What's Your Favorite Compressor, and Why?

[Theo Verelst]

In the digital domain, if you multiply the attack curve of the envelope with the signal going through the VCA, the maximum frequency is going to by the highest components in the attack added to the highest frequency components in the signal, so possibly you'll have to work at 96kHz to prevent aliasing.

 

Theo V

[Anderton]

In the digital domain, if you multiply the attack curve of the envelope with the signal going through the VCA, the maximum frequency is going to by the highest components in the attack added to the highest frequency components in the signal, so possibly you'll have to work at 96kHz to prevent aliasing.

 

Theo V

 

I assume this is more of a problem with limiting because the goal is to have a super-fast attack. If the attack was slower, it seems like you wouldn't exceed the capabilities of a 44/1 or 48 kHz system to reproduce accurately...yes?

[Theo Verelst]

For correct Digital to Analog conversion, aliasing errors need to be prevented, that's why the vca input signals need to have limited frequency content. when put together, or aliasing will occur (unless you want that..), You could do the compression at a higher sample rate (some Ladpsa plugins do that) and then low pass filter to get back the original sampling frequency.

 

Theo

[KuruPrionz]

In the digital domain, if you multiply the attack curve of the envelope with the signal going through the VCA, the maximum frequency is going to by the highest components in the attack added to the highest frequency components in the signal, so possibly you'll have to work at 96kHz to prevent aliasing.

 

Theo V

 

Will going to 96 kHz change the frequencies that are currently generated at a lower setting like 48 or 44.1 kHz or will it simply create additional higher frequencies because it can?

If the latter, using 96 kHz may not change anything or could even generate more frequency anomalies, true?

 

I've recently added a couple of analog compressors to my rack and will work on using some compression going in and avoiding compression plugins if possible. The same holds true for limiting, which I can also do in the analog realm prior to conversion.

 

Thoughts?

[Anderton]

Will going to 96 kHz change the frequencies that are currently generated at a lower setting like 48 or 44.1 kHz or will it simply create additional higher frequencies because it can?

 

I think the issue is that 96 kHz can accommodate higher frequencies outside the range of human hearing, so they don't fold back into the range of human hearing, like they would at 44.1 or 48 kHz. It's basically the same reason for oversampling virtual instruments and amp sims, both of which can generate crazy high harmonics that fold back down into the audio range.

 

FWIW some people prefer the sound of foldover distortion vs. no foldover distortion with some material.

[RABid]

I have a few compressors on my UA system but to be honest. I"m a compressor idiot with no ear. I can read reviews that say this compressor is great on drums and this is good for bass. But I don"t hear it myself. I"ve always had trouble with the sound of my mind overriding the sound in my ears. So if I use compression it is usually for balance and I stick with presets on Glue.

[KuruPrionz]

Will going to 96 kHz change the frequencies that are currently generated at a lower setting like 48 or 44.1 kHz or will it simply create additional higher frequencies because it can?

 

I think the issue is that 96 kHz can accommodate higher frequencies outside the range of human hearing, so they don't fold back into the range of human hearing, like they would at 44.1 or 48 kHz. It's basically the same reason for oversampling virtual instruments and amp sims, both of which can generate crazy high harmonics that fold back down into the audio range.

 

FWIW some people prefer the sound of foldover distortion vs. no foldover distortion with some material.

 

I get the part about 96 kHz being capable of reproducing higher frequencies. Won't it also reproduce lower frequencies?

In other words, are the artifacts at a frequency because that is the frequency they generate or do the artifacts completely shift to higher frequencies when that is possible?

 

I get that it could be an "it depends" question also. Not trying to stir up poop, just curious.

[Anderton]

I get the part about 96 kHz being capable of reproducing higher frequencies. Won't it also reproduce lower frequencies?

 

Yes, but the only frequencies we care about are the ones that meet or exceed the clock frequency, because those are the ones that can create foldover distortion. Make sense?

[KuruPrionz]

I get the part about 96 kHz being capable of reproducing higher frequencies. Won't it also reproduce lower frequencies?

 

Yes, but the only frequencies we care about are the ones that meet or exceed the clock frequency, because those are the ones that can create foldover distortion. Make sense?

 

Isn't foldover distortion an example of the way that multiple frequencies can create additional harmonics both above and below the original frequencies or is that something else entirely?

 

I ask partly because I saw a video (quite a while ago, no link, sorry) made by someone who was adamant that the only good distortion was zero distortion and that lower harmonics created by high frequency limitations remained "detectable" ( I don't think he was claiming he could actually hear them) until the clock speed was 192 kHz and therefore that was the only currently acceptable clock speed.

 

I will admit I thought he was maybe a bit over-zealous and I'm not trying to "win", I honestly can't hear any difference between 48 kHz and it's doubled frequency 96 kHz.

Maybe too many years playing louder than I should have?

 

Or, others have much better monitoring systems than I do and listen much louder perhaps. Or, they consider information they obtain through other softwares that provide visual information to be audio information?

Or something... Happy Holidays!!!

[The Real MC]

Guys:

Compressors don't operate in the frequency domain like EQs or filters or delay based effects do.

The threshold circuits sense signals in the amplitude domain.

The detectors operate in the time domain (limited by processor load).

The control law between the detector output and the output attenuator element operate in the time domain and define the transient response of a compressor (IE how it sounds with different instruments).

Unless you're encountering aliasing or excessive processor loading, compressors operate regardless of the frequency of the sampling system.

[KuruPrionz]

Guys:

Compressors don't operate in the frequency domain like EQs or filters or delay based effects do.

The threshold circuits sense signals in the amplitude domain.

The detectors operate in the time domain (limited by processor load).

The control law between the detector output and the output attenuator element operate in the time domain and define the transient response of a compressor (IE how it sounds with different instruments).

Unless you're encountering aliasing or excessive processor loading, compressors operate regardless of the frequency of the sampling system.

 

 

Plugins of all types do have their anomalies, that is what we are talking about. Analog compressors are a different animal. I am starting to use them going in, more as limiters but sometimes as a tone all it's own.

Space and $$$ considerations mean both have their advantages and uses.

[The Real MC]

Guys:

Compressors don't operate in the frequency domain like EQs or filters or delay based effects do.

The threshold circuits sense signals in the amplitude domain.

The detectors operate in the time domain (limited by processor load).

The control law between the detector output and the output attenuator element operate in the time domain and define the transient response of a compressor (IE how it sounds with different instruments).

Unless you're encountering aliasing or excessive processor loading, compressors operate regardless of the frequency of the sampling system.

 

 

Plugins of all types do have their anomalies, that is what we are talking about. Analog compressors are a different animal. I am starting to use them going in, more as limiters but sometimes as a tone all it's own.

Space and $$$ considerations mean both have their advantages and uses.

 

I was referring to plugins. I was taking the core components of the analog deal and referring which domain they would operate in a plugin.

[KuruPrionz]

Guys:

Compressors don't operate in the frequency domain like EQs or filters or delay based effects do.

The threshold circuits sense signals in the amplitude domain.

The detectors operate in the time domain (limited by processor load).

The control law between the detector output and the output attenuator element operate in the time domain and define the transient response of a compressor (IE how it sounds with different instruments).

Unless you're encountering aliasing or excessive processor loading, compressors operate regardless of the frequency of the sampling system.

 

 

Plugins of all types do have their anomalies, that is what we are talking about. Analog compressors are a different animal. I am starting to use them going in, more as limiters but sometimes as a tone all it's own.

Space and $$$ considerations mean both have their advantages and uses.

 

I was referring to plugins. I was taking the core components of the analog deal and referring which domain they would operate in a plugin.

 

Right and to an extent completely correct. In the end, compressors end up in the frequency domain because they are used to compress sounds, which all have tones. In other words, while they may not affect frequencies directly - in the digital domain they can be and seem to be affected by frequencies as a result of A/D and D/A conversions. This is not a function, this is a disfunction.

 

Problems are created by the cutoff frequencies of various A/D formats and further exacerbated by a phenomenon that two frequencies can create additional harmonics that are the sum and the difference of the two frequencies.

This phenomenon can be used for good or evil. The British organ builders learned to get lower frequencies from pipe organs by careful use of 2 higher frequencies. This is good.

A/D and D/A convertors are merciless and evil. Depending on which frequency you are converting to from analog, any plugin can create artifacts.

 

If I am not mistaken, those artifacts are what Theo mentioned. I would like it if he would return and provide more insight into his observations. Which, it appears, is what launched the present direction of this particular conversation.

 

My own thoughts are that there have been many plugin compressors written into code over the last couple of decades and some of them may very well have relatively ancient code limitations. In other words, you could use them at 96 kHz but perhaps they were created for 44.1 kHz and can't really compress higher rates. I don't know if my theory makes any sense or not to be honest, that's why I ask questions of those who have more specific knowledge. Or maybe I'm just insane...

 

I can set up your guitar better than pretty much everybody else, that's my saving grace. We all have our "thangs" that we excel at!

[techristian]

I have a Behringer HW compressor that I bought for the TV , but my daughter won't let me use it! (Who;s the boss in this house?)

 

Seriously I use Cool Edit 2000 for final mixes. I have a few DYNAMIC CURVES that I like.

 

Dan

[Sundown]

All of my compressors are virtual (I"m a keyboard player / songwriter who occasionally records guitar and voice). I have a pretty broad collection but I find myself using 1176 emulations frequently (either UAD or Waves). After that it"s the Waves Renaissance Compressor or C1, which I recently resurrected back into the rotation. I find the 1176 adds great tone to everything from pianos, to guitar, to synth patches. If I want dynamics control without a lot of coloration I"m more likely to use the Ren Compressor or C1. I know the Ren Compressor is meant to mimic vintage compressors but I find that it does a good job of semi-transparent dynamics control at modest ratios and thresholds. Occasionally I"ll use the Waves Hybrid Compressor (H-Comp) as well.

 

I have virtual versions of the dbx160, the LA2A, the LA3A, and the Fairchild. Occasionally I"ll use the dbx160 on extreme settings with a drum loop or sampled drum hits, but I almost never use optical compressors. I know folks rave about the LA2A but I"ve never really found a fit for it despite trying many times. I know the FET EL Distressor gets a lot of publicity but I don"t record live drums and I don"t think I would use it very often. I"ve listened to the demos and I don"t like that squashed sound.

 

I suspect we"re talking about channel compression here, but I feel I have to mention what might be my favorite compressor of all, which is the SSL G-series buss compressor. I have the UA recreation. That plugin compressor is magic, especially for a keyboard player when you"re trying to glue together disparate tracks. Even at modest settings it adds a wonderful sound and I probably use it on every single track I mix.

[Claus H]

"Favorite" here is more like 'What I have".

 

Since I do things like acoustic guitar/vocals and Fender bass, I have an HHb Radius 30 (which I bought years ago along with its EQ brother).

The Radius is actually TL Audio from England, re-branded. It's two-channel, with tubes, and it's a leisurely, softer compressor, definitely not meant for aggressive slam.

It thickens things up nicely going in, so I can just add the final overall compression on my PC when the mix is ready, nothing too heavy needed for my purposes.

 

I also still have my old DBX 286 in the rack, and while it's good on my bass, it's quite noisy, so I am using it less and less.

 

As far as dreams go, I have been thinking about an LA2A clone, either Audioscape or Regular John...just have a great channel and call it done.

 

Claus.

[Anderton]

I think that reduced to essentials, my preference is one optical-based (or software equivalent) for the "smooth" sound, and one fast (e.g., FET type). That covers most of my bases.

 

FYI I came up with a useful trick to give any compressor, hardware or software, lookahead when working with a DAW. The link goes to an article that's based around adding lookahead in Studio One to the Fat Channel plug-in, but it's a workaround that works with pretty much everything