Does the NSM flag make sense in this case?

[kotyczka]

Hallo, I would like to achieve something like Seamless
Branching. I've got a movie and want to skip a certain
part depending on some language selection (in the menu).
Looks like this:

C1 - start language 2
C2 - start language 1
C3 - main movie
C4 - end language 2
C5 - end language 1

I have authored a PGC1 containing all these cells in
sequence. At the cell borders the video is faded to black
and the audio is faded to zero too.
PGC1 can not be called from a menu. Therefore I have
created PGC2 and PGC 3 which reuse the cells from PGC1:

PGC2 = C2, C3, C5 (for language 1)
PGC3 = C1, C3, C4 (for language 2)

I have chosen that special sequence because C2 is shorter
than C1 and C4 is shorter than C5. C2 and C4 are about
10 seconds each.

It is clear that in PGC2 and PGC3 one cell has to been
jumped over. Software players have no problem with the
jumps, it plays really smooth.
The problem is the stand alone players. (I have not a
stand alone myself). A friend however reported that his
stand alone pauses for 7 seconds at each of the jumps.

Now my question: Does it make sense to set the NSM flag
at the jumps in PGC2 and PGC3? Right now the NSM flag
is not set. I did create PGC2 and PGC3 with DVDRemake,
where by default the check box "seamless playback" is
checked, which means no NSM flag.

If remember correctly I read somewhere, that there are
players, which can handle such a jump better if the
NSM flag is set. However I cannot find where I read
about this.
Is it correct that the NSM flag works for the beginning
of a cell? I.e. if I jump from C3 to C5 in PGC2,
do I have to set the NSM flag at C5?

[blutach]

Yes, the seamless joint flag will need to be unset. Perhaps you can re-author with angles, in which case the interleaving will take care of what you need.

Regards

[kotyczka]

Yes, the seamless joint flag will need to be unset. Perhaps you can re-author with angles, in which case the interleaving will take care of what you need.

Regards

Unfortunately interleaving is not an option.

C1 is about 40 secondes
C2 is about 10 secondes

C4 is about 10 secondes
C5 is about 40 secondes

The video in the cells is really different.
Or is it possible to interleave even cells
of such different length?

But let me ask again, if I want to jump from C3 to C5,
which are the PGC1 names of the cells actually,
if I name them correctly it should be
C1(PGC2) - reuses C2(PGC1)
C2(PGC2) - reuses C3(PGC1)
C3(PGC2) - reuses C5(PGC1)
then I have to clear the "seamless playing"
for C3(PGC2), and not for C2(PGC2).
Is that correct?

[r0lZ]

IMO, it is prudent to clear the seamless flag when there is a gap between two cells (unless they have been mixed as multi-angle or interleaved). Some players (including some standalones) can certainly play the cells seamlessly, but since you have included fades, the little pause should not be noticeable. And it is true that a player that cannot ignore the seamless flag may pause for a long time, and even hang, if the seamless flag is set.
You can probably leave the seamless flag set when the cells are contiguous.

BTW, I recommend to set also the STC Discontinuity flag after a gap.

[kotyczka]

IMO, it is prudent to clear the seamless flag when there is a gap between two cells (unless they have been mixed as multi-angle or interleaved). Some players (including some standalones) can certainly play the cells seamlessly, but since you have included fades, the little pause should not be noticeable. And it is true that a player that cannot ignore the seamless flag may pause for a long time, and even hang, if the seamless flag is set.
You can probably leave the seamless flag set when the cells are contiguous..

r0lZ, thank you for your attention. So I seem to do it right.

BTW, I recommend to set also the STC Discontinuity flag after a gap.

Could you please tell me how to do so. So far I have never
read about the STC Discontinuity flag. I prefer DVDremakePro,
how to set the STC Discontinuity flag in that software?
Or do I have use some other software to set this flag?

And again, just to make sure, could you please explicitely
confirm, that the flags have to be cleared/set for the cell
_after_ the gap.

TIA

[r0lZ]

The STC (System Time Clock, aka SCR, System Clock Reference) discontinuity flag informs that there is a break in the continuity of the time codes. It must always be set when the VOB ID of the cell changes (as the STC is reset to 0), and when there is a gap in non-interleaved cells (as otherwise, the player could display completely wrong durations).

I don't know how to set it with DRM, but it is easy to set with PgcEdit. ;-)
(And, BTW, PgcEdit highlights the Seamless flag in pink or yellow when it is not what it should probably be.)

[kotyczka]

I don't know how to set it with DRM, but it is easy to set with PgcEdit. ;-)
(And, BTW, PgcEdit highlights the Seamless flag in pink or yellow when it is not what it should probably be.)

I just tried it in PGCedit, and really, the cells after the gaps
were marked yellow, because the seamless flag is wrong there.
So I will remove it.

The STC discontinuity flag however was already set correctly
after each gap, DRM is obviosly intelligent enough to set
_that_ flag correctly.

Thank you very much for your help. So I will correct the DVD,
send it to my friend and see, if his stand alone player will
make a shorter pause than 7 seconds. Hope so.
I'll report his answer here when I receive it.

[kotyczka]

Now, this is really weird.

My friend reported me this:
He has a LG Bluray standalone player and he uses VLC as
a software player. Both they show the same behaviour:
There are short gaps (about 10 seconds), where both
the standalone and VLC wait for 7 seconds. At those gaps
I omit one cell in the original PGC, say I jump from
PGC1, cell3 to PGC1 cell5.
And there are other gaps (about 40 seconds) where both
the standalone and VLC play continuously. There I jump
over PGC boundaries, say from PGC1 cell 4 to PGC2 cell 2.

And it makes no difference at all, if the cells after
the gaps (in the PGCs reusing cells from PGC1, PGC2, ...)
are flagged as seamless or not.

For testing I installed VLC in a fresh virtual PC and
indeed the longs gaps work fine while the short gaps
cause a break.

Any other software player (like PowerDVD, WinDVD, Showtime,
MPC) plays continuously and any other standalone, which I
could test, did too.
I have really no idea, why my friend has a software player
and a stanalone, which show the strange behaviour identically.

And furthermore I have no idea, if it is an authoring error
or not and if I could do anything against it.

Any more ideas?

[r0lZ]

For the best compatibility, you should remove the seamless flag of a cell following a gap, and set the STC Discontinuity flag when you jump overt a cell. But normally, the VOB files should also have been muxed specifically to allow the gap, and it should have created a new VOB ID (which implies always the STC Discontinuity).

The players usually try to jump over the gaps as fast as possible, so they may not pause at all if the hardware is fast enough, and/or if they have large buffers.

I have no idea why some players pause for a long time when they jump to a cell that is NOT the first cell of a new VOB (thus to a cell with a Cell ID > 1), but since that kind of jump is not theoretically allowed, we cannot blame them. When you jump to another Title, you jump probably to the first cell of the VOB (Cell ID 1), and the players are happy. ;-)

[jinjin_jp]

I had similar experience long before, and resolved by suggestion in this forum. It is demux and remux each cell.

It is when add cells which is captured from TV program and authoring.

Old method;
*(a)Capture TV program.
*(b)Authoring (a) by "MovieWriter"
*(c)Cut (b) by "DVDShrink"
*(f)There is some VIDEO TS files, and create new cell by "PgcEdit".
*Replace cell of (f) to cell of (c).
=> Its results; Some hard player pause or stop, and soft player(PowerDVD) doesn't sound out.

New method
After (c) of above method.
*(d)Demux cell of (c) and check audio delay by "PgcDemux".
*(e)Remux files of (d) by "Muxman".
*(f)There is some VIDEO TS files, and create new cell by "PgcEdit".
*Replace cell of (f) to cell of (e).
=> Its results have no problem.

I found and thought strange that;
audio delay of demux/remux is -100 to -200 msec, and it changes to -10 to 10 msec after demux/remux.

Regards.