Closed mikeoliphant closed 3 months ago
The WaveNet check to make sure there is no "head" element in the json has a couple of problems that are currently canceling each other out under normal circumstances.
This line:
https://github.com/sdatkinson/NeuralAmpModelerCore/blob/028e648772f9931b01b8286eed3a1cfe4d9f9bc2/NAM/get_dsp.cpp#L175
has two problems. First, the logic is reversed, and it looks like it should incorrectly be setting "with_head" to "true" when it is null.
This is currently being compensated for by the fact that having:
"head": null
in the json actually doesn't make config["head"] null - it is rather a json object that references null.
So, everything currently works ok, but the check isn't really doing what it is supposed to do.
It turns out that the existing code is actually causing problems on some systems:
https://github.com/mikeoliphant/neural-amp-modeler-lv2/issues/69
The WaveNet check to make sure there is no "head" element in the json has a couple of problems that are currently canceling each other out under normal circumstances.
This line:
https://github.com/sdatkinson/NeuralAmpModelerCore/blob/028e648772f9931b01b8286eed3a1cfe4d9f9bc2/NAM/get_dsp.cpp#L175
has two problems. First, the logic is reversed, and it looks like it should incorrectly be setting "with_head" to "true" when it is null.
This is currently being compensated for by the fact that having:
"head": null
in the json actually doesn't make config["head"] null - it is rather a json object that references null.
So, everything currently works ok, but the check isn't really doing what it is supposed to do.