Re: [ARSCLIST] MP3 bit rates and usage factors for Web pages

[Tom Fine <tflists@xxxxxxxxxxxxxxxxxxx>]

I, too, have found 96bps acceptable for mono voice or old-media music
(cylinders/78's). Make sure to encode as mono. It's a good compromise format
because anyone who wants the content will wait for the download and the
sound quality is definitely OK with the limited frequency response of those
sources. I'd never do it that low for stereo anything or any kind of
high-fidelity music source material. I've found that encoders make a big
difference. Surprisingly, MusicMatch always does a good job for me (there's
an option in the FILE menu to convert WAV files to MP3). The new MP3 save-as
in Sony Soundforge 7 and higher is fine, too (before Sony bought Soundforge,
the MP3 encoding was very limited due to not paying the German owners for
full rights). iTunes MP3 encodes leave something to be desired for my ears,
almost like Apple is trying to throw the curve so people will use their
proprietary formats. I suspect but do not know that they haven't ponied up
to the German owners of the MP3 format and are using a reverse-engineered
encoder based on LAME. I have not had good results with LAME -- it's slow
and the sound is not great on my Windows boxes, with various front-ends
tried.

One man's results, mileage may vary, etc.

-- Tom Fine

----- Original Message ----- 
From: "Colin Schlachta" <colin.schlachta@xxxxxxxxxxxxxxxx>
To: <ARSCLIST@xxxxxxxxxxxx>
Sent: Tuesday, December 06, 2005 9:58 AM
Subject: Re: [ARSCLIST] MP3 bit rates and usage factors for Web pages


I wanted to add that I agree with David regarding the 96 bps resolution.
Here at the museum, I listened to cylinder recordings at various bps
levels and also selected 96 bps to be the resolution selected for our
mp3s.  The mp3s that I make are however only 20-25 second mono samples
of the wax cylinder transfers.  This is so that anyone on the web can
view the scans of musical/vocal transcriptions online and also can hear
part of the original recording from where the transcriptions came from.
The use of 25 second mono samples was simply the museum's way of dealing
with both copyright and museum server space issues.


Colin Schlachta
Audio-Visual Archives
Canadian Museum of Civilization
100 Laurier Street
Gatineau, Quebec
J8X 4H2
Fax: (819) 776-7055
Tel: (819) 776-8466



-----Original Message-----
From: Association for Recorded Sound Discussion List
[mailto:ARSCLIST@xxxxxxx] On Behalf Of David Seubert
Sent: Monday, December 05, 2005 11:52 PM
To: ARSCLIST@xxxxxxxxxxxx
Subject: Re: [ARSCLIST] MP3 bit rates and usage factors for Web pages

I've been out of town and I'm just catching up with ARSClist, but here
are some of my thoughts related to Richard's question and the discussion

that followed:

1) If rights allow you to, put it online and let people download it.
Yes, you lose some control, but in cases of real infringement beyond
what you state people are allowed to do with the material, you can
always turn the lawyers loose on them. We recently licensed a recording
from our collection to Norton for the new book that just came out on
Richard Feynman and we charged them a lot of money. That particular
recording is not online, but we do have Linus Pauling material from the
same collection online that is free to download. If somebody wanted the
Pauling material for similar commercial purposes, we would still charge
them a lot of money even though it is online. Free access online does
not ruin the market for commercial exploitation nor does it mean you
give up your rights to control what people do with things. Some people
may sell bootlegs on ebay, but legit users like Norton are going to
license it properly.

2) There is perceived value to higher bitrate recordings. For our
cylinders, we determined that 96kbits was sufficient. However, we used
128kbit because people perceive 96kbit mp3s to be of low quality, even
if they capture every last bit of frequency response. You could also use

VBR which is becoming more common.

3) Streaming is great, but it requires extra infrastructure. It is
helpful to be able to stream material so people don't have to download
large files to see if the material is actually of interest to them, but
it requires a dedicated server and a separate set of files. We've found
that given a choice people download more than they stream by a wide
margin. I can't give you exact figures, but people want to grab it and
go. Restricting to streaming might be a possibility if there are rights
issues that don't allow you to provide downloads. Yes, people can
capture audio streams with software, but that's breaking the law. People

also knife plates out of books which is also illegal, but we don't shut
our doors as a precaution. We should do our best to maximize use and
minimize risks and get on with things.

4) If copyright and/or donor agreements allow you to, give them the
whole recording online, not an excerpt. Why are we obsessed with
control? Do we want to evaluate the character or scholarly credentials
of each user before deciding to grant them access? That's been the
history at some institutions (I'm not naming names), but it is
antithetical to my way of thinking and is a waste of the archivist's and

the researcher's time. Over the past two weeks people have downloaded
over 200,000 mp3 files from our cylinder website. That is more use of
our historical audio collection (78s and cylinders) than the collection
has seen in its 30+ year existence by several orders of magnitude.
Granted, giving away music online is like handing out dope at a Grateful

Dead show and you won't find this kind of traffic for an oral history
collection, but the principle is the same. I'm just really glad that I
didn't have to process each one of these requests individually!

David Seubert
UCSB


1856-2006 - Musée canadien des civilisations / 1856-2006 - Canadian Museum
of Civilization

150 ans de culture, de collections et de découvertes / 150 Years of
Knowledge, Collections and Discovery

Journée portes ouvertes le dimanche 14 mai 2006. / Open House on Sunday May
14, 2006.


Ce courrier électronique est confidentiel et protégé. L'expéditeur ne
renonce pas aux droits et obligations qui s'y rapportent. Toute diffusion,
utilisation ou copie de ce message ou des renseignements qu'il contient par
une personne autre que le (les) destinataire(s) désigné(s) est interdite. Si
vous recevez ce courrier électronique par erreur, veuillez m'en aviser
immédiatement, par retour de courrier électronique ou par un autre moyen.


This e-mail may be privileged and/or confidential, and the sender does not
waive any related rights and obligations. Any distribution, use or copying
of this e-mail or the information it contains by other than an intended
recipient is unauthorized. If you received this e-mail in error, please
advise me (by return e-mail or otherwise) immediately.=