Wavosaur

[Joel]

I am using a VST (TLs-3127-LEA.dll) to do ‘hard limiting’ as in CoolEdit2.1. When I have VST:“Processing” checked, w/preset=”Peaks Delight” and turn up the gain to say 30, I hear what sounds like the start of signal clipping & the “Clip” light goes on, which I would expect. However the sound contains more Noise than CoolEdit would have added & is unacceptable.

At this point, “Processing” gets Unchecked, & “Apply” is depressed. A processing image to show that Wavosaur 1.0.0.4 is processing, comes&goes then the selected signal has in fact been changed & is larger with just a small amount of soft clipping (as expected). However the sound is garbled and totally unusable. What am I doing wrong? Wrong VST?? (I am new at audio processing)

[Rex]

Hello,

When you are in "processing" mode you can turn the knobs while you are playing the sound, and tweak the parameters until you get the sound you want.

with TLs-3127-LEA , make sure to set it on "limiter" (the peak delight preset is in "compressor" mode) and don't forget to adjust the "reduction" knob too.

You may use TLs-Maximizer, its sounds very good and is faster to obtain results. it works well as limiter

[Joel]

Thank you. Don’t know how I could have missed it. This vst will be usable.

However when I compare the wave forms the CoolEdit waves are never clipped even at 555db increase (totally ridiculous but to show the theory) , the vst waves go to square waves, when the gain is set to around 60.

Also is there another vst that has more analytical control & feedback. In cooledit instead of a ‘pretty’ interface, they have the tools to let me know what is going on that I may or may not hear or see. I can for example ask for the stats to tell me if I apply a certain db gain, that it will result in X% of samples being clipped. Very handy to know. I have noticed that many of the VSTs are pretty but at the expense of usefulness. I don’t think that digital I/O Is horrible. If you could point to there there are more analytical VSTs, I would appreciate that. thanks....

[Rex]

I don't use Cool Edit and i don't know how this hard limiting is working nor how it looks.

There must be some VST effects with analytics information, i'll do a search about that.

I know "func shaper" but it's a waveshaper : http://www.rs-met.com/freebies.html

Don't forget that VST works in 32 bits float (or 64 bits), so the "X% of samples being clipped" is not very useful, you can have values > 1.f that are not "clipped" until you convert to 16 bits.
it's a good idea to normalise in Wavosaur before saving to 16 bits.

[Joel]

Thanks again. I m learning much, as I go w/u’r help. I just found out how to control the vst via the input&Output gains together. Working much better.

However u brought up another ? which I had no idea I am dealing with. U said:
“…VST works in 32 bits float (or 64 bits), so the "X% of samples being clipped" is not very useful, you can have values > 1.float that are not "clipped" until you convert to 16 bits.”

My primary input signal is pcm wav @8bits/8kbs (voice/audioBooks w/smallFileSize) on Vista64. I tried audacity first but it could not handle 8bit. I just mention this because u were talking about 32/64 internal word representation that Wavosaur was using. Do I need to be concerned? & could u point to where there might be any further reading/info on this subject. Thank you again….

[Rex]

All VSTs work in 32 or 64 bits, so when you use VST plugins in Wavosaur to process your sound you have 8 bits pcm -> 32 bits floats (-1 +1) conversion and vice versa

you have to be aware that some plugins can go beyond the "limits" (-1 +1) without clipping, but the clipping will occurs when converting from 32 bits to 8 bits, you can avoid that with the normalisation feature of Wavosaur, to bring back the range to -1 +1