Some of the emulations go silent after the GUI window is resized. The Rhodes is affected, potentially also the Mini and others.
I think that there is a discrepancy between the GUI display and the engine parameter for some of the gains, also some potential errata with MIDI channel reselection after resize. For example, the Rhodes error is due to the fact that the GUI gain control shows zero on startup but the engine has been given an explicit gain level and does produce a signal. When the GUI is resized the libbrighton interface correctly sets the value of the visible controller to zero, the value that it actually had, and that suppresses output. A gimpy fix is to configure a non-zero gain on the controller and I think each case may need to be dealt with separately.
A more correct fix would be to suppress engine updates on resizing since we don't actually anticipate any of the values should really change. We have the BRIGHTON_NO_DRAW flag active on resizing and can piggyback an update suppression when it is set (tested and it does work) however that is also required per emulation. Alternatively per device (pot, button, scaler, etc).
This will block the release of 0.20.5 as it introduces the window rendering options for scaling improvements.
Log in to post a comment.