Computer Audio Asylum

Technical suggestions on getting the best sounding results

65.19.76.104

In Reply to: RE: You'll have to ask a MS/Audacity expert posted by VinylNewbie on April 15, 2008 at 06:36:41

I don't know the details of how to use Audacity, as I don't use it. But I do have some experience editing with Sound Forge and the proper sequence to do format conversions so as to preserve the maximum amount of aural information. (I have been doing cassette and open real digital transfers for the past several years.)

What you want to do is convert the input file to 24 bits or (better) 32 bits, and do all of your processing, whatever it is. The very last step of your processing should then be to convert the resulting information back into 16 bit format, using the appropriate dither function, which would generally be triangular dither and possibly with noise shaping. (I don't know what the options are with Audacity. You may want to experiment with different dither methods, as the best results may vary depending on the recording. If you can't hear any differences, don't worry, but in that case you would best go with triangular dither with no noise shaping as it will be free of distortion, if not the lowest noise.) The reason for doing the reduction to 16 bits last is so that all of the earlier processing is more accurate, with the only time your processing is impacted by 16 bits in the very final step.

Note that if your original material was 16 bits and you are not doing any processing of the music but just editing the beginning and ending of the tracks, you may want to keep everything in 16 bit format rather than going to 24 bits or 32 bits. That will keep things bit perfect. However if you do so much as a level adjustment or a mix, then you want to be using 24 or 32 bits, as you will be doing processing. (If Audacity makes this choice for you automatically, then you would not have this option.)

One more thing: Depending on your soundcard and your software, you may get better results recording at a high rate (e.g. 96 Khz in 24 bits) and then reducing by software to 44.1Khz, or you may get the best results by having the soundcard do the sample rate conversion. (My guess is that with a 24 bit sound card the computer will do as good or better a job, but with a 16 bit sound card the soundcard may do a better job.) You will have to try it and see. If you are doing down conversion to 44.1 Khz in software, then this conversion should be done before reducing the word length to 16 bits.

Note that when doing a sample rate conversion it is possible for the signal peaks to increase, so if you are not careful a sample rate conversion may result in some peaks clipping. This would be a big No-No! You can check the final result to see if it clipped, and if so then you should undo the sample rate conversion, reduce the volume, and redo the sample rate conversion. If you are going to normalize levels of an album (or an entire track if that's what you want) you should do it after the sample rate conversion to 44.1Khz and before reducing the word length to 16 bits. This will preserve maximum sound quality.

Lastly, and I discovered this the hard way, if you are transferring music that plays continuously track to track (e.g. a live recording) you will need to take control of your burning software and prevent it from adding additional lead in or lead out silence. In general, you will get the best sounding transfer if you don't add any digital silence between tracks even if the music is not continuous. That way, the surface noise or tape hiss will be constant and less noticeable.

Once you take control of the track to track spacing, then you are responsible for conforming to the Red Book. Each track on a CD must be an exact multiple of 588 samples. (A sample is two 16 bit words, one for the left channel and the other for the right channel.) 588 samples comprise a frame, and there are 75 frames per second. So you want to adjust the length of each track so that it is an integral number of frames in length. That will make it possible for the burning software to produce the proper result. You can test this process by arbitrarily breaking up a continuous track into two wave files and then burning a test CD. If you get it right, a regular CD player (or computer) will play from track to track continuously. If you get it wrong, you will probably hear a glitch.

If this sounds complicated and confusing, then it would be a good idea to practice until you are familiar with these concepts and how to use your editor. Then when you are comfortable, you can transfer the bulk of your albums. That way you will get the best results with minimum frustration. However until you have a lot of experience with your tools, don't expect to transfer 1 hours worth of music in anything approaching 1 hour. If you do, you will almost certainly become frustrated.

Tony Lauck

"Perception, inference and authority are the valid sources of knowledge" - P.R. Sarkar

Follow Ups:

great info., thanks! nt. - Tom Schuman 04:29:10 04/16/08 (0)

Post a Followup:

FAQ

Post a Message!

Forgot Password?

Moniker (Username):
Password (Optional):
	Remember my Moniker & Password (What's this?) Eat Me
E-Mail (Optional):
Subject:

Message:	(Posts are subject to Content Rules)

Optional Link URL:
Optional Link Title:
Optional Image URL:
Upload Image:
E-mail Replies:	Automagically notify you when someone responds.

To view your new posting or follow-up, click on the RELOAD or REFRESH button on your browser.