Analyzing PAL Source

[norlane]

1.I load my Index for PAL DVD into VDub through Avisynth script.
MPEG2Source("E:\video file\xxxx\xxxx.d2v")

I have trouble actually detecting Speed Up, Extra Frame Added, Fields Shifted,
Real Progressive and Feild Shifted Progressive.
What techniques and can i use in VDUB and what to look for by eyeball.
___________________
2 Also any tools like M2VInfo.This utility is written to help analysing the FLAGS
behaviour and values in an Mpeg-2 Video stream.
EXAMPLE
You can stop the process, because actually only the first GOP information is needed to determine the pattern of the FLAGS values and behaviour.
A sample of the dump.txt is as follow:Type 1 tff 1 rff 1 temp_reference 2
Type 3 tff 0 rff 1 temp_reference 0
Type 3 tff 1 rff 0 temp_reference 1
Type 2 tff 1 rff 0 temp_reference 5
Type 3 tff 0 rff 0 temp_reference 3
Type 3 tff 0 rff 1 temp_reference 4
Type 2 tff 0 rff 1 temp_reference 8
Type 3 tff 1 rff 1 temp_reference 6
Type 3 tff 0 rff 0 temp_reference 7
Type 2 tff 0 rff 0 temp_reference 11
Type 3 tff 1 rff 0 temp_reference 9
Type 3 tff 1 rff 1 temp_reference 10
Notes:
• Type-1 = I Frame
• Type-2 = P Frame
• Type-3 = B Frame
• temp_ref = DISPLAY ORDER
• rff = Repeat_First_Field flag
• tff = Top_Field_First flag

Now, what we need to do is to reconstruct the DISPLAY ORDER. I do this from the dump.txt above by rearranging the Frames according to the temp_ref. order
B B I B B P B B P B B P
Then, let's put the T_F_F and R_F_F in order too
0 1 1 0 0 1 1 0 0 1 1 0
1 0 1 0 1 0 1 0 1 0 1 0
Now we have the following sequence in DISPLAY_ORDER:
B B I B B P B B P B B P
0 1 1 0 0 1 1 0 0 1 1 0
1 0 1 0 1 0 1 0 1 0 1 0
Now, since we only want to know the CORRECT TELECINE sequence... just take the first 5 frames from the sequence above, and assume its an A B C D E sequence:

A B C D E
Apply the T_F_F and R_F_F values to the sequence above, and correctly follow the first T_F_F value, so we know which STARTING FIELD. I got this:
AlAu AlBu BlCu ClCu DlDu ElEu El
The STARTING_FIELD from above sequence is Al = Frame A lower field. Separate the sequence above into 1 frame (containing 2 fields), and we can calculate the TELECINE sequence within this .M2V as follow:

W S S W W W
So, the WSSWW sequence is the EXACT TELECINE sequence taking place in this particular .M2V
• The rest of the flags are also included, like FPS, DROP_FRAME etc. These are the determining factors to calculate the correct TELECINE Sequence used in a particular Mpeg-2 Video. There are 5 valid TELECINE sequences: WSSWW, SSWWW, SWWWS, WWWSS, WWSSW.
• W = Whole Frame
• S = Separate Field
From the dump.txt, we can now determine the TELECINE sequence to reconstruct the 2:3 pulldown into the 24fps avi. By this we MIMIC the actual DISPLAY_ORDER of the Mpeg-2 Video PLAYBACK into the AVI domain. In short, we CONVERT the 24fps avi into 29.97fps NTSC avi. It is important to apply the specific TELECINE sequence AND the correct STARTING FIELD. When the reconstruction gives lower_field as the BEGINNING of the DISPLAY_ORDER, we have to conform the conversion AS IS. Both the SPECIFIC TELECINE sequence and STARTING FIELD is IMPORTANT to create a 100% video and audio sync.
At the time of writing this document, I can only use Adobe After Effects to correctly convert the 24fps into 29.97fps while at the same time apply both determining factors above. Yes, you can do this directly in any ENCODER, but the applied conversion DOES NOT conform to the STARDARD TELECINE transfer. Encoders will apply a "4th frame repeated" calculation to get from 24fps to 29.97fps (30fps drop frame). Such conversion will be like this:
Current Encoders TELECINE creates: AA BB CC DD DD EE FF GG HH HH
Correct TELECINE Tranfer would be: AA AB BC CC DD EE EF FG GG HH
As you can see, even the STARTING FIELD factor is nowhere to be applied (panasonic encoder has this option, though). If this error is added to the whole 1 hour of Video conversion, Audio syncing WILL be screwed. This stated, I prefer to use Adobe After Effects (albeit it's a difficult program), UNTIL a new method can be found.

Thankx

[Mister Hatt]

I'm not sure what the point of this thread is other than repeating in laughably inaccurate terms the method I described a while back in some other thread. You've failed to explain how to determine what a source actually is. Rather you're only saying "here is what telecine is for people who don't get it and because I am writing it so horrendously you probably still won't get it." You could also do with some demo footage or something to actually show what you mean as well. Hybrid hard/soft material is a good example of demo stuff, although your technique would fail miserably on such a source. Additionally, if you IVTC properly or the Surefine Sugoi Pro Clever Black Magic method, you won't have any desync.

Sure is a laffo thread though.