API 312 clone, how about adding a great but simple limiter?

For me, the 312 project is to get great audio into my DAW (Protools Digi002 rack). So I'm trying to keep the signal path into the DAW as simple and high quality as possible. Then do my EQ and compression in Protools. For now at least, I'm actually quite happy with the Bomb factory compressors, however, I'm already getting the itch to clone a LA2A.

But for now, I'd like to get the strongest signal onto the 24 bits of the DAW as possible. So I'm thinking it might be nice to add a high quality limiter circuit to my API 312 preamp clones. I'm thinking of something in the quality range of a 1176 or SSL, but what if you stripped out everything that make those circuits so versatile, and just have it do one thing, limit the highest peaks as transparently as possible, you shouldn't even hear it working. I would think that this would reduce the complexity of the circuit significantly. I'll be making 10 pres in one 4u or 5u rack with the preamps in 5 matched pairs, I'd want each pair to have the limiters side-chained together. Of course I'd want true bypass for just pure 312. Absolute quality would be paramount.

What do you guys think? Would you recommend a classic circuit for stripping down, or start fresh? I already see a major limitation is the +16v/-16v power supply of the 312s. Still, any ideas?

Kelley

This is something that I would like to attempt as well.

I've always been interestred in the Apogee soft lim (?name) cct. I only ever heard once but, but it seemed to be really transparent. AFAIK it is also an analogue limiter.

Anyone have a schem?

Peter

bounce

Well, you can't really make it simple, if you wan't highest quality, because it needs a welldone sidechain detector circuit. A really simple way would be to make an opto comp. That can be done with only a handfull of compoents. But it's not good as a (protection) limiter.

I've played around with this one and it's pretty simple. But once you ad trannies, PSU and a decent output stage, it's not so simple anymore

http://sound.westhost.com/project67.htm

Someone also posted a link to a nice D&R FET comp. I liked that one! If not FET, then VCA. Still not simple.

Unless you do diode clipping I doubt you can make it simple & neat.

Don't underestimate Apogee's SoftLim feature! They are apparently using huge amounts of DSP power just for this function!

As for the sidechain of a brick-wall limiter, that would be simple: You just use a comparator, comparing your rectified (output) signal with a preset threshold. If anything goes ofer, then full-on-GR untill under thres again. Light the "lim" LED with this as well. Attack time should be as short as possible. Release timing takes some tweaking, but look at the SSL "Auto" release for ideas.

Optos are too slow for overload protection limiting. Vca's degrade sound quality. Fet's add noise. Take your pick on a compromise.

You could add a zero-crossing detector and a sample-and-hold, making it possible to only change gain when zero crossings occur. This reduces compression-generated distortion to almost nothing - on the cost of possibly having one half-wave clipped on threshold transition.

Jakob E.

[quote author="gyraf"] Vca's degrade sound quality. Fet's add noise. Take your pick on a compromise.

Jakob E.[/quote]

Hey Jakob,

While most VCA´s does degrade sound quality..I´d be hard pressed to say that about the Vca designed GML 8900 series 3 compressor..Soo Bad designed vca´s does audio noo good..good design..hmm..well just work..;-)... :grin:

Kind regards

Peter

You're right - could you get us the GML schematic?

:razz:

naa.. what I mean is - within reach for us DIY'ers, there are certain compromises that has to be taken. And I'll bet, that even the GML VCA sounds considerably worse than their linear gain stages..

Jakob E.

[quote author="gyraf"]You're right - could you get us the GML schematic?

:razz:

naa.. what I mean is - within reach for us DIY'ers, there are certain compromises that has to be taken. And I'll bet, that even the GML VCA sounds considerably worse than their linear gain stages..

Jakob E.[/quote]

*GGGG*...aaaahhh...Jakob...me thinks GML schematics is a no go ;-). Whose Liniar gain stages are we talking about..??? Anything GML..???. I mean I have to agree with somebody else around here..(the name slips my mind at the moment) first time I worked with a 8900..I throught it was not working, or I had forgot to push a bottom somewhere..I swear..up to 40 db of gain reduction without hearing it work...its an amazing piece of "artwork"..scary..really scary...I´m gonna buy me the first one later this year..have been saving up my money...been a dream of mine for years to own one of these monsters..already own 4 ch of GML preamps..jesus Ch... do they sound goood..;-)

Kind regards

Peter

Hi Kelley,

If you can accept the opto compressor/limiter slow attack time, you can do some simple experiments with Forsell´s opto compressor http://www.forsselltech.com/SideChain.PDF or some of the opto compressor application that Kev have on his great DIY pages, like http://recording.org/users/kev/KROC_schem5.PDF

You can try to apply the photo resistor direct over the R3/20kohm feedback resistor on the API 312 card and feed the side-chain input from the 2520 op-amp output or from one of the secondary on the 2503 output transformer.

Ok, this is not the best way to do a "Optimize/limit" function for the digital recording, but it can be a fun experiment to start with.

--Bo

[quote author="Bo Hansén"].....you can do some simple experiments with Forsell´s opto compressor http://www.forsselltech.com/SideChain.PDF[/quote]This looks promiseing, it certainly looks simple enough. This is the one being discussed int the Forssell compressor thread, I'll follow the thread with the idea to dovetail it into the 312 circuit. Thanks!
Any other ideas?
Kelley