Opto compressor experiments

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

RuudNL

Well-known member
Joined
Apr 26, 2009
Messages
3,106
Location
Haule / The Netherlands
At the moment I am experimenting with an opto compressor.
My prototype sounds good, especially when the ratio is not very high (say 2:1 and the threshold is low).
But...the problem is that my opto compressor is slow. And this is probably why it sounds pretty good!
I used VTL5C1 and VTL5C2 vactrols, without a very big audible difference.
Now, looking at the specifications of the vactrols I use, I see reaction times of a couple of milliseconds.
But some manufacturers of opto compressors claim attack times of 0.5 mSec!
How do they do this? As far as i know the cadmium sulfide cell usually needs time to get to the new resistance value.
Who knows more about this subject?
 
2cents:
One method is to drive the vactrol with a gate signal from a comparator (used e.g. in the Studiosound preamp limiter, probably not helpful in your circuit ), another useful thing is to add a little dc-bias to the sidechain to overcome the "led-turn-on-threshold-dead-zone" ;).

But attack speed is always an issue with vactrols...VTL5C1 is already one of the fastest
 
RuudNL said:
At the moment I am experimenting with an opto compressor.
My prototype sounds good, especially when the ratio is not very high (say 2:1 and the threshold is low).
But...the problem is that my opto compressor is slow. And this is probably why it sounds pretty good!
I used VTL5C1 and VTL5C2 vactrols, without a very big audible difference.
Now, looking at the specifications of the vactrols I use, I see reaction times of a acouple of milliseconds.
But some manufacturers of opto compressors claim attack times of 0.5 mSec!
How do they do this? As far as i know the cadmium sulfide cell usually needs time to get to the new resistance value.
Who knows more about this subject?
Ted Fletcher designed an opto compressor where the vactrol was under servo control; one LDR being in the FB loop of the LED driver. He claimed very small attack times, but I would think the release time was altered.
Now, if you look at the response time graphs of vactrols, you can see they are faster in the first millisecond than later, so by carefully balancing the resistor values, you can find an operating point where speed is increased.
For example, the VTL5C1 decreases from its obscurity value to about 2 kohms in 0.5ms. That is enough to produce about 20dB Gain Reduction (assuming it works against a 20k resistor). Indeed it takes some serious LED driver to supply 40mA.
When you complain about your comp being slow, is it the attack or release time that's slow?
 
Well, I am not complaining (it sounds pretty good), but the attack time could be faster.
Especially when I monitor the signal on a PPM meter (integration time 10 mS.), I can see serious overshoots.
 
Ted Fletcher also use a bridge configuration to allow more compression. probably nothing to do with attack but interesting.
 
RuudNL said:
Well, I am not complaining (it sounds pretty good), but the attack time could be faster.
Especially when I monitor the signal on a PPM meter (integration time 10 mS.), I can see serious overshoots.
If your goal is a compressor that is capable of dealing with fast transients, opto is not the best choice.
In fact, dealing with <10ms transients is not easy, whatever the technique employed (except digital). The best results are achieved using compound techniques, such as cascaded compressors, compressor with delay in the audio path and combination clipper/limiter.
 
RuudNL said:
At the moment I am experimenting with an opto compressor.
My prototype sounds good, especially when the ratio is not very high (say 2:1 and the threshold is low).

How are you changing ratio?

Cheers

Ian
 
ruffrecords said:
How are you changing ratio?
I use (DC) feedback from the LED in the optocoupler to the LED driver circuit. So: changing the gain in the LED driver stage.

abbey road d enfer said:
If your goal is a compressor that is capable of dealing with fast transients, opto is not the best choice.
At the moment I am getting pretty close. I increased the value of the resistor in series with the cadmium sulfide cell  (in the attenuator) , so that less change in resistance from the CDS cell is needed for the same amount of gain reduction. This improved things a lot, the attack time is now pretty fast. In fact I think it is good enough for what I want to use it for. Even the clicking sound of two screwdrivers hitting each other is efectively compressed! (And that is a very short peak!)
 
RuudNL said:
At the moment I am getting pretty close. I increased the value of the resistor in series with the cadmium sulfide cell  (in the attenuator) , so that less change in resistance from the CDS cell is needed for the same amount of gain reduction. This improved things a lot, the attack time is now pretty fast. In fact I think it is good enough for what I want to use it for. Even the clicking sound of two screwdrivers hitting each other is efectively compressed! (And that is a very short peak!)
That is good news. How is your PPM reacting to that? Does it still show overshoot?
 
On a standard DIN PPM (10 mS integration time) I don't see any overshoots now.
I will do some real measurements to get an idea of the attack/release times.
 
john12ax7 said:
Where did you get the vactrols? Thought they were no longer for sale due to rohs requirements.

https://www.ebay.com/itm/Xvive-Audio-VTL5C1-Opto-Coupler/331860686515?hash=item4d447036b3:g:l7wAAOSw2GlXKNOR:sc:USPSFirstClass!32828!US!-1


Many of the series are there too.....

I've bought chips from here and they are legit.....
 
Today I measured the attack time of my opto compressor prototype.
Attack time is in the range of 1 or 2 mSec. Fast enough for me! (Measured at 6 dB gain reduction.)
 
rogs said:
One of the Silonex optos - the NSL-32R3 - has similar attack and release times to the Vactrol, and being a current device can be easier (and cheaper!) to find..
Are you positive? My understanding is that Silonex has been absorbed several years ago, and now the only Vactrols available are those of Perkin Elmer Electronics.
 
rogs said:
One of the Silonex optos - the NSL-32R3 - has similar attack and release times to the Vactrol, and being a current device can be easier (and cheaper!) to find..

http://www.farnell.com/datasheets/1674205.pdf?_ga=2.146972542.899153935.1540673651-1951666110.1540673651

I used the NSL-32R3 on an optocompressor that I designed  for very good attack and release but Vactrol was better for it for its sound look even if with slower response times. In any case with Vactrol my optocompressor was  fast enough so I left it as reduction element.
I used the vactrol on the first sector of its resistance characteristic for fastest response.

EDIT: is very probable that several fast optocompressors uses this subterfuge for getting faster attack and release.
 
Back
Top