do you know how to dither?

  • Thread starter Thread starter kristian
  • Start date Start date
kristian

kristian

New member
I cannot for the life of me get dither to sound good. i set the quantisation to 16 bit, as im planning on burning the cd. (one problem i see is it doesnt have any options for me to drop the file from 48 to 44), then i've tried dither depths of 1, 1/2, and 2. I also usd noise shaping 1, 2, and off, and all combinations of those but it sounds like assmaster. is it perhaps the 'vegas dither' algorithms? do i have to shell out for some real dither programs?
 
You say ditha I say dither

I'm not familiar with VEGAS' dither algorithms. I use UV22, TPDF, HPTPDF, and SNR2. these all sound great. Quality dither algorithms shouldn't affect the final output much. If you are hearing artifacts, I would definitely find a better algorithm.
 
kristian said:
I is it perhaps the 'vegas dither' algorithms? do i have to shell out for some real dither programs?

It IS the vegas dither algorithyms. No matter what I try in dither( as well as most of their other plugins) they always end up adding phantom dirt to the audio. I just gave up and by default now record at 44.1 khz

Also try this. Record a vocal into ptfree, listen to it several times, and then import that file into vegas.

BIG difference.
 
Dither does not change the sampling rate. Dither is designed to preserve some of the extra dynamic range when going from a higher bit-rate to a lower one. If you're not recording at 24 bits, you shouldn't need to do any dithering.
 
this is my problem. im at 24, so im moving it to 16. good that gets me straight. but then i have a 48kHz sampling and i dont know if i am resampling that through the dither to 44.1... when renderig it seems like it would stop you if you shouldnt render a bunch of 24/48 as 16/44. feckarse. It really sounds poor. like when i used to use my SB Live! as a soundcard.
 
Resampling is a totally different operation than dithering. I recommend sticking with 24/44.1 when recording and dither down to 16-bits after you've finished all processing.
 
well its too late for that, i have 20 minutes of music recorded at 48. hmm.. i think ill check if its downsampling for me. i still think the dithering algorithms are crap because it shouldnt sound that bad.
 
As a test, why don't you just resample something to 44.1 but leave the bit depth at 24 bits (no dither). I think you'll find most of the quality loss is the resampling.
 
give the mana prize! yea thats right, i kept 24 bits, and downsampled, and it sounds like crap. i rendered as a 16/48 file = dithering but no downsampling and bobs your uncle. it sounds good. now i have a problem. what the hell do i do to get a good desampling program/algorithm?
 
I don't know if this is possible, but it might be worth a try. If you can play a 48k file and record at 44.1 at the same time, you might try connecting your outputs on the card directly to the inputs. Play the 48k file, record a new 44.1 on the input. Leave everything at 24 bits until the very end.

The idea here is that an extra trip through the D/A and A/D converters might be less damaging than just resampling if your converters are pretty good (and if it's a 24 bit card, they should be decent converters).
 
thanks guys...

You just told me what that dither button on my mastering compressor plugin is used for....cool. :)

Where's that link for the dither article??

zip >>
 
Back
Top