Yay! After several month of beta-testing and bug hunting, I'm proud to release version 1.0.0 of the TDR Feedback Compressor.
Get it here: Tokyo Dawn Labs - Feedback Compressor
Tokyo Dawn Labs - Feedback Compressor 1.0.0 release!
-
- KVRian
- Topic Starter
- 1169 posts since 24 Feb, 2012
Yay! After several month of beta-testing and bug hunting, I'm proud to release version 1.0.0 of the TDR Feedback Compressor.
Get it here: Tokyo Dawn Labs - Feedback Compressor
Fabien from Tokyo Dawn Records
Check out my audio processors over at the Tokyo Dawn Labs!
Check out my audio processors over at the Tokyo Dawn Labs!
-
- KVRAF
- 1702 posts since 22 Apr, 2009 from Belgrade
-
- KVRian
- 622 posts since 14 Jun, 2006 from Finland
Thank you for this plugin!
-
- KVRer
- 5 posts since 14 Jul, 2010
This comp is super precise! I appreciate a lot that there is a 64 bits version. Thank you very much! (I ditched density, hahah)
-
maxxxter maxxxter https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=1
DELETED
-
- KVRian
- Topic Starter
- 1169 posts since 24 Feb, 2012
Thanks everyone! It seems that the 1.0.0 release can be considered as bug-free?!
Fabien from Tokyo Dawn Records
Check out my audio processors over at the Tokyo Dawn Labs!
Check out my audio processors over at the Tokyo Dawn Labs!
-
- KVRer
- 11 posts since 17 May, 2007
I've been using this in my latest mixing-project in every song for over a month now, and haven't encountered a single bug! 
Daw is Cubase 6.5.3 on winXP.
Too often 1.0 releases are buggy - probably meaning that they were released too soon.
But you were patient, great!
Daw is Cubase 6.5.3 on winXP.
Too often 1.0 releases are buggy - probably meaning that they were released too soon.
But you were patient, great!
-
- KVRian
- Topic Starter
- 1169 posts since 24 Feb, 2012
Thanks for the flowers! Really happy to hear you found good use for the compressor!Quarius wrote:Too often 1.0 releases are buggy - probably meaning that they were released too soon. But you were patient, great!
Fabien from Tokyo Dawn Records
Check out my audio processors over at the Tokyo Dawn Labs!
Check out my audio processors over at the Tokyo Dawn Labs!
-
- KVRAF
- 3828 posts since 15 Mar, 2002 from Underworld
Thank you, Fabien, for this freeware gem! It's slowly becoming my go-to compressor and I have a rather nice "palette" to choose from... Love it! It's only when I need a more character compressor that I use something else, but I wouldn't say your compressor is so "characterless", however it is rather transparent no matter what you do with it. 
Cheers!
Love it! It's only when I need a more character compressor that I use something else, but I wouldn't say your compressor is so "characterless", however it is rather transparent no matter what you do with it. Cheers!
It is no measure of health to be well adjusted to a profoundly sick society. - Jiddu Krishnamurti
-
- KVRer
- 11 posts since 22 Nov, 2010
Hello Everyone! I'm just dropping by to give a giant +1 on the feedback compressor becoming my "go to" !!! Wow, after using for a few months I am still thrilled with it. This thing definitely ROCKS!
Thank you everyone at Tokyo Dawn Labs
-
- KVRian
- Topic Starter
- 1169 posts since 24 Feb, 2012
Oh! Wait until you've heard its successor:
Fabien from Tokyo Dawn Records
Check out my audio processors over at the Tokyo Dawn Labs!
Check out my audio processors over at the Tokyo Dawn Labs!
-
- KVRAF
- 2138 posts since 8 Feb, 2007
Hey ! don't tease us ! please tell us more...
Professional technicians are assessed by the abilities they possess.
Amateur technicians are assessed by the tools they possess - and the amount of those tools, with an obvious preference to the latest hyped ones.
(Gabe Dumbbell)
Amateur technicians are assessed by the tools they possess - and the amount of those tools, with an obvious preference to the latest hyped ones.
(Gabe Dumbbell)
-
- KVRAF
- 3828 posts since 15 Mar, 2002 from Underworld
This looks fabulous, Fabien. I can bet it will sound fabulous, too. This one will be more "colourful" I think? It doesn't have the "colour" buttons and we can choose between Lim/Comp. Speaking of "colour" buttons, I almost always use them at "200%". 
This one reminds me of a "classic" design like SSL, but you will probably tweak it to your own taste. I wonder what it will sound like. I am hoping for a more colourful one that can even pump if needed. I suppose limiter will not be a brickwall one, but also a "classic" design? Like LA2A or LA3A maybe?
Happy New Year! Cheers!
p.s. I would like to point out one little niggle that I find with your Feedback Compressor - it doesn't react to the mono/stereo input changes [in Reaper], so it always consumes the same CPU even when I'm compressing mono audio which I do in about 90% of the cases. Only busses are stereo when I mix, and master buss. 90% of the channels are mono. How about a stereo/mono switch? and "link 0-100%" box?
This one reminds me of a "classic" design like SSL, but you will probably tweak it to your own taste. I wonder what it will sound like. I am hoping for a more colourful one that can even pump if needed. I suppose limiter will not be a brickwall one, but also a "classic" design? Like LA2A or LA3A maybe? Happy New Year! Cheers!
p.s. I would like to point out one little niggle that I find with your Feedback Compressor - it doesn't react to the mono/stereo input changes [in Reaper], so it always consumes the same CPU even when I'm compressing mono audio which I do in about 90% of the cases. Only busses are stereo when I mix, and master buss. 90% of the channels are mono.
How about a stereo/mono switch? and "link 0-100%" box? It is no measure of health to be well adjusted to a profoundly sick society. - Jiddu Krishnamurti
-
- KVRian
- Topic Starter
- 1169 posts since 24 Feb, 2012
Hey Dux & Tp3,
first of all, the fb comp II is a radical revamp of both the sound "engine" and control scheme. That's why I created a new version branch "II". The original fb comp will still be maintained in the future (it's fine as it is IMHO).
FB Comp vs FB Comp II, what happened?
New compressor "engine":
- Compressor now provides *super* fast reaction times down to a sixth 44100kHz sample (!). Some parts of the side-chain run at nearly 0.5Mhz!
- Practically "invisible" oversampling in the signal path. Algorithm is bit-transparent without processing.
- Unharmonic distortion reduced by about 25dB (intermodulation and aliasing)
Updated control scheme:
- The previous version offered the possibility to tune the main detector. I never was really happy with it, most of all because it had a strong effect on the attack - in some cases even stronger than the attack itself. So, I removed the control and fixed it to "peak" mode.
- The compressor uses to distinct detectors circuits: A peak detector and a RMS detector. The highest of both takes control of the compression. The Peak Release and RMS Release controls allows to tune them independently (much like the previous fb comp).
However, there is a difference. In the previous version, both detectors thresholds were fixed to pre-set values. Now, there's that fancy knob called "Peak Crest" which controls the threshold of the peak detector compared to the RMS (the main Threshold knob controls the RMS detector's threshold). This is extremely cool, it allows to control the crest factor of the output signal (within reasonable limits). It somehow replaces to old "detector" control, but sounds much more useful.
The new "Peak Mode" switch "Lim" and "Cmp" is actually just a ratio control for the peak detector path. "Lim" fixes the peak path ratio to 7:1, "Cmp" set the peak ratio equal to the RMS path ratio.
New option "Precise"/"Eco" allows to switch the compressor to an economic mode. This is not a creative parameter and should only be used in cases of (cpu) overloaded projects.
The manual will extensively describe these topics.
Now to your questions:
- The colour section has been removed. Most of all because I wasn't happy with it and also made the implementation of the actual compressor difficult. IMHO, the concept is much cleaner without colour section. I already started working on a dedicated color/smasher plug-in instead.
- So, no, it's not dirtier. But the actual compression now sounds very crispy, warm and musical by its own.
- Pumping is possible too, but only at very specific settings. I'm afraid the compressor is designed to avoid such things. Pumping comes from bad comp design.
- Brick-wall: Oh, it works as expected if attack is set to the shortest value (of course it's not truly brick wall, but still effective).
- Analogue emulations/approximations are a difficult topic. Instead of thinking about Urei/Teletronics/Fairchild, have a closer look at truly high-end units such a the GML compressor, the Weiss compressor or the Crane Song STC-8. I only posses the last three (I'm into mastering) and have have no interest in buying noisy one trick ponies anyway. Much like the last three compressors, the FB comp II tries to combine the strength of these famous vintage units, but without the excessive distortion, restricted operational range and noise of these restricted designs.
In short, if you're looking for the actual compression effect, I can assure you that the fb comp II can easily approximate everything these old boxes do, and much more! Of course, without all the ugly degradation. Clean, accurate and aliasing free compression can actually sound fat and crispy like nothing else.
- Stereo Link: The compressor is mainly designed for mastering purposes, so, stereo stability is absolutely essential. There is only one way to guarantee full stereo center stability, one needs to make sure that both sides a compressed equally. That is, the side-chain output is actually mono and drives the gain of both channel in equal manner. So, to answer your question, it is already *mono*. A dedicated mono version could definitely run faster in theory, I need to check out how much work it means for me.
The strange "Key Stereo Diff" is not a conventional "coupled/detached" stereo-link control (as found in the API comp and a few others). Instead, it allows the compressor to ignore stereo anti-phase information (0%), which can effectively change the width in a signal-dependent manner. 100% means that no width manipulation is allowed (i.e. the stereo image is perfectly maintained).
I have plans for a mix oriented comp which will offer true dual mono compression (and a mono version). I like to keep the concepts as focussed and clean as possible, I'm sure you understand!
The post is already too long, give me a day or two to finish the beta package as soon I recover from NYE
first of all, the fb comp II is a radical revamp of both the sound "engine" and control scheme. That's why I created a new version branch "II". The original fb comp will still be maintained in the future (it's fine as it is IMHO).
FB Comp vs FB Comp II, what happened?
New compressor "engine":
- Compressor now provides *super* fast reaction times down to a sixth 44100kHz sample (!). Some parts of the side-chain run at nearly 0.5Mhz!
- Practically "invisible" oversampling in the signal path. Algorithm is bit-transparent without processing.
- Unharmonic distortion reduced by about 25dB (intermodulation and aliasing)
Updated control scheme:
- The previous version offered the possibility to tune the main detector. I never was really happy with it, most of all because it had a strong effect on the attack - in some cases even stronger than the attack itself. So, I removed the control and fixed it to "peak" mode.
- The compressor uses to distinct detectors circuits: A peak detector and a RMS detector. The highest of both takes control of the compression. The Peak Release and RMS Release controls allows to tune them independently (much like the previous fb comp).
However, there is a difference. In the previous version, both detectors thresholds were fixed to pre-set values. Now, there's that fancy knob called "Peak Crest" which controls the threshold of the peak detector compared to the RMS (the main Threshold knob controls the RMS detector's threshold). This is extremely cool, it allows to control the crest factor of the output signal (within reasonable limits). It somehow replaces to old "detector" control, but sounds much more useful.
The new "Peak Mode" switch "Lim" and "Cmp" is actually just a ratio control for the peak detector path. "Lim" fixes the peak path ratio to 7:1, "Cmp" set the peak ratio equal to the RMS path ratio.
New option "Precise"/"Eco" allows to switch the compressor to an economic mode. This is not a creative parameter and should only be used in cases of (cpu) overloaded projects.
The manual will extensively describe these topics.
Now to your questions:
- The colour section has been removed. Most of all because I wasn't happy with it and also made the implementation of the actual compressor difficult. IMHO, the concept is much cleaner without colour section. I already started working on a dedicated color/smasher plug-in instead.
- So, no, it's not dirtier. But the actual compression now sounds very crispy, warm and musical by its own.
- Pumping is possible too, but only at very specific settings. I'm afraid the compressor is designed to avoid such things. Pumping comes from bad comp design.
- Brick-wall: Oh, it works as expected if attack is set to the shortest value (of course it's not truly brick wall, but still effective).
- Analogue emulations/approximations are a difficult topic. Instead of thinking about Urei/Teletronics/Fairchild, have a closer look at truly high-end units such a the GML compressor, the Weiss compressor or the Crane Song STC-8. I only posses the last three (I'm into mastering) and have have no interest in buying noisy one trick ponies anyway. Much like the last three compressors, the FB comp II tries to combine the strength of these famous vintage units, but without the excessive distortion, restricted operational range and noise of these restricted designs.
In short, if you're looking for the actual compression effect, I can assure you that the fb comp II can easily approximate everything these old boxes do, and much more! Of course, without all the ugly degradation. Clean, accurate and aliasing free compression can actually sound fat and crispy like nothing else.
- Stereo Link: The compressor is mainly designed for mastering purposes, so, stereo stability is absolutely essential. There is only one way to guarantee full stereo center stability, one needs to make sure that both sides a compressed equally. That is, the side-chain output is actually mono and drives the gain of both channel in equal manner. So, to answer your question, it is already *mono*. A dedicated mono version could definitely run faster in theory, I need to check out how much work it means for me.
The strange "Key Stereo Diff" is not a conventional "coupled/detached" stereo-link control (as found in the API comp and a few others). Instead, it allows the compressor to ignore stereo anti-phase information (0%), which can effectively change the width in a signal-dependent manner. 100% means that no width manipulation is allowed (i.e. the stereo image is perfectly maintained).
I have plans for a mix oriented comp which will offer true dual mono compression (and a mono version). I like to keep the concepts as focussed and clean as possible, I'm sure you understand!
The post is already too long, give me a day or two to finish the beta package as soon I recover from NYE
Fabien from Tokyo Dawn Records
Check out my audio processors over at the Tokyo Dawn Labs!
Check out my audio processors over at the Tokyo Dawn Labs!
-
- KVRAF
- 14658 posts since 19 Oct, 2003 from Berlin, Germany
You know what would be great? If the compressor could be setup to "feed forward" or "feed back". Much like the old reaComp by Cockos.
That, and team up with vladg again to port over to Mac. Then you have a clear winner and I have an excuse to use your comp rather than everything else at a friends place.
That, and team up with vladg again to port over to Mac. Then you have a clear winner and I have an excuse to use your comp rather than everything else at a friends place.
