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.

 

Go Back   Doom9's Forum > General > Subtitles
Register FAQ Calendar Today's Posts Search

Reply
 
Thread Tools Search this Thread Display Modes
Old 12th August 2016, 13:45   #281  |  Link
ketsuban
Registered User
 
Join Date: Nov 2015
Posts: 3
I have found by mere chance a possible minor bug involving the time accuracy of some frames. My suspicions are mainly focused on the fact that it is possible that somewhere an inexact rounded fps value is used instead of the exact quotient in the case of non-integer fps such as 23.976 (24000/1001) or 29.97 (30000/1001). The particular example where I have found this issue is the following one, with a 29.97 (30000/1001) fps interlaced video:

The first video frame where it is intended to display that particular line, assuming, by the spec convention, that the frames start to be indexed by 0, is the frame number 37477. The time set in Aegisub for that purpose in format H:MM:SS:CS is 0:20:50:45, namely, at t=20*60+50.45=1250.45 s. Therefore, multiplying by 1001/30000 we have 37476.023976024, and therefore, according to the Aegisub criteria, rounding up the value the subtitle will appear in the frame starting from ceil(37476.023976024)=37477. Indeed, this is the desired result in Aegisub and the subtitle is displayed correctly there. However, after rendering this subtitle in avs2bdnxml, the subtitle appears however starting by the frame 37476, one frame before expected. This result is obtained if we use instead for the computations the inexact value 29.97, since 1250.45*29.97=37475.9865, and therefore ceil(37475.9865)=37476, which is coherent with my initial assumption about the possibility of using somewhere the inexact value. If the time value is changed by increasing an unit the CS to 0:20:50:46 then the issue is corrected in the BDN+XML output, because the result is still correct in Aegisub since in this case we have 1250.46*30000/1000=37476.3236763237 (approx.), hence ceil(37476.3236763237)=37477, and on the other hand 1250.46*29.97=37476.2862, and thus ceil(37476.2862)=37477 as well.

I have checked a bit the source code and it looks like you are doing the things well, so the problem might be in the output values provided by the AviSynth engine, although I am not completely sure. In any case, albeit I have only reported here this issue involving InTC values, I guess it happens as well in the OutTC case. If you could check this issue, confirm and fix it (if possible), I would be very grateful.

By the way, many thanks for considering my last request.
ketsuban is offline   Reply With Quote
Old 22nd August 2016, 19:28   #282  |  Link
Minister
Registered User
 
Join Date: Nov 2011
Posts: 11
I had formerly posted about an issue I had running this app, but it turned out to be a syntax error on my part. Sorry for making this unnecessary post.

Last edited by Minister; 22nd August 2016 at 19:35. Reason: Found a mistake in my script syntax. Post wasn't necessary.
Minister is offline   Reply With Quote
Old 21st February 2024, 09:30   #283  |  Link
KSSW
Registered User
 
Join Date: Sep 2022
Location: China
Posts: 23
I wonder if the quality of PNG output is very high?
KSSW is offline   Reply With Quote
Reply


Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 09:32.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.