Something fishy with 8-bit export

[cpeterson]

Before I just defer to audition (which would require me to tear about the linux box and put my windows drives back in), can somebody tell me if I'm doing something wrong here?

I want to take 8-bit binary data, record to an analog source, and then re-record it onto the computer so I can use some software I've written to visualize the data loss.

I know that audacity upsamples everything to 16-bit PCM -- that shouldn't be a problem, so long as it can downsample it back down to 8-bit data.

But it can't; the exported file that's supposed to be 8-bit unsigned data is huge, way bigger than the original file!

I'm using the 1.3.3 beta for MacOS X, and here's how I'm doing my export:

First, I downsample the recording to the same sample rate as the source. So, even though the recording was done at 44100hz, I resample to 8k to match the original. 8.3 million samples in the original, 8.3 million samples in the recording. They line up.

Now:

file->export selection. Format: wav / aiff / uncompressed.

With these set for the options:
Format: other
Header: RAW header-less
Encoding: unsigned 8 bit PCM

Now, what I WANT audacity to do here is, for each sample, convert it to unsigned, reduce precision to 8-bit, and save it in a file with no header, no padding, no alignment.

Why, then, would the resulting file be 8,379k when the original 8-bit data was 1,512k?

[cpeterson]

Ack! Sorry for posting, I figured it out minutes later.

Audacity uses the Project Rate when it exports, not the sample rate of the audio track, even if there's only one track in the project.

Incidentally, here's what it looks like when you record an image to a cassette tape. Obviously I'm having sign errors (among other things) but this made me laugh.