File-Based Audio Done Right at John's

[Clive]

I've never understood why upsampling is so often done to a non-multiple sample rates. Probably my logic is too simple but 88.2 from 44.1 seems so easy to implement and any additional benefit of going to 96 must be so small. Surely it's not worth the possible detrimental side effects of poor conversion.

[MartinT]

Mathematically, the formulae don't care. It's as easy in maths to calculate interpolations as it is exact multiples. In fact, software like SOX and streamers probably use the same algorithms regardless.

Also, an upsample from 44.1 to 88.2 doesn't involve simple doubling. It's much more complicated than that.

[Clive]

Mathematically, the formulae don't care. It's as easy in maths to calculate interpolations as it is exact multiples. In fact, software like SOX and streamers probably use the same algorithms regardless.

Also, an upsample from 44.1 to 88.2 doesn't involve simple doubling. It's much more complicated than that.

I understand that, I'm thinking more that doubling the sample rate suggests to me there will only be one correct algorithm whereas more complex algorithms for non-multiples will take us down the path of which one sounds best. Rather like downscaling a digital photo isn't so simple and different algorithms produce differing qualities of results, also a lot depends on the viewing medium. The same could well be true for audio and dacs.

[NRG]

It would be interesting to see the spectrum analysis of 16/44.1 up-sampled to 16/88.2 in something like Audacity vs the original file...think I'll have a go when I get home tonight...over on AudioTalk, Stevetheshadow is using the Mac Midi setting to up-sample to 88.2 and is reporting a nice improvement so there maybe some milage in this.

[NRG]

Interesting reading:

http://www.noiseaddicts.com/2011/02/...-88-2-vs-44-1/

[Gazjam]

Get what your saying Martin re: Wav format, and whilst its true both Wav and Aiff are "uncompressed" PCM formats (as opposed to lossless), I think they store data on the disc differently?
One has the most significant byte first and the other one last...does this make a difference anywhere except on paper though?
Don't know don't really want to think about it too much, (less FUD the better), but I listened to the same tracks in both formats and there was worthwhile improvement to my ears (and a couple of mates) using AIFF.
iWas so convinced of it I converted all my Flacs to Aiff.

Looking at it logically, I think it may come down again to the "less is more" way of thinking?
Perhaps the way AIFF files are stored on disk requires less processing, hard drive crunching or memory overhead before playback, albeit probably very slightly less.
Thinking along the same lines of why Wav is potentially better an Flac, despite the fact bit for bit a Flac file is exactly the same as a Wav...But only AFTER its been converted to PCM.
Not offering this as a reason, just as a theory, may be barking up the wrong tree totally.

Thing is though, for me it all comes down to what sounds better, even if its just a slight improvement it's worth having.
I've found everything, no matter how small, can make a difference.

One I'd like to try out is Broadcast Wav, apparently used by many recording studios as the "best" recording format, though e added features like timecode data etc might not translate to better audio quality.
Also there's the question of compatibility...but could be fun to give it a bash anyway.

[MartinT]

I'm with you there, Gaz. I have definitely found that adding noise shaped dither makes all the difference (and, of course, makes the individual interpolations mathematically uncertain).

[NRG]

Gaz, there should be no difference in the playback of AIFF vs WAV. Both are a wrapper / container for PCM data, yes AIFF is stored little endian and WAV big endian...this makes no difference! Yes the headers are different but we are not playing the headers. There must be some other reason for your perception of better playback with AIFF.

For reference I've not heard any difference with FLAC vs WAV either when first decoded on my squeezecentre server and transmitted to the squeezebox, decoding on the touch though is a slightly different matter but in reality its all but inaudible.

Oh, forgot theres two versions of AIFF one little endian and one big endian...

[nat8808]

Thanks Gaz and Neal. Most interesting. I'm definitely going to play with 88.2 upsamples. 176.4 is not an option as the Bushmaster doesn't decode it.

ooo... Whilst I know I'll probably not get around to buying a Bushmaster until his superdooper one is made in a couple of years, but that's a shame. Wonder why, receiver chip?

[Gazjam]

Yeah..just pick your favourite flavour of distortion.