Conversation
a mix of double and int16_t . MacOS, Linux and Android already do this, and ASIO also supports it. Change recording format to be 24-bit WAV, to get the full precision of the mixed audio! This patch gives the user the full resolution of the audio device, both when receiving audio and transmitting audio. Cleanup LSB/MSB sample processing in ASIO driver while at it. LSB and MSB indicate little-endian and big-endian data format presumably. Signed-off-by: Hans Petter Selasky <hps@selasky.org>
Switch all Jamulus audio sample processing to use floats instead of a mix of double and int16_t
# Conflicts: # ChangeLog
|
I think the recording should be tested since there were also changes needed in that module. |
I want to get the Mutex changes tested first. Didn't have much happening tonight, so I'd like to wait further trials. |
|
Wow, there are so many things which do not work as expected anymore with the float code... I think it will take quite some time until we can finally merge that code to the master. I'm just in progress of fixing the level meter. |
| } | ||
| }; | ||
|
|
||
| // Might want to comment that use of double in factors below avoids nasty conversion surprises. |
There was a problem hiding this comment.
Could also mention that FACTOR32 and FACTOR24 are the same because...
# Conflicts: # ChangeLog
# Conflicts: # ChangeLog
|
@pljones I have just merged the code to the master. hselasky has changed your jam recorder to record in 24 bit instead of the previous 16 bit resolution. Can you please check if the new code still works ok? Since the jam recorder already creates very large files, with 24 bit the amount will even be higher now. If it is too much data, we could consider going back to 16 bit. But this is your decision (since I do not use the jam recorder). |
4 participants
This is basically the pull request for the floating point support from hselasky. I did:
TODO is still the testing on all platforms.
This is work in progress and the Pull Request will most probably updated with new commits in the near future.
Feel free to review the code. If you have time to test the new code, please tell us your test results (audio quality, stability, are all existing features work as they did before the change, etc.).