Sign in with
Sign up | Sign in
Your question

MP3 encoders.... what's good for low baud rate encoding ?

Last response: in Home Audio
Share
Anonymous
April 21, 2005 11:41:38 PM

Archived from groups: rec.audio.pro (More info?)

Hey all,

I have, at the advisement of several, put Win Lame on a couple of PCs
to use for converting .wav files to higher rate MP3s and have been happy.

I've been asked by my church to take over the web posting of the audio
taken from the morning services. Win Lame is just great for the 128K
stream, but they need a 24 or 32K stream for the dial-up people and
after much manual experimentation with Lame, the results are seriously
sub-par.

One friend from another church has been using one of the Nero products
to do a 20K encode that doesn't sound bad at all, but he has a huge suite
of tools that were loaded along with the MP3 conversion stuff. I don't want
to stink up either my personal PC or any of my workstations with unnecessary
gobbledygook that might cause conflicts.

What's a good, simple, non-invasive encoder for doing very low baud rates?


Thanks,

--
David Morgan (MAMS)
http://www.m-a-m-s DOT com
Morgan Audio Media Service
Dallas, Texas (214) 662-9901
_______________________________________
http://www.artisan-recordingstudio.com
Anonymous
April 22, 2005 2:50:08 AM

Archived from groups: rec.audio.pro (More info?)

David Morgan (MAMS) wrote:
> What's a good, simple, non-invasive encoder for doing very low baud rates?

What about Ogg Vorbis [1] and Speex [2]? Both are open source.

[1] <http://vorbis.com&gt;
[2] <http://speex.org&gt;

Johann
--
..|()|-|/-\|\||\| |3|_||2|</-\|2|)
April 22, 2005 2:50:09 AM

Archived from groups: rec.audio.pro (More info?)

In article <426811f9$0$7515$9b4e6d93@newsread2.arcor-online.net>,
Johann Burkard <johannburkard@nexgo.de> wrote:

> David Morgan (MAMS) wrote:
> > What's a good, simple, non-invasive encoder for doing very low baud rates?
>
> What about Ogg Vorbis [1] and Speex [2]? Both are open source.
>
> [1] <http://vorbis.com&gt;
> [2] <http://speex.org&gt;
>
> Johann
> --


Couple years ago I tried everything available on the Mac and found to
my surprise iTunes had the best sound at tiny sampling rates.



David Correia
Celebration Sound
Warren, Rhode Island

CelebrationSound@aol.com
www.CelebrationSound.com
Related resources
Can't find your answer ? Ask !
Anonymous
April 22, 2005 2:50:10 AM

Archived from groups: rec.audio.pro (More info?)

"david" <ihate@spamo.com> wrote in message news:210420051813340177%ihate@spamo.com...
> In article <426811f9$0$7515$9b4e6d93@newsread2.arcor-online.net>,
> Johann Burkard <johannburkard@nexgo.de> wrote:
>
> > David Morgan (MAMS) wrote:
> > >
> > > What's a good, simple, non-invasive encoder for doing very low baud rates?

> > What about Ogg Vorbis [1] and Speex [2]? Both are open source.
> >
> > [1] <http://vorbis.com&gt;
> > [2] <http://speex.org&gt;
> >
> > Johann

> Couple years ago I tried everything available on the Mac and found to
> my surprise iTunes had the best sound at tiny sampling rates.
>
> David Correia
> Celebration Sound
> Warren, Rhode Island
>
> CelebrationSound@aol.com
> www.CelebrationSound.com


I'm in the PC world here.... and Win Lame will output Ogg Vorbis, but
no joy in quality at 32K and below.

DM
Anonymous
April 22, 2005 4:10:53 AM

Archived from groups: rec.audio.pro (More info?)

"David Morgan \(MAMS\)" <mams@NOSPAm-a-m-s.com> wrote in
news:SlT9e.27102$Zn3.19414@trnddc02:

> What's a good, simple, non-invasive encoder for doing very low baud
> rates?

Have you prepped your signal for low bit rates?

1) mono
2) band limited to 100 Hz to 5 kHz
3) massively compressed
Anonymous
April 22, 2005 5:06:14 AM

Archived from groups: rec.audio.pro (More info?)

"Carey Carlan" <gulfjoe@hotmail.com> wrote in message news:Xns963FCD513E4F1gulfjoehotmailcom@207.69.189.191...

> "David Morgan \(MAMS\)" <mams@NOSPAm-a-m-s.com> wrote in
> news:SlT9e.27102$Zn3.19414@trnddc02:
>
> > What's a good, simple, non-invasive encoder for doing very low baud
> > rates?

> Have you prepped your signal for low bit rates?
>
> 1) mono

All it would take is a virtual switch to go mono... but I was trying to avoid that.
I'm only a couple of years into this whole MP3 thing, so if this is something
that's really going to clean up my problem, I can try that.

> 2) band limited to 100 Hz to 5 kHz

Gaack !! I'm spending waaay too much 'free' time on this already, but
the answer is, probably not. I could lo-pass at 5K during the conversion,
but wouldn't that totally destroy the clarity?

> 3) massively compressed

Yes, I'm fairly heavily compressed after using a Compellor and a Dominator
in front of the computer, and RNCs on the vocals and sub-groups. (Something
I preach heavily against... but this is also what feeds the live broadcast, which
is what the CD and eventual net archive is derived from).

Then, I use SoundForge for a little more post work to further elevate the spoken
word and comp/limit there as well. When I burn the Audio CDs for the church,
I use CD Architect and a *little* bit of WaveHammer added there as well.

Massive enough? :-\ (I don't recommend this personally).

After this, I use the Architect file, saved as a .wav file, to bust the service into
it's segments for naming and encoding.

I'll grab some files and run them through Win Lame with heavy lo-pass and
convert to mono and see what happens.


Here's what I'm coming up with currently..... (only the past two weeks have
been my post prod MP3 work, though I've been FOH / broadcast for a few
years now). The first song is all you need to bother with...

http://www.cathedraloflight.org/MP3/05-04-10/Full96k.m3... 128K WinLAME


Here's what my friend's NERO MP3 PRO can do with my 128K files....

http://www.cathedraloflight.org/MP3/05-04-10/Full20k.m3... 24K NERO MP3 PRO


Here's the best I can get my WinLAME to do at 40K, stereo, lpf at 16,500,
no ATH, using .wav files... and the church want's it even smaller for dial-up....

http://www.m-a-m-s.com/01_Rejoice_ON_A_WONDERFUL_DAY.mp... 40K WinLAME


So... how does Nero take my 128K MP3s and make them sound so much better
at *HALF* of the baud rate I'm using to encode data rich .wavs ???


Thanks again,

--
David Morgan (MAMS)
http://www.m-a-m-s DOT com
Morgan Audio Media Service
Dallas, Texas (214) 662-9901
_______________________________________
http://www.artisan-recordingstudio.com
Anonymous
April 22, 2005 5:19:54 AM

Archived from groups: rec.audio.pro (More info?)

"David Morgan \(MAMS\)" <mams@NOSPAm-a-m-s.com> wrote in
news:a6Y9e.27151$Zn3.21838@trnddc02:

>> 1) mono
>
> All it would take is a virtual switch to go mono... but I was trying
> to avoid that. I'm only a couple of years into this whole MP3 thing,
> so if this is something that's really going to clean up my problem, I
> can try that.

Mono will remove the stereo portion and make the file smaller.

>> 2) band limited to 100 Hz to 5 kHz
>
> Gaack !! I'm spending waaay too much 'free' time on this already,
> but the answer is, probably not. I could lo-pass at 5K during the
> conversion, but wouldn't that totally destroy the clarity?

Any sound editor should be able to do this with minimal effort. If you
want more clarity, move to 8 kHz, but 5 kHz is more than you get from a
telephone.

>> 3) massively compressed

<snip>

> Massive enough? :-\ (I don't recommend this personally).

Yup.

> So... how does Nero take my 128K MP3s and make them sound so much
> better at *HALF* of the baud rate I'm using to encode data rich .wavs

Because the Nero file is sampled at 16 kHz and the LAME file at 44 kHz.
Anonymous
April 22, 2005 5:29:27 AM

Archived from groups: rec.audio.pro (More info?)

David Morgan (MAMS) wrote:
> "david" <ihate@spamo.com> wrote in message news:210420051813340177%ihate@spamo.com...
>>Couple years ago I tried everything available on the Mac and found to
>>my surprise iTunes had the best sound at tiny sampling rates.

> I'm in the PC world here.... and Win Lame will output Ogg Vorbis, but
> no joy in quality at 32K and below.

I don't know what you expect, but RealAudio always sounded acceptable to
me (but never good), even at low bitrates.

Johann
--
Durch den Einsatz von Cookies wird die Sicherheit verbessert und Ihre
Privatsphäre besser geschützt.
(<http://www.tubetown.de/ttstore/cookie_usage.php&gt;)
Anonymous
April 22, 2005 5:29:28 AM

Archived from groups: rec.audio.pro (More info?)

"Johann Burkard" <johannburkard@nexgo.de> wrote in message news:42683750$0$10497$9b4e6d93@newsread4.arcor-online.net...

> David Morgan (MAMS) wrote:
> > "david" <ihate@spamo.com> wrote in message news:210420051813340177%ihate@spamo.com...

> >>Couple years ago I tried everything available on the Mac and found to
> >>my surprise iTunes had the best sound at tiny sampling rates.

> > I'm in the PC world here.... and Win Lame will output Ogg Vorbis, but
> > no joy in quality at 32K and below.

> I don't know what you expect, but RealAudio always sounded acceptable to
> me (but never good), even at low bitrates.
>
> Johann


I am trying to eliminate the level of aliasing(?), or severe phasing
and 'beating' of hi-mid and high frequencies at low baud rates.

My friend with the Nero package gets a fairly smooth sounding
stereo MP3 at 20kbps !! There's just the slightest hint of light
chorusing effects. If I could get even close to that without installing
a complete package of redundant audio tools along with it, I'd be
thrilled.

The church decided several years ago, due to the 'spyware' issue and
the serious involvement (read, 'guilt') of all Real Networks software in
that regard, to eliminate anything "Real" from all of the computers. We
used to use Real for everything, but complaints from end users about
either the spyware issue or having to install the software at all, put a
stop to our using it. We just need a plain old MP3 that *any* player
can interpolate.

I certainly removed all traces of "Real" from my computers several
years ago, when Steve Gibson tried to take them to federal court over
the spyware issue. Sure... I miss a few web streams because of it,
but it's usually nothing that can't be found elsewhere.

Thanks again,

--
David Morgan (MAMS)
http://www.m-a-m-s DOT com
Morgan Audio Media Service
Dallas, Texas (214) 662-9901
_______________________________________
http://www.artisan-recordingstudio.com
Anonymous
April 22, 2005 5:34:42 AM

Archived from groups: rec.audio.pro (More info?)

"Carey Carlan" <gulfjoe@hotmail.com> wrote in message news:Xns963FD9044D97Egulfjoehotmailcom@207.69.189.191...

> "David Morgan \(MAMS\)" <mams@NOSPAm-a-m-s.com> wrote in
> news:a6Y9e.27151$Zn3.21838@trnddc02:



> > So... how does Nero take my 128K MP3s and make them sound so much
> > better at *HALF* of the baud rate I'm using to encode data rich .wavs
>
> Because the Nero file is sampled at 16 kHz and the LAME file at 44 kHz.


Cha-ching...!! <light bulb comes on>

OK, then my next step should be to change the WinLAME output module
to encode at 16khz. (Like I said, I'm relatively new at this). However,
WinLAME calls that, MPEG-2... will it still play back the same on all
MP3 players?

I'm off to continue experimenting.... continued thanks Carey,

DM
Anonymous
April 22, 2005 5:34:43 AM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote

> Cha-ching...!! <light bulb comes on>
>
> OK, then my next step should be to change the WinLAME output module
> to encode at 16khz. (Like I said, I'm relatively new at this). However,
> WinLAME calls that, MPEG-2... will it still play back the same on all
> MP3 players?

I highly recommend mono too! At low bit rates, you ay a lot in quality to
get stereo.

As far as compression goes, it concerns me you are compressing 2 or 3 times
according to what I read??? Just do it once, or do a compression step at 2
or 4:1 and then a limiting step at 10 or 20:1 ratios. I also always
recommend slow attack and release in general, but for very low bit rate, I
don't have enough working experience to advise you precise settings,

Julian
Anonymous
April 22, 2005 5:45:14 AM

Archived from groups: rec.audio.pro (More info?)

David Morgan (MAMS) wrote:
> Hey all,
>
> I have, at the advisement of several, put Win Lame on a couple of PCs
> to use for converting .wav files to higher rate MP3s and have been happy.

Lame has not put a lot of effort into the low bit rates. I
think the winner there lies with the dark force, WMA.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 22, 2005 11:54:27 AM

Archived from groups: rec.audio.pro (More info?)

"Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote in message news:116h3ef9ijmjtaf@corp.supernews.com...
>
> "David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote
>
> > Cha-ching...!! <light bulb comes on>
> >
> > OK, then my next step should be to change the WinLAME output module
> > to encode at 16khz. (Like I said, I'm relatively new at this). However,
> > WinLAME calls that, MPEG-2... will it still play back the same on all
> > MP3 players?
>
> I highly recommend mono too! At low bit rates, you ay a lot in quality to
> get stereo.
>
> As far as compression goes, it concerns me you are compressing 2 or 3 times
> according to what I read??? Just do it once, or do a compression step at 2
> or 4:1 and then a limiting step at 10 or 20:1 ratios. I also always
> recommend slow attack and release in general, but for very low bit rate, I
> don't have enough working experience to advise you precise settings,
>
> Julian

Thanks Julian... and all....

I believe that my real source of dissappointment was in not fully understanding
all the facets of MP3 in general. I under-estimated WinLAME capabilities based
on that weak experience. Once Carey mentioned the encryption rate of NERO,
it was fairly easy to understand why the top end phasiness would disappear...
couple that with his other suggestion of some serious Lpf, and I have a fairly
tolerable 32k stereo result now coming from LAME using 16k instead of 44.1.
A little muddy, but maybe I'll try a Hpf as well before it's over.

As to the compression, 85% of it was done in the analogue domain and was
spread around over a lot of source inputs, as well as on the stereo buss outs
feeding the broadcast and Sound Forge in the PC.

The compression and limiting done in the digital domain on the original 2-track
master .wav was done for further dynamic control only on the spoken portions
of the service when elevating them in gain to mesh better with the music levels.

Plus, there was a *very light* amount of limiting and maximization laid over the
edited stereo mix before burning the CDs, saving and converting the service to
MP3.

DM
Anonymous
April 22, 2005 12:30:43 PM

Archived from groups: rec.audio.pro (More info?)

"Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote:
>
> As far as compression goes, it concerns me you are compressing 2 or 3
> times according to what I read??? Just do it once, or do a
> compression step at 2 or 4:1 and then a limiting step at 10 or 20:1
> ratios. I also always recommend slow attack and release in general,
> but for very low bit rate, I don't have enough working experience to
> advise you precise settings,



No, Dave's doing it right. It ain't about making it pretty in this
case, it's about mitigating the effects of data loss. Slow attack and
release with gentle compression may sound nice, but it won't reduce the
dynamic range very much.

Besides, "compress early and often" is often a good idea anyway. That's
the approach I use for broadcast. Depending on the source material, it
may not always sound as pretty as one "light" pass followed by a
limiter, but intelligibility will be better. Sometimes, maybe even half
the time, it will actually be better. Again, depending on the source
material, multiple "layers" of compression can be *less* obvious because
the level stays more consistent.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
Anonymous
April 22, 2005 5:04:48 PM

Archived from groups: rec.audio.pro (More info?)

I've just tried encoding at 32 Kbps with WMA, LAME (mp3) and Fraunhofer
(mp3) CODECs. In all cases, the Fraunhofer CODEC produced the best results.
I don't have an Ogg player installed right now so I couldn't test Ogg files.

Bill.

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote in message
news:IyX9e.43081$qO6.36538@trnddc05...
>
> "Johann Burkard" <johannburkard@nexgo.de> wrote in message
> news:42683750$0$10497$9b4e6d93@newsread4.arcor-online.net...
>
>> David Morgan (MAMS) wrote:
>> > "david" <ihate@spamo.com> wrote in message
>> > news:210420051813340177%ihate@spamo.com...
>
>> >>Couple years ago I tried everything available on the Mac and found to
>> >>my surprise iTunes had the best sound at tiny sampling rates.
>
>> > I'm in the PC world here.... and Win Lame will output Ogg Vorbis, but
>> > no joy in quality at 32K and below.
>
>> I don't know what you expect, but RealAudio always sounded acceptable to
>> me (but never good), even at low bitrates.
>>
>> Johann
>
>
> I am trying to eliminate the level of aliasing(?), or severe phasing
> and 'beating' of hi-mid and high frequencies at low baud rates.
>
> My friend with the Nero package gets a fairly smooth sounding
> stereo MP3 at 20kbps !! There's just the slightest hint of light
> chorusing effects. If I could get even close to that without installing
> a complete package of redundant audio tools along with it, I'd be
> thrilled.
>
> The church decided several years ago, due to the 'spyware' issue and
> the serious involvement (read, 'guilt') of all Real Networks software in
> that regard, to eliminate anything "Real" from all of the computers. We
> used to use Real for everything, but complaints from end users about
> either the spyware issue or having to install the software at all, put a
> stop to our using it. We just need a plain old MP3 that *any* player
> can interpolate.
>
> I certainly removed all traces of "Real" from my computers several
> years ago, when Steve Gibson tried to take them to federal court over
> the spyware issue. Sure... I miss a few web streams because of it,
> but it's usually nothing that can't be found elsewhere.
>
> Thanks again,
>
> --
> David Morgan (MAMS)
> http://www.m-a-m-s DOT com
> Morgan Audio Media Service
> Dallas, Texas (214) 662-9901
> _______________________________________
> http://www.artisan-recordingstudio.com
>
>
>
Anonymous
April 22, 2005 5:41:43 PM

Archived from groups: rec.audio.pro (More info?)

Carey Carlan wrote:
> 3) massively compressed

Keep in mind that lossy encoding changes the peak values. I would
recommend keeping some dB headroom. I find that -1.3 dBFS gives me
enough headroom to do bitrates down to 160 kBit/s.

Johann
--
Nein, ich habe den Sachverhalt geradegerueckt, denn wer daemliche
Ideen hat, sollte die Kommentare nicht scheuen.
(*Tönnes faselt in <culhti$qgv$00$1@news.t-online.com>)
Anonymous
April 22, 2005 5:46:53 PM

Archived from groups: rec.audio.pro (More info?)

Bill Ruys wrote:
> I've just tried encoding at 32 Kbps with WMA, LAME (mp3) and Fraunhofer
> (mp3) CODECs. In all cases, the Fraunhofer CODEC produced the best results.
> I don't have an Ogg player installed right now so I couldn't test Ogg files.

Also don't forget mp3PRO [1].

[1] <http://www.mp3prozone.com/&gt;

Johann
--
Hier versucht einfach eine Bande von Faschisten, die Kommunikation
anderer zu beschränken. Und beleidigt absichtlich und fortgesetzt andere
Usenet-Benutzer. (Michael "chiap zap" Enezian in
<3DB2AAB7.DA34E152@stop.hier_keine_mail.net>)
Anonymous
April 22, 2005 6:03:06 PM

Archived from groups: rec.audio.pro (More info?)

>
>
>>Couple years ago I tried everything available on the Mac and found to
>>my surprise iTunes had the best sound at tiny sampling rates.
>>
>>David Correia
>>Celebration Sound
>>Warren, Rhode Island
>>
>>CelebrationSound@aol.com
>>www.CelebrationSound.com
>
>
>
> I'm in the PC world here.... and Win Lame will output Ogg Vorbis, but
> no joy in quality at 32K and below.
>
> DM
>
>
Well, there is iTunes for PC
Anonymous
April 23, 2005 5:41:37 AM

Archived from groups: rec.audio.pro (More info?)

"Lorin David Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote

> No, Dave's doing it right. It ain't about making it pretty in this case,
> it's about mitigating the effects of data loss. Slow attack and release
> with gentle compression may sound nice, but it won't reduce the dynamic
> range very much.
>
> Besides, "compress early and often" is often a good idea anyway. That's
> the approach I use for broadcast. Depending on the source material, it
> may not always sound as pretty as one "light" pass followed by a limiter,
> but intelligibility will be better. Sometimes, maybe even half the time,
> it will actually be better. Again, depending on the source material,
> multiple "layers" of compression can be *less* obvious because the level
> stays more consistent.

I respectively disagree. That's not the approach I use for broadcast. I
know why and when I am compressing and how much and how fast at each step of
the way. I generally both compress and limit vocal mics. At one station I
put the mics through a Symmetrix vocal processor at a ratio of 2:1. Then it
goes through a cheap Behringer (low budget station) with a 4:1 ratio and a
peak limiter, but I use a very slow attack and release, 'cause the Berhinger
will crunch. I do compress and limit the feed to the DAWs to prevent
clipping also. Then I use a good Omnia digital modulation processor with
carefully selected settings. I get a lot of loudness without sounding
artificial. At another station I use an Orban 8200, and I'm pretty happy
with the loudness and naturalness of that station. Listen to KEXP's
compressed and uncompressed (yes, uncompressed 44/16!) digital streams at
kexp.org. I put an Orban 6200 on those.

Maybe the reason so much radio sounds so bad is guys who simply compress
early and compress often.

Julian
Anonymous
April 23, 2005 12:11:23 PM

Archived from groups: rec.audio.pro (More info?)

Carey Carlan wrote:
>>
>> 3) massively compressed


"Johann Burkard" <johannburkard@nexgo.de> wrote
>
> Keep in mind that lossy encoding changes the peak values. I would
> recommend keeping some dB headroom. I find that -1.3 dBFS gives me
> enough headroom to do bitrates down to 160 kBit/s.




The amount of compression on a track doesn't necessarily affect the peak
level. You can have a massively compressed file with a peak level
of -60dBFS. Depends how much make-up gain you apply,if any.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
Anonymous
April 23, 2005 1:05:00 PM

Archived from groups: rec.audio.pro (More info?)

On Sat, 23 Apr 2005 01:41:37 -0700, "Julian Adamaitis"
<nospamJulianPA@Access4Less.net> wrote:

>Maybe the reason so much radio sounds so bad is guys who simply compress
>early and compress often.
>
>Julian
>

No. The reason radio sounds so bad is that people - like you -
compress. End.

d

Pearce Consulting
http://www.pearce.uk.com
Anonymous
April 23, 2005 1:18:51 PM

Archived from groups: rec.audio.pro (More info?)

"Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote:
>
> Maybe the reason so much radio sounds so bad is guys who simply
> compress early and compress often.



Pretty tough talk without knowing what my work sounds like..

I don't do radio, I do TV. I also obviously take a different approach
to music than to dialogue, and my approach to compression for promos is
different than for program material. The settings are not "one size
fits all," but I do tend to take a similar approach in concept -- that
is, more than one stage of compression, with varying thresholds and
ratios, and no peak limiting at all (save for the obvious safety limiter
at the xmitter).

I have no reason to believe your product sounds anything less than
excellent, but you may want to consider that yours is not the only
viable approach. I stand by my assertion that two or three stages of
compression usually sounds better than one stage followed by limiting.
Where and how much varies according to the content and application.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
Anonymous
April 23, 2005 1:32:18 PM

Archived from groups: rec.audio.pro (More info?)

"Johann Burkard" <johannburkard@nexgo.de> wrote in message news:4268e2f0$0$7517$9b4e6d93@newsread2.arcor-online.net...
> Carey Carlan wrote:
> > 3) massively compressed
>
> Keep in mind that lossy encoding changes the peak values. I would
> recommend keeping some dB headroom. I find that -1.3 dBFS gives me
> enough headroom to do bitrates down to 160 kBit/s.
>
> Johann


I'll do some comparisons on the pre and post MP3 files and check for
that... but I'd never start with .wavs greater than a half dB down at the
maximum. I'd also rarely ever be looking at program material with less
than a 14dB difference between peak and average, so my stuff still has
some dynamics that might be susceptible.

What would you say the encoding math does to cause peak values to
change, and is there any symptom repeatable enough to expect often
and watch out for?

Are you also indicating that the lower the baud rate the more subjected
it becomes to this change in peak volume?

DM
Anonymous
April 23, 2005 2:45:32 PM

Archived from groups: rec.audio.pro (More info?)

Up yours jerk. When you get a brain, let me know.

"Don Pearce" <donald@pearce.uk.com> wrote in message
news:426a0f9b.41006375@news.plus.net...
> On Sat, 23 Apr 2005 01:41:37 -0700, "Julian Adamaitis"
> <nospamJulianPA@Access4Less.net> wrote:
>
>>Maybe the reason so much radio sounds so bad is guys who simply compress
>>early and compress often.
>>
>>Julian
>>
>
> No. The reason radio sounds so bad is that people - like you -
> compress. End.
>
> d
>
> Pearce Consulting
> http://www.pearce.uk.com
Anonymous
April 23, 2005 2:53:09 PM

Archived from groups: rec.audio.pro (More info?)

"Lorin David Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote in message
news:%poae.58630$vt1.33936@edtnps90...
> "Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote:
>>
>> Maybe the reason so much radio sounds so bad is guys who simply
>> compress early and compress often.
>
> Pretty tough talk without knowing what my work sounds like..

Please notice I didn't specifically say YOUR work sounds bad. I have no
idea if that is true or not. You should not take that personally sir, as I
in no way meant it to be a judgment on you personally. I assume you were
joking and you do put a little more thought into what you're doing than
simply "compress early and compress often". Unfortunately not everybody
doing broadcast does...

Julian
Anonymous
April 23, 2005 3:12:43 PM

Archived from groups: rec.audio.pro (More info?)

David Morgan (MAMS) wrote:

> What would you say the encoding math does to cause peak values to
> change, and is there any symptom repeatable enough to expect often
> and watch out for?

IME many kinds of filtering of audio waveforms can potentially change
peak values to change. For example, a Linkwitz-Riley crossover can
cause unexpectedly high peak values to be developed in either of the
two bandpass channels.

My theory is that the phase shift associated with the processing
changes the phase relationships between components of various waves in
ways that things line up or not in different ways.

One of the things that perceptual coders sometimes seem to do is
"lose" phase relationships within a complex wave. Hey, losing phase
information can have limited or imperceptible effects on sound
quality, and obviously simplifies the coded that that is actually
transferred.
Anonymous
April 23, 2005 3:16:06 PM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote

> I assure you that after 30 years of playing with similar toys, that I am
> at
> least as alert. I'm also more than certain that Mr. Schultz is keenly
> aware of everything that he does as well. This puts us all on the same
> playing field. We know where, when, why, hopefully how, and how much.
> Good.... we're friends now.

Glad to hear that. You responses to me were so vague about why you were
doing things I wondered if you understood compression at all or if you were
merely turnign all the knobs up to 11. Several levels of compression can be
a good thing if done right. I had no information from your post that you
knew what right was.

>
>> I generally both compress and limit vocal mics.
>
> Great. I don't recall ever peak limiting a vocal mic after compressing...
> Does this mean one of us is wrong? Nah.... 'course not.

Yes, there is more than one way to do it right. Limiting after compressing
first creates a "soft knee" limiter. Much better sounding than a plain old
limiter. Some limiters have this built in.

> A ratio is meaningless number without an input source or a threshold
> level to justify it. What's on the output side and why? Could you change
> the ratio if you changed the release time and get the same result? It's
> odd how many ways there are to do things... sometimes to do the same
> thing.

You are absolutely correct. I meant to get into this too, but I thought I'd
already said enough, and form the looks of the response, I said more than
enough.

> The Behringer will 'crunch' if it's in the signal path.... period. <g>

Nah. The reason people say that is they don't use slow attacks and releases
with it. It is a cheap box, but works well with slow response.

>> Then I use a good Omnia digital modulation processor with
>> carefully selected settings.
>
> Never heard of it. Time to learn something new again... I love it in
> here!
> It does sound like something to be careful and selective with, and to be
> sure that you have a good one and not a bad one. <g>

Its a broadcast product made by Teleos. If you don't do broadcast you
wouldn't have heard of it.


>> I get a lot of loudness without sounding artificial.
>
> My bullhorn gets a lot of loudness without being artificial....
> I swear.... it sounds just like a bullhorn. And if you ever
> forgive me for all this, we'll be acquaintances for life.

All kidding aside.. The thing that makes stuff sound weird from over
compression are 1) too much (threshold) 2) too much (ratio) and 3) too fast
(attack and/or release).

>> At another station I use an Orban 8200, and I'm pretty happy
>> with the loudness and naturalness of that station. Listen to KEXP's
>> compressed and uncompressed (yes, uncompressed 44/16!) digital
>> streams at kexp.org. I put an Orban 6200 on those.
>
> Mr. Orban still posts here on occasion... he'll like that. ;-)

I saw Mr. Orban last week at the NAB show.

> When you first suggested that I might be overcompressing (which my
> entire history here will easily show that I am against), I assumed that
> you
> were worried that I was doing it all in the digital domain thus creating
> the
> inevitable math errors, but now I'm not so certain you read my description
> to Carey.

I can't say my newsreader even contains the entire thread and I can't say I
read every post I do have fully. I was concerned that your responses to me
seemed to be somewhat vague about why and how much you were compressing.
I'm glad to hear you understand more about compression that compress early
and compress often! Keep up the good work.

> I think you missed the point.... we're sorta' on the topic of miking large
> ensembles.... that's loud bands on the fly. In my case, done primarily
> for
> a live audience, done well enough to send a clone of that to a broadcast
> on the fly while recording it live to stereo, and then tweaking a little
> later
> for a web archive.

I did miss that point.

> Peace (and I mean it)

Damn, never meant to destroy any peace I merely said I "respectively
disagree" with the statement that "compress early and compress often".

Julian
Anonymous
April 23, 2005 5:29:38 PM

Archived from groups: rec.audio.pro (More info?)

On 4/23/05 4:11 AM, in article Lqnae.58624$vt1.32785@edtnps90, "Lorin David
Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote:

> Carey Carlan wrote:
>>>
>>> 3) massively compressed
>
>
> "Johann Burkard" <johannburkard@nexgo.de> wrote
>>
>> Keep in mind that lossy encoding changes the peak values. I would
>> recommend keeping some dB headroom. I find that -1.3 dBFS gives me
>> enough headroom to do bitrates down to 160 kBit/s.
>
>
>
>
> The amount of compression on a track doesn't necessarily affect the peak
> level. You can have a massively compressed file with a peak level
> of -60dBFS. Depends how much make-up gain you apply,if any.

Don;t think it's a compression comment but one of absolute max-value dBfs.
This isn;t bad practice anyway, even for release tracks. Anything I send out
with a client doesn;t get past -2dBfs.

Another issue I recall coming across ages agop with switch56 ISDN live
broadcast with COMREX gear was that they seemed reluctant to advise a lot of
compression, somehting about keeping a fair crest factor made the system
happier and sound better.
Anonymous
April 23, 2005 5:33:05 PM

Archived from groups: rec.audio.pro (More info?)

On 4/23/05 4:41 AM, in article 116k2i24b1ng16b@corp.supernews.com, "Julian
Adamaitis" <nospamJulianPA@Access4Less.net> wrote:

I generally both compress and limit vocal mics. At one station I
> put the mics through a Symmetrix vocal processor at a ratio of 2:1. Then it
> goes through a cheap Behringer (low budget station) with a 4:1 ratio and a
> peak limiter, but I use a very slow attack and release, 'cause the Berhinger
> will crunch. I do compress and limit the feed to the DAWs to prevent
> clipping also. Then I use a good Omnia digital modulation processor with
> carefully selected settings. I get a lot of loudness without sounding
> artificial. At another station I use an Orban 8200, and I'm pretty happy
> with the loudness and naturalness of that station. Listen to KEXP's
> compressed and uncompressed (yes, uncompressed 44/16!) digital streams at
> kexp.org. I put an Orban 6200 on those.
>
> Maybe the reason so much radio sounds so bad is guys who simply compress
> early and compress often.

JULIAN.... Didn't think there were any of you types left!
Keep the flame.
Is FMT still abenchmark for broadcast quality these days?
Anonymous
April 23, 2005 5:33:06 PM

Archived from groups: rec.audio.pro (More info?)

"SSJVCmag" <ten@nozirev.gamnocssj.com> wrote

> JULIAN.... Didn't think there were any of you types left!

What type is that?

> Keep the flame.

No flame intended.

> Is FMT still abenchmark for broadcast quality these days?

What is FMT?

Julian
Anonymous
April 23, 2005 6:07:37 PM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com>

> I never did respond to you. (Take me with a grain of salt, here). You
> just
> kinda jumped in and said to Lorin that we were doing it all wrong.

Your responses that I read just listed a bunch of different gain control
products without much reference to your rational or reason why you needed so
many gain control stages.

> That's the wonderful thing about Google... a quick search on my name
> and 'compression' would have shown that I've been preaching against
> it's over use around here for at least 8 years.

I too have crusaded against too much compression especially in the broadcast
field. I think we find more and more to agree on as we talk.

> Not to initiate another debate, but limiting in most cases is just lopping
> off
> the tops of otherwise smooth waveforms, creating near squares. There
> aren't too many inexpensive limiters that can do much more than that.
> I'd use higher ratio compression if at all possible, except in the event
> of seriously high transient peak source material. Whatever I do, I try
> not to ruin the source material before it ever gets well into the chain
> by creating near square waves from the git-go. There's nothing that
> can fix it after that happens.

Just a unclarity in semantics. We never discussed what you mean as limiting
as to what I mean. I think we both do this part the much the same. I never
use extremely high limit ratios either where I can control them. Where I
can't set the ratio, I use very little or none of it.

> (EG: I prefer fast attack times and modest release
> times on the majority of my applications, and that comes from 30 years of
> recording and live sound... different, I think, from your experience).
> But we're
> here to share what we do know, so don't let this put you off... just be
> up for
> justifying your recommendations with a little more content.

Me too until I attended a mastering session where the engineer used
extremely slow attack and release. I mean 500 ms attack and 4 sec release.
Not that I advocate that for every or even many situations, but it opened my
eyes to new possibilities. For example, try slow attack and FAST release
for a nice tight sound with lots of definition on a bass instrument.

> The only problem was that in the basis of your disagreement, you really
> provided no specific facts or examples... only that you knew, 'better'.

I just said I disagreed with the statement "compress early and compress
often". I don't think it takes a lot of rhetoric to claim that to be an
oversimplification.

> A little advice from having hung around here for a while.... If you're
> going
> to offer some advice or lodge a disagreement, be prepared to say more
> than, "I think that's wrong and I'd do it differently".... *especially* if
> you
> haven't read the entire thread. ;-)

First of all, it isn't always possible to tell if the entire thread is there
or not. I thought I DID have the entire thread until you pointed out some
info I didn't have.

Secondly I thought I did say some of the reasons WHY I did things the things
I did.

Thirdly, what I "respectively disagreed" with was the statement "compress
early and compress often". And even then I only disagreed in that I claim
there is more to it than just that. I thought a few basic examples was all
that were needed to deny such a wide sweeping generalization. Compress
early and compress often, maybe yes that can work, but more important is
your settings, not simply doing a lot of it and doing it often. Give me a
break here pal! I too have been making a living in audio for 29 years now
and I don't think a full thesis on my compression philosophy is needed to
argue there is more to compression than "compress early and compress often."
I was prepared to discuss specifics, but most others (you excepted) were
more interested in blowing off steam than discussing specifics.

Julian
Anonymous
April 23, 2005 10:08:42 PM

Archived from groups: rec.audio.pro (More info?)

On Sat, 23 Apr 2005 10:45:32 -0700, "Julian Adamaitis"
<nospamJulianPA@Access4Less.net> wrote:

>Up yours jerk. When you get a brain, let me know.
>
>"Don Pearce" <donald@pearce.uk.com> wrote in message
>news:426a0f9b.41006375@news.plus.net...
>> On Sat, 23 Apr 2005 01:41:37 -0700, "Julian Adamaitis"
>> <nospamJulianPA@Access4Less.net> wrote:
>>
>>>Maybe the reason so much radio sounds so bad is guys who simply compress
>>>early and compress often.
>>>
>>>Julian
>>>
>>
>> No. The reason radio sounds so bad is that people - like you -
>> compress. End.
>>
>> d
>>
>> Pearce Consulting
>> http://www.pearce.uk.com
>

I thought you might have some cogent, rational argument to advance.
I've no doubt the programmes you engineer are a match in quality.

d

Pearce Consulting
http://www.pearce.uk.com
Anonymous
April 23, 2005 10:08:43 PM

Archived from groups: rec.audio.pro (More info?)

"Don Pearce" <donald@pearce.uk.com> wrote

> I thought you might have some cogent, rational argument to advance.

When you come up with something more than

"No. The reason radio sounds so bad is that people - like you -compress.
End.",

I'll enter into discussion with you. You're the one who is not posting a
cogent or rational argument, sir.

Julian
Anonymous
April 23, 2005 11:17:01 PM

Archived from groups: rec.audio.pro (More info?)

"Arny Krueger" <arnyk@hotpop.com> wrote in message news:mPydnfLb_fN3-PffRVn-sA@comcast.com...
> David Morgan (MAMS) wrote:
>
> > What would you say the encoding math does to cause peak values to
> > change, and is there any symptom repeatable enough to expect often
> > and watch out for?
>
> IME many kinds of filtering of audio waveforms can potentially change
> peak values to change. For example, a Linkwitz-Riley crossover can
> cause unexpectedly high peak values to be developed in either of the
> two bandpass channels.
>
> My theory is that the phase shift associated with the processing
> changes the phase relationships between components of various waves in
> ways that things line up or not in different ways.
>
> One of the things that perceptual coders sometimes seem to do is
> "lose" phase relationships within a complex wave. Hey, losing phase
> information can have limited or imperceptible effects on sound
> quality, and obviously simplifies the coded that that is actually
> transferred.


I've definitely heard plenty of frequency related changes in my recent
experiments. The phasey, beating sort of thing definitely seems to
be louder that the basic levels using 44.1 encryption, but it lessens
dramatically when going to lower encryption rates like 16K as Carey
had me try earlier (for the tiny baud rate stuff).

DM
Anonymous
April 23, 2005 11:17:02 PM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote

> I've definitely heard plenty of frequency related changes in my recent
> experiments. The phasey, beating sort of thing definitely seems to
> be louder that the basic levels using 44.1 encryption, but it lessens
> dramatically when going to lower encryption rates like 16K as Carey
> had me try earlier (for the tiny baud rate stuff).

You'll hear even less if you go to mono.

Julian
Anonymous
April 24, 2005 1:18:58 AM

Archived from groups: rec.audio.pro (More info?)

"Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote in message news:116leaog86pgdd2@corp.supernews.com...
>
> "David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote
>
> > I've definitely heard plenty of frequency related changes in my recent
> > experiments. The phasey, beating sort of thing definitely seems to
> > be louder that the basic levels using 44.1 encryption, but it lessens
> > dramatically when going to lower encryption rates like 16K as Carey
> > had me try earlier (for the tiny baud rate stuff).
>
> You'll hear even less if you go to mono.
>
> Julian


I'm giving that some serious thought... but I had some non-pannable
stereo efx returns going and a couple of stereo miked sources panned
in the mix, which might go south on me if collapsed to mono. I'll be
experimenting, thanks....

DM
Anonymous
April 24, 2005 1:18:59 AM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote

> I'm giving that some serious thought... but I had some non-pannable
> stereo efx returns going and a couple of stereo miked sources panned
> in the mix, which might go south on me if collapsed to mono. I'll be
> experimenting, thanks....

I figured you probably had some reason for rejecting mono. Let us know.

Julian
Anonymous
April 24, 2005 2:42:39 AM

Archived from groups: rec.audio.pro (More info?)

"Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote in message news:116le8rnglurg9f@corp.supernews.com...

> I too have crusaded against too much compression especially in the broadcast
> field. I think we find more and more to agree on as we talk.

Indeed... it's just a magnitude or three higher in difficulty to discuss when
the topic of compression is the subject matter. There probably isn't any
other facet of audio that's more susceptible to the minute specifics of the
source material (which is constantly changing) than compression. I have
to say that I dislike most of the 'primers' and the 'recommended settings'
data that's out there because when dealing with compression, you just
*have* to be there, to understand in even the slightest fashion, what
needs to be done, if anything, to the source material.

> Just a unclarity in semantics. We never discussed what you mean as
> limiting as to what I mean.

Agreed. It's unfortunate that we'd both have to write a couple of books
here on the subject, before we could remotely understand each other's
philosophies and applications.

I caught your last reply to Don, and I think we're probably dealing with our
terminology from two widely diverse perspectives. I'm not worried about
overmodulation, but you have to be. You may not have to be worried
about singers that suddenly start channeling Aretha Franklin in the middle
of a live mix, but I do.

> Me too until I attended a mastering session where the engineer used
> extremely slow attack and release. I mean 500 ms attack and 4 sec release.
> Not that I advocate that for every or even many situations, but it opened my
> eyes to new possibilities.

I'd be willing to wager that this was followed by some serious peak limiting
and volume maximization processing. The long release doesn't sound
appealing to me... but I wasn't there to understand *why* it worked on
your source material.

> Give me a break here pal!

Done... ;-)

> I was prepared to discuss specifics, but most others (you excepted) were
> more interested in blowing off steam than discussing specifics.

Like I said <g>, you gotta' be there in *every situation* to get a grip on the
specifics... or as you said, we'd be writing thesis' and books on compression
for years to come before we'd have the slightest idea what the other is actually
experiencing and why we might do what we do.


Cheers,

--
David Morgan (MAMS)
http://www.m-a-m-s DOT com
Morgan Audio Media Service
Dallas, Texas (214) 662-9901
_______________________________________
http://www.artisan-recordingstudio.com
Anonymous
April 24, 2005 5:43:00 AM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote (responding to
Julian):
>
> I respect your opinion. I understand it's based on experience and
> preference.


Some of the disparity in approach may be the result of the difference in
applications. In radio, the music content is already mixed and
mastered, so some of the dynamics control is already done.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
Anonymous
April 24, 2005 6:01:23 AM

Archived from groups: rec.audio.pro (More info?)

"Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote:
>
> Me too until I attended a mastering session where the engineer used
> extremely slow attack and release.


I think you and I are trying to accomplish very different things with
our dynamics control devices. I find very little use for slow attack on
a limiter. If I'm using a limiter at all, it's to catch a transient. A
slow attack won't catch that, so it defeats the purpose.

I'd rather use a couple stages of compression with progessively higher
thresholds and ratios to even things out without clipping or chopping.
Kinda like how the SuperNice Mode on the RNC works.



> try slow attack and FAST release for a nice tight sound with lots of
> definition on a bass instrument.

Have you not found that to be an ideal recipe for distorting the bass
instrument? Or maybe I should ask what you mean by "fast release?"



> I just said I disagreed with the statement "compress early and
> compress often". I don't think it takes a lot of rhetoric to claim
> that to be an oversimplification.

Uh, yeah, it would be safe to say that one cannot summarize a dynamics
control regime in five words without skipping a few details! <g>. It
was, however, written in the context of a specific application, and
there was some discussion of how it was being applied. Besides, it was
intended to be a semi-humourous "rule-of-thumb" kind of remark, not a
mantra. And it works for me more often than not (though there have
obviously been more than a couple "nots").

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
Anonymous
April 24, 2005 6:01:24 AM

Archived from groups: rec.audio.pro (More info?)

"Lorin David Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote

> I think you and I are trying to accomplish very different things with our
> dynamics control devices. I find very little use for slow attack on a
> limiter. If I'm using a limiter at all, it's to catch a transient. A
> slow attack won't catch that, so it defeats the purpose.

As I said, the application in question was for mastering a session that was
already mixed and individual tracks has compression, limiting etc. If I
understand what and why the engineer did what he did (which is doubtful!) my
impression was he was using this to merely even out volume differences
between tracks and remove some of the dynamic range in individual cuts. He
would re-adjust the threshold for each cut as he went. It was more of an
AGC than compression or limiting.

> I'd rather use a couple stages of compression with progessively higher
> thresholds and ratios to even things out without clipping or chopping.
> Kinda like how the SuperNice Mode on the RNC works.

That's pretty much what I said a couple of posts back.

>> try slow attack and FAST release for a nice tight sound with lots of
>> definition on a bass instrument.
>
> Have you not found that to be an ideal recipe for distorting the bass
> instrument? Or maybe I should ask what you mean by "fast release?"

Fast enough release that it doesn't boom and sustain forever. I'd use to
ear to judge what poitn that is. It sounds tighter sometimes if the release
underemphasizes the sustain slightly. If you have a player with a lot of
string attack, you may want your attack time to be slow enough to let it all
through, because if it is too fast, it will be lost. Or if you
intentionally want to loose the pops then set it for a fast attack.

> Uh, yeah, it would be safe to say that one cannot summarize a dynamics
> control regime in five words without skipping a few details! <g>. It was,
> however, written in the context of a specific application, and there was
> some discussion of how it was being applied. Besides, it was intended to
> be a semi-humourous "rule-of-thumb" kind of remark, not a mantra. And it
> works for me more often than not (though there have obviously been more
> than a couple "nots").

Sorry. I don't know most of you guys well enough to tell when you're
kidding yet!

Julian
Anonymous
April 24, 2005 6:01:24 AM

Archived from groups: rec.audio.pro (More info?)

"Lorin David Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote

> I think you and I are trying to accomplish very different things with our
> dynamics control devices. I find very little use for slow attack on a
> limiter. If I'm using a limiter at all, it's to catch a transient. A
> slow attack won't catch that, so it defeats the purpose.

I also agree. A limiter is of very limited value with a slow attack. It
becomes more of an AGC in that case.

Julian
Anonymous
April 24, 2005 11:02:55 AM

Archived from groups: rec.audio.pro (More info?)

"Lorin David Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote in message news:EQCae.89805$7Q4.77732@clgrps13...
> "David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote (responding to
> Julian):
> >
> > I respect your opinion. I understand it's based on experience and
> > preference.
>
>
> Some of the disparity in approach may be the result of the difference in
> applications. In radio, the music content is already mixed and
> mastered, so some of the dynamics control is already done.

Have you had any requests for CD versions and Radio versions of
mixes? I recently had my first. I'm all too familiar with the vocal up
and vocal down versions, but now there seems to be a new need for
restoring some of the dynamic range people are removing by over-
compressing... apparently because it's screwing with the way in which the
high-dollar broadcast levelling/limiting gear does it's job. Hypercompressed
material doesn't sound as good going through the broadcast processors
as does the same material with a little more of the dynamic range left in
so that the broadcast devices can do what they were designed to do.

Comments... either of you guys ??
Anonymous
April 24, 2005 11:02:56 AM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote

> Have you had any requests for CD versions and Radio versions of
> mixes? I recently had my first. I'm all too familiar with the vocal up
> and vocal down versions, but now there seems to be a new need for
> restoring some of the dynamic range people are removing by over-
> compressing... apparently because it's screwing with the way in which the
> high-dollar broadcast levelling/limiting gear does it's job.
> Hypercompressed
> material doesn't sound as good going through the broadcast processors
> as does the same material with a little more of the dynamic range left in
> so that the broadcast devices can do what they were designed to do.
>
> Comments... either of you guys ??

Interesting idea. I've heard it mentioned over the years, but never had any
first hand experience with it. You'd almost need one of those broadcast
processors to tell what effect if any your mastering has. With certain
material, like techno with extremely high transients, it might sound better
if you limit more rather than less, so you can control that part yourself
instead of giving it to the downstream broadcast device to mangle. Other
places you might want to compress less to accommodate the extra compression
it will get later. There are a vast range of settings on Broadcast
processors from "classical" to "open pop" (my favorite) to "rock" to
"grunge". It seems you'd have to make 3 or 4 masters for which degree of
processing the station in question uses. The problem with these devices is
that they have so many settings and the user has so much control to modify
the settings some really awful results are possible by people who are merely
trying to be loud without having enough experience to hear when they have
gone too far.

Julian
Anonymous
April 24, 2005 11:30:47 AM

Archived from groups: rec.audio.pro (More info?)

"Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote in message...

> I don't know how it works in the UK. Here in the US, I understand that it
> will be required to get a different modulation processor for my digital
> signal on the 2 stations I will be converting to HD. The analog radio
> processor is not adequate for reasons I don't completely understand.
> Perhaps if Mr. Orban is reading he can explain. I for one am not too happy
> about having to spend up to $10,000 for a second processor.

I'd definitely check into that... and I'd love to have a shot at hearing the
reasoning behind it. Honestly, it sounds to me like just another push
for even tighter control over potential overmodulation in order to step
up broadcast output (volume levels) to yet another plateau. I'll bet
they're thinking that more expensive, read-ahead delay type limiting
will keep a tighter reign on over-modulation while increasing levels
to a greater degree than traditional analogue devices.

Maybe you have a chance to speak out here... quality versus loudness
could definitely come into play, and somehow I just don't believe this
has anything at all to do with the fact that station output programming
will be derived from HD. Then again, maybe there are some side-band
issues, or something, in HD that have to be addressed seperately.

Scott D. ?

> I again invite you to listen to the web streams I processed for kexp.org
> with an Orban 6200.

I wanna' hear something done with the Behringer... ;-)

Just kidding... I've got a road trip coming up, so maybe we'll talk again
on Wednesday night.

Peace,

DM
Anonymous
April 24, 2005 11:30:48 AM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote

> I'd definitely check into that... and I'd love to have a shot at hearing
> the
> reasoning behind it. Honestly, it sounds to me like just another push
> for even tighter control over potential overmodulation in order to step
> up broadcast output (volume levels) to yet another plateau. I'll bet
> they're thinking that more expensive, read-ahead delay type limiting
> will keep a tighter reign on over-modulation while increasing levels
> to a greater degree than traditional analogue devices.

The bandwidth is tightly limited to what is left over from existing analog
FM to 96kbps which is then divided up between sometimes 3 programs! Talk
about low bandwidth challenges.

> Maybe you have a chance to speak out here... quality versus loudness
> could definitely come into play, and somehow I just don't believe this
> has anything at all to do with the fact that station output programming
> will be derived from HD. Then again, maybe there are some side-band
> issues, or something, in HD that have to be addressed separately.

I wish I could express an opinion, but instead I was "informed" of the
standard while at the NAB last week.

Julian
Anonymous
April 24, 2005 11:33:22 AM

Archived from groups: rec.audio.pro (More info?)

"Bob Cain" <arcane@arcanemethods.com> wrote in message news:D 4adiq03iu@enews4.newsguy.com...
>
>
> David Morgan (MAMS) wrote:
> > Hey all,
> >
> > I have, at the advisement of several, put Win Lame on a couple of PCs
> > to use for converting .wav files to higher rate MP3s and have been happy.
>
> Lame has not put a lot of effort into the low bit rates. I
> think the winner there lies with the dark force, WMA.


Possibly... but I'm trying to cook both formulae on the same stove. ;-)

Honest to gosh... I didn't realize that I could leave 44.1 behind as an encryption
rate!! Taking the same encode function in LAME down to a 16Khz output
renders a pretty acceptable stereo, 32K baud rate.

Like I said, I'm new at this. I've had LAME for around a year now, but this
church web archive thing is the first time that I've had to give consideration
to creating anything less than a 160K MP3.

DM
Anonymous
April 24, 2005 11:43:31 AM

Archived from groups: rec.audio.pro (More info?)

"Lorin David Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote in message news:T5Dae.89808$7Q4.63865@clgrps13...

> "Julian Adamaitis" <nospamJulianPA@Access4Less.net> wrote:
> >
> > Me too until I attended a mastering session where the engineer used
> > extremely slow attack and release.

I'd really love to have been present to see under what set of circumstances
that a four second release time would have been employed. If there was
any density at all to the program material, it seems like the compressor
would have entered a state of steady compression and never come out.
At first thought, this would drastically change the 'appearance' of the original.

> I think you and I are trying to accomplish very different things with
> our dynamics control devices. I find very little use for slow attack on
> a limiter. If I'm using a limiter at all, it's to catch a transient. A
> slow attack won't catch that, so it defeats the purpose.

I didn't want to go there, but I didn't understand that such an option existed
on a peak limiter. If the objective is to catch and stop the peaks, which
are predominately the result of transients, then a slow attack is utterly
worthless. Julian is confusing me a little as to whether he's being a
purist on me or whether he's trying to avoid station overmodulation.
I don't think you can successfully be both in that situation. ;-)

> > try slow attack and FAST release for a nice tight sound with lots of
> > definition on a bass instrument.
>
> Have you not found that to be an ideal recipe for distorting the bass
> instrument? Or maybe I should ask what you mean by "fast release?"

Anything less than a couple of hundred milliseconds, and my experience
has been that distortion occurs when the following note strikes while the
compressor is still in the fast release stage... a sure fire recipe for erratic
pumping as well (depending on another dozen factors, of course). The
same tends to also apply on low frequency material if the attack side of
the equation is too fast, at least in my experience.

If I were letting the peaks get by, but wanted to set a limited RMS output
level, I'd just use a higher compression ratio (on a compressor that will
handle the load as transparently as possible). One reason I still like the
old ASHLY CL-52E so much is that it's very nearly an RMS limiter - even
at modest ratios it will practically brick wall it's output, and very little in the
way of transients gets by with a faster attack.

Oh well, it's late and I'm rambling... just wanted to say hello before slipping
down to Austin for a couple of days (after I finish my newly acquired MP3
archiving task for the church) tomorrow night.

DM
Anonymous
April 24, 2005 11:43:32 AM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote

> I'd really love to have been present to see under what set of
> circumstances
> that a four second release time would have been employed. If there was
> any density at all to the program material, it seems like the compressor
> would have entered a state of steady compression and never come out.
> At first thought, this would drastically change the 'appearance' of the
> original.

Possibly true depending on the threshold setting. As I said, he spent a
long time readjusting the threshold for every cut.

>> I think you and I are trying to accomplish very different things with
>> our dynamics control devices. I find very little use for slow attack on
>> a limiter. If I'm using a limiter at all, it's to catch a transient. A
>> slow attack won't catch that, so it defeats the purpose.
>
> I didn't want to go there, but I didn't understand that such an option
> existed
> on a peak limiter. If the objective is to catch and stop the peaks, which
> are predominately the result of transients, then a slow attack is utterly
> worthless. Julian is confusing me a little as to whether he's being a
> purist on me or whether he's trying to avoid station overmodulation.
> I don't think you can successfully be both in that situation. ;-)

I've been in situations of studio mixing and broadcast engineering and I may
lapse between the 2 without acknowledging it. In general, my experience is
a sum total of both professions as well as some live sound and semi-pro
mastering.

Also when I say limiter, I could mean anything with a ratio of 10:1 or more.
At 8:1 or 10:1 a slow attack can be useful. When you say peak limiter, I
assume you mean something that catches peaks, which can't be done with slow
attacks. It seems most combinations can be useful somewhere.

Julian
Anonymous
April 24, 2005 5:57:53 PM

Archived from groups: rec.audio.pro (More info?)

"David Morgan (MAMS)" <mams@NOSPAm-a-m-s.com> wrote:
>
> Have you had any requests for CD versions and Radio versions of
> mixes? I recently had my first. I'm all too familiar with the
> vocal up and vocal down versions, but now there seems to be a new
> need for restoring some of the dynamic range people are removing by
> over-compressing... apparently because it's screwing with the way in
> which the high-dollar broadcast levelling/limiting gear does it's
> job. Hypercompressed material doesn't sound as good going through the
> broadcast processors as does the same material with a little more of
> the dynamic range left in so that the broadcast devices can do what
> they were designed to do.
>
> Comments... either of you guys ??


All I know about that is what Bob Orban has written here. Given the
source, I tend think there may be some validity to that claim! <g>

I have no control over our "final stage" processing, nor have I ever
seen how it's set. I don't even know which box we're using.

I have, however, frequently compared recordings from Master Control to
what I recorded in the control room. Based on those, I got the
impression that our processing seems to be set to "fairly benign."
That's not all that surprising, as our Chief Engineer leans towards
erring on the side of safety (he thinks anything that *peaks*
over -20dBFS is too hot). Since I was once called into the Principal's
office for putting the transmitter in jeopardy with excessive levels, I
guess we're not doing much peak limiting either!

I don't have much occasion to put commercially mastered material to air,
so it's hard for me to judge. All the music I deal with is either live
or from our network library. The network stuff isn't mashed like a
commercial CD, so I don't know what pre-crushed material sounds like
through our chain.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
!