Not sure what you're trying to say here - though you are getting to my point in that the SL4000 Quad bus comp was...."Almost near Necessary".....for the many signal paths, around 134 switchings per channel/Bus/Routing module, then add the A,B,C,D routing and it was doing a slightly different task in that if you've used a 4000/E/G then you understand that when you are pushing the Channels to Drive the Console Hard... (Often pinning meters) the Console Master Quad Comp...IS. near a requirement in keeping Signal Level and Flow under Control?

The point regarding you G/SSL build....Sorry, I have no idea what you are talking about save you've built a unit with various Mods etc? Save the fact once you add more stuff to a unit it ceases to be that unit, though if you want to add as much stuff to an SSL outboard Comp it matters none to me? I mean people might dig Fet LA2a's though it's a near defeating the purpose scenario IMO..!

I think theres a 500 series version of the classic comp that was made by the legendary, lost son of groupdiy.com - Igor. You can find it on groupdiy.com, and unless I am mistaken, the user Ramshackles (double check that spelling) has boards for sale.

The VECA I linked to owes its heritage to the SSL bus comp (particularly Jakob/Gyraf's work on the GssL), but I have changed so many things that I would consider it a unit in its own right.

I would especially note the M/S balance control for the SC blend (I don't know of any other piece of gear that does this!?). Its something I tried for fun, but I was really surprised at how much control it gave when I sat down with a prototype and tweaked it for the first time.

Simply put, its lets you control how hard the center image of your program material is detected without altering the M/S balance in the signal path. This adds a new dimension to your sound sculpting possibilities.


Gustav

Great Post Gustav,
I think going back in the annuls of History there have been a few (2-3 Commercial compressors) offering M/S Control. Though these have mostly been Orban and other Broadcast Peak Programme type Compressors, and not the type favoured by Engineers for Music and scoring etc, IMHO (The SSL XLogic 5.1 G Bus Compressor doesn't even offer this) M/S options even when mixing material can often give a greater sense of depth and perspective as to what one might wish to achieve in a final Mix. Having one on a 500 Series Compressor would be a nice addition.

What I would really love is to take the Best from "All of the ever made" Series of SSL Bus Type Compressors and pick and choose various features, throw them into a 19" or 500 series module and sell it with a Dual Sidechain/Full or Signal path S/C, even offering even DBX 202c VCA's to THAT 2180A's and have the option of "Crush", M/S, Side-chaining Thru as opposed to just a Send/Return Side-Chain (I.E Full Frequency Pathway) and everything else.....that would make one damn Hot SSL Bus Compressor in a 3 Slot unit or 19". In the end though...Man...can Jakob can be proud that he designed a Compressor from the SL-4000/E Master Module that now counts up to over 25 separate units tracing heritage back to that original slant on the SL 4000 Bus Compressor design!

Hi,

For those who own one, I have a question. Does it lower the signal around 10dB?

I have it on the Inserts of my Main channel and I notice the signal comes back around 10dB quieter, of course even when no compression is happening and in bypass mode.

Do any of you experience something similar?

I didn't build it myself, so don't know details about how it was built.

Thx!

It´s depending on how the output of your GSSL is configured. With the change of a resistor it can be adapted to balanced or unbalanced output, resulting in a difference of 6dB. Give it to a good tech and point him to this: http://gyraf.dk/gy_pd/ssl/unity_gain.jpg

thanks jensenmann, actually it has the 27k resistors. The problem happens only when used in the Main Inserts on my Mackie 1642vlz3. If used in series (mackie output -> gssl -> speakers) no attenuation happens. I thought it has to do with wiring? though the cables im using seem to be fine (pin2 is tip and return respectively, pins1 and 3 go to sleeve). So i have no idea why this is happening.

Guessing here but I'm pretty sure the Mackie inserts are unbalanced. Connecting unbalanced and balanced works fine but you get a -6db gain reduction. So it could be working as expected.

Might be more to the story, maybe someone else can fact check me and chime in.

youre right, after some hours of reading this is expected behaviour due to the unbalanced inserts. Thanks!!

Being the guy who did the True Power Sum detection for VCA comps like the ones we build. That was ported over and called Turbo after we presented it publicly for the DIY field. We are not the only ones who use TPS, just the ones who released the methodology after having to design it without any real documentation to guide us. So, ours may sound a little different, don't know.

Anyway, it does true up stereo image and also makes the compressor more balanced WRT stereo compression.

The GSSL uses a single 2150 VCA for a mono (mixed voltage) detector which makes it really aggressive in the mids and LFs.... Not a bad thing to say the least! The TPS method sums the current of 2 detectors not the input signal voltage.

The single detector widens stereo imaging, compressing Mid more than Side.

My advice:
Keep the Gyraf detector available if you build it with Turbo.
I went back and forth with other experts on the mono detector in the GSSL, but none of those were musicians or mixers, just chip pushers...

I really liked the original and would probably use it more for audible compression. Once Turbo is in the picture it gets much smoother and transparent which many people do not like.

Me, I like both in their particular strengths.
Bottom line is keep that 2150 detector available or build 2, one with and one w/o Turbo.

The Gyraf can sound VERY similar to it's original design, but there are so many parts available for detectors and VCAs that will change the behavior that I would strongly recommend building it using the exact parts called out in Gyraf's documentation.

My Chameleon Labs 7721 stereo VCA compressor has one IC detector for each channel, but when I apply a constant -12 dB signal to the R channel and toggle a quieter -13 dB signal on/off on the L channel, the GR on the R channel is increased during the presence of the L signal. What sort of detection summing scheme does that imply? True Power Sum means RMS sum? I get that TPS is different than mono-sum, I guess because--like you said--you're summing the voltages from the detectors, not the audio signals. But from my experiment, I would guess that my compressor is NOT simply reacting to the louder of the two signals. Is that correct or is it more of a delta/loudness change that is being listened for? Is that the difference then, between the detector voltages and the audio voltages? The detectors are delta? If that's crazy, just tell me and I'll go back and settle for the simple phase-agnostic aspect/result of TPS over mono-sum detection.

And forgive me for asking this, because I know it's been covered, but this is NOT the reason why the original FX G384--which is considered less transparent, etc.--is more aggressive, correct? All the SSL designs had fully stereo sidechains, right? It's rather something in the...gain structure? I actually don't know. Paging Dr. Foote!

Hi,
Here is how some/most stereo linked compressors are connected internally.

The inputs each go to their respective detection circuits.


The detector outputs are then summed to create a single Ec signal (the voltage that the VCA uses to operate dynamically)

That voltage is applied to both VCAs.

This is the classic setup for stereo compression where we want to preserve stereo imaging.

The behavior you cite sounds correct, although it does not point to true power summing. You would see similar behavior with the V sum connection as well.

The GSSL used a single detector with the audio inputs summed as a voltage that in turn feeds the detector input. Instead of 2 detectors outputs summed as in true power or V sum of the 2 outputs.

That voltage is applied to both VCAs evenly.

The resulting compression compresses mainly MID signal, and compresses SIDE signal a lot less. This gives a widening effect.
This is a usable behavior, unless you want clinically correct compression of all audio signal, both MID and SIDE.

The GSSL Turbo uses TPS with 2 detectors.
I would not really want to eliminate the original GSSL mode, as it sounds GREAT when it works.

Caveat,
I got nearly zero sleep (again) last night.
I read this over and looks OK, but if I got something wrong, let me know and I will correct it.

I hope this answered some of your question.

Best,
Roger

Thank you, Roger. I appreciate your appreciation of the mono-sum detection scheme. It actually explains why everyone says the Obsidian sounds so wide. I have no doubt that that sort of comp would be useful.

I just tested my Chameleon Labs 7721 VCA and compression on this unit changes from 6.5 dB when presented with a stereo signal to to 5.5 dB when presented with that same signal on only the L or R channel. So it is closer to a 4k console compressor than a standard GSSL*, but still responds slightly more to the center image. So that is very cool to know. Dustin from RJR was telling me that compression and gain reduction are not the same thing. I was like WHAT? But I think I understand what he means now: summing control voltages (i.e., "compression" amounts) from two detectors can yield a different result (i.e., different "GR" amounts) than summing the audio signal voltage itself (mono-sum detection, GSSL style) and feeding that to one single detector.

Also for anyone interested in the Chameleon, my meter is way underreporting GR and 4 dB of GR on my 7721's meter was actually 6.5 dB of GR as measured with bx_meter in the DAW. You have to barely move that 7721 meter for most applications. On most mixbus settings it will wiggle almost imperceptibly, but it's still visual feedback lol I really like the sound, though. It seems similar to the C1 from what I've heard online. It's open with a very slight midrange emphasis that I find elegant and useful.

* = GR on the GSSL in this video DOUBLED from 6dB to 12dB when presented with the stereo signal during the same test https://www.youtube.com/watch?v=EQD8...nel=expataudio

The GSSL had ac voltage sum ahead of the single detector...
Unique, to say the least. Jakob probably worked for months on that, you know, integrating his detector into the old SSL circuit...

I don't know if he came up with that or not, just musing...

I often thought about 2 other schemes:
1- detector output voltages summed summed by a virtual ground summing node.

2- I built a bunch of MS boxes and some use multiple mono compressors to process the MS. But I wasn't very wowed by this kind of MS usage.

I thought you could have a detector setup where you use TPS, the Ct signal but have the inputs blendable between discrete L/R signals and an increasingly mono signal.

That would (should?) allow a variable widening (Erland effect?)

May or may not be original....

I am also working on a new compressor topology with 2 complete side chains like a P4DMS has, feeding the common VCA packs differentially.

I thought you could have SC1 set to nice smooth RMS/Feedback and SC2 set to explosive Attack/Release (peak) mode.

This would all be processed by the same VCAs.

Best,
Roger

That is cool stuff to consider! Thank you for sharing. Yeah, compressor sidechains could be a little more involved (on average) in the future. I'd be for it. The sidechain on the Gyraf G24 still impresses me, conceptually. More important than any of that though--to me--is a simple mix knob. Absolutely love having those on compressors.

And regarding continuous control over TPS/mono-sum detection, I figured since I do have two detectors in my stereo VCA comp, I could use some pre-emphasis/de-emphasis on the side signal to alter the overall mid-signal sensitivity, which I think would work.