Jump to content

Same Compressor Settings, Different Chart


Recommended Posts

Was going through this article on plug-in settings that approximate "A Day in the Life" vocals and noticed that when I set my compressor to the exact settings as in the article, the graph looks different.


I've always wondered why there's a strange 45 degree slope above 0 dB on my compressor's graph, but I guess it's not that important since I shouldn't be operating in that region anyways.


I also noticed that the slope of the curved section and the y-intercept are different as well. Is this just a graphical difference and not a sonic difference? Is it because he used Logic Pro and I used Logic Express? I know at the end of the day the settings should be finalized by ear, but I'm just curious why supposedly the same product appears to be behaving differently.


In the attached image, the left window is from the article and the right window is mine. Thanks!


Link to comment
Share on other sites

If you adjust the level of a test tone feeding the compressor (set to something like 30.0:1 and a low threshold) you can see that for input peaks over 0dB the compressor stops compressing. The display is alerting us to this.


But I have long been curious - why this behavior on compressor's peaks? Analog compressors don't do this and Logic's behavior is unexpected. I suppose that is why there is a separate limiter paired with the compressor along with output distortion settings included. But still - why, when the going get tough does our compressor stop compressing? :)



Analog compressors don't do this and Logic's behavior is unexpected.


Actually - upon some deep pondering :shock: I realized that there probably a limit to the amount of compression that a hardware (analog) compressor can deal with. The lamp in a (opto) compressor can only go so bright, the control voltage input to a VCA cannot go beyond it's range. So instead of burning out a lamp or exceeding the specs of a transistor, Logic does the most graceful, elegant thing and passes the signal peaks on unchanged to the limiter/distortion sections of the compressor.


Anyway that is probably what the 45 degree above 0dB is about.

Link to comment
Share on other sites

So I changed to Platinum mode, switched to Peak mode, then switched back to FET. That fixed the y-intercept issue. That's really strange that a disabled parameter would affect the graph. Seems like a bug to me.


The shape is still off--I have to reduce the knee to 0.3 to get it to resemble the curve in the article. Weird. Maybe a version difference.

Link to comment
Share on other sites


This topic is now archived and is closed to further replies.

  • Create New...