What's Better Than A Good Compressor?

Now I've built a bunch (gaggle ?) of compressors, I have lots to choose from ..

Ross, g1176s, gSSLs, dLA2As, dOACs, Weaks, Poormen 660s and soon to be available, Fairchild style and 175/Gates/RCA style.  As well as a (dwindling) bunch of commercial units TC Electronics and Yamaha digital, ART hybrid vca/opto/tube.

And so, given that compression is obviously a mania for me in the studio - vocals, guitars, bass, drums, mix stems and mix

I've recently come to see that there's only one thing better than a good compressor ..

And that is TWO good compressors 

(plug-ins need not apply!)

----

Seriously, I've been tracking some guitars with my diy channel strip of Neve+Pultec+1176  THEN into an LA2A
and I think I've gone nuts    The sound is so good. It's like the LA2A picks up where the channel strip leaves off.

Where alone, the channel strip pushed hard runs out of puff .. just a little .. and the LA2A kind of lifts it onto a magic carpet of solid, overextended glissando compression and just runs away with it 

Luckily the noise floors of the various bits of my chain are pretty good and the overall hum at good strong levels is very acceptable.

Probably around -63 dBu or so 'at rest' with a guitar plugged in and volume turned down. 
Anything below-62dBu is not really noticable, even in silence.
For a resultant output probably at +10dBu at full belt under GR. Which would have to be at least 15dB or so all up!

It's so good I am reaching for and getting the music of Delaney and Bonnie.
So much stuff there it's insane!

Banging away on my jazzmaster with noiseless pups in casual, effortless, sustained non somniference
while on a break at my bench makes me realise,

I am truly 'a fortunate man'.

8)

In a head to head comparison, a good musician will beat a good compressor every time.
The difference is, if the musician contains tubes, they might be more of a hinderance than a benefit....

Every time eh?

Oh to be young and so sure of music!

great post,
I'm glad you have discovered the use of 2 compressors.
I have been doing this for a few years while mastering with very satisfying results.
Having one compressor set to slow attack quick release on a high ratio and another to fast attack slow release with a low ratio (no more than 2db of GR on either compressor at anyone time) usually benefits most mixes.
as you say one compressor picks up where the other left off, and if you mix and match your flavours of compressor you will son find a go to chain to suit your tastes.

Thanks! That's exactly what I meant  8)

The sort of controllability it gives me over the dynamic, as a player, when the compression works for me ..
is like having a beautiful tail wind pushing me along .. even while I'm drawing breath for the next passage.

Good compression inspires me to try for things that normally I wouldn't pull off.

I guess that makes me one lucky but middling musician 

etheory said:

In a head to head comparison, a good musician will beat a good compressor every time.

Click to expand...

It's not like any musician can have the attack of certain compressor at their fingertips. Maybe in some artsy fartsy jazz circles the use of compressor is shunned, but anywhere else, whether virtuoso musicians or not, their use in sound shaping is always welcome. It's only in the very lowest echelon of shit musicianship where compressors are still used as mere devices to correct uneven dynamics.

It's kind of *amusing* - I have a setup of my diy units in test or 'in the dock' at my bench, integrated into a Motu and PC.

And I have a 'home' studio (just a room really with the racks of diy+commercial  into a few Motus and a few PCs.

I start up a song in the studio and it writes on the network discs as well.

Pick it up the next morning at the bench and bang away trying new/different parts.
Could  be bass or steel guitars or some raging lead thing out of a champ into mic.

I'm finding the things I do playing at the bench are the things that are ending up being the 'keepers'

So that's amusing to me.

The studio is more of the 'startup' room
The bench is where different stuff gets tried

Like 2 compressors inline which I wouldn't normally do.

And seperate monitors for the playback and instrument stems. One surroundf hifi, one se tube.

Sounds quite stunning - the seperation of the mix and the tracking instrument as well as the new combos of day ..

It really has gotten to the point where how involved I play is being pushed along by the quality of the sound coming out.

Freaky really.

Still, after so many years of working away at the various aspects of music and technishe, shouldn't be surprising.

But it is none-the-less.

alexc said:

I'm finding the things I do playing at the bench are the things that are ending up being the 'keepers'

Click to expand...

you need to be a pretty hard boiled session professional to overcome performance anxiety. Some strange psychological effect when you know rec is on, even in the privacy of a home studio. Like in the studio we often let the artist practise a very long time before the actual take. Except that we record the whole session and in all likelyhood most of those "practise" takes end up used instead.

The 1176 into an LA2A is a known classic (or an LA2A into 1176) but I only tried it recently too...and wow it is quiet a tone, used it on lead vocals and its ultra smooth, I need to do it more.

I designed a compressor/limiter/etc back in the early '80s  (LOFT/Phoenix Audio Lab) taking a somewhat different approach to compression.

Instead of the typical "above threshold" limiting or compression that only reduces gain above threshold, I made it a full range compressor. In effect for 2:1 compression a 100 dB dynamic range signal would be reduced to 50 dB... Of course you don't want the -80 dB noise floor cranked up to -40 dB, so I added a trick downward expander that removed compressor gain smoothly to hit unity gain at the expansion threshold then expand down from there. IIRC I layered on a more convention above threshold limiter on top of the full range gradual compression, with a de-esser too.

This was not conventional for the early '80s (or now) so wasn't immediately embraced by the market. Around that time I left the company over management disagreements so the product was never advertised or promoted, as my remaining partners slowly went out of business (I was president and told them what I thought we needed to do to avoid failure, but it was 3 against one and they declined my direction, so I voted with my feet.).

With several decades of hindsight, I still like the compressor concept, but if you squint your eyes and look at the transfer function, the expander section removing gain at low level is not all that different from a soft knee above threshold comp.

========
Another observation from a different product design, I did a single ended NR for AMR, that used a combination of downward expander and sliding LPF. Either one of these SE noise reduction techniques when used alone can be audibly distracting, but the combination of both allowing combining a little of one with a little of the other, did wonders to clean up dirty tracks. without sounding worse than the noise it was removing. 

Lesson= a little of multiple strategies together is better than a bunch of one.

=====
One on my mental projects that hasn't fallen into place (yet) is to make a digitally controlled analog compressor. I first had the idea back in the 80, but didn't have close to the technology in my tool kit back then. My goal (then) was to mimic the behavior of most available compressors, with presets then give the customer access to go fully custom.  Back then I was going to save the preset data to the recording tape leader along with the slate tones, a quaint archaic idea today.

My issue then, and a still a little today is interface. My approach involves soft control of far more variables than any conventional compressor. This would easily and quickly over whelm most users and simple hardware designs, so it needs a smart computer interface where the user can easily rely upon presents and just control a subset of knobs when they are desirable, but have the capability to drill down into subtle inner details when that extra control is desired.

Presets mean the standard settings can always be recalled, but the power and flexibility of full control of almost everything still sounds desireable to me.

You like an XYZ comp. that would be preset #22.  8)

=====
I still don't have the computer side programing tools to make this, and have real work to do in more green field areas. The world doesn't really need yet another comp, but a comp to end all comps is an interesting "compcept" (pun intended).   

JR

Thanks JR - that was a very interesting read. Still going thru some of the points you raise.

I like to use a seperate downward expander on some of these classic compressor chains.
The best I have for that is a tc electronic finalizer multi band expander section.

Really keeps the noise of the silent passages thru those fixed makeup classic compressors under control.
The multiband part really works - I can make it discriminate on just the typical hum frequency region.

Very handy kind of thing to have indeed.

"What's Better Than A Good Compressor?"

A good microphone.

desol said:

"What's Better Than A Good Compressor?"

A good microphone.

Click to expand...

Well I can surely understand that. So much of the captured sound is about the mic and space.

Absoultely - respect. 

I don't have so much engineering experience with mics,  either building or using 
I've used some, I built one. Mostly a 57 on my amp and some AKG sdc on the acoustic.
Now a modded c12 cheapie but goodie

But for sure, all the pro studio guys I  knew were dubious about my outboard ramblings.

For them it was all the space and mic. And the band. And their record company  8)

Mic pres, not so much at all. Desk was just fine. 
Admittedly they were pretty good desks. SSLs mostly.

Compressors - only really to the extent of capturing without injury.
Of course, they all get silly about the SSL bus comps!
More interested in gating than compression.

Still - those guys were running a business with clients to keep happy and a need to focus on the 'big picture'.

All respect - I admired those studio owning guys of my youth very much.

But me, as a player - I needed to bring the gear and insist  on the plugging thereof.

After all, I was the paying guy  And the dough came from the EE work!

alexc said:

Thanks JR - that was a very interesting read. Still going thru some of the points you raise.

I like to use a seperate downward expander on some of these classic compressor chains.
The best I have for that is a tc electronic finalizer multi band expander section.

Really keeps the noise of the silent passages thru those fixed makeup classic compressors under control.
The multiband part really works - I can make it discriminate on just the typical hum frequency region.

Very handy kind of thing to have indeed.

Click to expand...

in regards to the points JR made and what you just said above, I understand we are talking about hardware here, but Flux have 4 dynamics processors, PureComp, DeComp, PureExpande and DeExpander.
They neatly combine all 4 into one handy plug called Solera. As I said, it is a plug, but it really does cover all your dynamic needs.
http://www.fluxhome.com/products/plug_ins/solera2

All of their plugs are very high quality.

JohnRoberts said:

One on my mental projects that hasn't fallen into place (yet) is to make a digitally controlled analog compressor. I first had the idea back in the 80, but didn't have close to the technology in my tool kit back then. My goal (then) was to mimic the behavior of most available compressors, with presets then give the customer access to go fully custom.  Back then I was going to save the preset data to the recording tape leader along with the slate tones, a quaint archaic idea today.

My issue then, and a still a little today is interface. My approach involves soft control of far more variables than any conventional compressor. This would easily and quickly over whelm most users and simple hardware designs, so it needs a smart computer interface where the user can easily rely upon presents and just control a subset of knobs when they are desirable, but have the capability to drill down into subtle inner details when that extra control is desired.

Presets mean the standard settings can always be recalled, but the power and flexibility of full control of almost everything still sounds desireable to me.

You like an XYZ comp. that would be preset #22.  8)

=====
I still don't have the computer side programing tools to make this, and have real work to do in more green field areas. The world doesn't really need yet another comp, but a comp to end all comps is an interesting "compcept" (pun intended).   

JR

Click to expand...

I really like this concept JR!

Wouldn't AD-side-chain conversion-latency & DA-control-voltage conversion-latency be a problem? One would need very fast AD converters.
I would vote for a concurrent feedforward and feedback AD-side-chain design with an external AD-side-chain input.

If you could control the unit(s) through your DAW, you could use plug-in templates for your favourite model besides an 'all parameter model'. So in fact use the same workflow as the DAW-processing-plugins, but control the external analog device (through MIDI or HUI over ethernet).

grT

Balijon said:

I really like this concept JR!

Wouldn't AD-side-chain conversion-latency & DA-control-voltage conversion-latency be a problem? One would need very fast AD converters.
I would vote for a concurrent feedforward and feedback AD-side-chain design with an external AD-side-chain input.

If you could control the unit(s) through your DAW, you could use plug-in templates for your favourite model besides an 'all parameter model'. So in fact use the same workflow as the DAW-processing-plugins, but control the external analog device (through MIDI or HUI over ethernet).

grT

Click to expand...

Actually conversion for side chain use does not need to be audio quality. Allowing aliases to fold back down do not cause measurement errors except for extreme case of signal exactly equal to sample rate.

I've actually done some work for a different kind of product but in dynamics you don't really want instant gain changes (clicks) so there is some desire to LPF (slow) the gain changes.

I did some promising work with digital pots for gain elements, and while I'd be tempted to synchronize gain changes with zero crossings, you can push gain changes to a DPOT faster than you ever need. For VCA you can push serial data to DAC then LPF that output. SInce the DAC  can be modest word length (!0B?) also pretty fast to update.

I am not up on how to communicate between hardware audio platform and DAW... It seems working with a DAW just do the super computer in digital domain, why bother with analog domain?

My idea is kind of a philosophical shift to open up the box to give user full control of several tens of parameters, but with a friendly interface so not too scary or unwieldy (presets).

JR

JR - thats an interesting idea for a comp, I have an old dictation preamp/compressor/gating thing that came out of an old courthouse I think and its circuit sounds like it might be close to what you designed in the 80s.

Sort of expansion below the threshold and smooth compression all the way up. It sounds really good and unique and while it can mangle sounds it still sounds very nice, "tightens" the audio, or maybe "focuses" the audio would be a better description.

It almost makes me think that people like the sound of analog tape so much on some stuff because of the effects of the noise reduction and not necessarily just the tape...

Iggy Pop once said his new record sounded wrong, no matter what they did, until they figured out he was used to the sound of Dolby A.  Then Iggy happy. 

I like some intelligently predicted GR on the way in, then some fine tuning again on some things on the way out.  Then there's the tracks that don't need any either time. 

Compression IMO is often a medium consideration.  Squashed dynamic range is useful for listening in your (my) car, not so much for the quiet living room.

I think there are cultural considerations for information density. A couple decades ago we would be visually overloaded by the amount of images pushed at us at the same time. There may be some overlap into audio to make it busier.

Bob Katz did some good work regarding metering and monitor levels when mixing for different media.

My old comp philosophy was to get the most reduction of dynamic range, with the least mangling of existing dynamics within the performance was to apply a gentler rate to if full range. After fixing the build up in noise floor, it starts to resemble more conventional above threshold approaches. So whatever.

Even my conventional compressor designs often involved overlaying a fast limiter on top of a more gentle compression so you can deliver protection against peaks separately from sound shaping.

YMMV Another mature category.

JR

What's Better Than A Good Compressor?
Well....sometimes no compressor at all !

I'm not saying that compression can't be good, but it's not the only solution by any means.
Sometimes you can get a better and less coloured result by simply adjusting levels.
For example, I recently did an article on Gold Star Studios and noticed from the pictures that at the time of the 'Wall of Sound' sessions their control room had no outboard gear at all (!).
No compressors - nothing.
Larry Levine, the engineer who did most of Spector's stuff, said that with vocals for instance - you had to learn what the singer was going to do and anticipate the levels; and leave the rest to tape saturation.

Another great example of level control that I often use is Macarthur Park the Richard Harris hit.
The 7 inch single starts with solo voice and ends with full orchestra - and yet the levels are pretty much the same. Not because they put a compressor over it, but because the mastering engineer chased the levels and reduced them by 9dB from the start to the end.

I actually like good compression - whether it's used as level control or as an effect, but it's ceratinly not the only answer.

(I'm currently looking at the optical compressor that Lexicon use at the front end of their PCM 42 delay by the way. Seems like a very simple yet transparent design).