GR metering

[NewYorkDave]

It seems everyone agonizes over the subject of accurate gain-reduction metering, especially in the LA2A. I'm not familiar with all the compressor circuits out there, so someone my be doing this already, but why not directly compare the input and output signals and (compensating for any fixed amplitude differences) display the voltage differential directly, on a suitable log scale? That seems a lot more accurate than trying to infer GR indirectly through the sidechain.

 

[gyraf]

Yep, it could be done with this "ratiometric" approach.

I once made a led/ldr-based VCA-circuit with rather precise log tracking like this - with a dummy 2150 vca, some log circiuts and a comparator, making the ldr track the 2150. But it quickly becomes complicated to go this route.

But accurate GR metering is really not that important IMHO - mostly they're used as rough indication of what is goung on. Mind you, that any VU-type meter is way too slow to indicate precise GR anyway. You will always have to set the unit by ear anyway.

Jakob E.

 

[Michael Krusch]

Dave, I thought about the same thing, because I was unhappy with the metering inside a vary-µ compressor.
But then I came to the same conclusion like Jakob and I adapted the 1176 metering circuit.

 

[SSLtech]

In periods of silence, wouldn't it become impossible to display the slow recovery of a long release time? -If so, then the meter would wave about a lot whenever a signal reappeared that allowed it to re-make a comparison.

In fact any dynamic variation in the signal would require some pretty serious math that shouts 'digital' to me in order to simplify the sampling and comparison.

It sounds like a simple idea, but I can't think of any way to do it simply...

...which could well be my own failing! :grin:

Keith

 

[Michael Krusch]

You have to put a noise gate in front of the GR detection.