So is H.264/2-pass (Nero HDTV AVC profile) worth it over ASP/1-pass (Nero Standard profile) despite the higher CPU time demands?
Sizes (both set to 700M CD size):
(audio included in each)
Source MPEG-2 20Mbps CBR: 9.03 GB (9,697,421,312 bytes)
Nero MPEG-4 Standard ASP: 688 MB (721,929,583 bytes)
Nero MPEG-4 HDTV AVC: 688 MB (722,209,028 bytes)
Encoding times:
Didn't time them exactly, but Standard/1-pass took about one hour and HDTV/2-pass took about five hours. This isn't 100% scientific but it should give you some idea. (I wasn't planning on posting results but I do have a few accurate comparison images to show below.)
Source:
Samsung SIR-T451 ATSC OTA HDTV receiver
Jay Leno 720p (#3057-12/9/05 Scarlett Johansson, Larry the Cable Guy, Fallout Boy)
S-Video
ATI TV Wonder USB 2.0
ATI Multimedia Center 9.08 TV record (MPEG-2 20Mbps CBR-No VideoSoap).
Nero Recode 2.2.7.2 (Nero 7 Ultra)
Notes: I had to use MPEG-2 20Mbps vs. uncompressed due to bandwidth and space issues.
Playback path:
(Media Player Classic 6.4.8.7 Unicode)
Nero Standard: Haali Demultiplexer ("MPEG4 video")->ffdshow MPEG-4 Video Decoder (XviD)->Video Mixing Renderer 9 (windowed)
Nero HDTV AVC: Haali Demultiplexer ("AVC video")->ffdshow MPEG-4 Video Decoder (libavcodec h264)->Video Mixing Renderer 9 (windowed)
CPU usage:
ASP: ~25%
AVC: ~50%
Frame-accurate comparisons:
Don't mind the green vertical bar at the left on the Standard encode. I had used a slightly different crop setting.
Time:
00:00:00: Standard / HDTV AVC
00:00:42: Standard / HDTV AVC
00:01:07: Standard / HDTV AVC
00:12:12: Standard / HDTV AVC
00:17:37: Standard / HDTV AVC
00:21:10: Standard / HDTV AVC
00:32:09: Standard / HDTV AVC
00:32:15: Standard / HDTV AVC
00:59:56: Standard / HDTV AVC
01:00:05: Standard / HDTV AVC
Conclusion:
Overall, little difference. But when you bring high-motion or complex scenes into the equation, HDTV AVC encoding really shines. Since I'm picky about video quality, I'd just let the HDTV AVC run overnight. It's worth the extra quality to me. Besides, they're the same size. It's hard to say if the difference is because of the 2-pass or AVC. I think it's the AVC. But I recommend you run 2-pass regardless. It shouldn't take much longer. This was a little less controlled than I had hoped it to be, but the comparison was an afterthought and I thought I'd share it anyway.
What do you think?
Sizes (both set to 700M CD size):
(audio included in each)
Source MPEG-2 20Mbps CBR: 9.03 GB (9,697,421,312 bytes)
Nero MPEG-4 Standard ASP: 688 MB (721,929,583 bytes)
Nero MPEG-4 HDTV AVC: 688 MB (722,209,028 bytes)
Encoding times:
Didn't time them exactly, but Standard/1-pass took about one hour and HDTV/2-pass took about five hours. This isn't 100% scientific but it should give you some idea. (I wasn't planning on posting results but I do have a few accurate comparison images to show below.)
Source:
Samsung SIR-T451 ATSC OTA HDTV receiver
Jay Leno 720p (#3057-12/9/05 Scarlett Johansson, Larry the Cable Guy, Fallout Boy)
S-Video
ATI TV Wonder USB 2.0
ATI Multimedia Center 9.08 TV record (MPEG-2 20Mbps CBR-No VideoSoap).
Nero Recode 2.2.7.2 (Nero 7 Ultra)
Notes: I had to use MPEG-2 20Mbps vs. uncompressed due to bandwidth and space issues.
Playback path:
(Media Player Classic 6.4.8.7 Unicode)
Nero Standard: Haali Demultiplexer ("MPEG4 video")->ffdshow MPEG-4 Video Decoder (XviD)->Video Mixing Renderer 9 (windowed)
Nero HDTV AVC: Haali Demultiplexer ("AVC video")->ffdshow MPEG-4 Video Decoder (libavcodec h264)->Video Mixing Renderer 9 (windowed)
CPU usage:
ASP: ~25%
AVC: ~50%
Frame-accurate comparisons:
Don't mind the green vertical bar at the left on the Standard encode. I had used a slightly different crop setting.
Time:
00:00:00: Standard / HDTV AVC
00:00:42: Standard / HDTV AVC
00:01:07: Standard / HDTV AVC
00:12:12: Standard / HDTV AVC
00:17:37: Standard / HDTV AVC
00:21:10: Standard / HDTV AVC
00:32:09: Standard / HDTV AVC
00:32:15: Standard / HDTV AVC
00:59:56: Standard / HDTV AVC
01:00:05: Standard / HDTV AVC
Conclusion:
Overall, little difference. But when you bring high-motion or complex scenes into the equation, HDTV AVC encoding really shines. Since I'm picky about video quality, I'd just let the HDTV AVC run overnight. It's worth the extra quality to me. Besides, they're the same size. It's hard to say if the difference is because of the 2-pass or AVC. I think it's the AVC. But I recommend you run 2-pass regardless. It shouldn't take much longer. This was a little less controlled than I had hoped it to be, but the comparison was an afterthought and I thought I'd share it anyway.
What do you think?