[Flac] uncompressed FLAC

[Gregory Maxwell]

On Sun, Mar 11, 2012 at 12:27 PM, Mike. <the.lists at mgm51.com> wrote:
> Some people feel that the
> extra CPU power required to de-compress a FLAC file causes the timing
> issues in the audio samples going to the DAC.

This isn't something you have to "feel", it's something you could
_measure_— far more precisely than anyone could hope to hear.

To whatever extent that were true, it would also be true of 1000 other
possible CPU load sources, including operating systems busywaiting on
the disk, differences in display device drivers, etc.    In fact,
considering how massively slow disk IO is compared to FLAC
decompression it seems just as plausible that the increased disk IO
(with the resulting PCI traffic and IO blocking) caused by
uncompressed playback makes audible glitches more likely.  (Though the
solution is still 'fix the darn software')

>They don't hear the FLAC file, but rather the inability of the computer to properly process the FLAC file in the time domain

This sounds like a view driven a pretty substantial misunderstanding
about how PC sound hardware is interfaced.   The software doesn't just
write out one sample at a time, thus somehow leaving the audio at the
mercy of the software timing.  Instead the software writes large
blocks of hundreds of samples at a time and the OS/sound driver
maintain a pipeline of blocks to be fed to the audio hardware in order
to prevent glitches. If the software is late enough to impact the
audio, you won't get mistimed samples— you'll get a great big click or
thump from the dropout.

The notion that the CPU load is impacting intersample timings is a
rather tall claim which requires equally striking evidence. The people
who claim to hear this claim to have spent dozens of hours listening
in order to find the difference— though of course never with rigorous
blind testing methodology— Borrowing a scope and getting an eye
diagram of the SPDIF output w/ and w/o FLAC would have been easier and
a hell of a lot more sure and convincing.   (I'd offer to do it
myself, but I'd just be discounted has having magical unicorn hardware
that happens to not exhibit the issue on sundays)

If the FLAC output really did sound different the a theory more
plausible to me is the idea that the player software was buggy and
managed to send the FLAC sourced data through a different signal path—
one that got some cruddy resampling, for example— but this theory is
mooted by people being satisfied by uncompressed flac.

Though of course the people promoting this misinformation don't want
to discover actual bugs like that— far better to promote non-existent
snake oil solutions to non-existent problems because it leaves the
reader beholden to receiving the latest 'wisdom'.