb-like dispersion effects.

Balance / Level
The Balance knobs shift the delayed signals to either side of the stereo field, and the Level knobs
set the output volumes for each delay.

Mix
The Mix knob adjusts the ratio between unprocessed (dry) and processed (wet) signal. This control
is only useful when Satin is used as an insert effect (as it usually is). When used as a send effect
you should keep the mix at 100%. The Lock feature can come in handy here: While browsing
through a bunch delay presets that were created for use as insert effects (with e.g. 50% mix),
locking Mix at 100% will save you the trouble of adjusting it each time.

Feedback
Delay feedback level. Note that the feedback path is global i.e. the sum of all delays is fed back
into Satin’s input. With each iteration, the feedback signal is affected by EQ and other parameters.

Limit
Although Satin has ‘natural’ tape compression (important for typical tape-delay sounds!), we added
a dynamic limiter circuit to the output of the feedback stage, to be on the safe side. Imagine having
dialed in a complex echo structure using all 4 delay taps: The tape should inhibit very fast build-
ups, but it will often saturate too quickly. The limiter ensures that levels remain moderate, and can
even create interesting 'breathing' effects when driven harder.

Low Cut / High Cut

Cutoff frequencies of a dual filter within the feedback loop. The displayed values are in Hertz (Low
Cut: 20 Hz to 2 kHz, High Cut: 200 Hz to 20 kHz). For ‘un-filtered’ delays, set Low Cut to minimum
and High Cut to maximum.

TOC intro presets studio delay flange service config tricks plots nks gloss 25
 LOWER PANEL

Flange Mode
Tape flanging is actually very different from what is offered by effect units that sweep a resonant
phasing effect via LFO. Real tape flanging is more of a one-shot thing, a dramatic whoosh effect
for those special moments in the song. Early examples can be heard in the middle of the song
Itchycoo Park (The Small Faces, 1967), or Bold as Love (Hendrix).

True flanging requires two tape machines running identical recordings. The machines are started
at approximately the same position, but while tape1 runs at a constant speed, tape2 is slowed
down or sped up manually by applying force on the ‘flange’ of the supply reel. As tape2 can be
running either behind or ahead of tape1, the delay between the two can approach ‘zero’ from either
side: That is what Satin emulates.
Tape2 is best kept silent for most of the song, assuming you don't want to hear phase artifacts all
the time. As the two tapes approach perfect alignment, the level of tape2 is faded in until the
volumes match. As tape2 moves beyond alignment, its volume is faded out again.
Obviously, doing this on real tape machines not only takes skill but also hours of practice. Satin’s
Flange mode can do it automatically, you just have to activate the Trigger (see below).

Speed ips
How quickly, in inches per second, the virtual tape is drawn past the heads. Faster tape speed
usually means better overall audio fidelity (especially with high frequencies) and less noise. Also,
the lower frequency limit moves up slightly. How these effects arise is described in the chapter
about Satin’s service panel. Note: The most common tape speeds are 7.5, 15 (default) and 30 ips.

Trigger
This button initiates an automatic flange sweep. The indicator moves forwards or backwards (tape2
is sped up or slowed down). Remote-controlled triggering is possible after midi-learning the button.
MIDI note-on messages received by Satin will also trigger the automatic flange.
Note: If the Flange is re-triggered while the indicator slider is still moving (as a result of a previous
trigger action), the direction of movement is reversed.

Fade
Satin’s automatic flange has separate fade-in and fade-out stages which can be set to different
durations. You can select absolute times (0.1s, 1s, 10s) or host-synchronized ‘note lengths’ ranging
from 1/64 to 8/1. These are only nominal values, as they can be modified via Multiply...

TOC intro presets studio delay flange service config tricks plots nks gloss 26
 LOWER PANEL

Multiply
The multiply control shortens or lengthens the values set by the Fade selector – as a multiplication
factor for each fade stage with a range from 0.1 to 2 times the Fade value.

Shape
These two knobs affect the curvature(s) of the automatic flange effect. Experimentation is key
here: The exponential curve has a slower rise during fade-in and a slower decay during fade-out,
while the logarithmic curve has a faster rise during fade-in and a faster decay during fade-out.

Range
Although it can emulate just about everything else a pair of tape machines can do when used as
an effect, a processor plug-in simply can't speed up one of the tape machines ‘into the future'.
However, it suddenly becomes possible if tape1 is already delayed ‘into the past’...
It works like this: Tape1 is set to a fixed delay. Tape2 moves from zero delay (tape1’s future)
through perfect alignment (tape1’s present) to twice the delay of tape1 (the past). Or the other way
round. In either direction, the two recordings cross in the middle for true through-zero flanging.
The Range knob determines the maximum delay between the two machines, and also sets the first
machine to a static value – exactly half of Range. For instance, if you dial in 10ms, the two tapes
will meet at 5ms (the static delay of tape1).
Note: Tape1’s static delay is also the latency of the flange process. Satin reports this latency to the
host software dynamically, but the Range parameter is not automatable. How your host application
reacts to reported latency is beyond Satin’s control – if there is a noticeable lag, please contact the
developer of your host application.

Phase Invert
Running two tape machines in parallel would normally mean twice the volume if the two signals are
perfectly aligned (as they are in the middle of the flange). Activating the Phase Invert button flips
the phase of tape2 so that the signals tend to cancel each other out instead.
Tip: For smoothest results, turn wow & flutter down to minimum.

Manual flanging
Instead of triggering the automatic flange, you can grab the slider and move it manually (of course
this control can also be midi-learned).

TOC intro presets studio delay flange service config tricks plots nks gloss 27
 SERVICE PANEL

Service Panel
Click on the Service button at the bottom of Satin’s window to open a panel containing several
‘under the hood’ controls (as well as a small but perfectly usable analyzer).

Satin’s service panel gives you access to much more ‘esoteric’ parameters
Note: The service panel is closed by default to save on CPU – the analyser actually uses a hidden
instance of Satin so it can accurately display frequency response.
To understand everything in service panel, it’s best to learn a few basics of tape technology...

Tape Machine Primer

Record head
Getting audio onto tape requires a ‘record head’, basically a ring-shaped core of ferromagnetic
material with a coil of wire wound around it. At the front end of the core, in contact with the tape, is
the so-called ‘head gap’, a narrow space between the poles which turns the head into a kind of
horseshoe electromagnet. An input signal is fed into the coil, inducing a magnetic field which then
emanates from the head gap and realigns the magnetic state of the tiny metal oxide particles
embedded in the tape being dragged past the head. Sounds easy enough, doesn’t it? Doesn’t it?
Besides being difficult to understand and even harder to explain, there are practical problems. One
of them is that you need quite a strong field to budge the normally random magnetic orientation of
the metal oxide particles at all – quieter parts of the signal will affect fewer particles and the fidelity
of the recording will be (or would be) relatively poor.

DC bias
In the early days of tape recording, the late 1930s, adding some direct current (DC) bias improved
fidelity by orienting the magnetic state of all particles in one direction. Unfortunately, DC bias
practically halves the usable dynamic range of the tape. Another disadvantage is that it leaves the
tape with a net magnetization, causing more noise during playback. Still, it was better than nothing.

AC bias
In 1941, two German engineers working for the national broadcasting organization RRG (Reichs-
Rundfunk-Gesellschaft if you must know!) accidentally discovered alternating current (AC) bias,
which resulted in much better sound quality. The story goes that one of their tape machines was
producing exceptionally good recordings. Mystified at first, tests revealed that its DC bias amplifier
was generating a high frequency oscillation – that part of the machine was definitely faulty!
The trick of AC bias is to have the audio signal ride (like AM radio) on a high frequency carrier
signal, a sine wave somewhere between 50kHz and 200kHz. This frequency is high enough to be
filtered out naturally by the repro head, but strong enough to budge all the oxide particles out of
their lethargy and allow the audio signal to ‘fine tune’ their magnetic orientation. The usable
dynamic range is not affected much: with AC, everything oscillates around zero so there is no net
magnetization on the tape.

TOC intro presets studio delay flange service config tricks plots nks gloss 28
 SERVICE PANEL

Magnetic tape
All tapes, from 1/8-inch cassettes to 2-inch studio reels, have at least two layers: Firstly, a smooth
film which gives the tape its physical strength and flexibility. Originally paper, soon afterwards
acetate (fire hazard!), since the 1960s mainly polyester.
Secondly, the magnetic coating: tiny iron or chrome oxide particles embedded in a polymer
‘binder’. As this layer is in direct contact with the record and repro heads, its physical properties
affect how smoothly the tape will run. The density, size and distribution of oxide particles in the
binder determine the audio quality and maximum dynamic range of the recording / reproduction.

Repro head
The repro head is physically very similar to the record head, but it works in reverse: When the tape
is drawn past the gap in the repro head, the magnetic pattern on the tape induces (via the head
gap) a corresponding voltage in the repro head’s coil.
The repro head has a kind of highpass filter effect: higher frequencies mean higher voltage (like a
bicycle dynamo: the faster you pedal, the brighter the headlamp). However, there’s also a lowpass
effect affecting the high frequencies: wavelengths shorter than the head gap suffer significant loss,
for reasons too complex to discuss here.
Bass frequency response is influenced by the so-called ‘head bump’ effect, a low frequency boost
caused by resonances having to do with the size of the repro head’s core relative to its gap. Note
that the frequency of the head bump follows tape speed (see response curves).

Equalization
All tape recorders require a lowpass filter behind the repro head, which fixes the following problem:
Tape magnetization is directly proportional to the recorded signal, but the voltage induced in the
repro head’s coil during playback is proportional to the rate at which this magnetization changes.
So for the same degree of magnetization on the tape, a signal with twice the frequency will be
played back twice as loud.
Although the lowpass filter is necessarily shallow (about 6dB/octave), its cutoff frequency needs to
be set quite low so it can affect a wide enough frequency range. Such broadband filtering results in
a significant loss of overall volume (typically 50 to 60dB), which therefore needs to be amplified
afterwards. So the simplest acceptable signal flow would look like this:

input tape amp output

record repro lowpass
head head filter
bias

Properties such as tape thickness or amplifier response, as well as HF losses caused by the repro
head gap, also need to be compensated. There are more idiosyncrasies to be considered (self-
resonance in the head coils, eddy current losses in the cores, tape-head spacing losses and so
on), but we don’t really need to learn such details. Suffice it to say that Satin has all the noticeable
quirks of real tape machines, but doesn’t make them all user-adjustable – lucky you!
On-tape HF losses and head gap damping can be partly remedied by a kind of pre-filtering called
‘emphasis’. This isn’t the perfect solution: The louder the input signal, the more saturation you will
get on tape, especially of those high frequencies boosted by the emphasis circuit.

TOC intro presets studio delay flange service config tricks plots nks gloss 29
 SERVICE PANEL

Not surprisingly then, attempts were soon made to fix the HF response at the end of the signal
chain – at the cost of the some extra circuitry and a bit more tape noise. With the two extra filter
circuits in place, the signal flow now looks like this:

input record tape amp repro out

EQ record repro LP EQ
head head filter
bias

Finding the optimum balance between record-EQ (emphasis) and repro-EQ certainly delivers the
best sound quality – but how would those tapes sound when played back on a different machine?
Wouldn’t each manufacturer try to get the best out of their own designs, and ‘hang the rest’? This
problem was solved by setting frequency curve standards such as NAB or CCIR (later IEC).
Satin’s service panel lets you tweak several properties at different points within the signal chain.
The service panel is divided into three sections, each containing related parameters...

Tape Parameters
The area on the left of the service panel contains five controls under the heading ‘Tape’:

the 5 ‘Tape’ parameters in the service panel

Hiss
Controls the level of tape hiss. Although considered a serious disadvantage of tape, a modicum of
hiss not only adds ‘retro’ authenticity and dither to the sound, but can also e.g. reduce the
artificiality of existing cuts in comped vocal tracks. Note that Satin’s tape hiss is stereo, just like a
2-track tape machine. See Auto Mute below.

Auto Mute
Soft-gating for both hiss and asperity: If no input signal is detected, the tape noise is muted.

Asperity
Imperfections in the surface of the tape cause a roughening / blurring of the recorded frequencies.
Although they also cause some constant mid-range noise, asperity noise is mainly perceived as an
enharmonic distortion of the recorded signal. How noticeable this turns out to be depends on the
frequency content of the original signal: if the signal contains very few high frequencies, you should
notice an accompanying noise ‘tail’ when you turn the parameter up. Like ‘hiss’, asperity noise is
also stereo. Strangely, asperity noise can make monaural recordings sound as if they were stereo!

TOC intro presets studio delay flange service config tricks plots nks gloss 30
 SERVICE PANEL

Crosstalk
For ‘perfect’ stereo or multitrack recordings the tracks should be completely independent at every
stage – there should be zero crosstalk between adjacent tracks. Although electronic circuits in a
single device can never be 100% isolated, crosstalk is seldom noticeable if the circuits are well-
designed. Wherever magnets are involved, however, crosstalk can become a real problem.
Adjacent tracks of tape machines are very prone to induction, eddy currents and other unwanted
interactions. There is no satisfactory solution.
Crosstalk was included in Satin because it can help to ‘glue’ stereo tracks together. As synching
audio buffers between multiple instances would have significantly increased latency, we restricted
the crosstalk to two (L / R) tracks for each instance. See frequency response plots.

Wow & Flutter

Who really likes wobbly recordings? The transport mechanism in tape machines should be steady,
and manufacturers of the more expensive models took great pains to ensure the smoothest tape
flow possible. A little variance in the recording and playback speed is inevitable, however, and the
‘wow & flutter’ effect can even contribute to the charm of tape recordings. Satin’s Wow & Flutter
parameter has a very wide range, from unnoticeable at minimum all the way up to ‘horribly wobbly’
at maximum. Note that the speed of wow & flutter follows the tape speed.
Tip when using Flanger mode: For smoothest results, turn Wow & Flutter down to minimum.

Bias
Arguably the most important control in this part of the service panel, bias is crucial for attaining a
clean ‘recording’. Bias adjustments are normally made to flatten the tape’s response curve.
Because any ‘optimum’ bias setting is necessarily a compromise, you’re free to use bias creatively:
lower values tend to improve the high frequency response at the cost of more distortion wherever
magnetization is low (i.e. around where the waveform crosses zero). Higher values have roughly
the opposite effect i.e. you get lower distortion but less highs.
Setting bias to 0.00 by double-clicking on the control will give you what we determined to be the
best compromise – a fairly flat response without too much distortion. Note that some engineers like
to run their machines moderately over-biased, which delivers slightly lower distortion and a cleaner
low-end at the cost of a more dampened treble response. See frequency response plots showing
the effects of different bias settings.

Repro Head Parameters

The central block in the service panel lets you adjust physical attributes of the repro head...

TOC intro presets studio delay flange service config tricks plots nks gloss 31
 SERVICE PANEL

Gap Width
Of course real tape machines don’t have a gap width parameter – you would need to find the right
screwdriver and physically exchange the repro head to do that. Gap width determines both the
upper and lower frequency thresholds, it affects the amount of high frequency ‘gap loss’ as well as
low frequency resonance (see Bump below).
At first glance, Gap Width appears to act like a simple tilt filter, but in fact it introduces various
complex resonances and fluctuations across the entire audio spectrum. In conjunction with the
Bias control, Gap Width can be used to balance the frequency response, as these two controls
have roughly opposite effects on the signal. For instance, if you have set the bias fairly high, try
setting a smaller gap (note that this will also shift the lower frequency threshold upwards a bit). Tip:
If you want to imitate the frequency response of classic studio tape machines, set the gap width to
somewhere between 2 and 3 µm (micrometers). See the frequency response plots.

Bump
Low frequency response is chiefly determined by the physical design of the repro head. However,
just like the corrective elements found in some loudspeaker units, low frequency resonance effects
in real tape machines can be dampened.
Satin’s Bump parameter (usually called “head bump”) controls the amount of low frequency
resonance allowed: lower values mean more correction and a fairly shallow roll-off below the
bump. Conversely, higher values mean less correction, and therefore more fluctuations and
resonance build-up. See the frequency response plots.

Azimuth
Azimuth error (measured in 'arc-minutes') is emulated when the knob is dialed off-center – the tape
head is skewed slightly away from the vertical. Although usually unwanted on real tape machines,
this can sound interesting, especially when using the delay mode or when spatial repositioning of
sounds is desired.
Ideally, a magnetic recorder has its record and repro head perfectly aligned, the angle between the
head gaps and the tape is exactly 90°. If either head is just a bit off, the magnetic pattern of the
recording on tape won't be reproduced correctly upon playback. A mismatch in azimuth results in a
'washy' sound, with poorer treble response and with stereo signals tilted to one side.
In Satin, the repro-head's azimuth can be adjusted per stereo channel (i.e. not across the many
tracks possible while using multiple instances). What happens is that the pattern transported along
the head is no longer time-aligned. Since all tape heads are stereo here, mono components of the
signal hits one of the gaps later, and this short delay seems to shift the stereo image to one side or
the other. Audio engineers call this the 'Haas effect': our hearing system determines the spatial
location of an audio source by the difference in time/phase the wavefront reaches our ears. When
summed to mono, this delay also creates a comb-filter effect, colouring the sound quite drastically.
Another factor related to azimuth error is the head gap itself. Originally perpendicular to the tape,
increased skewing leads to a wider effective head gap. The wider gap can't pick up short
wavelengths as well as narrower gaps, leading to even more treble loss. The wider the track width,
the worse this becomes. Note: Satin simulates a 1" two-track head arrangement here.
Azimuth error also depends on the tape speed: the slower the transport, the more apparent the
effect. Machines running at 30 ips are less prone to azimuth error as the resulting delay will be
much shorter than at, say, 7.5 ips. Also, the gap-widening effect is less significant at higher speeds.
Although azimuth affects all repro heads, in Delay mode (only) the angular change alternates left/
right, which means that if you set azimuth to a positive value e.g. '+2', then heads 1 & 3 will turn to
the right while heads 2 & 4 will turn to the left. As Delay mode uses multiple repro heads / taps,
adjusting the Azimuth value can deliver some rather interesting spatial smear effects!

TOC intro presets studio delay flange service config tricks plots nks gloss 32
 SERVICE PANEL

Analyzer
Typically, when a tape recorder needs to be realigned, calibration tones are sent to the machine
and the results are monitored with a voltage meter or a built-in metering system. This is a tortuous,
time-consuming job, but since we're talking software, things are a lot easier in Satin...

The analyzer continuously measures frequency response. As we weren't keen on having test-tones
sent to the audio-outputs, we gave the analyzer its very own copy of the entire DSP part!

Plot mode
Right-click in the analyzer window to select one of three plotting modes. The default is glow, which
shows inertia just like a real analogue oscilloscope. The fast option is a low-CPU variant with no
blurring. The eco option is even lighter on the CPU, but updates at a much slower rate.

Circuit Parameters
Finally, the service panel lets you virtually ‘replace’ parts of the tape machine’s circuitry:

Circuit: EQ options and headroom

Rec EQ / Repro EQ
The choice of EQ is crucial for achieving clean, undistorted tape recordings. However, results still
depend on a variety of parameters and internal factors – there is no such thing as 'the perfect EQ
curve'. Satin offers a variety of industry-standard EQ curves. Of course you can choose a particular
curve to mimic a specific machine, but you could also get more experimental...
The rec/repro EQ curves were originally meant to be used in matched pairs – the same standard is
chosen for both the recording and reproduction process (the repro curve compensates for the
recording curve by applying the inverse function). For Satin, we thought that allowing them to be
selecting separately would not only be fun, but it would also let you correct strong frequency errors.
For matched pairs, set the Repro option to Same as Rec.
Having separate EQ curve selection also means you can use Satin as a format converter: if you
have transferred a real tape recording to your DAW from a machine with an unsuitable EQ, you
can select your machine's EQ from the list as Rec EQ, then set the target curve as Repro EQ.
Note: While using Satin to convert formats, you should activate bypass tape.

TOC intro presets studio delay flange service config tricks plots nks gloss 33
 SERVICE PANEL

Flat Disables equalization in this part of the circuit. Although we can't think of any
professional unit that doesn’t have a corrective circuit, the Flat option is there for
whenever you need as much high-frequency saturation as possible (for instance
when taming transients). Please be aware that using no EQ during recording
quickly leads to treble distortion, hence some serious slew-rate limiting.

Note: The IEC curves (formerly known as CCIR) for magnetic recording exist in two flavours,
and these are usually the best choice for tapes running at fairly slow speeds.

IEC 7.5 ips IEC 7.5 ips cuts off frequencies above 2275Hz (-3dB point), which is rather low.
Of course the primary focus in the past was to reduce distortion as much as
possible – lower speeds generally mean more distortion. Using IEC 7.5 ips
came at the price of reduced signal-to-noise ratio i.e. more tape hiss, but you
can turn that down in Satin anyway. Although IEC 7.5 ips is the most drastic
option, it has the advantage of retaining transients well. It also has the least HF
loss and therefore the least amount of slew-rate limiting.

IEC 15 ips The main difference between this option and 7.5 ips curve is that the corner
frequency is set to 4550Hz – exactly twice the frequency of IEC 7.5. This is no
coincidence, in fact most of the underlying issues and factors are frequency-
linear. Since IEC 15 outweighed distortion and noise floor quite successfully, it
became the de-facto standard for mastering-grade tape recorders.

NAB A long-running contender to IEC 15 ips, the NAB standard has (theoretically) a
higher noise floor while preserving more HF energy. Its regular corner frequency
is 3150Hz, but... what’s up with the bass? There’s a huge 3dB boost at 50Hz,
which grows stronger at lower frequencies, finally reaching a massive +8dB at
20Hz. This behaviour was designed to compensate for the typical LF losses
inherent in standard equipment of those days. The NAB repro EQ curve corrects
the bass boost admirably (which the other standards don’t really), but the
tendency towards LF distortion remains. Despite all the drawbacks, NAB can
create a low frequency 'oomph' that is just plain sexy!

AES 30 ips In this standard, the treble is cut at 9100Hz during recording. This frequency
happens to be very suitable for mastering applications, provided a high-output /
low-saturation tape and the correct tape speed (30 ips) are used. Possible
drawbacks: AES 30 ips can be more sensitive to very high input levels, and
there is some risk of transient-smearing.

Same as Rec Sets the playback (repro) EQ to the same option as the recording EQ.
(repro EQ only)

Note: In accordance with the standards, all EQ curves attenuate at the relatively shallow slope
of 6dB per octave, so none of them should cause significant phase problems.

headroom
By default, the headroom of the record and repro circuit is set to 9dB above the 0dB tape level,
leaving some room before you start to hear the circuit distorting. The amount of this distortion is
indicated by the brightness of the blue ‘LED’ above the ‘Headroom’ label.
For maximum tape saturation and circuit transparency, we recommend setting high Headroom
values. If you're aiming for a gritty, brickwall-type distortion, you could try a low value – but please
be aware that you would lose much of the tape ‘liveliness’.

TOC intro presets studio delay flange service config tricks plots nks gloss 34
 CONFIGURATION

Configuration

Clicking on the cogwheel icon at the top right opens the global configuration pages where you can
adjust the appearance of Satin’s window (size and brightness) as well as assign MIDI continuous
controllers to any Satin parameter.
The 4 buttons are MIDI Learn [L], MIDI Table [≣], Preferences [tools] and Close [X]:
Tip: Right-click within the button area to set the currently selected page as default.

MIDI Remote Control

MIDI Learn

This page lets you assign MIDI CC (‘control change’) to individual parameters. The CC data can be
generated by hardware knobs / sliders, or from MIDI tracks in the host application. For how to
route MIDI into effect plug-ins, please refer to the documentation of your host application.

The translucent overlay displays all MIDI-learnable elements as selectable outlines. Some of
Satin’s controls won’t be immediately available in the MIDI Learn page – you will have to choose
between Studio, Delay or Flange mode first.
TOC intro presets studio delay flange service config tricks plots nks gloss 35
 CONFIGURATION

Try it: Click on the Input knob and send Satin some MIDI CC data – move a knob or slider on your
MIDI controller and the connection is complete. Switch to the MIDI Table page to check that the
new connection has appeared in the list…

MIDI Table

Selecting the ‘≣’ MIDI icon button opens an editable list of all MIDI CC assignments:

Parameter
The first field selects one of Satin’s parameters, which are sorted into sub-menus. Try it: Click on
the ‘ADD’ button at the bottom left of the MIDI table window and experiment with the Parameter
field. Then remove the assignment by clicking on the small [x] to the right of the line.
Channel / Controller
The next two fields specify the MIDI channel and CC number.

Mode
Specifies the range and/or resolution of values.
Normal .........................Full range, continuous
Integer .........................full range, whole numbers only
Fine ...............................0.01 steps between the two integers closest to the current value
Type
Specifies the type of hardware. By far the most common is Continuous 7-bit.
Encoder 12 .................‘relative mode’ endless knobs that repeatedly send the CC value 1 when
turned in the positive direction, or 127 (interpreted as -1) when turned in the
negative direction
Encoder 64 .................‘relative mode’ endless knobs that repeatedly send the CC value 65 when
turned in the positive direction, or 63 when turned in the negative direction
Continuous 7-bit .......7-bit MIDI CC (standard resolution)
Continuous 14-bit .....14-bit MIDI CC (high resolution, rare)

Remove
To remove individual MIDI CC assignments, click on the small [x] to the right of the line. To remove
all assignments at once, click on the Delete All button at the bottom of the window.

TOC intro presets studio delay flange service config tricks plots nks gloss 36
 CONFIGURATION

Preferences Page
Click on the ‘tools’ icon to open the Preferences page, where you can set several global
defaults to suit your computer mouse and monitor:

Controls
Mouse Wheel Raster
If your mouse wheel is rastered (you can feel it
clicking slightly as you roll the wheel), set YES
so that each little click increments / decrements
in sensible value steps.

Appearance
Analyzer
The appearance of the response display in the
service panel: eco is particularly CPU-friendly.
Default Size
Sets the default GUI size for each new instance.
You can temporarily change the GUI size
without entering the Preferences page – simply
right-click in the background.
Gamma
Adjusts the GUI brightness (darker, dark,
neutral, bright or brighter).
Text Antialiasing
Switches the smoothing of labels and values on
or off. Normally left on.

Presets
Auto-Versioning
If ‘on’, an index is automatically appended to the preset name and incremented each time it is
saved. Saving ‘Space’ three times in a row would give you three files: ‘Space’, ‘Space 2’ and
‘Space 3’.
Save Presets To
Choose the user folder option if you want all saved presets to land in the User folder instead of the
currently selected one.
Scan On Startup
This determines whether the preset library should be scanned and the database recreated when
the first instance of Satin is started, e.g. when you reopen a project.

TOC intro presets studio delay flange service config tricks plots nks gloss 37
 CONFIGURATION

Others
Base Latency
If you are certain that your audio system – hardware as well as software – uses buffers that are a
multiple of 16 samples in size (please refer to the respective documentation), you can safely
disable Satin’s base latency. Otherwise leave it set to the default ‘16 samples’, which will prevent
audio drop-outs. Note that switching Base Latency will only take effect when the host allows e.g.
on playback or after switching the sample rate. Reloading Satin will always work.

MORE ABOUT BUFFERS

Internally, Satin processes audio in chunks of n x 16 samples. The ‘block processing’
method significantly reduces the CPU load and memory usage of all our plug-ins.
For example. if the number of samples to be processed is 41, Satin will process the first 32
and keep the remaining 9 in a small buffer (16 samples is big enough). Those 9 samples are
then processed at the start of the next call… and so on.
The extra buffer is only necessary if the host application or audio driver processes ‘unusual’
audio buffer sizes. Many hosts process buffers of 64, 128, 256 or 512 samples (all multiples
of 16), in which case you can try switching it off so that Satin can work latency-free.

TOC intro presets studio delay flange service config tricks plots nks gloss 38
 TIPS & TRICKS

Tips and Tricks

Here are a few tips that we and our beta testers noted during development, as well as some old
tricks rediscovered by testing ‘typical use’ cases...

Central control from a single instance

Even if all the Studio mode instances of Satin in your project are independent i.e. no grouping was
required, you might like to assign each one to it’s own group anyway, and name them accordingly.
Having done that, you can add a ‘dummy’ instance (inserted into its own ‘dummy’ track) to control
all the others, which can then remain hidden. Tip: To save CPU, lock and activate ‘Bypass’.

Taming transients
In the golden age of tape, recording engineers were well aware of how their machines and reels
would compress and colourize the recordings. To add some colour to the ‘digital perfection’ of
today, there is a tendency to slap a lot of compression, EQ and other effects onto all the tracks –
much of which may be unnecessary when printing to tape (virtual or otherwise).
Extracting the record and repro EQ from the overall chain, we can directly operate on the main
places where the balance between distortion and minimum noise is achieved. And since there are
no noise issues to worry about in Satin, we are free to choose the curve that best suits our needs.
To find the 'right' one, we need to check the spectral content of the incoming audio. If we want to
preserve as much transient information as possible, we need to apply pre-filtering that rolls off the
HF as low as possible, then do the opposite during playback.
The more high frequencies we have while recording directly to tape i.e. without using filters, the
more tape saturation there will be. This is precisely the case when both EQs (record & repro) are
set to 'flat'. For transients, this means maximum slam. We can exaggerate it even more by adding
some HF pre-emphasis. We could also misuse the head-gap control as a tilt filter to compensate
for the added brightness (although we'd be messing with the entire response curve).
Always check the peak meters – you should notice a small gain reduction on the transients. What
remains is the problem of ear fatigue: our hearing adapts very quickly to the 'coziness' of milder
transients, and if you overdo this trick, you will end up with a massive HF smear.
For plots showing high-frequency saturation, go here.

Multitrack coherence (the ‘glue’ effect)

The more tracks you process with a real multitrack tape machine or multiple instances of Satin, the
more 'tape' sound you will notice. This cumulative effect is often described as 'glue'.
An important factor was already discussed above – the transients taming / smear effect. When
applied several times in parallel, the results become softer more easily i.e. you don't have to aim
very high on each individual track. The process is often barely noticeable when you solo a track,
the magic starts when they all add up. Here's one recipe that can improve coherence – do the
following on multiple channels, with the grouping feature enabled:
• Choose a suitable tape speed. Remember that faster tape speeds generally mean more treble,
but also some loss of ‘weight’ in the bass.
• Choose a tape type according to how much distortion and 'roundness' you would tolerate.
• Set both the Rec EQ and Repro EQ to a type that still retains plenty of transient information, but
also offers some slight taming of the occasional overload.
• Apply a minimal amount of tape hiss. A decent studio recorder has about -70 to -60 dB.

TOC intro presets studio delay flange service config tricks plots nks gloss 39
 TIPS & TRICKS

• Add some asperity / modulation noise to make things a bit more random & rough, especially in
the stereo field. Be careful with this, as bass frequencies can reveal a prominent ‘noise tail’ very
quickly. This probably isn't wanted on every channel, or at the same amount. If in doubt, leave
the asperity control out of the group by locking it.
• Add a minuscule amount of wow & flutter. Set it so that you don’t notice any warbling from pure
sustained notes or 'difficult' instruments such as pianos.
• Adjust the output gain so that the music appears to have the same volume when you bypass
the plug-in. Make sure you don't drive each instance too hard – as distortion is introduced very
gradually in Satin, there will already be plenty of saturation and extra harmonics when you
approach the 0dB mark.
• Keep comparing your results by switching to bypass. Listen to the transients, spectral balance,
fullness / richness, stereo width and any other differences. Try to spot compression effects or
even differences in the front-to-back localization.
• Review and qualify your impressions: Is the 'tape processed' sound really better? Have you
overdone it, have you worsened the signal integrity? Do you still have enough transients, is it
still crisp enough? If in doubt, simply back off a bit and compare again!

Tape reverb!
Satin’s 4-tap tape delay is capable of some interesting reverb-like effects. Try this:
• The delay times should not be integer multiples of one another. They should be either prime
numbers or unrelated fractions. To do the math, although the delay range is only 0 to 8 inches, it
is best to multiply this by a constant first. For instance, if we multiplied by 10, we would have a
range of 0 to 80. Prime numbers within that range are e.g. 5, 17, 37, 43, 53, 73. Those values
can then be divided back into the 0 to 8 range: 0.5, 1.7, 3.7, 4.3, 5.3 and 7.3 (inches).
• Fix sensible levels e.g. tap1 = 1.0, tap2 = 0.5, tap3 = 0.25 and tap4 = 0.125: an exponential
decay in our particular delay arrangement.
• Now set the balance knobs to arbitrary positions (it doesn't really matter much at this point).
From the 'Routing' menu, we will choose 'Cross' for maximum stereo width.
• Now for some modulation. For the most ‘dispersive’ effect, use faster modulation on the shorter
delays and slow modulation on the longer ones. Experiment!
• The feedback levels should be somewhere between 0% and 50% for our chosen tap levels,
otherwise the delays are likely to ramp up instead of down.
• In the feedback path, higher frequencies should be damped by lowering the ‘High Cut’ to a
natural-sounding value. Typically, the larger the simulated space, the less HF would remain. If
needed, the bass can be cut a bit, thus making the sound being less troublesome within a
complex arrangement.
• Bingo! You should have a nice 'tape reverb' already – save it?

Format conversion
Decoding old tapes: Imagine the following scenario: You find a stack of old reels in the attic. The
tracks were recorded with NR switched on, but that particular machine is no longer available to
decode the tapes. You know somebody who owns an otherwise compatible machine, but with a
different compander system (or no NR at all). Here’s how you can decode those tapes anyway:
• Make sure the tape machine's NR section is switched off.
• Record the tape directly to your computer.
• Insert an instance of Satin for each track.

TOC intro presets studio delay flange service config tricks plots nks gloss 40
 TIPS & TRICKS

• Select Studio mode, assign all tracks to the same group and activate the Bypass Tape
button(thus disengaging Satin’s tape processing, which we won’t be needing).
• In Satin's compander, set the encoder circuit to ‘none’.
• Set the decoder to the type you believe was used to encode the tape – if you’re lucky, this will
be handwritten on the tape box!
• Experiment with Satin’s input level, as this can affect the decoding process. The incoming signal
should not drive the VU meters into the red.

Correcting equalization: Real tape machines are calibrated to suit certain EQ standards. When a
tape is played back on a machine that is calibrated very differently from the machine used for the
original recording, it is likely to sound much too bright or too dull etc.. Your job is to correct the
'error' caused by using two different equalization systems:
Let's say the original machine was set to IEC/CCIR at 15ips (again: inspect the tape box for
handwritten details) and the machine you are using for playback has NAB calibration. We need to
apply the inverse curve of that frequency error...
• Set Satin to Studio mode and activate the Bypass Tape button.
• In the Service panel, switch the record EQ to NAB.
• Set the repro EQ to IEC15ips.
Tapes processed in this way (first IEC then NAB) have a bass loss and treble gain. Encoding NAB
first then decoding IEC has the opposite effect – a stronger low end and a weaker treble response.
See frequency response plots in the next chapter.

Enhancing vocal tracks

If you push a large professional multitrack recorder hard enough, you probably won’t need noise
reduction as such – at least not for the less up-front tracks like backing vocals. However, it is
precisely these tracks that come alive so beautifully when processed by NR encoders...
Most noise reduction systems use some sort of frequency-dependent dynamics processing, either
in the main signal path, in the detection side-chain, or in both. Some engineers misused encoders
for the so-called Dolby® Trick, with the brand name standing for a broader range of systems.
• In Satin’s compander, set the encoder to 'A-Type Mod', but leave the decoder off. Also try
bypassing the tape and experimenting with the Mix parameter. Owners of the original hardware
units generally used them in an AUX bus – with Satin, you can simply adjust the 'mix' control for
the amount of enhancement you want.
• Listen to how the quieter parts become more airy, while the frequency response remains flat
during the louder passages...
Back in the days of portable cassette players, many wired-for-sound joggers preferred listening to
their encoded tapes with the NR decoding switched off: The higher noise floor didn’t matter too
much – it was worth it for that extra treble!

TOC intro presets studio delay flange service config tricks plots nks gloss 41
 PLOTS

Frequency Response Plots

This chapter contains about 30 frequency response curves – for general reference purposes as
well as for what to expect while using Satin’s built-in analyzer. You will find several links to these
plots within the main body of this manual.

• Bias & Total Harmonic Distortion – shows frequency response balance vs THD. THD was
measured with a sine wave at 1kHz, which is automatically subtracted from the plots (the 3kHz
peak is the 3rd harmonic of the test tone).

• Bandwidth & Head Bump – shows how the head bump / contour effect and HF response vary
according to head gap and tape speed.

• Crosstalk – the right channel (red) is set to 0dB, the left channel (blue) is set to minus infinity
so that only crosstalk from the right channel appears.

• High-frequency rolloff without pre-emphasis – shows the natural HF damping without

applying treble boost at the input.

• High-frequency saturation – illustrates treble smear when the input levels are high, with and
without proper record/repro EQ. The test signal was a logarithmic sweep. Note that there are no
amplitude errors, but some ripple.

TOC intro presets studio delay flange service config tricks plots nks gloss 42
 BIAS & THD

BIAS & TOTAL HARMONIC DISTORTION

Bias -5, Modern tape at 15 ips, IEC15, -12dB input

Bias -5, Vintage tape at 15 ips, IEC15, -12dB input

THD with bias -5, Modern tape at 15 ips, IEC15, -12dB input

TOC intro presets studio delay flange service config tricks plots nks gloss 43
 BIAS & THD

Bias +5, Modern tape at 15 ips, IEC15, -12dB input

Bias +5, Vintage tape at 15 ips, IEC15, -12dB input

THD with bias +5, Modern tape at 15 ips, IEC15, -12dB input

TOC intro presets studio delay flange service config tricks plots nks gloss 44
 BIAS & THD

Bias optimum, Modern tape at 15 ips, IEC15, -12dB input

Bias optimum, Vintage tape at 15 ips, IEC15, -12dB input

THD with optimum bias, Modern tape at 15 ips, IEC15, -12dB input

TOC intro presets studio delay flange service config tricks plots nks gloss 45
 BANDWIDTH & HEAD BUMP

BANDWIDTH & HEAD BUMP

Gap 2μm, Modern tape at 7.5 ips, IEC15, input -12dB

Gap 3μm, Modern tape at 7.5 ips, IEC15, input -12dB

Gap 4μm, Modern tape at 7.5 ips, IEC15, input -12dB

TOC intro presets studio delay flange service config tricks plots nks gloss 46
 BANDWIDTH & HEAD BUMP

Gap 2μm, Modern tape at 15 ips, IEC15, input -12dB

Gap 3μm, Modern tape at 15 ips, IEC15, input -12dB

Gap 4μm, Modern tape at 15 ips, IEC15, input -12dB

TOC intro presets studio delay flange service config tricks plots nks gloss 47
 BANDWIDTH & HEAD BUMP

Gap 2μm, Modern tape at 30 ips, IEC15, input -12dB

Gap 3μm, Modern tape at 30 ips, IEC15, input -12dB

Gap 4μm, Modern tape at 30 ips, IEC15, input -12dB

TOC intro presets studio delay flange service config tricks plots nks gloss 48
 CROSSTALK

CROSSTALK

Modern tape at 15 ips, crosstalk -20dB

Vintage tape at 15 ips, crosstalk -20dB

RED = right channel at 0dB

BLUE = crosstalk only (by setting the left channel to -infinity)

TOC intro presets studio delay flange service config tricks plots nks gloss 49
 ROLLOFF

ROLLOFF WITHOUT PRE-EMPHASIS

7.5 ips

15 ips

30 ips

TOC intro presets studio delay flange service config tricks plots nks gloss 50
 HF SATURATION

HIGH FREQUENCY SATURATION

Flat EQ, Modern tape, input -20dB

Flat EQ, Vintage tape, input -20dB

TOC intro presets studio delay flange service config tricks plots nks gloss 51
 HF SATURATION

IEC 15 ips, Modern tape, input -20dB

IEC 15 ips, Vintage tape, input -20dB

TOC intro presets studio delay flange service config tricks plots nks gloss 52
 HF SATURATION

Flat EQ, Modern tape, input 0dB

Flat EQ, Vintage tape, input 0dB

TOC intro presets studio delay flange service config tricks plots nks gloss 53
 HF SATURATION

IEC 15 ips, Modern tape, input 0dB

IEC 15 ips, Vintage tape, input 0dB

Note: The 0dB plots were created using sine wave sweeps, the -20dB ones used a Dirac impulse.

TOC intro presets studio delay flange service config tricks plots nks gloss 54
 NKS

NKS
Satin supports Native Instruments “NKS extensions” format so that it can be integrated into the
Komplete Kontrol software or Maschine environments. Satin’s factory presets are optionally
installed as tagged .nksfx files. A few pages of performance controls mapped to common
parameters are automatically generated and saved together with each .nksf preset.

Saving as .nksfx is only possible in the VST2 version

Mac owners can use – temporarily if necessary – any host application that supports VST2

Saving in NKS format

While the native, h2p and h2p extended options cause Satin to save presets into the currently
selected preset directory, nksfx files go directly into the preset location used for Komplete Kontrol
or Maschine (so .nksfx files do not appear in Satin’s preset browser). To make them visible in
Komplete Kontrol, open its preferences and rescan the preset locations.

Batch conversion
First, right-click the [save] button and set the target format to nksfx. Via cmd-click (Mac) or alt-click
(Win), select all presets in the current folder you want to convert, then right-click any of the
selected presets and choose convert to nksf. Note: The original files are not affected.

What to do if Satin doesn't show up in Komplete Kontrol / Maschine.

First of all, make sure your NKS software is up to date: Komplete Kontrol V1.5+ or Maschine V2.4
are the minimum requirements for u-he.
In Windows, Komplete Kontrol must know the Vstplugins folder containing Satin: Open Komplete
Kontrol preferences, go to Locations and add your Vstplugins directory if necessary, hit Rescan
and check whether Satin appears.
Maybe the NKS preset folder is empty? If so, please reinstall Satin with the correct VST path and
the NKS-option checked. Here are the preset folder locations:
Mac MacHD/Library/Application Support/u-he/Satin/NKS/Satin/
Win C:\Users\YOU\Documents\u-he\Satin.data\NKS\Satin\
Perhaps the XML-File is missing from this location:
Mac MacHD/Library/Application Support/Native Instruments/Service Center/u-he-Satin.xml
Win C:\Program Files\Common Files\Native Instruments\Service Center\u-he-Satin.xml
A re-install with the NKS-option checked should also remedy this issue.

What to do if Komplete Kontrol / Maschine is unable to load Satin

Either Satin wasn't installed as VST2, or it wasn’t installed with the correct path. The default VST
path is fixed in MacOSX, but in Windows it can be freely assigned during installation:
Mac: Macintosh HD/Library/Audio/Plug-Ins/VST/u-he/
Win: <User VST Folder> / (path for the VST plug-in set during installation)
If Satin’s VST plug-in cannot be found in one of these locations, run the installer again making sure
that you set the correct path and have activated ‘VST’ as installation option.

TOC intro presets studio delay flange service config tricks plots nks gloss 55
 GLOSSARY

Glossary
A
ACE (u-he product) Any Cable Everywhere – a compact modular synthesizer
AC bias addition of HF alternating current to improve fidelity of tape recordings
AES Audio Engineering Society
aliasing ugly distortion caused by resolution limitations in digital systems
amplitude the level of a signal
asperity roughness or imperfection (of the tape surface)
attenuate to reduce the amplitude
a-weight(ed) measurement offset compensating for how the ear perceives loudness
azimuth see http://en.wikipedia.org/wiki/Azimuth

B
Bazille (u-he product) PD and FM based modular synthesizer
bias a shift / offset by way of addition – see AC Bias, DC Bias

C
ColourCopy (u-he product) bucket-brigade delay plug-in
compander compressor / expander pair
compressor device that reduces an audio signal's dynamic range
crosstalk interference between adjacent tracks / channels
cutoff (of a filter) frequency threshold above and/or below which frequencies in the
signal will start to be attenuated or boosted

D
dB (Decibel) a commonly quoted unit of gain / attenuation
dBFS (DeciBels, Full Scale) dB levels in systems with a defined maximum level
DC bias (to tape, obsolete) addition of direct current to the signal being recorded
Diva (u-he product) mix-and-match VA synth with zero delay feedback filters
default value assigned before user intervention, or a sensible starting point
dynamic range span / difference between the quietest and loudest parts of a signal

TOC intro presets studio delay flange service config tricks plots nks gloss 56
 GLOSSARY

E
EQ equalization, manipulation of frequency response
electromagnetic interrelation between electrical currents and magnetic fields
erase head the tape machine head that removes previous recording / magnetization
expander device that increases an audio signal's dynamic range

F
fidelity (audio) degree of similarity to the original signal
flange (effect) dramatically sweeping cancellation / enhancement of frequencies
flutter often irregular, rapid variation of signal properties (amplitude, phase,
frequency) caused by mechanical imperfections of tape machines. See wow
frequency rate of vibrations within an audio signal
frequency response amplitude fidelity across the entire frequency range

G
gain increase in volume
gap width size of the head gap
gap loss any reduction in fidelity occurring in a head gap
GUI graphic user interface

H
harmonic distortion overtones, integer multiples of frequencies in the signal
headroom available margin before exceeding the capabilities of a system
head bump low-frequency resonance caused by physical attributes of the repro head
head gap small space between the two magnetic poles of a tape head
HF high frequency
hiss HF noise caused by random magnetic orientation of particles in the tape
Hive (u-he product) CPU-friendly VA / wavetable synthesizer
Hz (Hertz, cycles per second) the standard unit of frequency

TOC intro presets studio delay flange service config tricks plots nks gloss 57
 GLOSSARY

I
IEC/CCIR a tape EQ standard – International Electrotechnical Commission /
Comité Consultatif International Radiotélécommunique
IPS (inches per second) the standard unit of tape speed

JK
junk in Satin: A preset file made invisible in the browser, but not deleted

L
linear in a straight line
LF low frequency
LFO low frequency oscillator
lowpass filter electronic circuit that allows frequencies below a certain threshold to
pass unhindered, while attenuating frequencies above that threshold

M
make-up increase of the output gain to compensate for processing losses
modular consisting of discrete, interconnectable units
mod / modulation governing the value of a parameter from a control signal

N
NAB (National Association of Broadcasters) a tape EQ standard
NR noise reduction

O
oversampling process of sampling a signal at a frequency higher than twice the
highest frequency in the signal, which reduces aliasing artifacts

PQ
parameter a measurable factor within a system, usually user-adjustable
pre-emphasis EQ applied before the record head to compensate for adverse effects
generated by subsequent parts of the system
pre-magnetization see bias
Presswerk (u-he product) compression / dynamics plug-in

TOC intro presets studio delay flange service config tricks plots nks gloss 58
 GLOSSARY

R
Repro (u-he product) bundle of 2 analogue synth emulations (mono and poly)
repro head the reproduction (playback) head of a tape machine
resonance in Satin: strengthening of certain frequencies caused by oscillation within a
system (e.g. head bump)
RMS (Root Mean Square) a standard statistical average of any varying quantity

S
saturation (tape effect) complex compression, distortion and colouration caused by
inherent properties of tape machines – gives tape recordings a special ‘warmth’

T
tape head the parts of tape machines which transfer signals to and from the tape
The Dark Zebra (u-he product) cinematic soundset, bundled together with ZebraHZ
threshold a value above or below which an action commences
transients high amplitude, short-duration components within a signal
treble alternative term for high frequency range
Twangstrom (u-he product) spring reverb plug-in

U
Uhbik (u-he products) effects available either as plug-in bundle or as individual rack-
extension (RE) units for Propellerhead Reason

V
VU meter standardized indicator used for monitoring audio levels

W
wow fairly regular, low frequency variation of tape speed caused by mechanical
imperfections of tape machines. See flutter

XYZ
Zebra2 (u-he product) flagship ‘wireless’ modular synthesizer, combining many forms
of synthesis in a single plug-in.
ZebraHZ (u-he product) special upgrade/update for Zebra2, made for running all
presets included in The Dark Zebra

TOC intro presets studio delay flange service config tricks plots nks gloss 59