Hey Schwa! Great to see you polishing your fine product here. :mrgreen:
Just a few questions:
Is the new vst version much different to the js one?
Do you only use it when rendering a mix down to 16 bit?
What’s the advantage over say the Waves L2?
Thanks! :ugeek:
I’ll let schwa handle most of this, but L2 obviously has limiting and loudness maximizing functions along with dither. The better apples-to-apples comparison is Psycho Dither to Waves IDR. I suspect that Psycho Dither simply kicks its ass.
Scott
Hi Bevoss! You left one of your s’s over at the Reaper forums.
The VST psychodither has a couple of major differences from the JS dither_psycho that’s packaged with Reaper (which is itself a pretty cool plugin 
– LOSER aka Michael gets a lot of the credit for that one).
On the audio side – for psychoacoustic noise shaping, the JS (like most or all other apps) uses precalculated coefficients, but psychodither works out its coefficients at run-time and keeps adjusting them as the audio signal changes, to leave the maximum dynamic range for the original audio material. Also, psychodither adjusts the dither itself as the signal changes. Pretty much anything it can do to get the bit reduction noise out of the way of the input audio, it does.
On the visual side, psychodither shows you the spectral density (power at frequency) of the audio it’s processing, and against that the density/power of the bit reduction noise. So you can see exactly how it’s pushing the noise into corners where your audio signal isn’t.
As for comparisons … my opinion is, this is the best bit reducer there is. I think the noise shaping is so good that for most material the best way to run this plugin is in “red” mode, which is minimal dither, and high order noise shaping. In that setup, there’s almost no noise power added to the signal. I don’t think there’s another plugin out there that can do this.
It’s appropriate for any time you need to render to a lower fixed-point bit depth. So, rendering from 64-bit float to 24- or 16-bit int, etc.
Thanks for the clarification.
I only meant the dither feature on the L2 actually Scott, sorry I wasn’t clear about that.
Can you please explain this quote more? I thought dither was what you have to do when you mix down from 24 bit to 16 ? Is minimal dither desirable ?
“As for comparisons … my opinion is, this is the best bit reducer there is. I think the noise shaping is so good that for most material the best way to run this plugin is in “red” mode, which is minimal dither, and high order noise shaping. In that setup, there’s almost no noise power added to the signal. I don’t think there’s another plugin out there that can do this.”
OK, here’s a somewhat oversimplified explanation. If you’d like any specific technical detail on any of this, please let me know.
When you reduce the bit depth of audio material, you are reducing resolution. Reducing resolution is removing information, and removing information creates artifacts (aka truncation noise). These audio artifacts are very quiet, but for complicated reasons, they are easier for our ears to discern than random noise.
There are 3 basic things you can do about these artifacts:
Nothing. The artifacts are very quiet after all.
Dither: lay down a blanket of white noise over the artifacts. This increases the total level of noise in the recording, but white noise is harder for our ears to discern than truncation noise.
Noise shaping: apply a specific kind of audio filter to the artifacts to change their character, making them harder to psychoacoustically discern.
The goal should always be to do the least harm to the audio material first, and treat truncation noise second. With that in mind, in some cases option #1 – doing nothing – may well be the best choice. If you are going to treat truncation noise, in my opinion a good noise shaping algorithm is preferable to dither, because dither does increase the overall level of noise in the recording.
wow, that’s a fantastic explanation. thanks very much. This would be great to have in the docs.
so if im understanding things correctly, one is to use the psycho dither on mixdown. or is this plugin stricly for mastering purposes?
i know you’re not supposed to dither more than once in the life of a song, so im wondering exactly where i need to put this plug in my chain/timeline.
i would really like it if i could use it on mix down to counter act ableton’s extreme quality loss that happens on 16bit renders.