- Joined
- Oct 25, 2013
- Messages
- 1,044
- Motherboard
- Gigabyte B550M DS3H AC
- CPU
- Ryzen 5 5600G
- Graphics
- Radeon Vega 7
- Mac
@shuhung, what was the speed difference between the true IOGVA transcode and the software transcode?
Software transcode, result is unacceptable for me.
IOGVAHEVCEncode, fast.
Basically, same transcoding performance as IQSV but system is unstable, IQSV work better on H264, RX4XX/5XX work better on HEVC.
On 10.14.1, if FCPX initiate IOGVAHEVCEncode function properly (without freeze issue), RX4XX/5XX HEVC transcoding performance should equal to HD630 IQSV H264 transcoding, both around 4.9X x 23.97 = 117.453 fps
This is fixed in 10.14.2 beta? Or does it work slightly better?
No, same issue on 10.14.2 beta 3 (18C48a).
Personally, Apple programmers seem to be heading in the wrong direction.
They don't use AMD H264 encoding function at all, just like on 10.14 dp 1 (18A293u) remove it from IOGVACodec, which greatly improve system stability.