Re: Sound file formats for journal (Al Bregman )

Subject: Re: Sound file formats for journal
From:    Al Bregman  <al.bregman@xxxxxxxx>
Date:    Sun, 16 Sep 2012 15:11:21 -0400

Dear list, What people seem to be forgetting in this discussion of wave formats is that the sounds that are being discussed are those that were used in experiments. Therefore, to the extent possible, the sounds should be uploaded and preserved in the exact form in which they appeared in the experiment. For example, if a particular version of MP3 was used, and its compression algorithm led to an artifact in the experiment, the reader would want to be able to discover this, by analyzing the signal in the MP3 form in which it was actually used, and could decide whether this might have influenced the result. However, in the ideal case, the sound accompanying the article would be a very high resolution recording made at the level of the transducers that presented the sound to the listeners. This would capture any distortions due to either the sound-producing algorithm or the sound reproduction system. But if this were required, it might be asking too much of the author. To summarize, If a 24 bit, 96 Kilosample/sec version of the nominal stimulus were submitted to illustrate an experiment in which the listener actually heard a lower resolution sound, this would be deceptive. Cheers, Al ----------------------------------------------------------------------------- Albert S. Bregman, Emeritus Professor Psychology Department, McGill University 1205 Doctor Penfield Avenue Montreal, QC, Canada H3A 1B1. Office: Phone: (514) three-nine-eight-6103, Fax: (514) three-nine-eight-4896 ----------------------------------------------------------------------------- On Sun, Sep 16, 2012 at 5:39 AM, Dan Stowell <dan.stowell@xxxxxxxx> wrote: >> The question is then articulated as what file formats should be supported >> for longterm storage, and which for transmission purposes? Transmission >> formats could / would change over time, but the 'raw' would always be >> resident in the files. > > As I noted before, the International Association of Sound and Audiovisual > Archives seems like a reasonable body to look to for the archival side of > this... their guidelines are here: > <> Here's the page where they > recommend WAV: <> > > Dan > > > > On 15/09/12 07:15, Kevin Austin wrote: >> >> A somewhat parallel discussion recently took place on the cec-cnference >> list stimulated by the announcement / speculation that Sibelius [music >> notation program] might cease development. To me, what became clear [again] >> is that the file format needs to remain "independent" of 'interpretation' in >> order to remain viable for the future. I have several boxes of disks with >> files which cannot be opened, some less than 15 years old. >> >> Given that bandwidth and memory are no longer major issues in computing, >> it seems to me that a 'raw' [equivalent] file format may be best to survive >> a couple of decades. The file, while large, would be independent of the >> program which would compress / expand it for transmission. mp3 was created >> to reduce file size in an era when bandwidth was an issue. I recall >> discussions as to whether the CEC [Canadian Electroacoustic Community] >> Newsletter and its jukebox [] should make files available in any >> other format than mono 8bit/12kHz sampling. The decision was to store the >> 'master' file as 16/44.1 stereo, and convert to 8/12 mono when the file was >> to be sent to the modem. [A high speed device that would transfer up to 2400 >> bits/sec.] >> >> With the later decision to store the files as 16/44.1 and stream at 320 >> kb/sec, the original files could simply be 'switched over' [or even made >> available] when the bandwidth problem was solved. For my classwork, the >> minimum acceptable audio format is 24/48kHz, with most students preferring >> to work at 32/96kHz, this in a way approximating a 'raw' format for audio >> >> >> The question is then articulated as what file formats should be supported >> for longterm storage, and which for transmission purposes? Transmission >> formats could / would change over time, but the 'raw' would always be >> resident in the files. >> >> Kevin >> >> >> >> >> >> >> On 2012, Sep 13, at 10:54 AM, Robert Zatorre wrote: >> >>> Dear list >>> >>> In an effort to enhance the Frontiers in Auditory Cognitive Neuroscience >>> journal, we would like to enable sounds files to be uploaded for reviewers >>> to be able to hear the stimuli used in a given experiment. Eventually we >>> would also like to have a means of having these sound files embedded >>> directly into the online journal article so that readers can hear the >>> stimuli used. (Of course this could apply not only to stimuli, but also to >>> other sound files that are part of the study, such as recorded >>> vocalizations, speech or musical sounds produced under some experimental >>> conditions, and so forth) >>> >>> My question for you all is what file formats do you think we would need >>> to support? The two obvious ones are wav and mp3, but perhaps there are >>> others that you may think are important or that have some advantages that >>> should also be considered. >>> >>> Thank you for your thoughts. >>> >>> PS feel free to send me your comments directly >>> >>> -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ >>> >>> Robert J. Zatorre, Ph.D. >>> Montreal Neurological Institute >>> 3801 University St. >>> Montreal, QC Canada H3A 2B4 >>> phone: 1-514-398-8903 >>> fax: 1-514-398-1338 >>> e-mail: robert.zatorre@xxxxxxxx >>> web site: > > > -- > Dan Stowell > Postdoctoral Research Assistant > Centre for Digital Music > Queen Mary, University of London > Mile End Road, London E1 4NS > >

This message came from the mail archive
maintained by:
DAn Ellis <>
Electrical Engineering Dept., Columbia University