GroupDIY

Project Specific Discussions => Mixers/Monitoring Systems => Topic started by: pucho812 on April 19, 2018, 06:59:03 PM

Title: compressors per channel
Post by: pucho812 on April 19, 2018, 06:59:03 PM
if you could pick one compressor to have on each channel of a desk what would you pick and why?
Title: Re: compressors per channel
Post by: ruffrecords on April 19, 2018, 07:22:45 PM
if you could pick one compressor to have on each channel of a desk what would you pick and why?
I like the ones I have per channel on my AKAI DPS24. Lots of ready made presets plus you can store your own. I even have my own Fairchild preset.

Cheers

Ian
Title: Re: compressors per channel
Post by: abbey road d enfer on April 19, 2018, 07:44:59 PM
if you could pick one compressor to have on each channel of a desk what would you pick and why?
Analog or plug-in? That makes a huge difference.
Analog (like built-in the channel) I would go for a simplified 160-type (VCA with adaptive-timing RMS detector); it's what I go for 9 times out of 10.
If it was plug-in, I wouldn't lose the ability to change at a whim. :)
Title: Re: compressors per channel
Post by: shot on April 19, 2018, 08:04:40 PM
Pico!
Title: Re: compressors per channel
Post by: pucho812 on April 19, 2018, 08:57:16 PM
This is for an analog console.  If that helps any
Title: Re: compressors per channel
Post by: abbey road d enfer on April 19, 2018, 09:11:21 PM
The pico is in essence a "simplified 160-type (VCA with adaptive-timing RMS detector)". Seems to me a good choice.
Title: Re: compressors per channel
Post by: EmRR on April 20, 2018, 06:50:57 AM
Yeah FCS (fancy pico) P3500 if i had to reduce to one and multiply. 
Title: Re: compressors per channel
Post by: guze on April 20, 2018, 05:49:30 PM
An opto or fet type, can be more economical in a "large" format, and not so "THAT//DBX". more unique, and very simple control(threshold and knee/ratio). You dont want 5 more knobs in a channel strip.
Title: Re: compressors per channel
Post by: abbey road d enfer on April 20, 2018, 06:44:56 PM
An opto or fet type, can be more economical in a "large" format, and not so "THAT//DBX". more unique, and very simple control(threshold and knee/ratio). You dont want 5 more knobs in a channel strip.
My suggestion has 3 knobs (threshold, ratio, make-up) and 1-3 switches (by-pass, soft/hard, meter). Can you have less than that on opto or FET-type? In addition, Fet compressors have a coloration and opto have a somewhat unpredictable transient response that does not make them usable on everything. I'm not saying that my proposal will be adequate on 100% of the channels, but I think a good 80% is probable.
Title: Re: compressors per channel
Post by: pucho812 on April 20, 2018, 07:17:10 PM
 will we need part matching with opto's and fets, or do we not care here?
Title: Re: compressors per channel
Post by: abbey road d enfer on April 20, 2018, 09:59:34 PM
will we need part matching with opto's and fets, or do we not care here?
You may care if you do stereo - then you would need a "link" switch. IMO it complicates matters. Stereo compression is not that frequent; I think it's one of the cases where the compressor built in the channel is not adequate and needs to be replaced with an external unit.
Title: Re: compressors per channel
Post by: pucho812 on April 21, 2018, 10:44:27 PM
You may care if you do stereo - then you would need a "link" switch. IMO it complicates matters. Stereo compression is not that frequent; I think it's one of the cases where the compressor built in the channel is not adequate and needs to be replaced with an external unit.

good point. 
Title: Re: compressors per channel
Post by: ruairioflaherty on April 22, 2018, 02:48:14 AM
My vote is for VCA, opto and FET are typically stronger flavors and better left to outboard. 

For me attack and release are essential, especially in modern production where so much of what we are doing is transient shaping and not dynamic control.  It's a huge part of creating depth in a mix, even more so now that verb is verboten (I exaggerate of course).

I'd suggest a 4 knob setup - threshold, attack, release and output gain and put the ratio on a 3 way toggle switch - 2:1, 4:1 or 10:1.  If space was really tight you could put attack, release and ratio on 3 way toggles all side by side.
Title: Re: compressors per channel
Post by: ruffrecords on April 22, 2018, 04:05:26 AM
Back in 1972, for my degree dissertation, I designed a compressor using some of those new fangled op amp things. I used  piece-wise linear approximation to generate the desired output level from the input signal. Then I compared that with the output and used the result to control a FET.

Anyway, a few years later I got into microprocessors and it occurred to me you could make a much better job of the piece-wise linear approximation in software. Taking this idea one stage further, there is no reason not to store attack and release and ratio information in a micro. Plenty of micros have EE memory so several presets could be stored. Seems to me a compressor with the ability to store several user defined settings would be very handy, whether it was per channel or stand alone. Why has nobody done this already - or have they?

Cheers

Ian
Title: Re: compressors per channel
Post by: zamproject on April 22, 2018, 09:24:15 AM
Anyway, a few years later I got into microprocessors and it occurred to me you could make a much better job of the piece-wise linear approximation in software. Taking this idea one stage further, there is no reason not to store attack and release and ratio information in a micro. Plenty of micros have EE memory so several presets could be stored. Seems to me a compressor with the ability to store several user defined settings would be very handy, whether it was per channel or stand alone. Why has nobody done this already - or have they?

Hi
It exist, at least I know one.
http://www.davehilldesigns.com/TITAN.html (http://www.davehilldesigns.com/TITAN.html)
Don't know how it sound, never use it

I design a system (fader automation) using ARM stm32, I have in mind since I play with this MCU card to try driving a VCA for compression (already done for "volume" automation via midi)
I just have to hook an AD an test it...
What accuracy do you think is requested ? I have a ready to work 12bit AD, which I'm able to perform 11bit without any jitter.
What would you let at analogue side chain, only the wave rectifier ?

For the OP, I have in mind the VCA compressor optionally fitted at fader modules in the Studer 900 series, but I just have a look back at the shemo and it's an advanced design... to many component for a "simple" unit

Best
Zam
Title: Re: compressors per channel
Post by: pvision on April 22, 2018, 10:15:43 AM
IIRC the channel compressor on the SSL 4000B used the channel VCA as the gain reduction element. The 4000E-on used a separate VCA in the channel dynamics section. Wouldn't surprise me if the AMEK Big used the channel VCA for the virtual dynamics

Examples of micro-controlled compressors include the FMR RNC Really Nice Compressor, the Distressor & others

The SSL 4000E (etc) channel dynamics was a great utility compressor but a little too clean & undramatic for my tastes. I'd put an 1176 per-channel but recognise that might be a little extreme for most. So a VCA compressor with minimal controls might hit the spot

SSL used it for gating / expansion as well as to allow compression & ducking by linking to other channels. Something to consider

Nick Froome
Title: Re: compressors per channel
Post by: abbey road d enfer on April 22, 2018, 10:39:20 AM
The SSL 4000E (etc) channel dynamics was a great utility compressor but a little too clean & undramatic for my tastes.
Isn't it what one would want on a channel comp? Too much of the same "colour" can be stodgy.

Quote
  I'd put an 1176 per-channel 
Each one his own...  :)
Title: Re: compressors per channel
Post by: ruffrecords on April 22, 2018, 11:56:09 AM
Hi
It exist, at least I know one.
http://www.davehilldesigns.com/TITAN.html (http://www.davehilldesigns.com/TITAN.html)
Don't know how it sound, never use it

Zam

That comes close but it has no way to store and recall presets, I would have though that would be an easy addition.

Cheers

Ian
Title: Re: compressors per channel
Post by: abbey road d enfer on April 22, 2018, 12:42:09 PM
Seems to me a compressor with the ability to store several user defined settings would be very handy
Depends in which hands and circumstances. I agree that "total recall" of a whole mix is a nice feature in a work in progress, so switching between projects is made easy.
The downside is a tendancy from many to use the big name preset (insert favourite singer/SE/mixer), which is counter-creative IMO.
It's not a new thing; I was hired in 1977 to engineer a session in a recently installed studio. When I left, they hurried to put on paper all the mixer's settings. I tried later to tell them how futile it was, but they were not convinced.
I should have reset the mixer completely flat; that would have enhanced my guru status  :-[
Title: Re: compressors per channel
Post by: EmRR on April 22, 2018, 01:11:17 PM
My vote is for VCA, opto and FET are typically stronger flavors and better left to outboard. 

For me attack and release are essential, especially in modern production where so much of what we are doing is transient shaping and not dynamic control.  It's a huge part of creating depth in a mix, even more so now that verb is verboten (I exaggerate of course).

I'd suggest a 4 knob setup - threshold, attack, release and output gain and put the ratio on a 3 way toggle switch - 2:1, 4:1 or 10:1.  If space was really tight you could put attack, release and ratio on 3 way toggles all side by side.

I like these thoughts but would add, as I understand it, attack and release in a dbx type circuit is peak detection with modified timing, differing greatly in sound and distortion profile from dbx 'overeasy' rms detection.  If I can only have one, I want it to be rms overeasy.  Especially if it's 8 talking heads in broadcast with more downstream processing.   

Title: Re: compressors per channel
Post by: ruairioflaherty on April 22, 2018, 01:25:42 PM
Makes sense Doug but I'd add that nobody is mixing broadcast on an analog board anymore!  If this is really more than a thought exercise for Pucho then we have to look at how people are really working these days. 

DAWs have presets covered, what I want is a a hands on compressor that I can dial in quicker than I can find the closest preset.  FWIW I hated using that Titan but I'm not a Swiss Army Knife kinda guy.

DAWs have recall covered, if I need fast recall (and almost all pro mixers I work with do) then I stay in the box and use minimal if any outboard.

It's a personal opinion but I think plug in compressors can control dynamics and hold things in place reasonably well, where analog shines is in shaping the sound.  I use both in my mastering daily but I almost always use digital compressors to make things smaller and more contained, never to shape transients or enhance groove.

Optos are a non runner these days anyway, parts are hard to find. 

A FET compressor that you could linearize enough to be a generally useful tool could work, and it would have the advantage that you could un-linearize it with a "fun" button.

Title: Re: compressors per channel
Post by: abbey road d enfer on April 22, 2018, 01:43:51 PM
I like these thoughts but would add, as I understand it, attack and release in a dbx type circuit is peak detection with modified timing, differing greatly in sound and distortion profile from dbx 'overeasy' rms detection.  If I can only have one, I want it to be rms overeasy.  Especially if it's 8 talking heads in broadcast with more downstream processing.
Overeasy and adaptive timing are not mutually exclusive; you can have one, teh other or both. I mentioned the possibility of adding a soft/hard switch earlier, soft being overeasy.
Title: Re: compressors per channel
Post by: EmRR on April 22, 2018, 03:03:29 PM
Overeasy and adaptive timing are not mutually exclusive; you can have one, teh other or both. I mentioned the possibility of adding a soft/hard switch earlier, soft being overeasy.

Right
Title: Re: compressors per channel
Post by: ruairioflaherty on April 22, 2018, 07:05:19 PM
Makes sense Doug but I'd add that nobody is mixing broadcast on an analog board anymore!  If this is really more than a thought exercise for Pucho then we have to look at how people are really working these days. 

DAWs have presets covered, what I want is a hands on compressor that I can dial in quicker than I can find the closest preset in a DAW or digital console.  FWIW I hated using that Cranesong Titan but I'm not a Swiss Army Knife kinda guy.

DAWs have recall covered, if I need fast recall (and almost all pro mixers I work with do) then I stay in the box and use minimal if any outboard.

It's a personal opinion but I think plug in compressors can control dynamics and hold things in place reasonably well, where analog shines is in shaping the sound.  I use both in my mastering daily but I almost always use digital compressors to make things smaller and more contained, never to shape transients or enhance groove.

Optos are a non runner these days anyway, parts are hard to find. and ROHS etc.

A FET compressor that you could linearize enough to be a generally useful tool could work, and it would have the advantage that you could un-linearize it with a "fun" button.
Title: Re: compressors per channel
Post by: EmRR on April 22, 2018, 09:49:57 PM
Makes sense Doug but I'd add that nobody is mixing broadcast on an analog board anymore!

So you would think, but I keep mixing press conferences for live broadcast on them as that's what's supplied by the contractor to the contractor to the contractor for sports events. 
Title: Re: compressors per channel
Post by: ruairioflaherty on April 22, 2018, 10:36:37 PM
So you would think, but I keep mixing press conferences for live broadcast on them as that's what's supplied by the contractor to the contractor to the contractor for sports events.

Cool, glad to hear that. Let me be more specific, no one is buying analog consoles for broadcast. There's definitely gear in the field in use.  Again, I'm wondering if Pucho is thinking about a real design or if it's just a mental exercise. 

I'd always prefer to mix live on an analog board.
Title: Re: compressors per channel
Post by: EmRR on April 23, 2018, 12:18:29 AM
Cool, glad to hear that. Let me be more specific, no one is buying analog consoles for broadcast. There's definitely gear in the field in use.  Again, I'm wondering if Pucho is thinking about a real design or if it's just a mental exercise. 

I'd always prefer to mix live on an analog board.

I've mostly ran digital mixers in the field for live and broadcast events in the last 4 years.  I can say the analog mixer in question for broadcast use was specified because the A1 is always a local, and the contractor must have a system that any pickup tech can drive, with no learning curve.  It would require a different operational protocol for them to go with a digital mixer in the future....like bringing a dedicated crew....I'd say there are definitely live event people who would not be able to go into that particular job on an unknown digital mixer without a crash and burn. 
Title: Re: compressors per channel
Post by: pucho812 on April 23, 2018, 04:00:13 PM
Well  let's see beyond thought process.   Shall we. Every topology will have pros and cons. This we know. 

What about a pwm compressor per channel and kill any vibe we have😂
 
Title: Re: compressors per channel
Post by: abbey road d enfer on April 23, 2018, 06:07:34 PM
Well  let's see beyond thought process.   Shall we. Every topology will have pros and cons. This we know. 

What about a pwm compressor per channel and kill any vibe we have😂
I wouldn't venture myself i this territory. PWM gain control involves questionable filters. And unless all operate on a common distributed clock, possible interference is an additional issue.
Title: Re: compressors per channel
Post by: ruairioflaherty on April 24, 2018, 02:10:03 AM
I wouldn't venture myself i this territory. PWM gain control involves questionable filters. And unless all operate on a common distributed clock, possible interference is an additional issue.

Agreed and PWM is not sexy in the consumers eyes.  Ask the average engineer (not us geeks) to name a PWM compressor and you'll be met with a blank stare, maybe 1 in 10 could suggest Pye….maybe.  Compare that to "Varimu" as a selling point.

I'd need a lot more info on the target audience to make any kind of useful suggestions. 

Title: Re: compressors per channel
Post by: zamproject on April 24, 2018, 08:43:04 AM
Agreed and PWM is not sexy in the consumers eyes.  Ask the average engineer (not us geeks) to name a PWM compressor and you'll be met with a blank stare, maybe 1 in 10 could suggest Pye….maybe.  Compare that to "Varimu" as a selling point.

I'd need a lot more info on the target audience to make any kind of useful suggestions.

mmm...more or less the same with varimu for "average" engineer ?
maybe easier to name one, but only 1 in 10 don't name Manley  :)
PWM daily user here, not one, but 10 loaded in the desk, which is not far from one per channel (when comp inserted)
But yes, it's not the easier design to fit at each channel. and high frequency switch section need special care.

Best
Zam
Title: Re: compressors per channel
Post by: abbey road d enfer on April 24, 2018, 11:16:15 AM

PWM daily user here, not one, but 10 loaded in the desk,
Which compressor? Can you describe what made you choose it?
Title: Re: compressors per channel
Post by: zamproject on April 24, 2018, 01:06:20 PM
Which compressor? Can you describe what made you choose it?

hey abbey

Studer x89 modular series console,  the desk nor the compressor modules is clean compared to modern digital standard, but the compression itself is pretty "transparent", pumping effect is low even at extreme setting
Also as a 3 button setting (ratio/release/gain) it's an instant and easy tuning.
For sure not always the greater choices, but always a good result to handle dynamic so it's my basic go to
When more envelop/fx way is requested, I patch VCA comp
Finally I choose them because  I have them and a lot... (not saying I have them so I have to use them anytime anywhere!)

By the way sw freq is about 300kHz, low pass -3dB at 30kHz and the high freq osc section is in a mu-metal box  ::)

Best
Zam
Title: Re: compressors per channel
Post by: abbey road d enfer on April 24, 2018, 04:45:52 PM
By the way sw freq is about 300kHz, low pass -3dB at 30kHz and the high freq osc section is in a mu-metal box  ::)
Yes, I checked that; the oscillator is free-running, no sync from a master clock. I would be worried about frequency-beating...
Low Pass filter is 7th-order-ish, clock rejection should be about -140dB.

Timing/pumping/transient response is confined to the side-chain; that's the beauty of a linear VCA. Other types have the gain cell element interfering with the side-chain, particularly optos, that's a reason why some love them, but broadcast designers don't like this element of surprize.
Title: Re: compressors per channel
Post by: zamproject on April 24, 2018, 05:18:12 PM
Yes, I checked that; the oscillator is free-running, no sync from a master clock. I would be worried about frequency-beating...
Low Pass filter is 7th-order-ish, clock rejection should be about -140dB.

what do you mean ? non stable clock ? or some king of interference/submodulation ?

yes rejection is good i never heard the clock :P

Best
Zam


Title: Re: compressors per channel
Post by: abbey road d enfer on April 24, 2018, 06:15:23 PM
what do you mean ? non stable clock ? or some king of interference/submodulation ?

yes rejection is good i never heard the clock :P

Best
Zam
You can't hear the clock at 300 kHz, but you could hear the beating tone between two slightly different clocks. Let's say one is at 301kHz and the other at 299, you could hear a 2kHz beating if you mixed both signals.
Title: Re: compressors per channel
Post by: zamproject on April 24, 2018, 06:51:16 PM
You can't hear the clock at 300 kHz, but you could hear the beating tone between two slightly different clocks. Let's say one is at 301kHz and the other at 299, you could hear a 2kHz beating if you mixed both signals.

Ok, get your point, that's what I'll call intermodulation then.
I can't see a situation where it can occur, clock is shared by parr (stereo modules whatever sc link engaged or not) so no issue with stereo material.
Maybe if I process the exact same signal via two different comp ? but it's not a use case
And the clock have to pass the filter I guess !

Next time I check and align the modules I'll set a Lissajous at the scope to tune all clock modules equal  ::)

Best
Zam
Title: Re: compressors per channel
Post by: abbey road d enfer on April 24, 2018, 07:24:14 PM
Ok, get your point, that's what I'll call intermodulation then. 
Not really; intermod involves non-linearities in the transfer function. Beating happens in linear systems. Check
https://en.wikipedia.org/wiki/Beat_(acoustics)

 
Quote
I can't see a situation where it can occur, clock is shared by parr
I don't see that on the schemo; the 7555's are free-running.

Quote
(stereo modules whatever sc link engaged or not) so no issue with stereo material.
As the name implies, it's only the side-chains that is linked, I don't see any connection going out or into the 7555.

 
Quote
Maybe if I process the exact same signal via two different comp ?
No, just having two channels routed to the same output.

Quote
And the clock have to pass the filter I guess !
Wet-finger calculation says clock atenuation may be about 100dB real world, which is probably enough.

Don't sweat it; Studer's designers have done it long ago. My comment was addressed to anyone who would want to design a PWM gain cell and put it in a mixer.
Title: Re: compressors per channel
Post by: zamproject on April 24, 2018, 08:09:46 PM
Ho !
I don't think we talk about the same compressor  :)
There is no 7555 in my 1977 Studer 289  !
I suppose you refer to the late 80' 96x series which have "PDM-VCA" compressor fitted at master modules;)
which is a PWM of course...

Best
Zam

ps:tks for the wiki link
Title: Re: compressors per channel
Post by: abbey road d enfer on April 24, 2018, 08:19:00 PM

I don't think we talk about the same compressor  :)
There is no 7555 in my 1977 Studer 289  !
I suppose you refer to the late 80' 96x series which have "PDM-VCA" compressor fitted at master modules;)
which is a PWM of course...
Exactly.
I'll look if I can find the 289.

EDIT; OK I've found the schemo; indeed there's a single oscillator for two channels, so there's no risk of beating. The low-pass filter is only 5th order so the clock must be about -80dB below nominal signal.
Title: Re: compressors per channel
Post by: zamproject on April 24, 2018, 09:08:34 PM
yes it's this one
sorry I don't point any schemo link guessing you'r a well documented guy

Nominal is -15dBu for this comp, internal level at insert, console setup @ +8dBu line level
so the clock noise should be around -95dBu, 103dB below line level if I'm right
My overall noise floor is way over that with all channel open, outboard insert and fx at send/return  :)

Best
Zam
Title: Re: compressors per channel
Post by: Whoops on October 25, 2018, 10:23:41 PM
I would like to have in each channel of a console an analog emulation of this compressor:

(https://img.wavescdn.com/1lib/images/products/plugins/full/renaissance-axx.jpg)

Title: Re: compressors per channel
Post by: abbey road d enfer on October 26, 2018, 03:46:54 AM
I would like to have in each channel of a console an analog emulation of this compressor:
No ratio control, no release control?
I think a custom implementation of THAT chips combining AN  (Application Notes) 115 & 113 would fit the bill.
http://www.thatcorp.com/datashts/dn115.pdf
http://www.thatcorp.com/datashts/dn113.pdf
Title: Re: compressors per channel
Post by: Whoops on October 27, 2018, 10:03:39 PM
No ratio control, no release control?

Yes,
Great isn't it?

It just works  really well.

It would be better if it didn't even had the attack knob
Title: Re: compressors per channel
Post by: abbey road d enfer on October 28, 2018, 12:39:55 AM
Yes,
Great isn't it?

It just works  really well.

It would be better if it didn't even had the attack knob
Then you could use the basic implementation of the THAT audio engine 4301. Fig 2 of http://www.thatcorp.com/datashts/dn107.pdf, with the addition of a simplified meter as in http://www.thatcorp.com/datashts/dn132.pdf