|
|
| fccHandler |
| Posted: Jul 5 2005, 04:08 AM |
 |
|
Administrator n00b
Group: Moderators
Posts: 3961
Member No.: 280
Joined: 13-September 02
|
http://fcchandler.home.comcast.net
Although the official VirtualDub 1.6.8 has been declared stable, VirtualDub-MPEG2 1.6.8 may not be. So far it has only been tested by one person... me! 
If you have any problems, please post them here.
--------------------
May the FOURCC be with you... |
|
| neuron2 |
| Posted: Jul 5 2005, 11:50 AM |
|
|
Advanced Member
Group: Members
Posts: 1244
Member No.: 5294
Joined: 18-July 03
|
. |
|
| KornX |
 Posted: Jul 5 2005, 03:33 PM |
|
|
Advanced Member
Group: Members
Posts: 41
Member No.: 2043
Joined: 18-January 03
|
Hi thx for the new version...
just wanted to ask when it'll come...
Just tested i and the first crash...
(But i think the xource file is shitty)...
KornX
--------------------
There are only 10 types of people;
Those who do and those who don't understand binary. |
|
| fccHandler |
| Posted: Jul 5 2005, 05:05 PM |
|
|
Administrator n00b
Group: Moderators
Posts: 3961
Member No.: 280
Joined: 13-September 02
|
| QUOTE (neuron2 @ Jul 5 2005, 07:50 AM) | | I see that you have implemented RFF handling. But I don't see a "force film" capability. Is that on your agenda? |
Yes it is. Eventually I plan to replace the "allow rff" checkbox with a "force film" checkbox.
| QUOTE | | Also, how about multiple source file handling? |
I'd like that too, and I'll look into it. But I fear it's going to be very difficult to do... 
--------------------
May the FOURCC be with you... |
|
| fccHandler |
| Posted: Jul 5 2005, 05:07 PM |
|
|
Administrator n00b
Group: Moderators
Posts: 3961
Member No.: 280
Joined: 13-September 02
|
| QUOTE (KornX @ Jul 5 2005, 11:33 AM) | | Just tested i and the first crash... |
A crash already? 
Next time please save a crashlog and post it.
EDIT: Hmm, I just saw a crash too. AFAIK it can only happen if the source is broken at the beginning AND the source uses "repeat first field" flags. I'll work on fixing this today.
This post has been edited by fccHandler on Jul 5 2005, 06:01 PM
--------------------
May the FOURCC be with you... |
|
| fccHandler |
| Posted: Jul 5 2005, 07:51 PM |
|
|
Administrator n00b
Group: Moderators
Posts: 3961
Member No.: 280
Joined: 13-September 02
|
OK, in my case the decoder is trying to work with B-frames which lack a previous reference (because of the broken nature of the stream). It turns out that this causes all kinds of unforseen problems, especially if those B-frames use RFF (meaning their ugly fields are needed to reconstruct other frames, and then those frames turn out ugly too).
I decided to simply get rid of broken B-frames at the beginning of broken streams. That fixed my crash, and maybe it will fix yours. If not, please post a crash log.
New build 23657
--------------------
May the FOURCC be with you... |
|
| neuron2 |
| Posted: Jul 5 2005, 10:25 PM |
|
|
Advanced Member
Group: Members
Posts: 1244
Member No.: 5294
Joined: 18-July 03
|
. |
|
| meilin |
| Posted: Jul 6 2005, 12:29 AM |
|
|
Unregistered
|
thx 
|
|
| fccHandler |
| Posted: Jul 6 2005, 04:17 AM |
|
|
Administrator n00b
Group: Moderators
Posts: 3961
Member No.: 280
Joined: 13-September 02
|
| QUOTE (neuron2 @ Jul 5 2005, 06:25 PM) | | Are you getting rid of the corresponding audio too? |
No, but it may not be necessary...
"File Information" reports if there is a skew difference between the earliest video PTS and the earliest audio PTS. Right now the user must ALWAYS check this, because it's almost never zero. (I hope to fix that eventually, so the skew is dealt with internally.)
But to be confident about the accuracy of the reported skew, I have to be sure that the earliest video PTS and audio PTS are actually attached to the earliest video and audio samples.
In the case of the video, I think it's OK. The change I made today discards orphan P and B frames so that the first frame (in display order) is a guaranteed I-frame. My experience thus far has been that I-frames always have PTS attached to them.
In the case of the audio, I'm not so sure. I thought I read somewhere that all audio PES packets have PTS. I'll have to investigate whether that's really true, and get back to you.
--------------------
May the FOURCC be with you... |
|
| fccHandler |
| Posted: Jul 6 2005, 08:13 AM |
|
|
Administrator n00b
Group: Moderators
Posts: 3961
Member No.: 280
Joined: 13-September 02
|
| QUOTE (meilin @ Jul 5 2005, 08:29 PM) | | thx |
You're welcome, my friend! 
--------------------
May the FOURCC be with you... |
|
| stephanV |
| Posted: Jul 7 2005, 09:14 AM |
|
|
Spam killer ;)
Group: Moderators
Posts: 4348
Member No.: 8917
Joined: 18-February 04
|
When i try to play a VOB (PAL) with the space-bar (input only) i get sound but a paused, garbled image (image appears correctly when i press stop). With enter (play output video) it works fine. With AVIs there is no problem at all.
I nailed this down to the output colordepth in options ---> preferences. It was set to 16 bit (my screen is 32) changing that to anything else would fix the issue. Thus remains the question: why does it work still with AVIs and with playing output video? Unfortunately, I don't have any MPEG1 files to test (although I could create one).
My graphics card is an Intel 82852/82855 GM/GME with Intel Extreme Grahpics 2 M drivers. DirectX related settings did not affect it.
Not a major problem for me, just wanted to let you know.
[edit]
The same thing happens with MPEG1, but it does not happen in vanilla VDub. 
--------------------
useful links:
VirtualDub, Input plugins and filters, AviSynth, AVI-Mux GUI, AC3ACM by fcchandler, VirtualDub FAQ |
|
| meilin |
| Posted: Jul 7 2005, 09:53 AM |
|
|
Unregistered
|
hoho
I have the same problem
|
|
| stephanV |
| Posted: Jul 7 2005, 09:59 AM |
|
|
Spam killer ;)
Group: Moderators
Posts: 4348
Member No.: 8917
Joined: 18-February 04
|
heres a screenshot of the phenomena:
screenshot
Note that the video is standing still when in dub and the image gets only updated correctly after pausing. It looks like a colorspace issue to me?
--------------------
useful links:
VirtualDub, Input plugins and filters, AviSynth, AVI-Mux GUI, AC3ACM by fcchandler, VirtualDub FAQ |
|
| KornX |
| Posted: Jul 7 2005, 11:03 AM |
|
|
Advanced Member
Group: Members
Posts: 41
Member No.: 2043
Joined: 18-January 03
|
fixed my prob with the broken file...
and so fast...
thx
KornX
--------------------
There are only 10 types of people;
Those who do and those who don't understand binary. |
|
| fccHandler |
| Posted: Jul 7 2005, 05:40 PM |
|
|
Administrator n00b
Group: Moderators
Posts: 3961
Member No.: 280
Joined: 13-September 02
|
| QUOTE (stephanV @ Jul 7 2005, 05:14 AM) | | When i try to play a VOB (PAL) with the space-bar (input only) i get sound but a paused, garbled image (image appears correctly when i press stop). With enter (play output video) it works fine. With AVIs there is no problem at all. |
Oops. VirtualDub 1.6.8 adds a lot of new colorspaces and, not only did I forget to test them all, I never even wrote the code! (I just looked now, and there's "TODO" written everywhere.)
The "decompression format" in Video / Color depth is the key. The 16-bit RGB modes don't work, although 565 is supposed to work, at least it did in 1.5.10. Now it's all psychedelic green.
AVIs look OK because the codec does the colorspace conversion (or perhaps Windows, or VirtualDub). In the case of MPEG my decoder does all the conversions. At one point I looked into using some of VirtualDub's internal conversion routines, but AFAIK they don't correctly upsample interlaced chroma in MPEG-2 video. (In fact, I'm not sure they upsample at all...)
Thanks stephanV for finding this. I'll fix it ASAP.
--------------------
May the FOURCC be with you... |
|
