15:56dwlsalmeida[d]: Forcing y_log2 == 4 actually fixes both the images I was working with btw
15:57dwlsalmeida[d]: https://cdn.discordapp.com/attachments/1034184951790305330/1303025214724247737/Screenshot_2024-11-04_at_12.54.07.png?ex=672a404c&is=6728eecc&hm=6c825cee759c9092948e560fb0b223f3e134fb98b96e11f4a3076b524bb555a0&
15:57dwlsalmeida[d]: so yes, figuring out an API to make this value the same for both Y and UV planes is definitely the answer
16:03dwlsalmeida[d]: https://cdn.discordapp.com/attachments/1034184951790305330/1303026752200769677/Screenshot_2024-11-04_at_13.02.46.png?ex=672a41ba&is=6728f03a&hm=68d4dc5a1e4d6c4d2d4a71bc859009d4f2a78bd4c4ae46319e0a688760317775&
16:03dwlsalmeida[d]: This fine sir is also being decoded alright:
19:28airlied[d]: dwlsalmeida[d]: looks like you are ready for the fun times of P-frames 😛