Ross Mod Help

[tocs100]

I want a friend to mod a Ross Comp to use as a small vocal de-sser. Can anyone give the components to get the attack AND release down to one millisecond? I'd also like LEDS which show when 3dB, 9dB, and 20dB of compression occurs. (I'm sure guitarists would like that 2nd mod too!)

Thanks!

[Mark Hammer]

Reducing the attack AND decay probably involves reducing the value of the 10uf cap as well.  Try 3.3uf.  I'm not saying it's going to sound pretty.

[tocs100]

Thanks: Can anyone be more specific?

[Jay Doyle]

How much more specific can he get? There is only one 10uf cap!

Methinks you are asking too much of this little circuit. Getting A/D down to that level requires a bit more than changing out a couple of components. This little circuit was meant as a rough and tumble way to get compression on guitar, and, more importantly, it had to fit in a stompbox (the original circuit was the Dynacomp which had to fit in an even smaller enclosure). Because the A/D depends upon the above refernced cap and the value of the shared collector resistor feeding the 10uf cap, you will have to change either or both of those, which will change the working of the FWR.

As for the LEDs, this would require redesigning the whole circuit (as would your A/D mod) because there isn't a clamp on the front end that tells us how big the input signal is, so therefore we have no solid signal size to compare the compression to (and dB is a measure of comparision).

Short answer: you are going to have a hard time doing what you want. I suggest you look online for an actual de-esser.

Jay Doyle

[Mark Hammer]

Electronic Musician was raving about the RNLA recently (Really Nice Levelling Amplifier), which comes with a really nice price, based on what I've seen.  If your time is worth $8/hr, I suggest you and your buddy look into it.  Beyond that, it you thought Behringer stuff was cheap, try 2nd hand Behringer stuff.  I see plenty of rackmount units around where I live for $100 and thereabouts.

De-essers involve inserting a bandpass filter in the sidechain so that gain reduction is confined to, or focussed on, that frequency component known as sibilance.  The design of the Dyna/Ross does not lend itself to that sort of task very well at all.

Far better would be an adapted Orange Squeezer.  The OS naturally has a pretty darn fast response anyways, so it's off to a good start.  If you perf it, you can use a dual op-amp, with one op-amp allocated to the normal gain stage function, and the second one allocated to a dedicated bandpass filter imposed between the output of the main gain stage and the rectifying diode.  In principle, that would work.

Still doesn't get you your metering, though.

[tocs100]

....Far better would be an adapted Orange Squeezer.  The OS naturally has a pretty darn fast response anyways, so it's off to a good start....

Still doesn't get you your metering, though.

Thanks a bunch. I've already got the signal highpassed from a prior box. I just need a very small and very fast comp, and standard comp pedals have no release control. (One of the Maxon's has one LED, but I.m not paying $235 for it!)

[Mark Hammer]

I've already got the signal highpassed from a prior box. I just need a very small and very fast comp, and standard comp pedals have no release control.

There will almost always be a cap to ground for smoothing the rectified signal in just about anything that has a sidechain.  The smoothing is a function of how much lag that cap introduces, which in turn depends on how quickly the cap discharges.  You can almost always speed up the discharge of the cap by means of a parallel resistor.  The Orange Squeezer has a 100k resistor in parallel with a 4.7uf cap in that smoothing circuit.  Values smaller than 100k will speed up the discharge of the cap even faster (although it's fast to start with).

The other thing to consider for de-essing is that one normally wants full-bandwidth compression to have some lag for tha sustain-ey type sound, whereas de-essing should be instantaneous (particularly given how it adresses very quick bursts of high-frequency energy).  Normally, compressors use feed-back control of gain, where the control signal is coming off a point after the gain cell.  This means that gain reduction is contingent on the modified amplitude characteristics, not the original ones, with all their inherent variation, and gain changes are more gradual because of that. 

De-essing is a little more like limiting in that it should work best when the control signal is derived from the input, so that ONLY short bursts of critical frequency content have their gain cut back for ONLY as long as that burst lasts (or as close to it as you can get).  What this implies is that standard stompbox compressors oriented at sustain, and using a feed-back mode of control, are not well-suited to the de-essing function, whether you can easily insert a bandpass filter or not.

I tinkered around with the basic OS circuit last year and whipped up an experimental unit that let you select either feed-forward or feed-back control signals.  A suitable bandpass inserted into the sidechain of that might work out.  You can find the Tangerine Peeler here: http://hammer.ampage.org/?cmd=lt&xid=&fid=&ex=&pg=12 (scroll down a bit). I never built and tested it, though it ought to work in principle.

[tocs100]

That looks great Mark, thanks.