Why are you using "use_mel_posterior_encoder"?

[Moon-sung-woo]

Hi I'm sungwoo Moon. First of all, thank you for your sharing your code.

I'm looking at your code and I'm wondering why you use 'use_mel_posterior_encoder'. In the paper vits1, it says that we use spectrogram like vits1, but I wonder if there is a difference in TTS performance.

Thank you.

[p0p4k]

In paper they said they use Mel spec, while vits1 uses linear spec. Am I missing something? 🧐

[lastapple]

In paper they said they use Mel spec, while vits1 uses linear spec. Am I missing something? 🧐

So is there a big difference between the effects of these two conditions?

[p0p4k]

I think better to use linear spectrogram. But it's not a big deal

[Moon-sung-woo]

@p0p4k Oh! i'm sorry. I read the paper wrong. Thank you