What is the standard "attributes" for an MP3 file?

[happydog500]

I do not use MP3. I recorded a 3 day convention and want to give a copy to a friend. I used MP3 since it's just talking (no music), plus I can put it all on one disc (8 if I use .WAV).

It would not play on her standalone home DVD/CD player, even though it had a "MP3" sticker. She did get it to play in her computer.

I checked the "attributes" I can convert the files to MP3. I have a choice anywhere from, 8 kBit/S, 8,000 HZ to 56 kBit/S, 24,000 HZ.

What is, or is there a standard MP3 attribute? What would be the best one to use for compatibility?

Thank you,
Chris.

P.S. Her DVD/CD player shows the titails of the files but doesn't play them.

[mats.hogberg]

There's no std, but I'd dare say 96 kbps stereo, 44.1/48 kHz should be the minimum. Seems like some players don't cope with low bitrates. If anything is std, 128 kbps would qualify. As it's just speech, low bitrates can be used.
I think you need the lame acm codec (or lame.exe) to encode @ higher bitrates than the std MS codec allows.

/Mats

[happydog500]

I don't have any of those numbers I can use in winamp to make MP3 files from my .WAVs.

[mats.hogberg]

Like I said, you prolly just have Windows std (Fraunhofer) mp3 codec, and that is severely limited.
Get something like CDEx (and lame) to get a wider range of MP3 options.

/Mats

[guns1inger]

Unfortunately, due to licensing issues, Windows ships with minimal MP3 encoding support. If you download and install the LAME MP3 Encoder you will get the full compliment of MP3 encoding levels.

Edit : Damn you, missed by less than a minute !

[mats.hogberg]

wav to mp3 can easily be made with Audacity too, but you still need lame_enc.dll

/Mats

[happydog500]

I don't need a wide variety. I was thinking that maybe her cd player can't read it because it was outside of what it would support. I wanted to get it to as close as a "standard" as possible, which winamp will do without downloading another piece of software.
Thank you for the replys,
Chris.

[turk690]

When creating an MP3 CD meant to be played on a DVD/CD player that supports the feature, that CD should be a true-blue ISO9660 and/or UDF-bridge CD-ROM that is duly finalized (closed). In some recording programs (like Nero), "finalized" is NOT the default selection when creating a CD-ROM. "No Multisession" should ALWAYS be checked; otherwise some players refuse to recognize the created MP3 CDs. This is also why it's a bad idea to use incremental recording programs (like the bothersome InCD that comes with Nero, or using Windows built-in CD recording service to drag-&-drop files to the CD-ROM drive icon) because they DO NOT finalize by default; it has TO BE deliberately done. Lastly, some players do not recognize directories more than a few deep (maybe 2) so will play only MP3s in the root or as far as the recognized directory level, and not more, or will downright refuse the CD altogether.

[happydog500]

Thanks for the reply. The CD has 30 tracks and it lists the name of all 30. I thought since it sees that, it should work (if it doesn't recignize it, it won't see the tracks).

One question I am looking for is if I have settings from 8 kBit/S, 8,000 HZ to 56 kBit/S, 24,000 HZ and EVERYTHING in between, what is one that is used most? There has to be some kind of standard for an MP3 Disc.

Chris.

[Midzuki]

Well, I have an LG 7352N, and according to its manual
it accepts MP3 files with bitrates between 32 kbps and 320 kbps,
but necessarily sampled @ 44.1 kHz. It can list VBR files, but
will not play them.

===

[Cornucopia]

"You're wanting to cross the Atlantic on 1/2 the usual amount of gas..."

IOW, you say you're willing to get some additional features that you might need, but don't want to download much. You've been told what could get you all the way there, but that seems to be TOO much for you.

Maybe this can clear things up:

"MP3" = Mpeg Audio, Layer 3 (yes there are layers 1 and 2 in use also)
There is an MPEG1 Layer 3 (defined by ~64kbps-320kbps, with 32kHz, 44.1kHz and 48kHz sampling)

and then there's the MPEG2 Layer3 extension (defined by ~8-48kbps, with 6, 8, 11.025, 12, 22.05, and 24kHz sampling).

Players that understand MPEG2Layer3 will also understand MPEG1, but not the reverse.

My guess is that most hardware players (especially the cheaper ones) use only MPEG1 specs, and so will IGNORE an MPEG2Layer3 file. Even if the filesystem "sees" it.

************

Go the whole 9 yards and DL the LAME codec, and you won't be sorry. It'll get you into the ballpark you probably need to be (~128kbps, stereo, 44.1kHz for music, ~48-64kbps, mono, 32kHz for speech).

Scott

[jman98]

Please don't use sampling rates less than 44.1 kHz. Using anything less won't save you any space and it will make the files sound worse. Why would you want to make something sound worse and not save any space?

96 Kbps and mono should be fine for a voice MP3 and I'd definitely use 44.1 kHz for this. Using mono will save you space over stereo.

[Cornucopia]

Actually, using 32kHz for SPEECH might make the file sound better. Bear with me...

Speech freq. response has upper limit of ~1/2 that of music. Everything above is "random/noise". Stuff lossy compression has trouble with.

For a given data bitrate, a "coherent" signal makes better use of lossy compression than an incoherent signal. Less "overhead" means better use of what's left. This is particularly true for (and similar to) VBR encoding.

Here's another similar trick: Take a recording that has speech and put it in an audio editor. At places where the speaker stops (sentence breaks, etc), replace with same length of silence. Encode BOTH the original and the "silenced" version with same bitrate compression (and same other settings)--the "silenced" file will always sound better.

jman98, I am with you as regards compatibility though--44.1 and 48k are much safer. And for hardware decoders that may be "optimized" for only 44 and 48, 32 may be hindered in its reproduction quality.

Scott

[olyteddy]

I transcode my Audible books using TotalRecorder and use 22 KHz sampling and 48 Kbps, Mono. That profile is close to the stream rate that comes out of the Audible player and sounds OK in my car.

[turk690]

to happydog500, as a direct answer to "what is the most used", then this has to be a constant bitrate of 128kb/s and a sampling rate of 44.1KHz. 44.1KHz is the most common sampling rate because it happens to be that of CDDA, where most people's (legitimate or otherwise) music files originate. 128kb/s is the most common constant bitrate because with current popular music, this is the sweet spot, i. e., most people can perceive lower bit rates starting to sound subpar, and higher bitrates than this not really making the resulting MP3 file sound any closer to the original than it already is. But as in most things else, MP3 is not a one-size-fits-all world. I have converted my CD collection to MP3 using 160kb/s; to me, this bitrate brings the MP3 a tad closer to the original on some complex files. For instrumental & classical (opera), I use 192kb/s. Unless it is purely speech (where a sampling rate of 32KHz and a bitrate of <96kb/s may make sense), you should choose 128, 160, 192, & 224kb/s at either 44.1 or 48KHz. It's not hard to perceive the deteriorating quality with lower bitrates, and the whole point of using MP3 compression & saving space is quite moot when using higher bitrates. Also, MP3 players of whatever media seem to behave best with these specs.