First a comment, then back to the original problem: actually, the file sizes ARE
different... I have tested it, using a single WAV source (Mozart?s 25th symphony
in D Minor, if that is of any interest...). MP3 files are smaller than wma, and
ogg smaller than both the others. I used dBpower AMP, and it indicated the
supposed bit rate for the ogg files. All conversions using CBR, even for ogg;
the VBR files were smaller still, but I guess that was expected.
OK, I got the idea about bit rate, and I understand that it can't be
compared across formats. So, the big question now is, is there a way to compare
the results? Lets think of something like creating different music files, and
comparing them back to the original DIGITALLY; can you calculate in any way
which file is closer to the original? BTW, is this a good indication of sound
quality (a MP3 might for instance catch ALL the silence perfectly, thereby being
"closer" to the original in some way, even if it does not render a
good sound...)? How about dividing the analysis into different frequency levels?
Its not that difficult to imagine a way, its just the goode-ole square
differences. Calculate the differences between source and converted file, square
and add it up. The file with the lower sum is the best conversion... I
understand there is a problem there because of the bit rates (most of the
comparison would be between music on the original and nothing on the lossy
file). You could convert the lossy back to wav, but then results depend on the
lossy-to-WAV decoder.
Anyway, I guess you got the picture. ok, it depends on ear sensibility, the
price of your equipment, and so on. But in the end, its all zeroes and ones.
There SHOULD be a way to compare the results when using different formats/bit
rates/settings... I just can't find it anywhere!!
Rodrigo
----- Original Message -----
From: Graham Mitchell
To: Rodrigo Escobar Nunes
Sent: Tuesday, September 28, 2004 12:44 AM
Subject: Re: [Vorbis] Bit rates, files size
> Does anybody has any idea how to compare mp3, wma and ogg in terms
> of bit rate and quality? I mean, I have tested it and the ogg files
> are always smaller if you set the same bit rate as compared to the
> other two, and smaller still if u use VBR. But how about quality? Is
> a 160bps ogg comparable to a what in a mp3? The same bit rate, lower,
> higer?...
Looks like you haven't read my "Introduction to Compressed Audio with
Ogg
Vorbis". :)
http://grahammitchell.com/writings/vorbis_intro.html#just_say_no_to_bitrates
In short, you're trying to compare apples and oranges. A 128 kbps mp3
will be
the SAME size as a 128 kbps Ogg Vorbis file, and as a 128 kbps WMA, because
the size of a file is determined by its bitrate (and length).
And to compare quality, I'd need to know what encoder you're using for
your
mp3s. A 128 kbps mp3 encoded with the Xing encoder or BladeEnc sounds pretty
terrible, whereas one with the LAME mp3 encoder sounds fairly decent.
In any case, Ogg Vorbis beats mp3 at any bitrate. Vorbis audio at 96 kbps
sounds as good as most 128 kbps mp3s, in my opinion. Vorbis audio at 112 kbps
sounds slightly better than your typical 128 kbps mp3.
Vorbis audio at 160 kbps would be comparable to at least 192 kbps mp3, I'd
say. But again, it's like saying "my Porsche at 5000 RPMs is as fast
as a
Mustang at 6000 RPMs." Just like talking about RPMs when you really mean
speed is silly, it's silly to talk about bitrates when you really mean
sound
quality. Vorbis sounds better at a smaller file size. Bitrate is a measure
of file size, not quality.
WMA also sounds better than mp3, but then this isn't the WMA mailing list,
is it?
--
Graham Mitchell - computer science teacher, Leander High School
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xiph.org/pipermail/vorbis/attachments/20040928/2ab8b513/attachment.htm