[flac-dev] Next generation WebM and FLAC

Wed Jun 20 13:21:23 PDT 2012

I've just realised this missed the list. Please use 'Reply to all' to CC
the list. Thanks.

(There's also an inline reply.)

On Jun 20, 2012 8:39 PM, "Dennis Brunnenmeyer"
<dennisb<dennisb at fullfidelitymusic.com>
@ <dennisb at fullfidelitymusic.com>fullfidelitymusic.com<dennisb at fullfidelitymusic.com>>
wrote:
>
> James...
>
> Actually, I am a part-time 'real' DJ as well as an electronics engineer
with a Master's degree, having studied digital signal processing as part of
my graduate program. But that's only a part of it. I am also an old-school
music lover that finds much of what is offered today on CDs or in MP3 files
to lack good sound quality, due in part to the trend to rely on compressors
to try and squeeze in more presence, and certainly due to the media
delivery and playback systems now in common use.
>
> One of our issues as volunteer DJs and engineers at our community radio
station is the difficulty encountered with ingesting (i.e, downloading and
putting to practical use) inconsistently tagged and lossy MP3 files with
little or sometimes no metadata. When we decided to begin creating a
digital music library on a file server nearly three years ago, we
immediately decided to employ FLAC as the desired storage format. Like a
curator in an art museum who hangs original artwork rather than
photocopies, we deemed it foolish to build a huge archive-quality digital
library based on lossy MP3 files, knowing that someday soon, CDs would
disappear. With only downloads to rely one, we'd be left holding the bag,
so to speak, with a motley and inconsistent collection of files that were
lacking in quality. Furthermore, they'd be nearly impossible to locate on
the server when searching for certain metadata that's almost never included
and when it is, is often incorrect or worse yet, non-compliant with any
'standard.'  It's far easier for us to rip CDs and tag them ourselves, but
even that is ridiculously time-consuming. We receive roughly 100 CDs a week
for consideration. We had to do something to try and rectify the issue
facing us as well as any quality-conscious listener.
>
> As for metadata, we wish to see a new and extended but standardized set
of vorbis comments that include parameters such as Album Artist; Performer
(vocal role, instruments played); Style (a subset of a limited set of Genre
tags) Disc Number, Total Discs, Track Number, Total Tracks, album Copyright
information, UPC (the retail bar code) and so on. The list is not long, but
it is important to the ability to host Google-like library searches for
such information as the release date, the composers, songwriter or
arranger's name and of course to find the tracks the DJ or home enthusiast
wants to play. Not only that, but radio stations are required to report
their airplay to organizations like SoundExchange, BMI and ASCAP and to
their web sites. Having the proper metadata in the files required to do
this makes generating reports simple...as long as the vorbis comments are
complete and follow the extended standard we wish to see implemented.
>
> MP3 files are lossy and for the most part unnecessary relics of the past,
when storage was expensive and drives were tiny by today's standards, and
when network connections were iffy dial-up links instead of today's
rock-solid broadband connectivity experienced by a large percentage of the
population. We're not against their distribution, but we would like to see
artists and labels offer well-tagged FLAC files as an additional,
higher-valued option. Oh, and we think it foolish if they don't included a
PDF 'booklet' of their liner notes along with the files. Some already do
that, but that too should be standardized for consistent appearance and use.
>
> In a nutshell, we'd like to see the industry move towards a well-tagged
FLAC standard for music distribution. Why not? It's certainly better than
what we have going for us today. Users should be able to download lossless,
properly tagged files directly into their music libraries....and it
provides a means to avoid the smearing and distortions associated with low
sample rate MP3 files and CDs, especially when formatted with higher sample
rates and bit resolutions, and played though proper equipment.
>
> As you can tell, I disagree with many of the points the author tried to
make in the referenced article, and for very good reasons. I also contend
that lossless support is not too much to ask for. Why would that be the
case?

I have no idea, see for yourself:

https://<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>
groups.google.com<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>
/a/<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>
webmproject.org<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>
/group/<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>
webm<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>
-discuss/<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>
browse_thread<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>
/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b<https://groups.google.com/a/webmproject.org/group/webm-discuss/browse_thread/thread/42fd53c71b2bcc74/5d2e322de440280b?#5d2e322de440280b>

James.

>
> Peace...
>
> Dennis...
> **********************************************
> Dennis Brunnenmeyer
> FULL FIDELITY MUSIC
> 15019 Rattlesnake Road
> Grass Valley, CA 95945-8710
> Office:   1 (530) 477-9015
> Mobile: 1 (530) 320-9025
> eMail:  dennisb /at/ fullfidelitymusic /dot/ com
>
>
> On 6/20/2012 11:24 AM, James Haigh wrote:
>>
>> On Jun 20, 2012 1:47 AM, "Dennis Brunnenmeyer" <dennisb<dennisb at chronometrics.com>
@ <dennisb at chronometrics.com>chronometrics.com <dennisb at chronometrics.com>>
wrote:
>> >
>> > James...
>> >
>> > The author is passionate but very wrong on a crucial and important
point. The Nyquist sampling theorem holds only for signals that are
band-limited. Everything else is an approximation. A truly band-limited
signal has to have existed (and been sampled) for all time in order to be
reproduced exactly. While that is a fact often overlooked by those not
trained in the field, there are practical situations where a signal's
components of short duration can be sampled satisfactorily—in fact, quite
well. That is not true for transient material such as the clash of cymbals
or a stream of short notes on very fast fiddle tune. Transients are far
from having existed for all time and impossible to reproduce exactly by any
sampling method, regardless of sample rate. Music is not a collection of
ideal sine waves.
>> >
>> > Secondly, I was not referring to music delivered to end-users such as
the kid with a iPod and ear buds listening to rap. High end recording,
mastering and editing systems with multi-track mixes benefit from higher
resolution and higher sample rate systems. The final product need not be
32-bit, of course. I agree with you on that point.
>>
>> Sorry, for being flippant before, but without mentioning any of this in
your previous email, you did sound like the kid with the 'beats' headphones
that thinks he's a DJ! :-p
>>
>> >
>> > There's a saying that every passionate audio engineer thinks they're
an expert, but none of them agree with each other.
>>
>> No no, you make good points, you probably know better than me, I'm no
expert. It's just that I was still thinking in 'web mode' where even
lossless support is too much to ask for.
>>
>> Do you know if ALAC supports 32bit? I've just found out that it's now
patent-free, and released under the Apache 2 licence.
>>
>> James.
>>
>> >
>> > Dennis...
>> >
>> >
>> >
>> >
>> >
>> >
>> > On 6/19/2012 3:54 PM, James Haigh wrote:
>> >>
>> >> On Jun 19, 2012 8:46 PM, "Dennis Brunnenmeyer" <dennisb<dennisb at chronometrics.com>
@ <dennisb at chronometrics.com>chronometrics.com <dennisb at chronometrics.com>>
wrote:
>> >> >
>> >> > Future higher fidelity music file capability at high sample rates.
Yes, the files get bigger.
>> >>
>> >> I think there's more important things than music for your dog!
>> >>
>> >> 16bit is just below the limit of the human hear, 24bit hugely exceeds
it. 32bit may perhaps be used for scientific research on bats and dolphins!
>> >>
>> >> Furthermore, I'm not aware of any 'future' music services for these
animals, and is certainly not a priority now. :-p
>> >>
>> >> Please read this:
>> >> http:// <http://people.xiph.org/~xiphmont/demo/neil-young.html>
people.xiph.org
<http://people.xiph.org/~xiphmont/demo/neil-young.html>/~<http://people.xiph.org/~xiphmont/demo/neil-young.html>
xiphmont <http://people.xiph.org/~xiphmont/demo/neil-young.html>/demo/<http://people.xiph.org/~xiphmont/demo/neil-young.html>
neil <http://people.xiph.org/~xiphmont/demo/neil-young.html>-<http://people.xiph.org/~xiphmont/demo/neil-young.html>
young.html <http://people.xiph.org/~xiphmont/demo/neil-young.html>
>> >>
>> >> James.
>>
>> --
>>
>> Dennis Brunnenmeyer
>> FULL FIDELITY MUSIC
>> 15019 Rattlesnake Road
>> Grass Valley, CA 95945-8710
>> Office:   1 (530) 477-9015
>> Mobile: 1 (530) 320-9025
>> eMail:  dennisb /at/ fullfidelitymusic /dot/ com
>> http:// <http://www.fullfidelitymusic.com/index.html>
www.fullfidelitymusic.com
<http://www.fullfidelitymusic.com/index.html>/<http://www.fullfidelitymusic.com/index.html>
index.html <http://www.fullfidelitymusic.com/index.html>
>>
>>
>> --
>>
>> Dennis Brunnenmeyer
>> FULL FIDELITY MUSIC
>> 15019 Rattlesnake Road
>> Grass Valley, CA 95945-8710
>> Office:   1 (530) 477-9015
>> Mobile: 1 (530) 320-9025
>> eMail:  dennisb /at/ fullfidelitymusic /dot/ com
>> http:// <http://www.fullfidelitymusic.com/index.html>
www.fullfidelitymusic.com
<http://www.fullfidelitymusic.com/index.html>/<http://www.fullfidelitymusic.com/index.html>
index.html <http://www.fullfidelitymusic.com/index.html>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.xiph.org/pipermail/flac-dev/attachments/20120620/d1f6d838/attachment.htm 