TC Electronic RESTORATION SUITE User manual

USER MANUAL

CONTACT DETAILS
If you would like to get in touch with us, please e-mail: [email protected].
Alternatively, you may contact the TC Electronic distributor in your area, or write to:
TC ELECTRONIC A/S
Customer Support
Sindalsvej 34
Risskov DK-8240
Denmark
www.tcelectronic.com
e-mail: [email protected]
USA:
TC Electronic, Inc.
742-A Hampshire Rd.
Westlake Village, CA 91361
www.tcelectronic.com
e-mail: [email protected]
© BY TC WORKS SOFT & HARDWARE GMBH 2003 - 2005. ALL PRODUCT AND COMPANY NAMES ARE
TRADEMARKS OF THEIR RESPECTIVE OWNERS. VST IS A TRADEMARK OF STEINBERG A.G., AUDIO
UNITS IS A TRADEMARK OF APPLE, COMPUTER. ALL SPECIFICATIONS SUBJECT TO CHANGE WITHOUT
NOTICE. ALL RIGHTS RESERVED.
TC Electronic is a TC Group company.
TC SUPPORT INTERACTIVE
The TC Support Interactive website is designed as an online support and information
center. At the site you can find answers to specific problems regarding TC software.
All resolved problems are stored in a database searchable by product, category, key-
words, or phrases. Under the section “My Stuff” you can login to check the status of
your questions, download materials such as product manuals, software updates and
new presets. This site has been specifically designed to meet the needs of our users.
We constantly update the database and you will find the site to be a huge resource of
information. Browse through Q&A’s and discover new aspects of your TC product. If
you can’t find an answer online, you have the option of submitting a question to our
technical support staff who will then reply to you by e-mail. The TC Support Team is
on constant alert to help you in any way they can.
EENNGGLLIISSHH

TABLE OF CONTENTS
WELCOME TO RESTORATION SUITE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
GETTING STARTED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
GLOSSARY OF TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
GENERAL PLUG-IN CONTROLS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
DECLICK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
DECRACKLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
DENOISE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
DESCRATCH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
DEThump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
PRESET MANAGEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
TUTORIAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
APPENDIX A: PROCESSING DELAY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
APPENDIX B: CPU OVERLOADS IN WAVELAB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
APPENDIX C: PATENT INFORMATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
EENNGGLLIISSHH

WELCOME TO RESTORATION SUITE
Thank you for choosing this hi-end restoration solution for POWERCORE. RESTORATION
SUITE brings you all of the key-components required to restore vintage or otherwise
damaged recordings – from broadband noise reduction to click or scratch removal!
TC Works have created this unique restoration tool-kit in collaboration with Noveltech
of Finland (www.noveltechsolutions.com). The vision of this product is to deliver a
complete solution, catering for all your restoration needs
The RESTORATION SUITE introduces Noveltech’s unrivalled breakthrough technology
for the removal of scratch and large artifacts from your audio material that goes way
beyond the reach of common solutions.
The Declick and Denoise algorithms are based on TC Electronic’s years of experience
in that field of restoration – now with extended functionality and POWERCORE sup-
port.
Together, Noveltech and TC Electronic will continue to expand the feature set of the
RESTORATION SUITE – our goal is to deliver the best and most complete restoration
solution to you!
Enjoy!
The TC Electronic Team
EENNGGLLIISSHH

GETTING STARTED
SYSTEM REQUIREMENTS
WINDOWS
• POWERCORE PCI or FIREWIRE running system 1.7 or later
• Windows XP
• PIII 1 GHz or faster
• 256 MB RAM
• VST compatible host application
• System must meet requirements of the host application
MAC OS X
• POWERCORE PCI or FIREWIRE running system 1.7 or later
• Mac OS X (10.2 or higher)
• G4 / 800 MHz or faster
• 256 MB RAM
• VST or AudioUnits compatible host application.
• System must meet requirements of the host application
POWERCORE REQUIRED!
The RESTORATION SUITE Plug-Ins require a POWERCORE PCI or POWERCORE
FIREWIRE to be available in your system. These Plug-Ins will NOT run without POW-
ERCORE.
CPU PERFORMANCE NOTE
RESTORATION SUITE is a set of hybrid Plug-Ins, utilizing native CPU processing and
POWERCORE DSP processing at the same time. This allows us to combine the key
advantages of both native and DSP processing technologies for optimal sound quality
and the best real-time results. Other than ‘pure’ POWERCORE DSP plug-ins, these
hybrid Plug-Ins might cause a significant CPU load on your system.
PROCESSING DELAY
In addition to the typical latency delay associated with DSP Plug-Ins, RESTORATION
SUITE also has an internal processing delay that will increase the latency.
RESTORATION SUITE / 1
EENNGGLLIISSHH

GLOSSARY OF TERMS
Artifact
An Artifact is an undesirable anomaly or glitch in the audio. Artifacts can originate
in specific recording environments, by using inadequate or primitive recording
techniques, playback errors, low quality systems, data reduction technologies and
even signal processing.
Click
Clicks can be recognized as periodic burst artifacts with a short duration and relatively
high gain. A click can be from 1 to 200 samples in length.
Crackle
Crackle is also a short duration artifact, but appears as an almost constant surface
noise, as you would hear in the background of a record. Crackle typically ranges from
1 to 20 samples in length.
Fingerprint
A Fingerprint is a short sample of the background noise that you want removed from
the audio signal. This is a unique spectral signature of the audio you wish to remove.
It is always best to get the Fingerprint of the noise you wish to remove as isolated as
possible.
Scratch
A scratch can be recognized as an aperiodic (occasional) burst artifact with a long
duration and relatively high gain. A scratch will be characterized as an artifact from
200 to 2000 samples in length.
EENNGGLLIISSHH
2/ RESTORATION SUITE

GENERAL PLUG-IN CONTROLS
The RESTORATION SUITE Plug-Ins are designed for consistency and ease of use; as
a result, the Plug-Ins share many common controls. This portion of the manual will
explain controls common to all or most of the RESTORATION SUITE Plug-Ins.
NOTE: Clicking the POWERCORE Logo on the plug-ins will NOT engage No Latency
Mode. RESTORATION SUITE does not support No Latency Mode.
KEY COMMANDS
In most host applications, RESTORATION SUITE Plug-Ins support keyboard
commands for certain functions. The following key commands are available for all
Plug-In parameters:
KEY COMMANDS IN MAC OS X:
Fine Adjust = Apple
Reset To Default = Alt
Context Key = Control + Mouse Button
KEY COMMANDS IN WINDOWS:
Fine Adjust = Alt
Reset To Default = Shift AND Control
Contextual Menu = Right Mouse Button
METERING OPTIONS
By clicking on the meters, you will have access to customize the visual feedback
provided by the meters. These parameters can be accessed in all RESTORATION
SUITE Plug-Ins.
LEVEL FALLBACK
With the Level Fallback parameter you can set the speed of the level indication to fall
back to the current level. ‘30dB/s’ is the fastest setting, ‘1dB/s’ is the slowest setting.
PEAK HOLD
The Peak Hold option will vary the time that the maximum peak value is displayed in
the meter. The hold time range, in seconds, is 10, 6, 1, None or Forever – which is
handy when you are not watching too closely. There is also a Reset option, useful for
those times when Peak Hold Forever is engaged.
RESTORATION SUITE / 3
EENNGGLLIISSHH

PEAK FALLBACK
With the Peak Fallback parameter you can set the speed of the peak level indication
to fall back to the current peak level (after the hold time is over). ‘Infinite’ will imme-
diately jump to the current peak, ‘1dB/s’ is the slowest setting.
CLIP HOLD
The Clip Hold option will vary the time that the plug-in displays clipping in the Clip
Indicator. The hold time range, in seconds, is 10, 6, 1, None or Forever. There is also
a Reset option.
TIP: The quickest method for resetting the Clip Indicator is to single click directly on
the red Clip Indicator LEDs.
AUDITION
Activate Audition to monitor the portions of your signal that are being removed. This
is extremely helpful when checking your Plug-In settings: the more of your original
signal you hear while in Audition mode, the more of that signal is removed by the
processing. If you can hear audio fragments in audition mode, your settings are too
extreme.
BYPASS
By using the Bypass button for the Plug-In instead of the host processing, you may
be interested to know that selecting the Plug-In Bypass still includes the processing
delay. This is not the case when using the standard bypass button of your host
application: this will cause the playback to “jump” back and forth; preventing you
from accurately comparing processed and unprocessed audio.
RESTORATION SUITE’s internal bypass function takes the latency into account,
allowing for hassle-free comparison before and after processing!
EENNGGLLIISSHH
4/ RESTORATION SUITE

WAVEFORM DISPLAY OPTIONS
Both the Declick and Descratch Plug-Ins have Waveform Options to customize how the
waveforms are displayed in the timeline of the Plug-In. Depending on the type of
material you are working on, you may find that you prefer one display to another, even
though no display method is better or more accurate than another.
TIP: When selecting one of the Waveform Display Options, this will be the default dis-
play behavior every time you open the plug-in.
NORMAL DISPLAY
Normal Display, the default display mode, is a high contrast scrolling waveform. A
clean waveform will appear black, while click and scratch artifacts will be indicated
with red lines. These lines reflect the amplitude of the artifact signal: large spikes
imposed on the waveform will indicate clicks, scratches and pops, while lower level
red lines—which in some cases appear to be a constant—will reflect crackle.
INTENSITY DISPLAY
Intensity Display is a low contrast scrolling waveform. A clean waveform will appear
gray, where higher amplitudes are displayed in darker shades of gray. Click and
scratch artifacts are displayed in much the same way; the intensity of the shade of
red will indicate the level of the artifact.
OUTLINE DISPLAY
Outline Display shows the outline of the waveform with red lines superimposed over
the waveform, indicating clicks and scratches. Again, these clicks reflect the
amplitude of the artifact signal: large spikes imposed on the waveform will indicate
clicks, scratches and pops, while lower level red lines will reflect crackle. This view
is best used with low-level signals.
RESTORATION SUITE / 5
EENNGGLLIISSHH

DECLICK
Declick removes not only unwanted click artifacts, but also more erratic crackle.
Declick is also a great tool for removing digital clicks, which can appear in audio
due to synchronization issues, digital overload or other sources. Despite its powerful
algorithm, Declick features an extremely user-friendly interface getting you to the
results you want to hear, fast!
THRESHOLD (0 TO 100)
This parameter determines Declick’s processing strength: high values will leave most
of the signal untouched, whereas lower values might have Declick remove more than
just clicks and crackle. Move the Threshold fader slowly up and down to find the
proper setting, where most of the clicks and crackle are removed, but the sound itself
isn’t affected.
TIP: Occasionally refer to the Audition Mode to confirm that you are not removing
important source signal.
EENNGGLLIISSHH
6/ RESTORATION SUITE
WAVEFORM DISPLAY
DECLICK PARAMETER
SLIDERS
LATENCY ADJUSTED
BYPASS/AUDITION
PRESET MANAGER
STATISTICAL DATA
IN/OUTPUT CONTROLS
AND OPTIONS

CRACKLE / CLICK SLIDER (0 TO 100)
This slider sets Declick’s processing sensitivity: if the slider is set to ‘crackle’
(value=0), Declick removes crackling noise, if it is set fully to ‘click’ (value=100),
the algorithm focus will shift towards click removal.
Tip: When the slider is set to ‘Click', you may want to use slightly higher threshold val-
ues than when the slider is set to ‘Crackle', where a lower threshold may be the better
the choice.
STATISTICS
Declick also offers statistical feedback based on your plug-in settings. ‘Clicks Per
Second’ will let you know the number of clicks detected in your signal over a period of
time, averaged per second. ‘Avr Click Length’ will let you know the average length of a
click in a sample value.
RESTORATION SUITE / 7
EENNGGLLIISSHH

DECRACKLE
Whereas Declick is also capable of removing a certain amount of crackling from your
audio material, Decrackle uses an algorithm dedicated to remove crackling, as well as
impulsive buzz. Undamaged signals will simply pass through without any degradation
of the signal.
The Decrackle plug-in will in almost all cases be used after the Declick process and
will remove signal like the surface noise of an old record. After applying Decrackle,
you should get a signal with a smooth background noise. Any kind of clicking and
crackling should be gone.
DISPLAY (ORIGINAL – PROCESSED – REMOVED)
The display on top of the Decrackle plug-in shows the audio signal as a sonograph. It
shows the signal as frequency content over time, where the color represents the level
of the particular frequency. Red indicates the highest level, black indicates that there
is no signal at that frequency at all. A sine wave of 10kHz at 0 dB will for example
show as red line near the top of the display.
The display can be switched to show either the original input signal, the processed
signal or to show only the parts of the signal that were removed by Decrackle. This 3rd
mode actually shows best how much the plug-in is working: The more you see in this
display mode, the more signal is removed. Therefore, this setting ("removed”) is the
default setting of Decrackle.
EENNGGLLIISSHH
8/ RESTORATION SUITE
WAVEFORM DISPLAY
DECRACKLE PARAMETER
SLIDERS
LATENCY ADJUSTED
BYPASS/AUDITION
PRESET MANAGER
IN/OUTPUT CONTROLS
AND OPTIONS

TYPE (CRACKLE – BUZZ – CORRUPT)
The Type parameter affects the removal process of Decrackle. The more distortion is
present in your signal, the more aggressive Decrackle needs to be set. The standard
setting is "Crackle". This will give the best results in most cases.
"Buzz" and "Corrupted" will alter the signal much more, and should only be used if the
standard setting can't cope with the signal. Typical “buzz” is for example created by
lighting equipment. A corrupt signal is a signal, where the disturbances correlate with
the actual signal.
THRESHOLD (0 TO 100)
The Threshold parameter sets the crackle detection level. At lower values, more crack-
ling is detected. Set to 100, the signal is only slightly touched. Compare the signal
with the bypassed process and with the plug-in in audition mode to find the setting
where you do not remove more signal of your music than necessary.
REDUCTION (0 TO 100)
Reduction determines the amount of crackle removal. The higher the value, the more
crackling noise will be removed. At 0, nothing will be removed.
CRACKLE SIZE (1 TO 20)
The Crackle Size is an estimate of how many samples are masked by a single crackle
signal.
TIP: Switch the Decrackle to "audition" mode and move the Threshold and Reduction
parameters until you hear only crackling, without any parts of the music in. Watch out
for rhythmic patterns: If the signal that is removed shows some rhythmic/periodic
behaviour, this is most likely already part of your music signal. Set the Threshold to a
higher value.
RESTORATION SUITE / 9
EENNGGLLIISSHH

DENOISE
Denoise is a broadband de-noiser featuring now three modes of operation. When the
plug-in is inserted, it opens up in "Automatic mode”, which does not require a finger-
print of the noise. It will start immediately to denoise the streamed signal. The second
mode is the classic "Fingerprint” mode, which is activated by pressing the "Learn”
button on the user interface. The third mode is the "Draw mode”, that will allow to
adjust any taken fingerprint manually. Let’s first have a look at the parameters avail-
able in the Denoise user-interface:
EENNGGLLIISSHH
10 / RESTORATION SUITE
SPECTRAL DISPLAY
DISPLAY OPTIONS
FINGERPRINT SETTINGS
PRESET MANAGER
IN/OUTPUT CONTROLS
AND OPTIONS
REDUCTION AMOUNT
AUTOMATIC MODE
SETTINGS
CHARACTER TYPE
REDUCTION SETTINGS
LATENCY ADJUSTED
BYPASS/AUDITION

SPECTRAL DISPLAY
Note: The settings of the display do not affect the processing of Denoise
FINGERPRINT
A static yellow line indicates the fingerprint. Signal levels below the fingerprint line
will be removed from your source material. By clicking on “Fingerprint” in the inter-
face, you can engage and disengage the metering of the current fingerprint used by
Denoise. This will not change the fingerprint: it is only a means to bypass metering.
ORIGINAL / PROCESSED SIGNAL
The signal spectrum of your original audio—or the source signal—is displayed as level
over frequency. By clicking on "Original” in the interface, you can engage and disen-
gage the metering of the original signal of Denoise. This will not bypass the original
signal itself: it is only a means to bypass metering.
When switching to "Processed”, you will actually see both the original and the
processed signal: The original signal is shown as single line, and the processing is
actually displayed as deduction from the current value of the original signal. If you see
a single line, this means that there is currently no processing going on. If the line gets
thicker at a certain frequency, this indicates processing at that frequency. To get a
feeling of how this display works, move the Reduction parameter up and down while
streaming audio – you will see that the line gets thicker if you move the Reduction
down (remove more), and gets thinner if you move reduction up (remove less).
LEFT / RIGHT
The left channel is displayed in green, the right channel in red. With the L/R knobs on
top of the display, you can switch the indication of the left or right channel on and
off.
AUTOMATIC MODE
AUTO
When the plug-in is loaded, it comes up in automatic mode and will immediately start
to denoise your material. By pressing the AUTO button, you can turn the automatic
mode on or off. If you switch it off, it will keep the current setting as fingerprint. If
you switch automatic mode on again, it will give you a warning that the current finger-
print setting will be altered.
The automatic mode will automatically detect the parts of the audio that are noisy,
and will set itself accordingly. The amount of reduction is set by the general parame-
ters of denoise (see: mode independent parameters).
RESTORATION SUITE / 11
EENNGGLLIISSHH

EENNGGLLIISSHH
12 / RESTORATION SUITE
TIME (100ms TO 9999ms)
The time parameter sets the time it takes Denoise to adjust the automatic fingerprint
to changes in the audio signal. Too fast settings might result in removing too much of
the instrument signals (in case they have a noisy character), too slow settings might
result in non-removed noise at the beginning of e.g. a new track (or the noise charac-
ter changing fast).
FINGERPRINT MODE
LEARN
When you hit the ‘Learn’ button, Denoise will begin taking your signal’s fingerprint.
In most host applications, when playback is not active while you press ‘Learn’, the
fingerprint will be taken as soon as playback begins. Learn automatically stops at the
end of the adjusted Capture Time. When there is no fingerprint in Denoise’s memory,
the Learn button appears with a transparency background and a light gray outline.
While the fingerprint is being taken, the button flashes. An active fingerprint— a fin-
gerprint that is in Denoise’s memory—will be indicated when the Learn button is
black.
NOTE: If you select Learn when there is an active fingerprint, you will receive a
warning asking if you wish to replace the current Fingerprint.
CAPTURE TIME (10ms TO 3000ms)
The Capture Time defines the fingerprint’s length. In most cases the default value of
500ms is enough for a good fingerprint, shorter times are not recommended for
standard noise reduction, but they are certainly possible.
TIP: Experiment with different ‘Capture Time’ values to remove periodic noise such as
hum, computer noise and room ambience.

DRAW MODE
When you select “Draw”, you enter a graphical editor where you can fine tune the
fingerprint of your noise sample. You can also enter Draw mode to create an artificial
fingerprint with your mouse. You can enter the Draw mode any time.
Editing a fingerprint in Draw mode is straightforward: there is a range of handles
that match the points of your fingerprint. Click and drag these handles to change the
values. A readout on the upper right corner of the Plug-In will indicate frequency and
gain of each handle.
- To create a new handle, click in an empty area of the Fingerprint.
- To remove a handle in Mac OS X: control + click on a handle
- To remove a handle in Windows: right mouse click on a handle
The fingerprint can be edited for left and right channels independently (e.g. if you
have some noise only on the left channel). By default, you are editing both channels
at the same time, but you can deselect a channel by pressing the L/R buttons below
the ‘Draw’ button. Please note that at least one button (L or R) needs to be selected.
When both channels are selected, the handles are still sitting on the left channel fin-
gerprint, but are actually controlling both channels.
Note: When entering Draw mode when no fingerprint is active, there is an empty
fingerprint line at 0dB with two handles; one each at the left and the right.
RESTORATION SUITE / 13
EENNGGLLIISSHH
HANDLES

MODE INDEPENDENT PARAMETERS
REDUCTION AMOUNT (-60dB TO 0dB)
The reduction fader sets the noise removal’s amount. Naturally, a greater reduction
value will reduce the noise further than a low value.
TIP: It is best to be careful with amount—typically, a reduction value of -6 to -15
should be enough—anything more can reduce the perceived ‘brilliance’ of your
material, and might even remove parts of the music signal.
CEILING (0dB to –99dB)
The ceiling parameter allows you to set an upper limit for the noise reduction process.
Levels that go beyond the ceiling will never be touched by the denoise process, no
matter what the captured (or automatically generated) fingerprint would remove.
Instead of setting the ceiling parameter, you could of course also manipulate the fin-
gerprint in draw mode.
BIAS (-20dB TO +20dB)
If the fingerprint does not perfectly represent your noise floor, you can make further
adjustments to the fingerprint with the Bias parameter. This will increase / decrease
the fingerprint’s noise level by up to 20dB.
TIP: Bias can be especially useful when you are using a fingerprint from another
recording, i.e. when you are editing an existing preset based on a Fingerprint from
another audio file. As the Fingerprints do not completely match, you may want to Bias
the Fingerprint to meet the needs of a different audio file. If your only chance to get a
good fingerprint is during the fade out of your material, you will want to increase the
Bias value.
ATTACK (10ms TO 999ms)
To remove as much noise from your signal as possible, use a short attack time
(10ms). Try increasing the attack time for "softer” processing, for example if you are
working with a less than ideal fingerprint.
CHARACTER
Generally, it is advised to leave the ‘Character’ parameter at default; the default
setting will work in most cases. When you consider the ‘Character’ parameter as an
extreme fine tune parameter for the character of the Denoise algorithm, you will be
able to very subtly adjust the behavior of Denoiser. Changing the ‘Character’ will have
an impact on the other parameters, please get to know the conditions described
below.
There is a psychoacoustic phenomenon where the reduction of noise will "trick” your
ear into perceiving that the processed audio is ‘darker’ or ‘less brilliant’ than the clean
signal. This happens simply because the high frequencies from the noise are removed
along with the noise. This is especially true of Type 1. By selecting Type 2, the
EENNGGLLIISSHH
14 / RESTORATION SUITE

Denoise algorithm will be tuned to maintain the high frequency content, but will be
perceived as less noise reduction.
TYPE 1 – Use less ‘Reduction’ and faster ‘Attack’ times; generally less than 50 ms.
TYPE 2 – Use more ‘Reduction’ and slower ‘Attack’ times; generally over 100 ms.
SELECTING THE RIGHT CHARACTER
How do these work together? Let’s assume the following example: you are cleaning a
dialog recorded in a small room. Using Type 1, you will find that when ‘Reduction’ is
too great, you may lose the perception of room ambiance and if the ‘Attack’ is set to a
longer value, the space in-between the dialog will slightly pump, sounding unnatural.
In switching to Type 2, you may find less breathing artifacts in the dialog and a more
consistent high frequency content, but you will also perceive less noise reduction.
By reducing the Attack time too low or the Reduction too much, you may set this
particular algorithm tuning to act too quickly, causing musical noise to appear, which
sounds like a sample and hold oscillator.
TIP: When adjusting ‘Character’ types, it is advised to store your current
configuration into one of the A or B temporary locations to adjust your work and
quickly compare the differences between the two ‘Character’ types.
RESTORATION SUITE / 15
EENNGGLLIISSHH

EENNGGLLIISSHH
16 / RESTORATION SUITE
DESCRATCH
Descratch is the first result from the ongoing collaboration between TC and Noveltech,
employing their patent-pending technology. This innovative new technology can be
used to reconstruct missing or damaged sections of audio material that may be up
to thousands of samples in length. Descratch can be used to eliminate disturbing
impulsive noise bursts resulting from a variety of sources, including scratches on the
surface of a vinyl recording – without adding audible artifacts! Based on accurate
mathematical modeling of the clean signal immediately before and after the damaged
portion of the signal, Descratch extrapolates the undamaged signals across the
damaged portion creating an accurate mathematical estimate of the missing signal
section. When Descratch is set up properly, the processing result is a clean signal that
does not add audible artifacts.
THRESHOLD (0 TO 100)
This parameter determines Descratch’s processing strength: high values will leave
most of the signal untouched, whereas lower values might have Descratch remove
more than just clicks and crackle. Move the Threshold fader slowly up and down, to
find the proper setting.
WAVEFORM DISPLAY
PRESET MANAGER
CPU METER
LINK CHANNEL
STATISTICAL DATA
SCRATCH SIZE
IN/OUTPUT CONTROLS
AND OPTIONS
LATENCY ADJUSTED
BYPASS/AUDITION