View Single Post
Old 28th July 2020
Lives for gear

Hmm.. interesting. During Pro-L2 development I toyed a lot with this idea but it is _extremely_ difficult to pull it off naturally over a wide range of audio material. You've come to the same conclusion I did, it seems as you need the "max gain" type of control and huge look-a-head values to get it to behave properly.

There are real reasons for why clipping/peak limiting is what most companies have chosen for ultimate loudness, especially considering perceived transparency. It's simply very difficult to get an upwards algorithm to follow the music in a natural way. Most sounds decay naturally from loud to quiet.. same for complete works of music. Most parts of the music flow from loud to quiet. Thus maintaining that 'illusion' is really difficult with an upwards algorithm. Even if you heavily clip a kick drum, once it's been through the digital to analogue conversion, which involves filtering, the actual physical wavefront still keeps most of the original shape intact and thus the illusion. Yes you get some momentary distortion with all the negative that brings but at least the overall energy and shape of the sound is pretty much intact. This same thing is incredibly tricky to achieve with upwards processing.

Did try a quick demo of the plugin and I have to say you've made a great algorithm! At low settings it remains quite natural. Well done!

However, for actual mastering grade loudness boosting of full program material, it simply can't compete with the established plugins/hardware, in my opinion. You need to do a LOT of tinkering with your settings and pretty much have to use the Punch button + distortion to get back some of the lost impact from the skewed illusion created by boosting all the low level material.

Also, for mastering you absolutely need some kind of control over left/right linking. It feels like the algorithm is now 100% unlinked which on some of my test material creates real problems.

Having said all that, as a mixing tool this could be pretty interesting. Now the question is, what makes this unique from all the other upwards capable dynamics tools out there? For instance DMG Audio Multiplicity has absolutely ridiculously transparent upwards capabilities (both compression AND expansion if so desired). Then there are all the various Melda Productions dynamics processes that offer tons of upwards compression possibilities.. though they come packaged in a very awkward User Interface.

Will need to test the demo a bit more until making up my mind. Having explored these avenues for literally months I can appreciate the algorithm you have going. It's good! For mastering it's not enough in my opinion but for general purpose easy to use and very clean upwards limiting it's very impressive and undoubtedly useful.