:-)

so am i ;)

Thanks for an interesting thread folks. I've really never understood compression and I'm beginning to see why.

I was thinking of building a desktop unit but found that the transformers are incredibly expensive. I know a guy that was building these really nice tube units for studios. One knob on the panel and dead easy to use.

Does Cakewalk ever issue white papers? I would like to learn some more about this stuff before I decide to build or buy something better. I don't use PC at all because it doesn't like me. The 64 bit producer plugs seem to work better for me. In fact, isn't the VC64 channel strip very similar to ProChannel?

More questions than answers I'm afraid.

Good diagram Mike, I can see why you like the Sonitus. IIRC the GUI shows the relationship between knee and the threshold very well. Looking through a few of the sample files I noticed that the Sonitus is all over the place but PC is only being used for EQ. Check out the Javier Colon track and you'll see it right away.

Hhmmmm.... COMPRESSION

Good thread this one...it really explains this whole compression issue to me rather well.

Couple of questions...

1-when would you leave compression out?

2-I know that I could go with what my ears tell me in terms of what works for me when it comes to compression but is there a ratio range that could be used as a good range?

CakeAlexS


So you are saying the Sonar X1 Power book is wrong here?
Could you submit an errata to Scott? (I've found a few fiddly issues myself which will hopefully go up sometime soon).

http://www.garrigus.com/powerbooks/SX1P/SX1PErrata.asp

All books have issues and mistakes.... so does software... and so do human beings :).
It's a lot easier to fix books and software. It's up to all of us to help each other improve as human beings..

Yes. As is the help file. But I have noticed that Scott seems to take quite a bit from the help files (rightly or wrongly)


 It states that the compressor will start working when the threshold is set to the level of the audio data. In fact the PC 4K starts working a good 10db before. Totally confusing for a beginner or even more advanced user. 

great diagram

This is definitely the reason. In DAW world, we are used to 0 dB as 0dbFS (full scale). It is different in analog. I didn't make any research into this specific compressor, but on a real one, threshold marked in dB probably means dBu, which is something totally different, it is a specific voltage level. You usually can go much higher than 0dBu, so when you set 0dB on that hardware, you can get pretty significant compression (depending on input level). I expect cakewalk made this knob to emulate this behavior. Yep we have now very simpler life now with DAWs.

Hi All,

Sorry for the confusion on this.  It is exactly right that the challenge, when modelling hardware, is that 0dB full scale in the digital domain represents hard-clipping, but on an analog desk, 0dB is "nominal-level" - kind of like a musical mf or f.  Clipping, OTOH, can be anything the manufacturer wants it to be (often around +20dB).  The original, analog-board threshold ranges make no sense for a DAW.  While many plug-in manufacturers have left the original ranges, they also include an Input Gain, and an Output Gain, on the entire channel strip to allow users to force the signal up into the dB range that the plug-in is looking for.  Since we didn't want to have these extra controls on every modeled ProChannel module, we opted to break from tradition and push the working threshold range down below 0dB-FS where it is sensible for a DAW.   As other posters pointed out above, a setting of 0dB should produce a slight amount of compression for input approaching 0dB, due to the soft knee starting to engage.

FWIW, there was a bug reported by users (in X1C and earlier) that, for the Bus Compressor, the tooltip range on the threshold knob made no sense - it still displayed the original analog range (+20dB max).  This was fixed in X1E/Expanded/Production Suite.  Important:  The compressor processing didn't change with this fix - it was only a tooltip issue.

So again, yes, the threshold range doesn't match the original hardware, but it's that way for the convenience of SONAR users.  One last note - the hardware doesn't act quite as "sensible", or predictable as plug-in dynamics processors.  In the electronics, there seems to be "interesting" behaviors for Threshold, Ratio and Internal Make-up Gain. Since this is circuit-modeled, our plug-in has the same "character".  :-)

Hope this helps.

Cheers,
~Bob

Hi Bob,
Thanks for the explanation. Really appreciate you folks chiming in.

Keep up the good work,
Dean