Premium Membership - All Voxengo Plugins at a Fixed One-Time Fee - Click Here For More Info
Forums     Plugins     Soniformer How about a Simple-Former?

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.






I was just thinking about how much I like the interface of Soniformer, and how it could be leveraged for a new product - a simple multi band TRACK/MASTERING compressor.

First, let me say that the best thing about Soniformer workflow for me is that you can pull down the graph bars and start out like you are setting up a single band compressor.  Then, you can add control points on, say, the threshold and move them around to hear what frequency ranges need more work.

This is soooooooooo intuitive.  Most other multiband compressors require the set up of ALL the controls in the bands separately, then move HP and LP cutoff points and re-adjust thresholds.  Not intuitive.

So how about this: A Soniformer style interface on a simple 3-4 band track compressor?  On a four band comp that would be 5 control points for each control bar.  Then, moving the control points horizontally would move the HP/LP cutoff points for each band.  Simple!

Of course, the number of control points would be fixed, and the control bars would be "stepped", not sloped.  And yes, there are other attack / release characteristics that would need to be addressed that are a non issue with Soniformer's 32 band's.  But, you could borrow features from Crunchessor for that.  Say, that would be a Crunch-Former - kind of catchy!  It seems like this type of design would fill the need for a simple multiband where Soniformer can't be used due to CPU issues.


Sounds like a good idea to me.

Actually the Maximiser in Magix Audio Studio is a bit like this but without the dragable control points.  It allows you to drag the lines between bands to set the freq. divisions.  It only has basic compressor controls so it is no comparison, but it does let you see the volume output (but not the in/out levels like in soniformer.) so it is the closes thing I have seen to what you are suggestion.


That's good to know, thanks for the reply.

EnzymeX, why would you want such compressor?  Do you want a different sound or less CPU demand?  A thing to consider is that compressor with lesser number of bands may not sound smooth enough.


CPU is the only concern - I love the sound of the Soniformer Demo!  The only reason I haven't ordered yet is I'm deciding whether to order this as part of a Voxengo software bundle or separately.

As you know, many times a single band comp is all that is needed for a track or bus use.  Other times, having two or three bands is useful for the following situations in a track or bus where Soniformer would be overkill due to CPU concerns:

-two bands: compressing the bass/lower mids separately from the highs (a la Endorphin - simple Bass / Treble balancing without a lot of fuss).

-three bands: limiting the compression to a sweepable mid band by bypassing the wet output on the top and bottom bands.

The more I think about it, if I needed four bands I would use Soniformer!  So maybe a three band soni-graphic comp - positioned as a simple track and bus compressor - would be the ticket to round out your range of offerings!  What do you think?  Thanks.

Thanks for the idea.  I'll think about offering such compressor.  However right now such plug-in does not fit my plans fully.

EnzymeX - here's a couple of comments to maybe think about discussing multi-band compressors if you haven't already.

I have Magix and true enough you can easily slide crossovers, Ozone lets you do that too.  If you test one of these type of multibands you'll see that they can more or less independently compress a mid band independent of the bass band or high band.  But what happens is the entire band (not counting some funny stuff at the crossovers) is compressed the same amount - that band might be 2 octaves wide (depending on how you set it) so all that stuff gets compressed when anything within that 2 octave range triggers the compressor - the entire 2 bands is compressed, not very precise and in some cases not smooth sounding.  If you try to make the range tighter, 1/2 octave for example, you run out of bands since you would need 20 1/2 octave bands if you wanted to cover the audio spectrum, also the normal multi-band compressor becomes less effective using tight bandwidths - even if it shows it is compressing a certain amount in the gain reduction lights an RTA placed inline will show it is fooling itself.  You can also hear that the compressor doesn't seem to be doing much (don't watch the compressor gain reduction lights - placebo effect).

The neat thing about Soniformer is that it does sound very smooth due to it's 1/3 octave bands.  If a very narrow resonance kicks in the compressor then the 1/3 octave band corresponding to that freq compresses it without disturbing adjacent bands.  If the sound energy is a bit wider then a couple of 1/3 octave bands may kick in - the band closest to the center freq of the wave crossing the threshold will get the highest attenuation with adjacent bands getting attenuated a lesser amount kind of 'feathering' or smoothing the compression that way too.  It even sounds smooth at all of the 1/3 crossovers (I'm not sure there are actual crossovers in Soniformer though).

I'm suggesting that cloning Soniformer and making a new variable band multi-band compressor would change the sound so it wouldn't be a clone-former at all.  It would be a 5 band multi-band compressor like the one Sonitus has (which is a good tool too).  Different types of tools gives a different sound !  I agree about Soniformer2 interface - it is very cool and quickly adjusts 32 bands.

Concerning cpu utilization - I'm not at home right now or I'd check this out.  I seem to remember an instance of Soniformer utilizing 7-12% of cpu on a P4 2.6GHz PC with a 44.1KHz 24bit wave, does that sound about right?  If I go up to 96KHz at 24bits I start getting into trouble - cpu utilization doubles and Soniformer is the least of my worries.  Elephant really eats the cpu (thank you for letting me turn off oversampling Aleksey!).  Other plugs chew up a lot of cpu too.  Add to that P4 normalization bug which pops up from somewhere when I add a lot of plugs...I went back to 44.1KHz because of all that.  I don't know what cpu you have or wave file types you're using.

When I get into cpu utilization issues I have Adobe Audition lock the track for me, Sonar4 does this too (there's a bug concerning this Cakewalk is looking into).  This frees up the cpu.  Also waiting to use buss effects during a later stereo mastering session is easier on the cpu too obviously.

Anyway - I just felt like talkin some audio since I'm out on the road and it's Saturday night !!!  Maybe a Voxengo 5 band multi-band compressor would be cool someday - who knows.  For a good single band (besides crunchessor) try the Voxengo Polysquasher too!  Goodnight :)

Strange, I find Elephant uses negligable CPU and Soniformer uses around 50%.

Of course I typically run only 1 of each... but I generally have Elephant on all the time (just in case the signal goes too loud) but I only use Soniformer in the final mastering.  I'd like to try 2 instances of Soniformer for the expander + compresor trick, but I don't have the CPU.

Maybe this is a denormals issue, I never though about it before...

I have a P4 3GHz, it seems strange we'd get opposite CPU usage.  Do you have HT on.  I turned mine off because it messes with FL Studio.

I don't have HT (hyperthreading - yes?) on my PC...I'll be home sometime next week so I can check my Soniformer2 cpu utilization better.  It would be interesting to see how it's humming these days - after going back to 44.1KHz life is a lot easier...

Yes, Hyperthreading.

I'd be interested to read of your results...  I would expect Soniformer to be the bigger CPU hog... if you just use an analogy of running 32 compressors and 32 band pass filters at the same time (though I don't know if this is how things are done internally.)

I don't get much CPU difference between Auto and 4x on Elephant, but I haven't really benchmarked it.

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.