Subj : Re: 3d printing
To   : Nightfox
From : Vk3jed
Date : Fri Sep 20 2019 09:54:00

-=> On 09-19-19 10:16, Nightfox wrote to Jamestyree <=-

 Ni> FLAC is lossless in that it doesn't remove any data during compression,
 Ni> as opposed to a format like MP3..  And it's lossless compared to the
 Ni> original digital audio (i.e., WAV file or CD audio track) - All of the
 Ni> digital data is preserved in FLAC format.  With MP3, some of the audio
 Ni> data (which many people are not likely to hear) is lost, which results
 Ni> in a smaller file size, but it doesn't 100% match the original data.

Agreed on all counts there.  FLAC should sound as good as the source PCM audio.

 Ni> I have a hard time believing that a lossless format such as FLAC can't
 Ni> sound as good as vinyl.  The Nyquist sampling theorem says that if the
 Ni> sample rate is at least double the highest frequency in the audio, the
 Ni> recording can faithfully reproduce the original sound.  Some people

That's correct.

 Ni> have argued that the standard CD sample rate isn't enough for some
 Ni> recordings - but these days there are higher definition digital formats

There is an argument that the ear and brain are subconsciously sensitive to
ultrasonic frequencies.  We may not directly perceive them, but it appears
their presence or abscence can affect the perceived audio quality.  Audio that
has the ultrasonic range heavily filtered (e.g. CD) has been said to sound
"harsh" by some people.  I recall reading about a CD player (in the 90s I
think) that deliberately allowed some of the aliased (> 24 kHz) energy to bleed
through, which some people thought sounded better in listening tests. 

Another possible culprit is severe phase shifting in the upper audio
frequencies caused by sharp analog filters, necessary at 44.1 kHz sampling
rates.  With sample rates as high as 192k, it's possible to do the first stage
of antialiasing filtering digitally, where sharp cutoffs can be achieved with
no phast shifting/distortion.  Then, the analog filter can be much simpler,
because it no longer has to filter out audio at 24 kHz, only from around 150
kHz (the DSP has removed everything else in between), which means no severe
phase shifting in the audible frequency range.

 Ni> (i.e., 24-bit 192khz recordings).  Also, I'd think it might depend on

In theory, higher sampling rates with good source material should be as good a
sound as you can get, if the rest of the chain is hi fi.

 Ni> the device/hardware you're playing the music on.  Record players and
 Ni> devices for playing FLAC/MP3s tend to have different hardware,
 Ni> different speakers, etc., which may have an effect on the sound.  Many
 Ni> PCs and laptops use fairly inexpensive audio codecs/hardware, and might
 Ni> not have the best speakers, which can definitely reduce the sound
 Ni> quality.

That's true, though the comments about the "harshness" of CDs comes from the hi
fi community from people who had both a CD player and a turntable on the same
system.  I'd like to see what they'd think of a good 96 or 192k recording on
quaity hardware.


.... Anything, when cooked in large enough batches, will be vile.
--- MultiMail/Win v0.51
 þ Synchronet þ Freeway BBS, Bendigo Australia.  freeway.apana.org.au

.