Hello,
I just begun using FB2000 a few days ago. It's great. I think it is the best sounding player out there by far.
MY QUESTION. for MP3 playback with my terratec DMX 6fire 24/96 sound card, should i be using "24-bit fixed point output (under play back section)??
Thanks
great forum.

Yes.

You might also want to try the 32-bit fixed point output, it is supposed to give better performance (although the quality should be exactly the same).

The whole 32bit deal with Terratec cards has absolutely nothing to do with output bit depth, it's just the way the audio driver transfers 24bit data over the PCI bus to the soundcard (24bit 4 byte mode, or 32bit). So, DO NOT use 32bit output, the cards don't support it.

will he also want to check off "dither"??

No. Although the difference would probably be inaudible.

Thanks for your help guys!!

Hi
I'm also new to foobar 2000. I'm using version 0.586 and I was wondering what the best "Output data format" for my Sound Blaster Live! 5.1 is. I suppose it's 16 bit but I've tried 32 bit fixed point and to me it's sounds better...
And what about Dithering, should I leave it checked or check it off?
Is there some manual or tutorial for fb2k newbies??

BTW. I don't think I'm ever returning to WinAmp again.

DMX 6fire supports it (32-bit fixed point).

Nope, DMX 6fire supports a max of 24bit. You may be able to select and use 32bit with this card, but it's being converted to 24bit somewhere (API, driver) before being sent to the card.

For Live! the best is to use 16 bit dithered output.

i'm a little foggy... when WOULD you check off dither? when the file is higher than the output of your soundcard?

(i'm using a sonica which supports 24bit out, fyi)

Chiefly when using lossless, or if you prefer the artifacting of truncation to that of dither (And, according to the tests that have explored this, some do.)

hmm... i do have a large lossless collection... but i also have a large lossy collection all mixed into a playlist... which way "should" i set it? is the difference basically transparent?

If you can't hear the difference, then it doesn't matter. (Though I suspect dithering would use negligibly more processor power.) I've got dithering turned off, because I can't tell the difference.

For listening, I'd leave dither on always. In practice there will be little difference, but in theory dither is better in case of lossy compressed files, and has negligible importance in case of lossless audio. Dither off would be useful in special cases, for example if you want to be absolutely sure that what comes out of the card is what is at a lossless file, although you don't hear a difference.

What should I use for my M-Audio Revolution 7.1? Someone here told me already, but I want to see if you all suggest the same thing.

24bit? 32bit floating? 32bit fixed? 64bit?

[sighs]

24-bit linear with dithering, no resampling.

IIRC, I saw somewhere that this card supports 32-bit (at least in ASIO, where it says: Int32LSB).

i have DMX 6fire

AD and DA converters are 24 bit, internal pci bus, ram, processor transfert is 32 bit...

DMX 6fire work at 32 bit fixed point foobar output.
i am not able to listen any differences with 24 bit output, of corse i use 16 bit files from CD! and i am not see differences of use time processor.


question : why convert from 64 bit to 24 (foobar) then to convert to 32 bit? ( pci bus ) and convert to 24 bit (terratec)?

so assuming you have set the bit depth to match your card (for me: 24 bit fixed for my ap2496)...

if de-selecting dither insures the lossless audio (which is all I listen to) to be left alone, what would the advantages be (if any) of using dither in this situation?

thanks in advance for any replies.

Because the second step (conversion to 32 bit for pci bus transport) is NOT the bitdepth conversion. It's just data repackaging that allows lower cpu usage during data transfer. I think that Ogg explained everything already in this thread.

Terratec works with 32 bit audio probably just because driver truncates 32 bit samples to 24 bit.

if you're using replaygain on those lossless files while you listen then you would still be better off with dither turned on, because you've altered the song's volume, which is what the dither option is there for. You've changed the original's volume, and the original's existing dither is now meaningless as it is now gone. Gone, because you've lowered the volume and the dither will be dropped below the signal floor. So you now need to dither it again relative to the new floor. So, if you are changing the volume, either on your own or via replaygain, you should have dither on. If you're not changing the volume on your own or via replaygain, you don't need to dither as it should already be dithered originally. At least that's the simple answer. I think it gets just a little more complicated if your source's bit depth differs from your sound card's output bit depth.

<edit> I think this only strictly applies if you've got a 16bit sound card and gets a bit fuzzier if you have a 24/32bit card