Add audio output support!#756
Conversation
This evolves the API from 2.x (and breaks it). Playback devices are now separate from the samples themselves. This allows for greater playback flexibility. Two sample sources are audioio.RawSample and audioio.WaveFile. They can both be mono or stereo. They can be output to audioio.AudioOut or audiobusio.I2SOut. Internally, the dma tracking has changed from a TC counting block transfers to an interrupt generated by the block event sent to the EVSYS. This reduces the overhead of each DMA transfer so multiple can occure without using up TCs. Fixes micropython#652. Fixes micropython#522. Huge progress on micropython#263
dhalbert
left a comment
dhalbert
left a comment
There was a problem hiding this comment.
Massive! I couldn't review all the DMA stuff knowledgeably, but tried to look for obvious stuff. Looks great!
| // TODO(tannewt): Should we have a way of sharing GCLKs based on their speed? Divisor doesn't | ||
| // gaurantee speed because it depends on the source. | ||
| uint8_t find_free_gclk(uint16_t divisor) { | ||
| if (divisor > (1 << 8)) { |
There was a problem hiding this comment.
Should this be divisor >= (1 << 8), or maybe clearer divisor > 0xff? Otherwise divisor=0x100 could be used for an 8-bit clock, which won't work
| #endif | ||
| } | ||
|
|
||
| void reset_gclks(void) { |
There was a problem hiding this comment.
It's possible that if we need to turn on clocks in a complicated way later (say to enabled crystaled clocks), checking these #define may not be reliable. Just something to watch out for, nothing to change now.
There was a problem hiding this comment.
Yup totally. I figure when we make the clock init more complex we can do it in this file and stop using ASF4 completely.
| #include "timers.h" | ||
|
|
||
| void audioout_reset(void) { | ||
| // Only reset DMA. PWMOut will reset the timer. Other code will reset the DAC. |
| audio_dma_stop(&self->right_dma); | ||
| #endif | ||
|
|
||
| // FIXME(tannewt): Do we want to disable? What if we're sharing with an AnalogOut on the 51? |
There was a problem hiding this comment.
Maybe forbid this up front? Can't do audio out if the AnalogOut exists?
| //| :class:`WaveFile` -- Load a wave file for audio playback | ||
| //| ======================================================== | ||
| //| | ||
| //| A .wav file prepped for audio playback |
There was a problem hiding this comment.
Could you add doc here about what formats of WAV files are supported? Looks like 8-bit PCM (always unsigned) mono and stereo and 16-bit (always signed) mono and stereo. Is that right?
There was a problem hiding this comment.
That's all I've tested so I added a comment about it.
| //| .. class:: RawSample(buffer, *, channel_count=1, sample_rate=8000) | ||
| //| | ||
| //| Create a RawSample based on the given buffer of signed values. If channel_count is more than | ||
| //| 1 then each channel's samples should rotate. In other words, for a two channel buffer, the |
There was a problem hiding this comment.
"should rotate" -> "should alternate"
|
Thanks for the quick review! I don't expect it to be perfect but am happy to follow up with fixes later. |
2 participants
This evolves the API from 2.x (and breaks it). Playback devices are now
separate from the samples themselves. This allows for greater playback
flexibility. Two sample sources are audioio.RawSample and audioio.WaveFile.
They can both be mono or stereo. They can be output to audioio.AudioOut or
audiobusio.I2SOut.
Internally, the dma tracking has changed from a TC counting block transfers
to an interrupt generated by the block event sent to the EVSYS. This reduces
the overhead of each DMA transfer so multiple can occure without using up TCs.
Fixes #652. Fixes #522. Huge progress on #263