Plotting compressor curves on PC

[Biggles]

I'm trying to find a way to measure and plot compressor/limiter curves on a PC.
This would involve amplitude sweeps and input / output plots
I use the excellent REW for eq frequency plots but can't find a way to plot amplitude other than painstakingly writing it all down and putting the data into a spreadsheet.
Any help would be greatly appreciated.
thanks

 

[TwentyTrees]

ainstakingly writing it all down and putting the data into a spreadsheet.

Must admit, that's the only way I've found to do it. Will be interested to see if there are better ways out there!

 

[gtroc71]

I found this great plugin a while ago while looking for the same type of thing.
Compressor Measurement Toolbox or CMT
Probably the closest for what you are after.
Best of luck

https://github.com/ilyaorlovru/cmt

 

[gtroc71]

Excellent tutorial on youtube

 

[12afael]

Plug in doctor can do it. I used to measure and graph it on excel

 

[user 37518]

Octave

 

[Biggles]

Thanks a lot for all your replies.
What I'm looking primarily is for hardware testing of my DIY units on the test bench and I'll check out all your suggestions. I'm sure there's some answers in there.
Cheers

 

[Bo Deadly]

Octave

While I agree that Octave is the best tool available for serious audio analysis, I don't think you can just post a link without explaining that it is a programming language that necessitates additional skills. Better still, post a .m script for generating compression curve plots so that the less experienced can put it to work. Better still, put it on github with examples and instructions.

 

[tobi_woerth]

I do that with PluginDoctor! You can send all kinds of signals to outputs of your interface and have the inout plotted. While Attack / Release is a bit too basic sadly the amplitude In/Out curve is really nice. Plus you can see harmonic distortion and frequency curve! Hope this helps. All the best, Tobi

 

[gswan]

It really is not that hard to measure on the bench with a signal generator and a couple of meters. If your meters measure in dBu then that's even better. Some higher-end signal generators also allow you to control them with simple software so you can set the compressor input amplitude values this way and then just record the outputs for each step.

 

[user 37518]

While I agree that Octave is the best tool available for serious audio analysis, I don't think you can just post a link without explaining that it is a programming language that necessitates additional skills. Better still, post a .m script for generating compression curve plots so that the less experienced can put it to work. Better still, put it on github with examples and instructions.

I think that the OP can read the first lines in that link to know what it is:

Scientific Programming Language

• Powerful mathematics-oriented syntax with built-in 2D/3D plotting and visualization tools
• Free software, runs on GNU/Linux, macOS, BSD, and Microsoft Windows
• Drop-in compatible with many Matlab scripts

And if that is not enough, the 'About' tab at the top clearly explains what it is: About

The OP asked for a tool, I provided one, I will not do all of his homework. Google, YouTube and a bit of reading are all that is needed from this point onward if there is a real interest. Just providing an .m script won't do any good, he must learn how to setup his audio interface and the basic operation of the software, again, reading will suffice.

But, I'll grant you that is not a bad idea to make some programs for different needs of the people here in the forum, I will probably do it in the future since I enjoy Matlab programming. Are you interested in participating?

 

[abbey road d enfer]

It really is not that hard to measure on the bench with a signal generator and a couple of meters. If your meters measure in dBu then that's even better.

It depends what your endeavour is. If you just want to check conformity of an existing product, it's acceptable, but if you're adjusting a unit, let's say in order to have the proper markings on the fascia, then doing it manually is a very lengthy and boring task. I use REW to do it but the display doesn't offer enough resolution for an accurate result. I miss the AP I had at work.

 

[warpie]

I never understood why compressor curves are not so common in measuring software. At the end of the day it's just a graph of Vin versus Vout, isn't it?

 

[abbey road d enfer]

I never understood why compressor curves are not so common in measuring software. At the end of the day it's just a graph of Vin versus Vout, isn't it?

That's correct. It seems doing Amplitude sweeps is not a priority for software designers. Compression is not supposed to happen in HiFi, that may be the reason.
I think I'll drop a line to John Mulcahy (REW software developper); he's probably capable of implementing it in a cinch.

 

[warpie]

That's correct. It seems doing Amplitude sweeps is not a priority for software designers. Compression is not supposed to happen in HiFi, that may be the reason.
I think I'll drop a line to John Mulcahy (REW software developper); he's probably capable of implementing it in a cinch.

John is great at listening to users feedback and REW is excellent IMO. I'm gonna be buying the enhanced version some time after summer because I'd like to measure more than one channel at once and also to support his effort.

It'll be amazing if he implements a compression curve graph!

 

[abbey road d enfer]

John is great at listening to users feedback and REW is excellent IMO. I'm gonna be buying the enhanced version some time after summer because I'd like to measure more than one channel at once and also to support his effort.

From time to time I make a little donation.

It'll be amazing if he implements a compression curve graph!

I sent him a request, but he seems to have an enormous backlog.

 

[Bo Deadly]

Bear on mind that the compression curve alone doesn't tell you anything about transient behavior. A really good analyzer would need to emit bursts and generate plots showing how the level changes in reaction to stimulus. So a complete solution would generate a compression curve and attack and release plots. And the plots should have multiple series for different burst levels and at different frequencies. And of course everything should provide accurate dB scales.

Using an Octave script with a USB audio interface would be near ideal for this. I only say "near ideal" because the interface would require calibration to generate absolute levels and not just relative.

 

[abbey road d enfer]

Actually John Mulcahy reminded me there is an amplitude sweep in THD measurement, but the process is quite slow.

 

[warpie]

So maybe someone experienced with octave could kindly code something for our lovely community?

Actually John Mulcahy reminded me there is an amplitude sweep in THD measurement, but the process is quite slow.

Is there? Never seen seen it. That's good news.

EDIT: But I can't seem to find it....

 

[abbey road d enfer]

https://www.roomeqwizard.com/help/help_en-GB/html/spectrum.html#steppedsine