Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion. Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules. |
4th October 2007, 17:49 | #1022 | Link | |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
|
|
4th October 2007, 22:45 | #1024 | Link |
Registered User
Join Date: Sep 2006
Posts: 2,197
|
Another problem: I tried to convert the 5.1 TrueHD 24-bit track of letters from iwo jima to FLAC (using eac3to v1.22; that track was the 2nd track on the disc btw, not the 1st). the thd track size is 1,32GB, the final FLAC has 3.2GB of size. movie length is 2h 20min. the strange thing is now that the flac file has a length of 3h 30min!
when playing this together with the .mkv movie, the fps of 23.976 are kept, but the sound is horribly slowed down (and of course, theres also no picture beyond the 2h 20 min mark). how can this be possible? is it possible that demuxed sources (and the remux to temp .evo) give some trouble because some information about playback, speed for example, can get lost? maybe it would help if you enable .evo input as source again, so we could test and compare this. just converted the 2nd track on the disc, DD+ 5.1, to ac3 just to compare and in this case the original length was kept at 2h 20 mins (no muxing to temp .evo took place). converted the DD+ 5.1 to flac, again to compare, length was 2h 20min (again no temp evo muxing took place). (there was a length difference between the converted ac3 and flac track, both from the dd+ 5.1 source btw., the ac3 track is 32ms longer. how can this be possible with the same source file?) on the other hand, that other DD+ 2.0 commentary track from another movie I converted to flac, there also didnt take place any temp evo muxing, length is fine, but still it speeds up the .mkv movie. p.s. when rebuilding the video .evo, powerdvd wasnt able to jump into the 2nd half of the rebuilt evo, from whereon the 2nd evo begins. the time would still continue, but no picture is shown. after remuxing to .mkv with gdsmux & mkvmerge this was possible though with ffdshow, I guess this is normal for powerdvd and files up to a certain size? Last edited by Thunderbolt8; 4th October 2007 at 23:27. |
4th October 2007, 23:29 | #1025 | Link |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
eac3to v1.23 released
http://madshi.net/eac3to.zip Code:
* bugfix: sometimes TrueHD decoding resulted in incorrect sampling rate |
4th October 2007, 23:40 | #1027 | Link | |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
|
|
4th October 2007, 23:46 | #1028 | Link | ||
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
Quote:
|
||
4th October 2007, 23:47 | #1029 | Link |
Registered User
Join Date: Sep 2006
Posts: 2,197
|
did use offsetpts before any rebuilding & demuxing of course, but it reported everything to be fine. the only case it was needed for me was indeed so far mulholland drive from studio canal.
oh, now I see it, that truehd 5.1 flac created with v1.22 only had 32Khz :P edit: ok tested it now with 1.23 and length is fine! thanks! guess only that commentary track video speedup remains for me atm :P btw. it also seems like the delay it that tiny bit better with +30 ms, the +30 ms the ac3 track was longer than the flac track. how does it come delay modification is apparently needed only for flac tracks? btw. dont know if that helps, but that normalization removal thing for all the tracks (trueHD and DD+) I recently converted always said -27db. no other value than this one. same for the bitrate for all the trueHD and dd+ tracks, always got the info that the current track contains more than 16-bit of information -> 24-bit. so far all the flac tracks I made, not only from trueHD, but also from DD+ were quite bigger than the original track sizes, so that normalization removal problem might be a more general one and not only restricted to the truehd tracks Last edited by Thunderbolt8; 5th October 2007 at 02:02. |
5th October 2007, 03:50 | #1030 | Link |
Registered User
Join Date: Jan 2007
Posts: 22
|
ok reinstalled my older version of Nero, I can't purchase Blu-ray/hd-dvd plugin, it keeps taking me to the product page of nero 8.
when I reinstalled nero 7 this is what I got. E-AC3, 5.1 channels, 1:00:13, 640kbit/s, 48khz, dialnorm: -27dB Remove Dialog Normalization information. Please wait... Decoding eac3 file to raw. Please wait... Disabling DRC for Nero E-AC3 decoding... The file size of the raw file doesn't seem to fit. The expected file size for 16 bit is 1.93 GB. The expected file size for 24 bit is 2.90 GB. The real file size is 0 Bytes. |
5th October 2007, 08:31 | #1032 | Link | |||
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
Quote:
Quote:
However, it is not "right" that your FLAC tracks are bigger than the TrueHD tracks. There's definitely something wrong there! Since I added dialnorm removal for TrueHD tracks, I've never seen FLAC files getting bigger than the original TrueHD file on my PC yet! What do the other eac3to users say? Are your FLAC files also sometimes (or often) bigger than the original TrueHD files? |
|||
5th October 2007, 08:32 | #1033 | Link | |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
|
|
5th October 2007, 08:50 | #1034 | Link | |
Moderator
Join Date: Feb 2005
Location: Spain
Posts: 6,881
|
Quote:
All ac3 encoders put a 5.33 ms. delay (256 samples at 48 KHz) with silence to preserve the initial value of uncompressed audio. This delay (only 5.33 sec) can force a new frame, at the end, 32 ms. long, the rest is filled with silence. Using Aften parameter -pad 0 the delay disappear, but the first 5.33 ms. is not the same than uncompressed value. |
|
5th October 2007, 08:53 | #1035 | Link | |
Registered User
Join Date: Sep 2006
Posts: 2,197
|
Quote:
what im doing is (for audio decoding) 1. checking both .evo files with offsetpts 2. unchecking all other audio/video/subtitle than the audio track(s) I want to use for my remuxes (evodemux). 3. giving them the correct endings (.thd for truehd tracks, .eac3 for dd+ tracks) 4. use them with eac3to with "eac3to source.thd/source.eac3 dest.flac" (5. put them in .mka afterwards, but this doesnt matter for the flac filesize). thats it basically. unless there some mistake in what im doing or all this occurs only because we have only used different movies so far, which apparently seem to have differences with their truehd tracks, there seems to be a difference in what we are doing or maybe also somehow with our eac3to versions (or maybe nero/sonic filters? is there any way how I can check if the correct filters are installed?) Last edited by Thunderbolt8; 5th October 2007 at 09:07. |
|
5th October 2007, 08:56 | #1036 | Link | |
Registered User
Join Date: Sep 2006
Posts: 2,197
|
Quote:
|
|
5th October 2007, 09:15 | #1037 | Link | |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
|
|
5th October 2007, 09:17 | #1038 | Link | |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
Can you send me a 10MB sample of a TrueHD file where the FLAC file ends up being larger than the TrueHD file? Before you send me that sample, please run the sample through eac3to and check if it reports that the audio data contains more than 16bit of information. Sending me the sample makes sense only if eac3to reports that. Thanks! P.S: You could also name the movies you've tried so far. Maybe I own one of them, too. |
|
5th October 2007, 09:22 | #1039 | Link |
Registered User
Join Date: Sep 2006
Posts: 2,197
|
im just demuxing them (from the 2 original .evo files), as .evo as input doesnt work any more (of course, I could first rebuilt a big audio file only, load it into evodemux and demux from there then, but this takes more time of course. could try it though, if you think its worth testing and comparing). edit: did it, both files, the direclty demuxed one and the rebuilt evo -> demuxed from there one both have the same big file size.
im not really familiar with that cutting stuff, how can I cut samples from trueHD/flac/other audio tracks? does it work that easily that I could just stop the demuxing process after 10mb or would this sample then be invalid somehow? movies ive tried so far (all hddvd): fear and loathing in las vegas (truehd) out of sight (dd+) eternal sunshine of the spotless mind (dd+) letters from iwo jima (truehd) Last edited by Thunderbolt8; 5th October 2007 at 10:58. |
5th October 2007, 11:04 | #1040 | Link | |
Registered Developer
Join Date: Sep 2006
Posts: 9,140
|
Quote:
The dd+ files are irrelevant. They're always 24bit. Only the TrueHD tracks are interesting. Fear and Loathing is really 24bit, I believe. The TrueHD track should be very big. Iwo Jima should be 16bit. So a sample of Iwo Jima might help, if the problem is reproducable. |
|
Tags |
eac3to |
Thread Tools | Search this Thread |
Display Modes | |
|
|