SSL clone mix control mod idea...is this possible?

would it be possible to make a control that would mix between the compressed signal and the bypassed uncompressed signal? a pot that when turned hard right would be fully compressed, and turned hard left would be uncompressed...similar to a wet/dry control on a reverb or delay.

any thoughts?

difficult or easy to do?

yea, but I would like to be able to do it on the comp itself if its possible.

I've found myself working lately on 32 channel API consoles only and when you want to do the mult thing with comps you run out of channels very quickly. I'm using the monitor section for a lot of that, but its not a lot of channels to really do a mix. It occured to me that it would be really cool to have this feature on the ssl to save a couple of channels if possible...

It would be very handy.

thanks..

Jakob?

Chris

I'm on to this! Shouldn't be any problem!
After I've built a couple of Gyraf SSL's for friends and had problems with space for the old 202 VCA, I started from scratch with a new version that will have atleast a switch for strait signal pass that also activates a second level controll for the compressed signal!

I think it would be easier, if you just use a 50k log stereo-pot, where you can adjust the "dry level" otherwise you would need a 4x pot...

connect the one side of the pot with pin3 of the input connector, the middle pin to the output of the vca with a 4k7 resistor in series...
third pin of the pot to ground...

that's how I would do it..

but I didn't try this yet... !!
so I cannot guarantee that this will really work...!!


mat

Im with Chris completely on this one. Im mixing on a 32 channel console also, except with a more limiting monitor foldback than an API and would jump at the chance to free up two channels from a parallel mult if I could keep it in the box when appropriate. Im not EQ'ing *all* my parallel busses, so it would be a nice feature to have when you want it. Even if its on a stepped switch that gave you some levels of resolution, would be better than nothing, if a pot wouldnt work.

my experimentation time is out the window right now, but if someone figures this out, Im all over it.

The more I think about it, a stepped switch might be the better way to go. Seems like you'd need all the resolution towards the top near unity, putting the top 2/3 of the uncompressed signal on a 12 step switch would probably work well for me. 1 dB steps down from unity, that would work fine for me!

dave

As the VCA's are current in/current out devices, and a halfway feed-forward design, a separate "direct" signal chain could easily be implemented by mounting a pot from before the 15K(27K) going to the VCA, directly to the inverting input of the post-VCA current-to-voltage converter.

Maybe a stepped switch would be best here, as it needs to be "off" in order to be bypassed.

Alternatively, a log pot between the mentioned point before the VCA and ground, with a 15K resistor from it's wiper to the post-VCA current-to-voltage converter.

I must confess, I have some real trouble trying to figure out how this "dry mix in" differs from straight compression with a lower ratio. But it does.

Jakob E.

This is a really good idea, because it would get around all kinds of the dreaded latency issues that you run into when splitting or duplicating channels on digital consoles and DAW's.

@jakob: what do you mean by :

"the inverting input of the post-VCA current-to-voltage converter"

:?: :?: pin 2 of the output ne5532 :?: :?:

jakob-

running drums in parallel is a pretty standard thing for me. One instance where this would be useful is say on 20:1 completely smashing the hell out of a close mic'd snare with a short attack so youve completely ruined the attack of the drum and turned it into that sponge rubbery snare skin sound. Bringing in an uncompressed signal with it will give you all the point that you need and the compressed half serves as the bottom and depth for the track. You cant do this with low compression, you need to do this in parallel. That said, I dont know that our simple control is going to grant us that ability anyway as in my example, you'd want the uncompressed signal louder than the compressed signal... Im usually bringing the compressed signal up behind the uncompressed signal but either way, its worthy of thought.

can anyone draw this one out?

dave

Dave is right, really what is needed simply is a parallel signal path through the unit that splits off the SSL input, and then combines the two at the output. So rather than a blend control, we need a parallel processed/unprocessed pair of signals, each with a seperate gain control at the output of the box. Simple as it seems, it would dramatically increase the flexibility of the unit, because mixing a smashed version of the signal in with the dry version gives a unique sound. The kicker would be to make sure that both are perfectly in phase.

[quote author="Steve Jones"] The kicker would be to make sure that both are perfectly in phase.[/quote]

err yeah
and the next thing you will ask for is the ability to mono the smashed signal and centre it.
:roll:
this will add more circuitry and have a summing stage ... and then what op-amps do you use ... and ... and ...
I like ultra simple DIY projects as it is often the simplicity and and clean signal paths that bring so much of the benefits.

Use the mixing desk ???

[quote author="Steve Jones"]Dave is right, really what is needed simply is a parallel signal path through the unit that splits off the SSL input, and then combines the two at the output. So rather than a blend control, we need a parallel processed/unprocessed pair of signals, each with a seperate gain control at the output of the box. Simple as it seems, it would dramatically increase the flexibility of the unit, because mixing a smashed version of the signal in with the dry version gives a unique sound. The kicker would be to make sure that both are perfectly in phase.[/quote]

Jonesy, just wanted to say the sex pistols rocked dude.

I don't know why the "kicker" would be making sure they're both in phase in what you described above...there would be quite a bunch of other "kickers" with your idea with that being the least of your worries...

just talkin about a blend control here...I think it might be best to just stick to the orginal idea without getting into summing and adding a bunch of additional circuitry...It kind of defeats the purpose then...

What I was asking about was if there is a way to combine the sound of the unit in bypass AS IS in the clone circuit (not from the input jacks directly connected to the output jacks) combined with the compressed signal and have a control to vary the amount...

A couple of different ways of doing it have been suggested that sound like they would work without too much modification...thats exactly what I was hoping for. thanks guys!

Chris

p.s. - I know you're not that Steve Jones

No definately not that Steve Jones the last time I checked, :grin: and no Kev, I'm not really asking for anything, just putting forward an idea to throw around based on Dave's post, as he has a good point. I am not proposing any kind of modification to Jakob's design, just a reasonably simple add on enhancement that would serve a really useful purpose, particularly for those using digital consoles.

Oops, I don't know how the triple post occured, maybe I need to ditch the Mac and buy a PC. :green: Sorry for hijacking the original concept, I got a little excited at the thought of what I could do with a built in parallel split on a drum subgroup!

> trying to figure out how this "dry mix in" differs from straight compression with a lower ratio. But it does.

Large peaks overwhelm the limited signal, burst out of the limiter.

Small peaks are controlled, large peaks come out large.

To me, it violates the whole point of limiting.

If you really want this break-out effect, use a "crummy" limiter. Both the LA2 and the 660 will bend-up at severely high input (the LA2 has a limit switch to get less of this "flaw"). Just a matter of nonlinear limiting action that can't handle large inputs.

I admit that given an SSL and a desire for this "poor limiting", a dry/wet mix seems simpler than messing with the SSL's near-perfect sidechain signal. And as you say, mixing at the I/V converter is probably simplest.

But watch phase. If dry and wet are opposite phase you get a null between low and high signals. If the dry signal has 6dB loss, and the limited signal is at 6dB GR, output cancels. When GR is zero output is around -3dB from unity, and when GR is much more than 6dB the output is -6dB from unity. That would be a "very special" effect, and probably have very few musical uses.

I think there are many ways to use a limiter... It also depends opon WHY you are using a limiter. On my session today, for instance, I tracked a snare with three mics, one hitting a 2500 like a monster, another crushing in a daking and a third all open and nice. The 2500 mic pretty much filled out the bottom, the daking mic was eq'd for maximum ring and both were brought up BEHIND the uncompressed mic to round out the snare and make it big and then the buss of all three went to an 1176 before hitting the tape deck. If you have a lot of limiters, you can use them strictly as effects on certain bands if you have the right EQ's which is what this setup was all about. When I mix, I'll mult the snare and do a similar dance. Its not about the rote task of limiting. Its about isloating a frequency, where like the shell ring in a snare and making that as loud as the stick attack on the skin, or at least being able to control it as such. Being able to do this inside the box on the SSL or any limiter would be just fine and dandy. Would it defeat the purpose if you were using it to just control peaks? Sure. But part of a multi band thing on a single source, its a great option to have. As for it complicating a simple design, thats all a matter of perspective. If you dont want to complicate your ssl, dont do it. for me, for the value that a mod like this has, Id not consider adding a summing stage as overcomplicating anything. Im sure a someone at neve suggested that the limiter side of the 33609 was overkill and complicated, but I sure am glad that feature is there... Regardless, it sounds like it can be done somewhat painlessly. I cant wait to get some free time to tinker with that, I started a record today, I think my next blast of free time is in six weeks, argh...

dave

Steve, I think you can delete your multiple posts if you want to...Unfortunately, I think what you are talking about isn't simple, you would need to have a summing amp to combine both of those signals like that. Thats like a whole other project in itself probably as complicated, if not more so than the whole SSL. I'm not sure you understand that..

It does appear there are a couple of different and simple ways to do a mod that would do essentially the same thing without having to go through all of that, but you would have to have the uncompressed signal be the clone's bypassed signal. (NOTE: that is not the same as going from the input XLRs directly to the output XLRs)...The signal IS going through the circuit, but uncompressed...follow? The effect is identical, its just that its happening within the circuit as it was designed. Its not a true bypass, but you could never have that anyway because you'd have to introduce all this other circuitry to combine and mix them like that...then you are also messing with the whole output section of the compressor as well and changing the sound...

[quote author="PRR"]> trying to figure out how this "dry mix in" differs from straight compression with a lower ratio. But it does.

Large peaks overwhelm the limited signal, burst out of the limiter.

Small peaks are controlled, large peaks come out large.

To me, it violates the whole point of limiting.

If you really want this break-out effect, use a "crummy" limiter. Both the LA2 and the 660 will bend-up at severely high input (the LA2 has a limit switch to get less of this "flaw"). Just a matter of nonlinear limiting action that can't handle large inputs.

I admit that given an SSL and a desire for this "poor limiting", a dry/wet mix seems simpler than messing with the SSL's near-perfect sidechain signal. And as you say, mixing at the I/V converter is probably simplest.

But watch phase. If dry and wet are opposite phase you get a null between low and high signals. If the dry signal has 6dB loss, and the limited signal is at 6dB GR, output cancels. When GR is zero output is around -3dB from unity, and when GR is much more than 6dB the output is -6dB from unity. That would be a "very special" effect, and probably have very few musical uses.[/quote]

what the :?: ...I'll be watching my phase PRR...Thanks for the tip or I might have forgotten. As for the "special effect"... go assist a top mix engineer for a couple of years or at least try it out before you comment on it...
you obviously haven't done either.


I thought this mod would be a cool idea and I have reasons for wanting it that have developed over the last decade of working. It would be very important and extremely useful for me. Some other people appear to also be interested which is wondeful and those that are interested can all benefit. lets stay on DIY please.

thanks,

Chris

I haven't tried it yet, but I think that this would work:



It would be a simple upgrade - just solder five wires to the PCB, and mount the four needed 22uF electrolytics directly on the 10KB potentiometer..

Can anyone see a problem in this approach?

If it was a "real" feedback compressor it surely wouldn't - but as I see it, this would do the trick...?

Jakob E.

This whole thread gives me an idea for a new DIY box: I, myself would love to have a device that makes this kind of processing available for just about any stereo device without using up busses on your mixer.

How about making a simple box with a pair each of 'master' inputs and outputs and a pair of sends and returns. The only control would be a 'blend' or 'mix' knob. The box would split the signals, send one pair of splits to the 'sends' at +4 level and receive the processed signal at the 'returns' and blend the unprocessed split with the processed returns with the 'blend' control to a +4 output. This way we could use it on pretty much any stereo processor we wanted to (SSL clone, 1178, 670 :wink: ).

Shouldn't be hard to do, and it would be pretty useful, I'd think.

Anyone?

JC

the problem is, you can only control how much of the dry signal is added, but you can't remove any of the compressed one. Seems to me you want a crossfade from the dry tot he compressed.

This could be done with one pot per side like a pan circuit. You have a resistor on each channel (the compressed and dry) then a pot with wiper to ground and each end to one of the channels, then a resistor on each side again into a summing stage. One of these for right and one for left, so you'll need a dual linear pot.

You could maybe pull this off between the VCA and the current to voltage converter. I guess you'd have to add another summing stage.

mark