Here’s a quick companion post to the previous one with some raw technical information on the file formats for the Bose Sleepbuds.

Each sound in the library needs a collection of background images, thumbnail images, a preview WAV and separate raw mono audio files for the Left and Right Sleepbuds.

Assuming we want to make a new sound called Sound Name then this is how the background files and thumbnail files should be names, and what size they are.

We need a Carousel Image, a Blurred Carousel Image and a square format Thumbnail

Filename	X Res	Y Res	Scale
Sound%[email protected]	563	1001	0.375
Sound%[email protected]	375	667	0.25
Sound%[email protected]	750	1334	0.50
Sound%[email protected]	1125	2001	0.75
Sound%[email protected]	1500	2668	1.00
Sound%[email protected]	563	1001	0.375
Sound%[email protected]	375	667	0.25
Sound%[email protected]	750	1334	0.50
Sound%[email protected]	1125	2001	0.75
Sound%[email protected]	1500	2668	1.00
Sound%[email protected]	255	255	0.375
Sound%[email protected]	170	170	0.25
Sound%[email protected]	340	340	0.50
Sound%[email protected]	510	510	0.75
Sound%[email protected]	680	680	1.00

Then there’s the audio. First is the preview WAV which in most examples is stereo, 32-bit floating point, 44.1 kHz sample rate. Then each of the L and R channels are extracted, down-sampled to 8 kHz, decimated to 24-bit and saved as a raw audio file with no header information.

The preview WAV is saved as Sound%20Name.wav and the raw audio files are saved as Sound%20Name-Left.bin and Sound%20Name-Right.bin

I was using Adobe Audition to create/edit the audio files. I couldn’t see any way to export raw audio, only WAV etc. What I had to do was export each of the Left and Right channels as WAV and, use sox to convert to raw and then rename as .bin – e.g:

sox Sound%20Name-Left.wav --bits 24 --rate 8192 --channels 1 Sound%20Name-Left.bin