1 Mar
2014
1 Mar
'14
7:43 p.m.
I've been using triangle dithering when I export at 16-bit in Ardour.
Not to start a religious war or anything, but I'd actually be interested
in hearing opinions about why that or some other alternative might be
better (shaped noise, etc.). How do you usually dither? Does it make a
difference?
--
+ Brent A. Busby + "We've all heard that a million monkeys
+ Sr. UNIX Systems Admin + banging on a million typewriters will
+ University of Chicago + eventually reproduce the entire works of
+ James Franck Institute + Shakespeare. Now, thanks to the Internet,
+ Materials Research Ctr + we know this is not true." -Robert Wilensky
1 Mar
1 Mar
7:51 p.m.
On Sat, Mar 1, 2014 at 6:43 PM, Brent Busby <brent(a)keycorner.org> wrote:
How do you usually dither? Does it make a difference?
I'll just throw this up, its a good overview of dither types, and *why* it
makes a cleaner mix (for humans):
http://wiki.xiph.org/Videos/Digital_Show_and_Tell#Dither
(that's the accompanying document for this video:
http://www.xiph.org/video/vid2.shtml )
I generally export to 32bit float .flac... so no dithering (or burning to
CD's :)
HTH, -Harry
2 Mar
2 Mar
3:43 a.m.
On 03/01/2014 08:51 AM, Harry van Haaren wrote:
On Sat, Mar 1, 2014 at 6:43 PM, Brent Busby
<brent(a)keycorner.org
<mailto:brent@keycorner.org>> wrote:
Hmm, I thought FLAC only did 24-bit???
--
David W. Jones
gnome(a)hawaii.rr.com
authenticity, honesty, community
http://dancingtreefrog.com
How do you usually dither? Does it make a
difference?
I'll just throw this up, its a good overview of dither types, and *why*
it makes a cleaner mix (for humans):
http://wiki.xiph.org/Videos/Digital_Show_and_Tell#Dither
(that's the accompanying document for this video:
http://www.xiph.org/video/vid2.shtml )
I generally export to 32bit float .flac... so no dithering (or burning
to CD's :)
4:38 a.m.
On Sun, Mar 2, 2014 at 2:43 AM, david <gnome(a)hawaii.rr.com> wrote:
On 03/01/2014 08:51 AM, Harry van Haaren wrote:
I think the FLAC spec
says it will handle anything from 4-32 bit-depth:
https://xiph.org/flac/faq.html#general__samples
That said, Audacity only has FLAC export options of 16 & 24 bit depths.
Ardour supports 8, 16 and 24. Still no 32 bit float support (at application
level).
I should correct my previous statement though: I *thought* I exported 32bit
float: but it turns out they're 24bits (from Ardour3), dithering set to
None. And cropping the resulting output in Audacity and exporting was to
16-bit PCM, so I was actually doing this all wrong (no dithering, 32 -> 24
-> 16).
A better workflow would be to:
A) Ardour export 32 bit float -> 16bit (with dither) -> Audacity 16bit in,
crop, 16bit out
B) Ardour export 32 bit -> 24 bit (no dither) -> Audacity 24bit in, crop,
export 16bit (with dither).
The important part being to not dither twice, since then you'll be adding
noise to the signal twice!
I'll be using option A above from now on I think, since it involves less
bit-depth changes.
Living and learning :) -Harry
I generally export to 32bit float .flac... so no
dithering (or burning
to CD's :)
Hmm, I thought FLAC only did 24-bit??? 
4:47 a.m.
On Sun, 2014-03-02 at 03:38 +0000, Harry van Haaren wrote:
A better workflow would be to:
A) Ardour export 32 bit float -> 16bit (with dither) -> Audacity 16bit
in, crop, 16bit out
B) Ardour export 32 bit -> 24 bit (no dither) -> Audacity 24bit in,
crop, export 16bit (with dither).
The important part being to not dither twice, since then you'll be
adding noise to the signal twice!
I'll be using option A above from now on I think, since it involves
less bit-depth changes.
Living and learning :)
When you need to add noise, consider to use proprietary noise :S.
-------- Oops,I used the wrong account: Forwarded Message --------
From: Ralf Mardorf <ralf.mardorf at alice-dsl.net>
To: linux-audio-user at lists.linuxaudio.org
Subject: Re: [LAU] Dithering...should we dither about it?
Date: Sun, 02 Mar 2014 00:50:58 +0100
Mailer: Evolution 3.10.4
On Sat, 2014-03-01 at 18:51 +0000, Harry van Haaren wrote:
I generally export to 32bit float .flac... so no
dithering (or burning
to CD's :)
That's good. I keep my recordings as 48 KHz 32bit float PCMs, but I
would be willing to use FLAC too, assumed somebody would be interested
in my recordings and I would be willing to share them. However, what is
dithering ;)? What are CDs?
I remember that a long, long time ago, when [... long story ...] there
were special "noise algorithms" available. It wasn't just steady noise
by what ever waveform. IIRC it was noise, that only was added to low
audio signal levels. Non-free-open-source-noise ;).
5:19 a.m.
On Sat, Mar 1, 2014 at 10:38 PM, Harry van Haaren <harryhaaren(a)gmail.com>wrote;wrote:
I think the FLAC spec says it will handle anything from 4-32 bit-depth:
https://xiph.org/flac/faq.html#general__samples
flac cannot handle 32 bit *floating point*. the link you provided reads:
"FLAC supports linear PCM samples with a resolution between 4 and 32
bits per sample. FLAC does not support floating point samples"
That said, Audacity only has FLAC export options of 16 & 24 bit depths.
Ardour supports 8, 16 and 24. Still no 32 bit float
support (at application
level).
Ardour supports 32 bit float export to file formats that allow it. FLAC
does not.
Note that technically WAV does not allow 32 bit float. You really need
"WAVEX" for that. Sensible audio file I/O libraries can handle it, however,
even though technically it violates the WAV specification.
9:56 a.m.
On 03/02/2014 04:38 AM, Harry van Haaren wrote:
On Sun, Mar 2, 2014 at 2:43 AM, david
<gnome(a)hawaii.rr.com
<mailto:gnome@hawaii.rr.com>> wrote:
24bit integer is equivalent to 32bit float in terms of resolution, and
pretty much identical as long as the float samples are clamped between
-1.0f and 1.0f.
On 03/01/2014 08:51 AM, Harry van Haaren wrote:
I think the FLAC spec
says it will handle anything from 4-32 bit-depth:
https://xiph.org/flac/faq.html#general__samples
That said, Audacity only has FLAC export options of 16 & 24 bit depths.
Ardour supports 8, 16 and 24. Still no 32 bit float support (at
application level). I generally export to 32bit float .flac... so no
dithering (or burning
to CD's :)
Hmm, I thought FLAC only did 24-bit??? I should correct my previous statement though: I
*thought* I exported
32bit float: but it turns out they're 24bits (from Ardour3), dithering
set to None. And cropping the resulting output in Audacity and exporting
was to 16-bit PCM, so I was actually doing this all wrong (no dithering,
32 -> 24 -> 16).
A better workflow would be to:
A) Ardour export 32 bit float -> 16bit (with dither) -> Audacity 16bit
in, crop, 16bit out
B) Ardour export 32 bit -> 24 bit (no dither) -> Audacity 24bit in,
crop, export 16bit (with dither).
The important part being to not dither twice, since then you'll be
adding noise to the signal twice!
actually, you *must* dither at every truncation step - if you don't, you
will lose information _and_ introduce signal-dependent requantisation
noise, which can never be removed again.
so the best approach is to only reduce the wordlength once, at the very
end of the chain, before going to CD.
some mastering people (bob katz among them) even go as far as demanding
dither at every level control in the chain.
however, if you are exporting from JACK's native 32bit float to 24bit
int _and_ you make sure that there is no sample larger than full-scale
(floats are funny :), then there is no actual loss of information, and
no dithering is required. all you do is map the 23 mantissa bits and the
sign bit of the float to the 24 bits of the integer.
--
Jörn Nettingsmeier
Lortzingstr. 11, 45128 Essen, Tel. +49 177 7937487
Meister für Veranstaltungstechnik (Bühne/Studio)
Tonmeister VDT
http://stackingdwarves.net
2:59 p.m.
On Sun, Mar 02, 2014 at 09:56:36AM +0100, Jörn Nettingsmeier wrote:
some mastering people (bob katz among them) even go as
far as
demanding dither at every level control in the chain.
In theory you'd need to dither every time a signal
is rounded to some representable value. That would
include intermediate results in filters etc.
In practice, as long as you stay in 24 bits or float,
most signals have enough noise on them (from analog
recording) or are complex enough to be considered
self-dithered (no correlation between the rounding
error and the signal).
I did some test to see the result of dithering years
ago. They are still on my website but there's no link
to them. Try
<http://kokkinizita.linuxaudio.org/linuxaudio/dithering.html>
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
3 Mar
3 Mar
11:04 a.m.
On Sun, Mar 2, 2014 at 1:59 PM, Fons Adriaensen <fons(a)linuxaudio.org> wrote:
<http://kokkinizita.linuxaudio.org/linuxaudio/dithering.html>
Interesting, thanks for sharing.
@Paul, apologies, I made an incorrect assumption there.
@Thorsten, indeed the trim functions (and add functions) could be used...
I'm a bit visual in my "silence adding" workflow.. and generally add some
extra compression / mastering in Audacity after export from Ardour: hence
the extra import-export.
Interesting discussion! Cheers, -Harry
2 Mar
2 Mar
10:54 a.m.
On 03/02/2014 04:38 AM, Harry van Haaren wrote:
A better workflow would be to:
A) Ardour export 32 bit float -> 16bit (with dither) -> Audacity 16bit
in, crop, 16bit out
B) Ardour export 32 bit -> 24 bit (no dither) -> Audacity 24bit in,
crop, export 16bit (with dither).
An even better workflow would use the "Trim silence at start" and "Trim
silence at end" options of format profiles, reachable via the Export
dialog ;)
The whole thing is designed for not having to edit the files at all,
after export.
--
Thorsten Wilms
thorwil's design for free software:
http://thorwil.wordpress.com/
1:34 p.m.
On Sun, 2 Mar 2014 03:38:07 +0000
Harry van Haaren <harryhaaren(a)gmail.com> wrote:
On Sun, Mar 2, 2014 at 2:43 AM, david
<gnome(a)hawaii.rr.com> wrote:
One of FLACs competitors does 32bit float since many years: wavpack.
I'm not quite sure why it is much less popular than FLAC.
Regards,
Philipp
--
JID: murks(a)jit.si
On 03/01/2014 08:51 AM, Harry van Haaren wrote:
I think the
FLAC spec says it will handle anything from 4-32
bit-depth: https://xiph.org/flac/faq.html#general__samples
That said, Audacity only has FLAC export options of 16 & 24 bit
depths. Ardour supports 8, 16 and 24. Still no 32 bit float support
(at application level).
I should correct my previous statement though: I *thought* I exported
32bit float: but it turns out they're 24bits (from Ardour3),
dithering set to None. And cropping the resulting output in Audacity
and exporting was to 16-bit PCM, so I was actually doing this all
wrong (no dithering, 32 -> 24 -> 16).
A better workflow would be to:
A) Ardour export 32 bit float -> 16bit (with dither) -> Audacity
16bit in, crop, 16bit out
B) Ardour export 32 bit -> 24 bit (no dither) -> Audacity 24bit in,
crop, export 16bit (with dither).
The important part being to not dither twice, since then you'll be
adding noise to the signal twice!
I'll be using option A above from now on I think, since it involves
less bit-depth changes.
Living and learning :) -Harry I generally export to 32bit float .flac... so no
dithering (or
burning to CD's :)
Hmm, I thought FLAC only did 24-bit???
3920
days inactive
3922
days old
linux-audio-user@lists.linuxaudio.org
10 comments
9 participants
participants (9)
-
Brent Busby -
david -
Fons Adriaensen -
Harry van Haaren -
Jörn Nettingsmeier -
Paul Davis -
Philipp Überbacher -
Ralf Mardorf -
Thorsten Wilms