Open iver56 opened 1 year ago
Here's a figure from the Fenton/Lee 2017 paper. We can see the peak at ~500 hz
However, in the implementation in pyloudnorm, the 500 hz peaking filter seems to have gain set to 0 dB, so it doesn't do anything:
https://github.com/csteinmetz1/pyloudnorm/blob/1fb914693e07f4bea06fdb2e4bd2d6ddc1688a9e/pyloudnorm/meter.py#L126
I'm not sure what exact value it should be set to, but from the plot I can see that it should be somewhere close to 3 dB
Here's a figure from the Fenton/Lee 2017 paper. We can see the peak at ~500 hz
However, in the implementation in pyloudnorm, the 500 hz peaking filter seems to have gain set to 0 dB, so it doesn't do anything:
https://github.com/csteinmetz1/pyloudnorm/blob/1fb914693e07f4bea06fdb2e4bd2d6ddc1688a9e/pyloudnorm/meter.py#L126
I'm not sure what exact value it should be set to, but from the plot I can see that it should be somewhere close to 3 dB