Voxengo Premium Membership - All Voxengo Plugins For a Fixed One-Time Fee
Forums     Plugins     Marquis Compressor Program-Dependent Release Contour

This topic was created before release of the latest product version, and it may contain details irrelevant to this version.  Replying is disabled for this topic.

Ok...found this on the internet which explains what I am attempting to relay to you.  This is how I believe that three filters are wokrking.  Notice how in the first graph, you have a line, then in the second, you have a parabola (curve) and in the third, you have an S-curve.  Thats what I can extrapolate from all this.

http://www.themathpage.com/aPreCalc/equation-of-a-line.htm

WARNING; Further Insanity Ahead: Notice the effects that occur with first, second and third degree polynomials.  My only guess is that this is how the filters are being applied.  Where in stead of X^3 we have Param1 * Param2 * Param3. (Am I close Aleksey?)

The graph in Marquis...only relfects overall contour.  Speculation abound though.


Some probably important bit of information (and I'm starting to release 'trade secrets' here ;-)

To sum up, with release contour you are roughly defining the 'ideal' release steepness - if incoming audio signal follows it, compressor algorithm will make no adjustment to the specified release time.  Otherwise release time may become longer.

This paragraph will be included into the user's manual.


That makes sense.  So you are setting a center (ideal) point for the deviation.

Any comment on my thoughts about the combination of the parameters?

Aleksey Vaneev: Some probably important bit of information (and I'm starting to release 'trade secrets' here ;-)

Aleksey Vaneev: To sum up, with release contour you are roughly defining the 'ideal' release steepness - if incoming audio signal follows it, compressor algorithm will make no adjustment to the specified release time.  Otherwise release time may become longer.

Aleksey Vaneev: This paragraph will be included into the user's manual.


Individual parameter values are not important by themselves.  Only the overall release curve shape is important.

On your note about converting line into S curve, I do not think this is applicable to Marquis.  I'm not using this kind of lines/curves.  At least, they cannot be 'transcribed' to audio attack and release stages.  So, I guess there's no need for more speculations here.


Aleksey Vaneev: Individual parameter values are not important by themselves.  Only the overall release curve shape is important.

Aleksey Vaneev: On your note about converting line into S curve, I do not think this is applicable to Marquis.  I'm not using this kind of lines/curves.  At least, they cannot be 'transcribed' to audio attack and release stages.  So, I guess there's no need for more speculations here.

My only reason for speculating was that you had mentioned that with only one parameter, you could only vary the contour linearly, but with three, you could curve it.  At any rate, the answers there are probably too revealing of your trade secrets.  I will restrict my concern of the PDRC function to overall curve.


Aleksey Vaneev: Individual parameter values are not important by themselves.  Only the overall release curve shape is important.

OK - then you think the visual release contour graph conveys enough information to tune the PDRC (including the min/max indication you mentioned you are adding in a previous post).  No additional cumulative values or release metering is needed (new meter type release, expected vs actual or something - haha).  Thanks Aleksey!

Elvenking - thanks for additional notes on your thoughts also.


I love elvenking

Geeez this is a looonnggg thread.

All I need to do is set the 3 params to 10 and watch the GR meter to make sure the transients i want to compress are being compressed quickly enough, i find the master release works fine.

Anyway, this is'nt about PDR Aleksey, just thought i'd say that after purchasing Marquis I havent looked back, it sounds fantastic mate, especially on feedback and round mode.

well done on a great piece of work man.

Ged Leitch,

BitHead Mastering


Ged, thanks for your support!

This is an old discussion, but I feel that for an update of marquis, the way the knobs work should be made clearer by correct labeling.  Obviously, we have a graphic representation of something and the knobs allow adjusting the graph.  For me it is quite clear that three knobs/values allows drawing more complex graphs than just one.  Maybe you should have given just this information to satisfy the askers.  But: At the left down corner we currently have a label min and at the right upper corner a label max.  However, actually we see a x-y-field.  So obviously we have 2 parameters that influence the release contour.  Otherwise it were impossible to draw a graph like this.  So why not give labels to the axises?  Actually, I realized that also the filter section lacks a label of the y-axis.  In this case, it does not matter much since everybody is used to it and know it means amplitude.  But in order to make sense the release contour graph giving the information what actually is represented is crucial - since it is something specific of marquis.  It is no wonder that people complain about it.  This is what I am asking for.

cheers, drjee

This topic was created before release of the latest product version, and it may contain details irrelevant to this version.  Replying is disabled for this topic.