Optical Comp

[analag]

[quote author="bcarso"]Opto compressor-limiters rely mostly on the intrinsic dynamics of the photoconductor. You simply can't speed them up over what they are.[/quote]

Exactly, an LED activated compressor such as the one shown here is meant for transparent squeezing. VCA type is much better suited for "effect compression" where we can fiddle with Threshold, Ratio, Attack, Release and gain make up all day.
Brad requests data and I've been searching the old mental archive, hopefully I might remember where I had found some. Here is a link to the other isolator manufacturers maybe they might be more accomodating to your inquiries.
http://www1.silonex.com/audiohm/constants.html
http://www1.silonex.com/audiohm/application.html
Do check out both, the second link has a nice opto comp with some tweakable paramters.

analag

 

[bcarso]

[quote author="analag"]
Here is a link to the other isolator manufacturers maybe they might be more accomodating to your inquiries.
http://www1.silonex.com/audiohm/constants.html
http://www1.silonex.com/audiohm/application.html
Do check out both, the second link has a nice opto comp with some tweakable paramters.

analag[/quote]

Glad to know Silonex is still at it. You are right, they are more likely to give the time of day than the EEG empire companies. I remember when they were pitching to Harman in the early 1990's as replacements for Vactrols. Instead, Harman eliminated optos in the automotive products altogether to satisfy the quality assurance gang and went to VCA structures, which frankly never sounded as good but were more statistically echt. Such is progress :razz:

 

[Kingston]

Opto compressor-limiters rely mostly on the intrinsic dynamics of the photoconductor. You simply can't speed them up over what they are.

Exactly, an LED activated compressor such as the one shown here is meant for transparent squeezing. VCA type is much better suited for "effect compression" where we can fiddle with Threshold, Ratio, Attack, Release and gain make up all day.

Speed up no, but slowing them down.. shouldn't that be possible with similar time constant structure as in PRR vari-mu? Big caps with slow discharge times?

And speeding up... wouldn't different leds have wildly varying characteristics, some of them having faster reaction times than others? Also, could this be further unfluenced by artificially changing the response of the opto element, for example by changing the led driving charateristics: more linear or more logarithmic, or even exponential way of seeing the signal? (I understand the tubes already play a big part in this opto vs. signal reactance)

I'm new to opto compression so excuse the possibly unconventional use of the terminology. Also, I have no idea how to achieve all that as analog implementation as my background is in audio plugin development.

 

[analag]

You can always use a S.S. sidechain such as this to achieve what you obviously want. It's not for me though, as I'm already getting the performance I need.

analag

 

[Kingston]

Yeah, I had a look at that earlier, not quite realising it can be easily adapted to here as well. Although if I was to build this, I'd probably start with your original schematic. Could be it really is quite optimal for my uses as well.

Heck, since it already works for both vocals and master bus, it must be pretty damn adaptive and good sound.

 

[analag]

If squeezing is what you're after, it does it like any Big Boy manufactured gizmo can. I notice how dynamics can slip by VCAs, nothing gets by this boy when he's on watch!!!!

 

[Kingston]

:green: :thumb:

 

[bcarso]

[quote author="Kingston"]

Opto compressor-limiters rely mostly on the intrinsic dynamics of the photoconductor. You simply can't speed them up over what they are.

Exactly, an LED activated compressor such as the one shown here is meant for transparent squeezing. VCA type is much better suited for "effect compression" where we can fiddle with Threshold, Ratio, Attack, Release and gain make up all day.

Speed up no, but slowing them down.. shouldn't that be possible with similar time constant structure as in PRR vari-mu? Big caps with slow discharge times?

And speeding up... wouldn't different leds have wildly varying characteristics, some of them having faster reaction times than others? Also, could this be further unfluenced by artificially changing the response of the opto element, for example by changing the led driving charateristics: more linear or more logarithmic, or even exponential way of seeing the signal? (I understand the tubes already play a big part in this opto vs. signal reactance)

I'm new to opto compression so excuse the possibly unconventional use of the terminology. Also, I have no idea how to achieve all that as analog implementation as my background is in audio plugin development.[/quote]

LEDs are fast! Read some references on photoconductors---that's where the action is. And yes of course you can slow the system down by slowing down the drive. You can also find optocouplers with faster response photoconductors, but the sound that is prized has much to do with the peculiar photoconductor characteristic. They also have the benefit of essentially zero offset voltage, pretty good linearity, and reasonably low noise, but if it is flexibility of attack/release you want you'd be better off with good VCAs.

Photoconductors also have a number if disadvantages, among them unit-to-unit variability, history dependence, and a high temperature coefficient.

 

[Kingston]

[quote author="bcarso"]LEDs are fast! Read some references on photoconductors---that's where the action is. And yes of course you can slow the system down by slowing down the drive. You can also find optocouplers with faster response photoconductors, but the sound that is prized has much to do with the peculiar photoconductor characteristic. They also have the benefit of essentially zero offset voltage, pretty good linearity, and reasonably low noise, but if it is flexibility of attack/release you want you'd be better off with good VCAs.

Photoconductors also have a number if disadvantages, among them unit-to-unit variability, history dependence, and a high temperature coefficient.[/quote]

I certainly wouldn't want this thread to stray further in to VCA design. I'm just musing on the opto design limitations I'm unfamiliar with. It's META dig time for me.

As far as history dependence on photoconductors, I always thought that was one of the great things about them, as a natural and cheap way to program dependency of the envelope detection. I certainly wouldn't file it under disadvantage, and I'd think that's one of the main attributes contributing to the transparency of the squeeze we all like. :grin:

 

[bcarso]

When I say history, I'm talking about time scales of hours, days, weeks, months....

This reference may, dare I say it, be illuminating. Note well the sections Speed of Response and Light History.

http://www.mcdelectronics.com/PC101.html

 

[Kingston]

:shock:

and



I had the nagging feeling history dependence was not the same thing as program depency of compression even before I posted. Decided to talk out of my ass anyway. I stand corrected. :wink:

 

[bcarso]

Not to worry :grin:

 

[bovox]

Hey Analog, nice design! but what is the missing resistor from anode
top of U3? or am I missing something that I thought was a WCF?
Cheers Bo

 

[analag]

There is no missing resistor because it's not a White Cathode Follower. How many variation of the cathode follower are you familiar with.

 

[Viitalahde]

Don't remember if I said thanks but I say it once more. :razz: This thing looks interesting and is on my to-be-tested list.

 

[analag]

It should not disappoint the mastering engineer. I think you will like it a lot, if I may say so

analag

 

[Viitalahde]

Damn, and I've been needing to build an opto compressor anyway. :? First I've been thinking of a tubed Forssell comp but this does look tempting to try.

This one has less knobs to tweak! :razz:

I'll have to look around what kind of junk I have, maybe I could put it together pretty quickly.

 

[analag]

By manipulating the volume and compression control you will be very surprised at how much sound control can be had from just two knobs. This compressor is strong yet smooth, that's my best description. I choose it over my knob infested other such things and not just because I made it.

 

[Viitalahde]

Well, I think I'm on it. Will prototype a 1-channel version first. Since I run everything unbalanced, I'm probably omitting the input transformer..

If it does what you say, it's exactly what I want: A compressor that just tightens things up a tad when pushed, not radically manipulating the rhythm/flow. :thumb:

 

[analag]

[quote author="Viitalahde"]
If it does what you say, it's exactly what I want: A compressor that just tightens things up a tad when pushed, not radically manipulating the rhythm/flow. :thumb:[/quote]

It's great without an input transformer but when I made the transformerless tube pre I needed balanced in so I stuck a CMLI15/15 from Cinemag in it. I like this compressor so much, I made a stereo version for my bus insert, it's one of my weapons I use against the local studios in my location. I like my sound digital clean but analogue warm and phatttt.