Correct update of time in stream record - fixes failure to obtain a rate for stream using rate helper
def124b44c39 — Chris Cannam 2 months ago
Tidy
7f409fe84c65 — Chris Cannam 2 months ago
No, we should be reporting the rate as stored - it was calculated using a block stream rate helper and so should be right
79ac5dd636a3 — Chris Cannam 2 months ago
Fix tests
a64214e80f45 — Chris Cannam 4 months ago
Update following matrix api change
14c63c3154cf — Chris Cannam 4 months ago
Add stateful version of channel selecting stream
5c1ebd46b69e — Chris Cannam 4 months ago
For our practical purposes it seems better to treat this as reducing the rate (random access adapters for example like it better) - I hope it doesn't break anything else
895c2ad024bb — Chris Cannam 4 months ago
Fix failure when source stream ends on block boundary
8e20474044b2 — Chris Cannam 5 months ago
Comments and debug output only
d289a6c0ddbc — Chris Cannam 5 months ago
Add an explicit mechanism to call finish (if we aren't reading to the end)
784a166b2ab4 — Chris Cannam 8 months ago
Provide a friendlier representation with times converted
Correct comment
Cepstral stream only needs magnitude as input, so make that explicit
Fix test
Correct summariser rate by using a blockstream rate helper
Add blocksize alteration to maps
fd344efd1ba4 — Chris Cannam 2 years ago
Fix derivative which wasn't working correctly; add silence threshold
9489a074a773 — Chris Cannam 2 years ago
Comment
8406412c9b80 — Chris Cannam 2 years ago
This seems conceptually simpler (I happened to be writing this elsewhere so thought I'd make it match)
d6d2c82e2f3e — Chris Cannam 2 years ago
Introduce PullResampler to make the callback-driven resample logic available to code other than direct users of ResamplingStatefulStreamFn
Next