Output both 24k and 48k spectograms

[Veldhoen]

Teng has trained two models. He is going to check which model is most appropriate. For now, export spectogram for both audio framerates.

[Veldhoen]

Output to separate files - so we can delete them later

[Veldhoen]

@Veldhoen Voorbeeld van tijd/video-verschil