I am curious now because I was having a discussion with a friend. I know for a fact that FLAC and APE definately sound better than mp3s, but seems to me that wav is a lossy audio format.

As far as I can tell from Googling, Wav isn't lossless, so i'm curious as to how accurate a lossless audio file is in relation to the original source.

.wav is the file extension given to PCM data. if you rip a cd without using any compressor, this is the file format you get. Hence, for all intents and purposes, .wav's are lossless. Monkey's and Flac and all other compressors use them as their input.

Wave files contain exactly the same data as that found on a CD (Raw PCM) plus of course some metadata (so that it can be saved as a file). They are therefore lossless with respect to the original CD.

All wrong!

WAV is a container. You can throw almost anything inside it: MP3, ADPCM, Vorbis, GSM, AAC, Speex, Acelp.net... and PCM, of course.

Just because the most usual content of a WAV file is PCM audio, which is uncompressed, doesn't mean it's always that content. And it makes no sense labeling a container lossy, lossless or whatever.

Just to confuse things, you can use a .wav extension to play a lossy file if it has an ACM wrapper around it. So for example, if you encode a file using the FhG/Radium ACM codec, you can actually play the file using either the .mp3 or .wav extension.

In general, though, when people refer to the wav file, they are talking about the uncompressed PCM data resulting from something like a CD rip.

ff123

way to confuse the guy! I don't think he wanted/needed the complete answer

I don't think he wanted the wrong answer either

Hi.

If you ask me, I'd say that normally .wav files are just raw PCM files.

To make an analogy, think of .bmp files. They are 'bitmap' files, which describe pictures by painstakingly writing down the colors of every single pixel in an image. The quality of the image depends on how acurate the colors are, and how much width/height is being used to represent the image.

Consequently, .bmp files are uncompressed raw images.

I'd say .wav files are similar, because .wav files are normally 'PCM' files, which describe sound by painstakingly writing down the amplitudes of the sound at each moment in time. The quality of the sound will depend on how acurate the amplitudes are, and how often the amplitudes are sampled.

In both cases, if you do the "sampling" right, then it's really difficult to tell between the original source and the digital "copy."

So are they lossless? I can't really say. I would say they are just 'raw.'

Comparing bmp to wav is plain wrong. BMP is simply a bitmap image format, WAV is an audio container.

The only graphical format you could use to compare to WAV is TIFF.


BTW, I wouldn't say they are "RAW" either. I would say they are uncompressed.

rjamorim: OK, OK.

Then what about comparing BMP and plain PCM? Would that be more apt?

Almost. The problem is that BMP can use compression (runlenght)
http://www.faqs.org/faqs/graphics/fileform...section-18.html

Oh my sainted aunt!

OK, then I definitely should have said something like "bitmap vs. plain PCM!"

How is that possible? If you put a vorbis stream inside a wav container, will it need the vorbis codec to play? And what will the filesize be? Same size the file inside it? Bigger?

Also, i heard the .mp4 container will support video as well... and objects (whatever that is...). How will you know what will popup when you double click a .mp4 file?

Maybe audio... maybe video... You'll have to click it to find out?

edit: typo

It will be equal to the size of the raw data contained within the WAV container (be it PCM, MP3, Ogg Vorbis or anything else), plus the size of the WAV header. A canonical WAV header is 44 bytes, although it may be extended to include other descriptive information.

- M.

Exactly. There is a vorbis ACM codec in some japanese page.



A little bigger, due to the wav header. Very little, really.



Yes, it will support video, and objects like VRML. It will also support scriptable actions (buttons, checkboxes, links...), still images (that you can do crazy things with as well)... Ever played with those MOV trailers that have all sorts of wacky features? MP4 has similar features, specially because MP4 was based on Apple's MOV technology.



You don't know, but blame Windows on that, since it doesn't check files by their contents (like Unix), it checks by their extensions.

For that reason, there's a trend going on of renaming audio-only MP4 files to M4A. This trend was started by Apple.



On Windows, yes. Unless someone comes with an Explorer shell that allows you to see the contents on the file properties page, or when passing the mouse over the file, or something like that.

Regards;

Roberto.

And how exactly do you know this?

The HA predilection for finely splitting hairs is interesting sometimes. Here's another approach:

Most people refer to shn, flac and ape as lossless because the original wav file can be completely reconstructed from any of those formats. So you're rather confused if you think flac or ape is better than wav (in the way 99.99999% of people refer to wav).

If you are a live music collector/trader, you want wav, shn, flac or ape because
- lossy formats (such as mp3) aren't tradeable
- lossy formats become obsolete (as the sweet spot for bit rates change and the encoder technology changes)

It's not especially useful to refer to wav files as lossless or not, as they are the baseline against which the others are measured - cd audio.

However, if you wanted to really split hairs - and this would be a more interesting set of hairs to split, I'd say - cd audio is obviously a somewhat lossy encode of the original studio masters. It's just we have all accepted that. But that's not really a view that comes up in practice.

i think this could be another possibility for new buffer overflow exploits on windows

-andy-

Thank-you for the insightful comment. >_<

I used to have one that if you passed the curosr over a mp3 file it gave all the info, such as bitrate, filesize, filename, tag info. have to think of what it was. it DID become a pain since the box was rather large when looking through a collection, and if renaming it hid the area. so i deleted it.
here it is
http://www.mutschler.de/mp3ext/

"MP3Ext is an extension for the Windows Explorer.
It enhances it, so you can get information in many ways:

A PropertyPage is installed, where you can fully edit the ID3-Tag
An icon-handler for individual icons is included too
Tooltips, when you simply leave the mouse over the MP3-file (only with IE4.0 Active Desktop)"

the icons were intersting since it would display different icons according to the bitrate per mp3. so by glancing through the folder one could tell the 160Kbps from the 192kbps, just by the icon it was assinged.

Yeah, I use this as well. I just don't install the mouseover info box part. All I use is the icons that display the different bitrates which is grand for organizing my mp3s.

It's a good job Roberto isn't a huge fan of Vorbis, if this is anything to go by, I expect he would hunt down and kill anybody who referred to it as ogg..

dBpowerAMP will do that too, for any audio file type it's registered with. I've done it with Ogg, MPC, MP3, WAV and FLAC so far. In addition to all the file and tag info, it also tells you the tag type(s) as well, which is kind of handy.

I had used dBpowerAmp in the past. but i stopped when it was not as up to date with oggenc as i would have liked it to have been. (3 years or so ago).

but it doesnt offer the sorted icons that mp3 ext does. which is a nice visual. or has that changed?

>but it doesnt offer the sorted icons that mp3 ext does. which is a nice visual. or has that changed?

It does not do icons for bitrates, but there is nothing to stop you using both programs together (obviously only one would supply the popup, which ever was installed last).

Actually, Monty himself wants people to say only Ogg instead of vorbis. If you search, you can find a very funny post where he explains why he wants people to only call his codecs "Ogg".

I personally don't use Ogg for everything that comes from Xiph, for the obvious reason that Ogg is not an audio codec, so it might end up leading to confusion. To each it's own name...

You could compare wav with avi ! Because avi also is a container - the video can be uncompressed or divx or whatever you want! Also you can throw wav into the avi-container which may be pcm or mp3 or whatever you want!
Praise me rjamorim !

All wrong!

Technically, there is no WAV container. RIFF is the container type. WAVE is the format field, and WAV is the extension.

And then he wonders why corporate stiffs don't take Ogg OR Vorbis very seriously. Way to complicate things further!!

Rjamorim, I did a search and couldn't find much. Do you have a link for that? Or have we gone off-topic enough?

MP3 as a lossy format, I believe, removes some high frequency data and, at least in my opinion, occasionally muddles the bass and midrange. This is, of course, subjective, and that opinion must be considered "to my ears", "on my equipment", and the like.

Besides this, wow, thanks for all the replies. I didn't think my first post here would hit 2 posts let alone 2 pages.

Anyway, the nature of the debate I had with my friend was whether wav could fully reproduce exactly what was on CD data as it's a digital -> digital conversion which, in theory would mean there is no need for loss. The debate started because I felt that wav was in fact, lossless (at least for my intentions), otherwise, there would've been no point for people to bother creating a lossless audio codec because data would already have been lost in the transfer from the CD.

On this same line of discussion then, what is the best method for ripping a wav to be as close, if not exactly, the same as the data on a CD? Would there be any noticeable difference if I went to Soundforge, ripped it to 16/44.1k files and had Soundforge resample and convert up to 24/48k or 24/192k? If not, is there something that actually rips to a wav at the 24/48k+ level?

There is always a potential to hear a difference at that level. By increasing the number of bits as well as the sampling rate, you will ironically be performing a lossy conversion.

The best way to rip a wav exactly the same as the CD is to use a program like Exact Audio Copy. The Wav output will essentially be the same as what's on the CD. I say essentially because if your CD has errors, then EAC will attempt to compensate for them.

In fact EAC tries so hard to insure perfect wavs, it can do things that us non-obsessive people would find silly (like another thread on how someone didn't understand why their CD quality was 99.9% rather than 100%).

As for something ripping a wav to 24/48, I don't know if there is, because almost all CDs are encoded at 16/44.1. Any attempt to rip a wave like that could potentially be worse than 16/44.1 and at best, will not sound better

You cannot rip a CD higher than 16/44.1. Anything higher is just resampling, and losing quality (though likely undetectably).

True, but i dont want a whole big list of shell ext's.
The reason's i had stopped using your very good product was at the time, then some 3 years back. it did not work well with Large fonts, and i was not able to include my own preferred ogg Serial number scheme.
which then, it was important to serialize all of them in a set manner based on certain rules. and dBpowerAMP had no option for changing the serial #
( might have changed now ? - i have used it in some time)

I have set my younger brother up with dBpowerAMP, and he uses it alot of the times now to convert his oggs.