Stereo Tool

Hans,

Working with Hip-Hop as a format on one of my stations, I deal with tons of hypercompressed material. I have had dialog with you before about how I like to use Declipper via Audition to fix up some of these prior to dropping the .wav in my automation system.

I am curious as to when you plan to include a feature similar to 'Undo' in the Omnia.9 into Stereo Tool. I know you mentioned you had a disabled feature called Natural Dynamics, which you said was too complex.

Anything on the board for something simpler that will handle this content? It has spread to CHR and is creeping into Country now. Waveforms clipped to death before we even process them. I know it will only get worse.

What are your plans on this?

This is an interesting topic and I guess there are couple of approaches to this?

Although maybe not what quite Chris mentions, what about having some sort of density awareness that could modify parameters on the fly?

Just so you know, there is already a density detection driven multiband expander in StereoTool. It's called Natural Dynamics.
However, it's a *very* CPU intensive filter and is hidden by default in the ST GUI.

I do use this in my presets and believe me it brings even a Core i5 processor to its knees. Unless you have a modern processor that has large L2/L3 cache sizes, forget about using it.

_________________
visit website

I still want to make something like this, and I have some new ideas that might work well - but I don't really know when I'll do it yet.

I did the same thing on my modern crushed music station.

Until we installed stereo tool. Went back and reloaded the original files.

When I turn Natural Dynamics on my CPU usage goes up to around 50% but I get buffer issues. I don't really understand the buffers so I have no idea why it does that. Even if I put the buffers on the Input and the Output up to the maximum settings for each I still have the stuttering as they exhaust and refill.

Clipper

About CPU usage:
If you have 2 cores, most of time you can see almost 100% (90% usually) without buffer issues!
If you have 4 cores most of time buffer issues will start even at 45-50.

All that because ST uses 2 cores and not even 100% of them!

So, CPU for current ST is better to be stronger per core then more cores!

Example - i can run Full FM preset (declipper + composite > strictness 16) with E5700 CPU, but not with Q6600 or Q9300.

Hans gave me directions on Natural Dynamics, but his warning made me shy away from trying it.

I may try it on an i7 box with a very high core frequency. In fact, I'll disable all but two cores to keep TDP down and clock the remaining ones higher.

Hans, can you share any of what you are thinking of doing? I really want to pitch your software as a replacement to our aging fleet of processors. I just know this topic will come up when I do so.

Might be a good idea to do this soon. I'll check what to do next as soon as version 7.23 is finished (I'm working on some smaller issues, made some good progress the last 2 days so it might be coming soon).

Reason for using 2 cores: There are 2 channels! So it makes sense to process the channels separately. Only the composite clipper is an exception, but the 1 channel has so much data that I can split *most* of the processing steps into 2 halves.
I could probably try using 4 instead, that would however mean that I have to split the audio into 4 parts and the parts need to be merged together at the end. And for that I need to do some things that will cause some effect on the audio. Not sure how bad it would be... Anyway, the composite clipper is the only thing that can be split into 4 sections, for the rest 2 is the maximum.

How do I enable the Natural Dynamics in the Winamp version? (I know about the CPU load and stuff, but I'm curious. ).

I've edited the .sts and changed Enabled to 1 in the Natural Dynamics section, but my CPU is staying the same and I do not hear any differences.

(using version 7.23).