diff options
author | Robin Gareus <robin@gareus.org> | 2015-03-30 19:52:21 +0200 |
---|---|---|
committer | Robin Gareus <robin@gareus.org> | 2015-03-30 19:52:21 +0200 |
commit | f74acc6e7921e82842b1c33eb7da00204c6f3039 (patch) | |
tree | 3c051e6706d7233430a86025bed7aab0ac0fb306 /libs | |
parent | 8ca0680871b4cb768b7ce217b914479944965c83 (diff) |
note to midi devs.
Diffstat (limited to 'libs')
-rw-r--r-- | libs/ardour/midi_ring_buffer.cc | 12 |
1 files changed, 12 insertions, 0 deletions
diff --git a/libs/ardour/midi_ring_buffer.cc b/libs/ardour/midi_ring_buffer.cc index c2e040d947..7e88f5ff45 100644 --- a/libs/ardour/midi_ring_buffer.cc +++ b/libs/ardour/midi_ring_buffer.cc @@ -165,6 +165,18 @@ MidiRingBuffer<T>::skip_to(framepos_t start) ++count; + /* TODO investigate and think: + * + * Does it makes sense to keep track of notes + * that are skipped (because they're either too late + * (underrun) or never used (read-ahead, loop) ? + * + * skip_to() is called on the rinbuffer between + * disk and process. it seems wrong to track them + * (a potential synth never sees skipped notes, either) + * but there may be more to this. + */ + if (ev_size >= 8) { this->increment_read_ptr (ev_size); } else { |