Simple resampling tool?

[stk]

Hi,
Can anyone recommend a simple way to batch resample a tonne of wavs from 24 down to 16bit, with the least amount of loss?

thanks, sk

[shamann]

Foobar2000?

[Juanjo]

For free, AudioMove or Voxengo r8brain... I think both do batch resampling:

http://www.lcscanada.com/audiomove/
http://www.voxengo.com/product/r8brain/

[Mr. Tunes]

goldwave has batch functions too

[Reverse Engineer]

http://www.voxengo.com/product/r8brain/

[stk]

http://www.voxengo.com/product/r8brain/

Tried r8 free, unfortunately it only does bitdepth changing (ie 24 to 16) if also changing the sample rate (which I don't want to do).

Will try out the other suggestions, thanks!

sk

[DuX]

Tough decision, mate. r8 is the best resampler there is on PC. Sounds like you need a dithering tool, otherwise you're gonna lose mucho quality. Now I know there was a command line tool called "dither", but I don't know where.

Thinking it over, Foobar2000 have a decent dither and can do batch conversion. It's certainly better than this dither from 2000.

Cheers!

[nitex]

Found the converter from the "Shibatch Audio Tools" pretty nice. Download it (*.tgz can be extracted by WinRAR for example) and copy ssrc_hp.exe into the folder next to your wave files. Then use a batch file like below (placed next to the waves too), to do the mass conversion. You can give it the filename start.bat or start.cmd.

http://shibatch.sourceforge.net/download/ssrc-1.30.tgz

Code: Select all

@echo off
cls

md converted

for %%a in (*.wav) do call ssrc_hp --rate 44100 --bits 16 --twopass "%%a" "converted\%%a"

Oh, and if you really wish, this command line app has some dithering options too. Add the parameter below with the desired algorithm to the commandline call in the batch file.

--dither [<type>]
0 : no dither
1 : no noise shaping
2 : triangular spectral shape
3 : ATH based noise shaping

[stk]

Thanks again - haven't had a chance to try these suggestions yet, but will, likely this weekend.

I know next to nothing about dithering, but I assume it's the lack of this which causes that fizzy digital noise on a cybal tail when one reduces a quiet 24 bit sample down to 16bit?

sk

[93143]

I know next to nothing about dithering, but I assume it's the lack of this which causes that fizzy digital noise on a cybal tail when one reduces a quiet 24 bit sample down to 16bit? :shrug:

That sounds about right, although you really should boost the volume or normalize before doing the conversion, because dither only goes so far.

Dither is low-level noise added to a high-bit-depth sound so that when the bit depth is reduced, transitions between the least significant bits are represented by changes in the average position of the noise (the signal 'dithers' between the two values) rather than just a step. It can sound significantly better on quiet stuff.

You've probably seen low-bit-depth pictures that appear speckled across the colour transitions - that's analogous to dither. It looks bad to someone who's used to 24-bit picture formats and colour-optimized GIFs, but simple step transitions between the limited number of colours available would look a lot worse.

[stk]

I see, thanks for that.

sk