poor quality export

[shinnen]

Hi,
I copied a section of my voice. It sounded find when I played it in audacity, but when I exported it to mp3 it has a sort of crackly sound, but only in the voice, not in the empty spaces. It was as if part of the voice was distorted. Anyone have any idea what might be causing this?
Thanks,
.... john

[steve]

If the original recording was slightly "clipped" (recorded at too high a level causing distortion), then after compressing to MP3 the distortion is likely to sound a lot worse.

[shinnen]

Thanks for replying Steve.
I suspect that this is what happened.
So what do I adjust to change this..... the little input volume slider (- <<<<<<< +)?
Roughly were should it be?
I guess there's no 'easy' way of removing this distortion? I like the recording, but can make another.
..... john

[steve]

When you make the recording you should see a red recording meter flickering up and down with the sound. Ideally you want the level to go reasonably high, but never all the way to 0dB (max). In Audacity 1.3 (not sure about Audacity 1.2 as I've not used it in ages) you can grab the recording/playback meters with the mouse and drag it out of its "docking" position, then stretch it to full screen width (and optionally "dock" it back in the main Audacity window) which makes it much easier to read.

There are various ways of adjusting the recording level depending on what your equipment set-up is - I usually use the Windows Mixer (double click on the loudspeaker icon near the clock, then look for the "Recording settings" section).

If, after you have recorded, the level looks a bit to low, you can use "Normalize" or "Amplify" to boost the level. I would always advise leaving a little "headroom" and only boost the level to around -0.3dB rather than to 0dB.

Repairing clipped audio is fiddly and often not very successful, so re-recording is probably the best option.

[shinnen]

Thanks again Steve,
I tried it once more, after reducing the recording volume, and still had the same problem. But I did notice two things.
1. The recording is a mixture of speaking and chanting; and the chanting is where I see the distortion (after
coverting to mp3).
2. When I convert to wav, instead of mp3, there is NO distortion. Unfortunately the wav file is about 10 times the size of the mp3 and I can't send it via email, with splitting or some other means.
Is there a way around this problem?
...... john
P.S. I think I've solved the problem. I exported the recording to ogg, and, low and behold, NO distortion.
How widespread is ogg. Is it a good default choice for compressiong files, better than mp3?

[steve]

Could you export a short section of the "chanting" part as a WAV file (select a few seconds, then use "Export Selection" from the File menu), then upload the file to a file sharing website (for example http://www.yousendit.com/ ) and post the link to the file here. I'll have a look at it and see if I can tell what is going wrong with MP3 export.

How widespread is ogg. Is it a good default choice for compressiong files, better than mp3?

Ogg is among the best audio file compression formats. All computer platforms are able to support, or at least convert Ogg files, though I think that it is only Linux that fully supports Ogg by default. Some Windows programs do not support Ogg, but Ogg to WAV conversion programs are readily available (free) for Windows, Mac and Linux.

Some MP3 players support Ogg files, but I think that the majority do not. For personal use, Ogg is an excellent choice, but for distributing on the web MP3 is possibly a better choice because of its widespread use. If I am sending a file by e-mail, I will quite often send it as an Ogg file, partly for the better sound quality, and also to help "spread the word".

For Windows users, Foobar2000, WinAmp and VLC are among the excellent free media players that support playing Ogg Files.

[shinnen]

Hi Steve,
Thanks again for your expertise.
I should explain that I wanted to sent this to a friend in Holland. I sent her the ogg file and she said that she couldn't open it with Windows Vista. So I sent her MPC HomeCinema and told her to install it. Well....... to make long story short... she's not very computer savy. Then I discovered from a search, that CDex (I think 'you' orginally put me on to) will convert wav to mp3. (My thinking was that maybe Audacity doesn't do so hot a job of exporting to mp3). Well.... it turns out that CDex does a good job of converting wav to mp3 so .... problem solved.
Thanks for the media players suggestion. I had VLC. It was good, but my hard drive crashed and burned and I haven't reinstalled it. I did try Foobar, at the suggestion of someone here, but I just couldn't get the hang of it.
I'll let you know when I upload the file segments for your examination.
...... john

[shinnen]

Hi Steve,
Here are the files. It's the gravelly sound in the export, ChantingMP3.mp3 that I don't like. I've also sent you the aup and data files, the wave export (chantingwave.wav) and the ogg export (chantingOgg.ogg).
http://shinnen.ca/ChantingMP3.mp3
http://shinnen.ca/ChantingOgg.ogg
http://shinnen.ca/ChantingRaw.aup
http://shinnen.ca/ChantingWave.wav
http://shinnen.ca/e0000d78.au
http://shinnen.ca/e0000d78.au
http://shinnen.ca/e0000fa6.au
http://shinnen.ca/e0000ff1.au
Please let me know if any of these fail.
.... john
P.S. I've already sent these to Koz via email.

[kozikowski]

First thing, the channel is distorted all the way through, not just during your presentation. I amplified the "silent" portions...

http://www.kozco.com/tech/audacity/ChantingWave2.wav

So they're not so silent after all.

You used some magic words at the beginning. You "copied" your voice to Audacity? You did not perform live into a microphone and capture in Audacity? Where did the voice come from?

Koz

[steve]

There are a peculiar mixture of characteristics in the sound audio that you are encoding. Whilst the sounds are to a large extent lowish frequency (with a couple of very prominent peaks in the frequency spectrum), there is also a fair bit of noise (as indicated by koz). Also, where you have cross faded voice samples, the frequencies are "interfering" with each other, causing low frequency undertones.

The compression rate that you have used for the MP3 is relatively high (low quality) and at a constant bit rate. I think that it is a combination of all of these things that is causing MP3 compression to perform surprisingly poorly. If you use VBR and "joint stereo" for the MP3 encoding, I think that you will find that the MP3 quality is very much better.