Preparing Chapters for an Audio Book

I often use N-Track to prepare the tracks for audio CDs of lectures I give on local history. Below is the proceedure I use. I wonder if there is a better way.

1. I record the lecture and, at my computer, download it from the device and convert it to one long .wav file.

2. I run N-Track, use “File/Settings/Preferences” to set the sampling rate to match that of the .wav and then use “File/Import wave file” to load the lecture as a single track.

3. I set the master volume and master channel effects to give the best sound quality possible.

4. Working through the big track, I cut out unwanted material and also cut a very small section each place where I want a chapter break to occur on the finished CD.

5. I zoom way out until I can see the entire track and then drag the chapters created in step 4 down to create a track for each chapter.

6. I mixdown each track individually to a final .wav. There will be one .wav per chapter on the final CD. This takes a few minutes, but I have found no way in N-Track to say, “Mix down each track to its own wave file.” Also, I want the finished waves to be in mono, and have to be careful to keep clicking this option each time I enter the mixdown menu.

7. When the mixdown phase is complete, I save the project, get out of N-Track, and then use burner software to assemble the tracks and create the final CD.

How’s that sound? Is there a better way to go?

Hi, In some cd burning software e.g Nero you can open up a long Wav file get a waveform view and put markers in for track breaks (with or without a 2 second gap between tracks). In which case you could do your editing and level control in N-track then mixdown to one long Wav file. Then open that in Nero and put in the track split points.

Nick

I have a Python program that splits a long track into sections divided by silence, one wave file per section. It’s a DOS command-line program – no GUI – but if you’re interested I could provide it for you. You’d have to install Python, which is a scripting language similar to Perl.

Let me know if you’re interested. My program does a lot of stuff you wouldn’t be interested in (I wrote it to chop samples up for making soundfont instruments), so I’d have to modify it for your purpose. But it would be a handy general purpose utility I could post on the web for anyone to use.

If your original wave file was “foobar”, it would chop it into “foobar_1”, “foobar_2”, etc. You could specify on the command line the amount of silence for cutting and the decibal level to be considered silence.

Cheers
Jeff

To easily split long wav files, I regularly use a program called CDWave (http://cdwave.com). I use it to split long church service recordings, before I burn them to CD. Any bits not required can be ‘unticked’ after the ‘split’ marking process. If you remember to rename the bits sequentially, you can ensure that the finished CD retains the files in the correct order.
By the way, I usually normalise (in n-Tracks) the spoken parts to - 3 db, and the singing/musical parts to -1 db: this seems to sound more realistic than normalising all tracks to the same level. n-Tracks is also used to ‘trim’ the start and end of each track; music can also be faded-out very effectively. Remember to save the .sng file before mixing down!
A note about normalising: if, after normalising, the file seems to have the ‘wrong’ amplitude, it is probably because of one or more large transients, which may not be seen when the wave file is being displayed in full. If you suspect that this is a problem, zoom-in to the beginning of the waveform, and then spend a bit of time working along it to the end. If you find a few big transients, and they can’t be removed, I usually select the part of the wavefile BEFORE the faulty bit, and then normalise it first. The rest of the file, AFTER the fault, can then be normalised to the same level. Hope that this helps.