PDA

View Full Version : Detected no frames and/or video to encode.


Mystiqq
12th September 2006, 11:34
Found couple of old threads about this issue but there was nothing in them to fix or workaround this issue.

http://forum.doom9.org/showthread.php?t=98873&highlight=Detected
http://forum.doom9.org/showthread.php?t=109273&highlight=Detected

This happened when i "cut up" Cars trailer (from DVD) to make it smaller for my encoding quality test (personal curiosity).

I tested both, VobBlanker and DVD Shrink, "cut" functions and both results in the same error message. I even demuxed and remuxed with Muxman, but it didnt do nothing (suprise).

In VobBlanker (2.1.1.0) i used the "prev/cut" and used the "keep movie" option. When i try to mark any spot in the movie, i get message...

The selected GOP is an open one, so if this mark is the init of the playback, the first picture can not be properly decoded, so VobBlanker will mark it as a broken GOP. Do you want to continue?

In DVD Shrink (3.2) i used the re-authored mode and then "cut it" using the "Set Start/End frames" option.

Any ideas?

jdobbs
12th September 2006, 12:38
I don't think I understand where DVD Rebuilder comes into this?

Mystiqq
12th September 2006, 12:56
I don't think I understand where DVD Rebuilder comes into this?

DVD Rebuilder is the one giving the error, just like the threads to the older posts about this same issue.

-----------------
- DVD-RB v1.10.6
- CCE 2.70.2.0 encoder selected.
- Detected no frames and/or video to encode. Aborting.

What im doing is that ive got the trailer and i want to make it even smaller to make the encoding process faster. Ive already done tests with the whole trailer and they work fine, but when i try to use the cutted video, it gives me that error.

Is it just not possible to cut the video and process in DVDRB?

jdobbs
12th September 2006, 15:42
Maybe I'm still misunderstanding.. but I think the answer was in the threads you referenced. If you want to process very small VTSs you have to set VTS_MIN_SIZE to some small value (like 1). By default DVD-RB does not process VTSs that are under about 50MB.

You would add this to the "[Options]" area of the REBUILDER.INI file:

VTS_MIN_SIZE=1

But a note to all those who use this... For most instances, however, the default is a better choice. Processing small VTSs just increases the processing time and contributes almost nothing to quality. That's why the default is set as it is. The exception is DVDs that have a very large number of small VTSs.

Just to clarify. This isn't an "issue" -- that's how it's supposed to work.

jdobbs
12th September 2006, 15:48
By the way... it usually isn't a good idea to cut using a break point at an open GOP. You're going to either lose frames or get some terrible blockiness at the beginning of that section. An open GOP requires something outside itself in order to correctly build pictures. The "quality test" isn't going to be very accurate.

Adding: I'm not completely familiar with what you're doing, so this may be off-the-wall... But usually doing quality tests on small segments is a lost cause -- every encoder/transcoder will do well when working with small video segments... the only time you see differences is when the work is "challenging". Of course, I'm assuming a lot in saying that...

Mystiqq
12th September 2006, 15:58
Thanks for the reply.

I actually read about that option but i just cant remember why on earth i didnt test it... Anyways, that did the trick.

Does the cutting affect the frames anywhere else besides in the beginning?

edit:
So encoding a part of the video with same (relative) compression doesnt result in same kind of/exact quality as it would with bigger file and same compression? If anyone can understand what im saying... :P

Thanks again.

jdobbs
12th September 2006, 17:06
So encoding a part of the video with same (relative) compression doesnt result in same kind of/exact quality as it would with bigger file and same compression? If anyone can understand what im saying... :P
No it wouldn't. For example, there would be less diversity in the segment and as a result less latitude in the decision as to how bits would be distributed.

Also you have to remember that compression levels alone mean nothing. If you have an original that was encoded at a very high bitrate and reduce it 40% it might look great. But if you have an original that was encoded at the edge of what was acceptable, it might look terrible at even 90%... Compression levels are inexorably linked to source complexity and bitrate.

The times that really matter are at the points that are very challenging... that's also where you'll see the significant differences between things like within-the-compressed-domain-transcoders and encoders. On easy stuff they will almost be identical.

The same is true for a good encoder and a bad encoder. If you ran an easy-to-encode segment against CCE and the (terrible) MPEG reference encoder you probably wouldn't see any difference. But take a challenging source and it becomes very obvious.

Encoders are a lot like people in that respect... their true colors only show through when under stress.

Mystiqq
12th September 2006, 18:56
Interesting. Never thought about it that way.

edit:
DVDRB rocks! ;)

blutach
19th September 2006, 10:01
The cutting will affect where you made the cut and where the open GOP is. If you, for example, cut the credits with VobBlanker (http://jsoto.posunplugged.com/vobblanker.htm) and at the cut got this message, it is likely that the very last frame (which you might not care about anyway) might not be fully and properly decoded (as it might depend on later frames which you have cut away).

Obviously, try to cut at a closed GOP, but this is not always possible due to the way things are muxed. So, cuts at the end of a clip should be a bit after the video you want to watch and vice versa for the beginning of a clip (cutting out the 1st 30 seconds of studio intros for example).

Regards