I'm demo-ing compressors at the moment, mainly for the purpose of creating big punchy pumping drums - rock/big beat/edm. I've used psp oldtimer for most things for ages, but it sounds a bit too woolly to my ears for this purpose.
However, I'm confused with the whole 1176-style approach - ie very fast release times. On any other compressor I've always gone for a decent length attack (around 30ms) in order to get a nice punchy sound. However 1176-style compressors seem to be considered one of the best for punchy drums, but the release times on these (eg the rocket/softube FET) are in micro-seconds...
Now demo-ing them, I am easily getting a good punchy sound, but can anyone explain why it is that 1176-style compressors can do this with such short attack times while other types of compressors (eg VCA?) would lack any punch below about 10ms?
So far though, I think I'm preferring the sound of Big Blue Compressor or Waves H-Comp for drums - but maybe that's just because I'm more used to this style of compressor. Is there any difference in approach I need to consider when setting 1176-style compressors?
Punchy rock/edm drums compression - 1176-style short release vs longer attack compressors
-
alligatorlizard alligatorlizard https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=276868
- KVRist
- Topic Starter
- 133 posts since 14 Mar, 2012
-
- KVRAF
- 6426 posts since 22 Jan, 2005 from Sweden
I'm in the process of doing the same thing - but for bass.
The "knee" also have a great impact on how it sounds - what happends and how after threshold is reached. Most important part if you ask me. The knee is giving the character of each compressor.
I think the range(I believe it was called) control of the Glue is adjusting the knee.
Others have settings of hard or soft knee etc. Some like ImageLine have a knee you can design.
The "knee" also have a great impact on how it sounds - what happends and how after threshold is reached. Most important part if you ask me. The knee is giving the character of each compressor.
I think the range(I believe it was called) control of the Glue is adjusting the knee.
Others have settings of hard or soft knee etc. Some like ImageLine have a knee you can design.
-
alligatorlizard alligatorlizard https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=276868
- KVRist
- Topic Starter
- 133 posts since 14 Mar, 2012
True, true - actually one of the things I'm liking about the big blue one is the knee-changing knob.
My question here tho is about about the difference between 1176 (and maybe also la2a?) style compressors, ie those with very fast attacks - and what is different about these that means you can still get a nice transient punch with these fast attack times. Or are you saying the knee shape is the difference here?
btw, after more demo-ing, favs are now FET for parallel, Big Blue for non-parallel (getting great punchy results even with low ratio, few db reduction, and a bit of tweaking of choke and gain knobs!)
In fact I'm really liking the big blue, but can't find many mentions of it or reviews - anyone else use this one?
My question here tho is about about the difference between 1176 (and maybe also la2a?) style compressors, ie those with very fast attacks - and what is different about these that means you can still get a nice transient punch with these fast attack times. Or are you saying the knee shape is the difference here?
btw, after more demo-ing, favs are now FET for parallel, Big Blue for non-parallel (getting great punchy results even with low ratio, few db reduction, and a bit of tweaking of choke and gain knobs!)
In fact I'm really liking the big blue, but can't find many mentions of it or reviews - anyone else use this one?
-
- KVRAF
- 6426 posts since 22 Jan, 2005 from Sweden
I'm assuming a bit about the knee, being perhaps the most transitional property, kind of.
I saw some topic on compressors at KVR that had some screenshots running a square wave through compressors - and that is rather revealing what might be different in the knee part.
So as test I guess one could run squarewave though and notice what does what to the sound.
Run it on tracks next to each other and study differences. It should correspond to the knee, what you see there.
But these hardware emulating plugins might also pick up dynamic properties from amp sections, I don't know. It's all a black box, so it's hard to tell what is what.
Mostly I'm just into listening right now - what works better for certain material.
I saw some topic on compressors at KVR that had some screenshots running a square wave through compressors - and that is rather revealing what might be different in the knee part.
So as test I guess one could run squarewave though and notice what does what to the sound.
Run it on tracks next to each other and study differences. It should correspond to the knee, what you see there.
But these hardware emulating plugins might also pick up dynamic properties from amp sections, I don't know. It's all a black box, so it's hard to tell what is what.
Mostly I'm just into listening right now - what works better for certain material.
-
- KVRAF
- 3080 posts since 17 Apr, 2005 from S.E. TN
I'm not ordinarily a fan of intentional distortion except geetar and occasional others. But if anybody makes a good tape emulator plugin, I would try that along with compression experiments, for big fat drums. If I ever get back to tracking live drummers.
The only reason for the doubt about "good tape emulator" is simply I've not tried a tape emulator for many years and dunno the market.
Assuming you can find a plugin that just nails an old 2" studor or mci, you don't have to overdrive the snot out of it and get obvious distortion. Just slow vu meter levels peaking to +3 maybe occasionally hitting +6 dB will get you more than halfway there before you hook up a compressor, and it ought to sound quite clean. No obvious distortion. It will just transparently carve off a few dB of brief transients and spit out real solid drums.
Dunno the many purposes compressors may have been inserted on drums in the old days, but my motivation was mainly to get a signal hot enough to swamp the tape noise while avoiding the drummer getting excited and smacking the drums hard enough to audibly distort the tape and ruin an otherwise good take.
Tape noise would build up, and if you compressed on the output of the tape at mix time it made the noise situation worse. So if you maybe set up a situation where you are running the clean recorded digital drums thru a tape emulator, raise the average level peaking up to +3 on a slow vu meter, then insert a compressor ahead of that and pretend you are trying to use the compressor to keep occasional "exuberant" drumming sections from overloading the tape-- Maybe playing that game would lead to duplication of how they got those sounds the first time around?
The only reason for the doubt about "good tape emulator" is simply I've not tried a tape emulator for many years and dunno the market.
Assuming you can find a plugin that just nails an old 2" studor or mci, you don't have to overdrive the snot out of it and get obvious distortion. Just slow vu meter levels peaking to +3 maybe occasionally hitting +6 dB will get you more than halfway there before you hook up a compressor, and it ought to sound quite clean. No obvious distortion. It will just transparently carve off a few dB of brief transients and spit out real solid drums.
Dunno the many purposes compressors may have been inserted on drums in the old days, but my motivation was mainly to get a signal hot enough to swamp the tape noise while avoiding the drummer getting excited and smacking the drums hard enough to audibly distort the tape and ruin an otherwise good take.
Tape noise would build up, and if you compressed on the output of the tape at mix time it made the noise situation worse. So if you maybe set up a situation where you are running the clean recorded digital drums thru a tape emulator, raise the average level peaking up to +3 on a slow vu meter, then insert a compressor ahead of that and pretend you are trying to use the compressor to keep occasional "exuberant" drumming sections from overloading the tape-- Maybe playing that game would lead to duplication of how they got those sounds the first time around?
-
alligatorlizard alligatorlizard https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=276868
- KVRist
- Topic Starter
- 133 posts since 14 Mar, 2012
Thanks for replies, lots of useful info - I've been trying a bit of sidechaining with these plugins, and as for tape dist, am using varying ammounts of fabfilter saturn on most things these days.
My question is specifically about attack times tho. Why with most compressors will you need a sizeable delay to keep it punchy (eg 30ms), but with FET/1176 style compression, which seems to be widely favoured with drums, the attack times are in micro seconds but you can still get a nice punchy sound (even without parallel compression). What is the difference in how they work?
My question is specifically about attack times tho. Why with most compressors will you need a sizeable delay to keep it punchy (eg 30ms), but with FET/1176 style compression, which seems to be widely favoured with drums, the attack times are in micro seconds but you can still get a nice punchy sound (even without parallel compression). What is the difference in how they work?
-
- KVRian
- 986 posts since 8 Mar, 2009
Play those drums on a large system and then tell me they are punchyalligatorlizard wrote:Thanks for replies, lots of useful info - I've been trying a bit of sidechaining with these plugins, and as for tape dist, am using varying ammounts of fabfilter saturn on most things these days.
My question is specifically about attack times tho. Why with most compressors will you need a sizeable delay to keep it punchy (eg 30ms), but with FET/1176 style compression, which seems to be widely favoured with drums, the attack times are in micro seconds but you can still get a nice punchy sound (even without parallel compression). What is the difference in how they work?
I suppose this is purely subjective cause my idea of punch is basically untouched transients which smack through speakers (especially in the low mids) and i don't associate that with the 1176 at all or any compressor for that matter.in fact, the 1176 is an envelope destroyer the thing is basically pure sustain so perhaps what you're hearing as punch is just the perceived loudness it adds as it pretty much reduces the transient envelope down to a few milliseconds,if you listen to it really low you can hear this then compare it to the dry sound at the same rms
I
-
- KVRian
- 908 posts since 10 Jan, 2010
got a link? i've been playing with something similar recently... though it's not strictly a square wave, it's actually a wave that's 100% max value (similar to a 0% PW pulse). you end up basically drawing the response of the compressor. pretty interesting to see what the knees actually look like in different compressors.lfm wrote: I saw some topic on compressors at KVR that had some screenshots running a square wave through compressors - and that is rather revealing what might be different in the knee part.
anyways, if someone else has already done this, wanted to see what they did. thanks!
-
- KVRAF
- 3080 posts since 17 Apr, 2005 from S.E. TN
I used to test analog compressors with a signal recorded on old analog tape, and view the results on memory oscilloscpe.
Then in digital, would make audio file of the test signal, suitable for testing either hardware or plugins. Run the signal thru the test device, record the result, and view the result in an audio editor. Much easier than expensive analog memory oscilloscope.
You can make the test signal to your preference and needs. I like to use cooledit pro / adobe audition, but many programs will do. My signal might be sine wave, several second amplitude steps at -24 dB, -18 dB, etc, up to 0 dB, or higher if your software can go higher than 0 dB in float format. The format can take it, but maybe some sine generator functions will want to protect you from yourself and not allow it.
Some dynamics processors are frequency sensitive, and it might be educational to make different-frequency test files and compare the results. Maybe a 50 or 100 Hz file, an otherwise identical 440 Hz or 1KHz file, etc.
To view release behavior, you need steps down in amplitude in addition to steps up in amplitude. So an alternate test file might be -24 dB, step up to -18 dB, step back down to -24 dB, step up to -15 dB, step back down to -24 dB, etc. The best file will have several seconds duration at each step, because some devices can have pretty long time constants.
Maybe your lowest level test amplitude would be -48 dB or lower. Just depending on the kind of device you wish to test, and your own individual nerd quotient.
When you view the processed result in an audio editor, when zoomed out, the sine is just a big black bar with squiggles at the step edges showing attack and release, easy to see the responses, and measure them if you wish.
Maybe a dc test signal would work on some portion of the plugin processors, but maybe not all plugins, and you definitely need a stepped ac signal to analyze the response of analog compressors, as most all of em will block dc and give no useful result with a dc test signal.
Then in digital, would make audio file of the test signal, suitable for testing either hardware or plugins. Run the signal thru the test device, record the result, and view the result in an audio editor. Much easier than expensive analog memory oscilloscope.
You can make the test signal to your preference and needs. I like to use cooledit pro / adobe audition, but many programs will do. My signal might be sine wave, several second amplitude steps at -24 dB, -18 dB, etc, up to 0 dB, or higher if your software can go higher than 0 dB in float format. The format can take it, but maybe some sine generator functions will want to protect you from yourself and not allow it.
Some dynamics processors are frequency sensitive, and it might be educational to make different-frequency test files and compare the results. Maybe a 50 or 100 Hz file, an otherwise identical 440 Hz or 1KHz file, etc.
To view release behavior, you need steps down in amplitude in addition to steps up in amplitude. So an alternate test file might be -24 dB, step up to -18 dB, step back down to -24 dB, step up to -15 dB, step back down to -24 dB, etc. The best file will have several seconds duration at each step, because some devices can have pretty long time constants.
Maybe your lowest level test amplitude would be -48 dB or lower. Just depending on the kind of device you wish to test, and your own individual nerd quotient.
When you view the processed result in an audio editor, when zoomed out, the sine is just a big black bar with squiggles at the step edges showing attack and release, easy to see the responses, and measure them if you wish.
Maybe a dc test signal would work on some portion of the plugin processors, but maybe not all plugins, and you definitely need a stepped ac signal to analyze the response of analog compressors, as most all of em will block dc and give no useful result with a dc test signal.
-
alligatorlizard alligatorlizard https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=276868
- KVRist
- Topic Starter
- 133 posts since 14 Mar, 2012
So, if you're doing parallel compression with fast attack times, the "punch" will have to come from mixing in the clean channel?
And if not using parallel method, you'll need a compressor with a longer attack time, as in this case the punch comes from letting enough of the transients thru?
Would I be right in saying these are the two main approaches? Each with slightly dif sonic results (hence why you'd choose one over the other in each situation), but both can create bigger/punchier drums?
And that the answer to my question about attack times might be: fast attack compressors result in punchy drums only if mixed in parallel with an unprocessed signal?
Btw, I just tried some parallel compression with psp Oldtimer ME (which I already own), and on fast attack/release settings, max gain, max ratio, am actually getting almost indistinguishable results from with the FET! Hadn't expected this - maybe I don't need a new compressor after all for NYC-comp?? Also, any extra dist/character/flavour I need can always come from fabfilter saturn.
And if not using parallel method, you'll need a compressor with a longer attack time, as in this case the punch comes from letting enough of the transients thru?
Would I be right in saying these are the two main approaches? Each with slightly dif sonic results (hence why you'd choose one over the other in each situation), but both can create bigger/punchier drums?
And that the answer to my question about attack times might be: fast attack compressors result in punchy drums only if mixed in parallel with an unprocessed signal?
Btw, I just tried some parallel compression with psp Oldtimer ME (which I already own), and on fast attack/release settings, max gain, max ratio, am actually getting almost indistinguishable results from with the FET! Hadn't expected this - maybe I don't need a new compressor after all for NYC-comp?? Also, any extra dist/character/flavour I need can always come from fabfilter saturn.
