MP3 or OGG?

Ubu · Post by **Ubu** » March 1st, 2009, 12:00 pm

Why I should download ogg files rather than mp3? 'Like the sound of "Ogg" but that's no reason to switch over now is it.

Can you give me some definitive help on this? Or is it six of one, half a dozen of the other?

icyjumbo · Post by **icyjumbo** » March 1st, 2009, 1:21 pm

I have no idea what the answer to your question is, but it occurs to me that the person who can best answer it is you yourself.

Download both Ogg and MP3 versions and see which you like better. Why do you like one better than the other? Is that true for every MP3 vs Ogg pair you try? If so, you've learnt something useful to you, and nobody can tell you that you've got it wrong. They might tell you that there was something you hadn't considered, but until they provide you with MP3 and Ogg samples that illustrate the difference, it won't matter to you. In fact, quite the reverse: you will already know that in all the cases that are important to you the choice you will have made will have been the right one.

If this sounds like the scientific method, that's because it is.

harvey · Post by **harvey** » March 1st, 2009, 2:59 pm

At the Wikipedia, see Comparison of audio codecs and Codec listening test.

I get my complied binary of LAME from the JT HZ Web site. They are
members of the Audio Engineering Society and have a page titled
Creating quality audio-files using Windows, L.A.M.E., Ogg Vorbis.

strogon14 · Post by **strogon14** » March 7th, 2009, 9:13 am

I wonder how the OGG files for LibriVox recordings are actually created. Since contributors are supposed to upload 128 kps MP3 files (in most cases?), I can only surmise that the OGG files are created by decoding the MP3 files and then encoding them as OGG Vorbis again.

Is this correct? If this were the case, the quality of the OGG files, though the format in theory allows better quality at the same bitrate as MP3, would actually be inferior to the 128 kps MP3 files and I don't really see the reason for having them.

Can somebody more knowledgeable about the LibiriVox publication process perhaps shed some light onto this?

Chris

Boomcoach · Post by **Boomcoach** » March 7th, 2009, 9:20 am

The Internet Archive (where the finished files are stored) will automatically create the ogg, 64 bit MP3 and streaming files when you upload a 128 bit mp3.

strogon14 · Post by **strogon14** » March 7th, 2009, 9:41 am

Ok, thanks for the info. Seems a sub-optimal process to me. But the only solution would be to allow upload of lossless encoded files like WAV or FLAC and create the MP3 and OGG files from them. And that would mean increasing the storage requirements at least sevenfold.

Steampunk · Post by **Steampunk** » March 7th, 2009, 12:00 pm

Keeping in mind that we're talking about pure voice recordings (usually) produced on non-professional equipment in (usually) non-studio environments. The difference between ogg and MP3, even with a few derivations one way or the other, is inconsequential.

In my opinion, if you have an ogg device that you're happy with, then ogg is fine. Likewise, if you have an MP3 device you're happy with that's ok, too.

Jim

strogon14 · Post by **strogon14** » March 7th, 2009, 12:36 pm

Steampunk wrote:Keeping in mind that we're talking about pure voice recordings (usually) produced on non-professional equipment in (usually) non-studio environments. The difference between ogg and MP3, even with a few derivations one way or the other, is inconsequential.

In theory, if the original material is encoded with comparable settings to either MP3 or OGG, the quality shouldn't differ much (although the OGG file should be slightly smaller at the same nominal bitrate). But in practice, I noticed that the OGG files at LibriVox have very often much lower nominal bitrates than the MP3 files at 128 kps. I have observed nominal bitrates at 80 and 112 kps. While the latter might be comparable to MP3, the former is definitely not, even for speech-only recordings. This sheds a bad light on an otherwise superior format, IMHO.

I can't check all the files, but I wonder, what is the standard for encoding the OGG files (i.e. bitrate, qulity setting, etc.) and if this can be influenced in any ways by the uploader? Or is it possible by a contributor/reader to provide his/her own OGG file, which was encoded from the original recording?

Chris

Post by **chocoholic** » March 7th, 2009, 12:42 pm

strogon14 wrote:I wonder, what is the standard for encoding the OGG files (i.e. bitrate, qulity setting, etc.) and if this can be influenced in any ways by the uploader? Or is it possible by a contributor/reader to provide his/her own OGG file, which was encoded from the original recording

Archive.org (where our cataloged files are hosted) does all the file derivations, including the OGGs. They require that we send them 128 kbps mp3's, then they take care of the rest. So this would be a question to ask them.

harvey · Post by **harvey** » March 7th, 2009, 1:59 pm

strogon14 wrote:In theory, if the original material is encoded with comparable settings
to either MP3 or OGG, the quality shouldn't differ much.

Here's a different view, from members of the Audio Engineering Society
(see my post above):

Ogg Vorbis is more efficient than MP3 and is the state of the art
in audio compression technology. Used properly, recent Ogg Vorbis
encoders deliver sound quality surpassing MP3 with all possible
bitrates.

strogon14 · Post by **strogon14** » March 7th, 2009, 2:07 pm

Well, if you read my posts carefully, I always said, that OGG is superior. With "comparable settings" I meant settings, that yield the same results. To put your quote another way, the same quality can be achieved through the use of OGG Vorbis compression at lower bitrates (and therefore yielding smaller files) than with MP3 encoding.

BUT, this will only be the case, if the OGG file is created from the ORIGINAL uncompressed recording (be it in WAV, AIFF or any other lossless format), NOT from an already lossy-encoded MP3!.

Chris

RuthieG · Post by **RuthieG** » March 7th, 2009, 3:28 pm

strogon14 wrote:BUT, this will only be the case, if the OGG file is created from the ORIGINAL uncompressed recording (be it in WAV, AIFF or any other lossless format), NOT from an already lossy-encoded MP3!.

In an ideal world etc etc... we would, I am sure, love all our recordings to be lossless. However, we are dependent for our file hosting on the Internet Archive, which demands 128 kbps files. We also have to store all our 'works in progress' on our own server, which does not have limitless space, and each of us has limited time (and probably limited bandwidth) to upload our recordings.

So it's not an ideal world and everything seems to work pretty well, considering.

Ruth

strogon14 · Post by **strogon14** » March 7th, 2009, 3:35 pm

I see, though, that some projects accept uploads in FLAC format (including projects where you, RuthieG, are participating). Is this decision left to the project coordinator or how do you go about this?

RuthieG · Post by **RuthieG** » March 7th, 2009, 3:39 pm

.wav or .flac are sometimes requested in dramatic works projects where the individual files are to be edited together to make the whole production. The final file will be uploaded as usual as an MP3.

Ruth

Post by **hugh** » March 7th, 2009, 6:26 pm

the other advantage of ogg vorbis is that it's an open, patent-free codec.

whereas there is a patent on mp3, and, in theory, the owner could start charging licensing fees to anyone using mp3 to distribute content. unlikely, of course, but impossible with ogg vorbis.