Yes, that's all true, but not relevant (well, not to me).  I want the device to be operating in 32 bit mode, regardless of how Core Audio actually delivers the data.

Paul Sanders.

On 04/04/2018 15:09, Philippe Wicker wrote:

AFAIK CoreAudio don’t allow a user to get the raw samples from the device, it always provide audio data with a 32 bit float format (single precision). Device native PCM data are converted at some point (within the driver? Or maybe in the HAL?) to this float format before being delivered to the user. It’s also true in the other direction, you have to provide 32 bit float to the device. These float data are converted to whatever the device needs at some point. If you need to get or send raw data then you are limited to 24 bit integer which is the largest bit-size you can encode in a lossless manner within a float (24 bits is the size of the float mantissa).


Do not post admin requests to the list. They will be ignored.
Coreaudio-api mailing list      (
Help/Unsubscribe/Update your Subscription:

This email sent to

Reply via email to