DVD9 layer break cell; where to put non-seamless flag?

[sdsumike619]

I use GEAR Pro for mastering to DLT and I use their layer break calculator spreadsheet. Based on their spreadsheet, I determined that my cell for the layer break is 20 with entry sector at 146,842 (see screen shot) The mastering facility reported to me that Eclipse detected an error because the layer break was not set on a cell flagged as non-seamless. They said the only problem with this error is that there might be a 1-2 second pause when the layer break occurs. I have the replicated disc and confirm there is a pause on tabletop players. No big deal.

My question is about setting the non-seamless flag with PGCedit. Should I set it at cell 20 or 19? In other words, does the layer break happen at the beginning of the chosen cell, or does it happen at the end of the previous cell? When I play the disc in the player the break in playback occurs right at 8:20 which is the end of cell 19 and the beginning of cell 20

I'm going to submit another set of DLT's to the mastering facility after using PGCedit to set the non-seamless flag. They said they'd run it through Eclipse again for me at no charge. But I need to know if it should be in 19 or 20. My guess is that it should be in cell 20..

Another question I have: Why is it that the title where the layer break occurs has a bunch of cells, while most other titles have only 2 cells. For example, title 8 is about the same length as title 5, but title 8 only has 2 cells, while title 5 has 45 cells?

[blutach]

Untick seamless joint in cell 19. But why not burn with ImgBurn and jut let it do the work for you?

Also, you don't need it to be non-seamless - I've not seen anyone report a non-seamless layer break as ruining their disk.

Another question I have: Why is it that the title where the layer break occurs has a bunch of cells, while most other titles have only 2 cells. For example, title 8 is about the same length as title 5, but title 8 only has 2 cells, while title 5 has 45 cells?

This is a function of the physical position of the project on the disc. Most times, the LB will fall in the "main movie", which is the longest title i.e. it spans the layers). In fact, if you have 2 titles about the same length and one has just one or two cells, and the order was reversed, you are likely to have problems setting a layer break.

Regards

[sdsumike619]

Yes, I'm familiar with ImgBurn, however, this is not for burning a disc. This is for writing DLT's to submit to a mastering facility for replication.

So the previous cell (19) is the one that should be set as non-seamless if I selected cell 20's entry sector for the layer break?

And yes you're right that it doesn't HAVE to be non-seamless. There's a 1-2 second pause while the laser focuses onto layer1 and it's not a big deal. HOWEVER, I did discover on a Sony player that if someone happens to play through the layer break, then rewind past it and play through it again, playback actually jumps 25 seconds to the start of the next cell. Granted this will be a rare case so I decided to go ahead and release this run of 1000. But for the future, I want to make sure the layer break is non-seamless so everything is smooth.

So it seems to be better to have one long title with many chapters than multiple titles with no chapters in them. From an editing and authoring standpoint it's easier to make each chapter its own title. I don't know, it probably doesn't matter either way as long as everything is within spec...

[blutach]

Sorry, I was wrong on that. Cell 20 is the one. Early morning here

Regards

[sdsumike619]

Thank you, that is what I was hoping to hear! So I will simply uncheck the box on cell 20, save it, then import to GEAR and proceed as I did before. The mastering facility will run the new DLT's through Eclipse and let me know how it turns out this time. But I think it will be fine. At least now I know for the future...

[r0lZ]

The pause occurs at the last frame of the previous cell. (In fact, during the refocus process, the player can not read the disc, and therefore it displays the last frame it has in its buffer for some time.)

Clearing the seamless flag forces most players to pause. So, I disagree with "... Eclipse detected an error because the layer break was not set on a cell flagged as non-seamless. They said the only problem with this error is that there might be a 1-2 second pause when the layer break occurs." It is true that it's theoretically an error, but the effect is usually exactly the opposite. The DVD-Video standard impose a non-seamless layer break, and therefore a pause with most standalone players, but the technique of the "seamless layer break" has been officially used by commercial companies, for example on the Superbit DVDs, and works usually well. When the seamless flag is set, the player can theoretically switch to L2 immediately, if it is fast enough. However, there is a complex trick in the muxing of the previous cell to "prepare" the player for the layer change, and since the seamless layer break is not allowed by the standard, there is no guarantee that it works well with all players. So, to be on the safe side, you should clear the seamless flag of cell 20, and accept the pause.

[sdsumike619]

Quite a bit of what you said went right over my head. But I looked at several previous DVD9 titles that I authored a little while ago. I opened them in PGCedit and they ALL had the seamless box unchecked where the layer break occurs. And there was no noticeable pause during playback on the set top players of those titles.

So, based on that, having the layer break cell flagged as NON-SEAMLESS will result in SEAMLESS playback (no pause) at the layer break.

Whereas having the cell flagged as SEAMLESS at the layer break CAN (depending on the player) result in a 1-2 second pause.

So I am sticking with what my experience has showed me. The terminology of seamless and non-seamless is what was confusing me. Common sense tells you that you would want to have seamless. But as I discovered, not in the case of a layer break.

[blutach]

No, there is no pause if the cell is marked seamless (the box is ticked). Non-seamless makes the pause.

Regards

[sdsumike619]

I have to disagree with you about that, at least with regard to layer break. PGCedit even tells you in the tool tip that the box should be checked UNLESS it's the layer break position. If it's a normal video cell, then yes it should be checked. I have 8 DVD9 titles supporting this and 1 title confirming that the pause is generated when the seamless box is NOT ticked.

[r0lZ]

The fact that the pause occurs when the cell is seamless depends exclusively of the player, as anyway, it's out of specs. But trust me: many players will pause as soon as the cell is flagged as non-seamless, even if the layer break is not at that position. (This is the case, for example, when the previous cell has a cell command: the next cell must be flagged as non seamless so that the player can take some time to verify if the cell has a cell command.) My old Sony pauses systematically during 2 seconds when the seamless flag is not set, even if it has nothing special to do. But it is true that most modern (and usually cheap) standalone players, and all software players I've checked so far, do not pause any more, because they have large memory buffers.

PgcEdit offers to clear the seamless flag, as it's imposed by the standard. But it has also an option (called Seamless Layer Break) to determine if the muxing is theoretically suitable to leave the flag set, and avoid the pause.

[r0lZ]

BTW for a complete discussion about the "seamless layer break" technique and its genesis in PgcEdit, see here: http://forum.doom9.org/showthread.ph...669#post819669 (and here: http://forum.doom9.org/showthread.php?t=102185)

Pay attention to the first sentence in post #36:

Originally Posted by jamos

Playback on my Sony Home player is now Flawless at the layer break no pauses or skips with the new disk.

I think this was the first time a DVD+R DL has been burned with the "seamless layer break" technique. As you can see, setting the seamless flag suppresses the pause.

Note that the screenshots in these discussions come from old versions of PgcEdit, where the "seamless joint" flag was not present and actually replaced by the "layer break" pseudo-flag, which was exactly the opposite. So, when the flag is set in the old versions, it is clear in the new ones, and vice-versa.

[Emulgator]

I can confirm, (using that technique for over 2 years safely now on DVD+R DL)
that if one follows DVD-Video standard and sets LB cell as not seamless, it will indeed be played as non-seamless
by majority of players, buffer will be emptied, LB tilt made, buffer refilled, so stutter.

While on the same player with the same compile if you do the non-conform thing with PGCEdit and/or ImgBurn and flag LB as seamless,
you will be rewarded with smooth play over the LB, player depending maybe, not violating anything but the spec...

Tested on Panasonic DMR-EX95, Panasonic BD-50, Toshiba RD-XV47 and some more...
You may still ignore Eclipse's error report in that case...

[r0lZ]

I must add that the pause between two cells is handled very differently by different players.

For example, I have a cheap MagnuM DVD recorder that can play seamlessly all DVDs it has recorded, but pauses on ALL cells of most commercial DVDs, including on chapter cells muxed normally and with the seamless flag set! That's obviously a bug.

Most modern players do not pause at all, even if the seamless flag is clear, but they MUST play what is in their buffers before beginning to read the next cell. It's what the seamless flag clear is supposed to tell to the player. Usually, when the player is fast enough, this results in a very short and almost imperceptible pause in the sound.

Some players respect the DVD-Video standard literally, and pause for one second or so when the seamless flag is not set, even if the hardware is capable to play the cell seamlessly. This is the case of my old and expensive Sony DVP-S725D.

The fact that the seamless or non-seamless cell is used for the layer break doesn't change that much, except that the slowest players or the players with small buffers will need to pause a little bit longer, even if they are told to play the cell seamlessly. Of course, if a DVD+R is badly burned, the player may require a very long time to refocus on the second layer.

[sdsumike619]

I'm sticking with what I said. It seems you guys are talking about burned discs. I'm talking about replicated discs pressed from an actual glass master.

I have authored and had replicated 9 DVD-9 titles so far that have been run through the Eclipse system at Synchromastering in Utah. 8 of the 9 titles (I checked every single one of them) have the layer break on a non-seamless cell (empty checkbox with yellow highlight)

I have two tabletop players at my disposal:

1. Cyberhome CH-DVD 300
2. Sony DVP-NS425P (2003)

On those 8 titles with the cell set as non-seamless there is NO pause at the layer break. On this 9th title seen in the screenshot in the first post of the thread, cell 20, the layer break cell is NOT non-seamless and on both players, I get the pause.

In addition, the audience we sell our DVDs to are not very tech savvy and I'm certain I would have received some reports out of 8,000 DVDs about stuttering playback at the layer break. But I have received no such reports. This further supports that the layer break should be on a non-seamless cell to avoid the pause.

The GEAR guide I follow:
http://www.gearsoftware.com/howtogui...breakpoint.php
also states that "ideally" it should be on a non-seamless cell which I have been following on those other titles, but screwed up on this one.

[r0lZ]

There is no difference between pressed and burned discs from the layer break point of view, except perhaps that the players are usually much more rapid when they read pressed discs. That could explain your experience. But the theory remains: seamless, as its name implies, means no pause, and non-seamless means that the players can (and even should) pause, or at least flush their buffers.

[sdsumike619]

rapid is not the word here, the pause is non-existent on both of my players when the layer break happens at a cell flagged as non-seamless. Another reason I trust GEAR's guide is because they are a member of the DVD Forum which means they have access to the actual DVD spec.

[blutach]

The reason a pause might be non-existent at a non-seamless break is that the movie fades to black or a defined new scene, which you wouldn't observe. In the early days, the LB was done almost at any old place, making it very discernible. These days, disc mastering has gotten better, making it "appear seamless", while it really isn't.

Regards

[sdsumike619]

Not the case here. All of my DVDs are 4 hours (they're full) and the layer break is always happening during a talking scene (it's mainly all talking, educational) When I first started, I preferred to have the layer break in a scene dissolve, but then when I learned that the layer break is invisible (no pauses) I quit worrying about where it was.

[r0lZ]

Please note that on many old DVDs, there is a note telling that the pause during the layer change is normal and is not a defect. That means that the OFFICIAL way to author the layer break (with the seamless flag clear) really produces a pause, even with pressed DVDs. That's a fact, easy to verify. As you may have noticed if you have read the other thread, Superbit DVDs, which are also commercial pressed DVDs, break the rule and set the flag to avoid the pause.

If you don't see a pause with your players, that means that you have good players, able to skip the pause. That doesn't mean that the pause is invisible with all players. Strangely, unlike the vast majority of the players, your players seem to be unable to skip the pause when the flag is set, perhaps because the DVD has not been muxed with the correct mux rate for the seamless LB technique.

Also, the fact that someone that has read the DVD-Video specs says something is not at all a guarantee that he is right. The specs are well known to be extremely badly written and ambiguous.

Anyway, if you are happy with the flag set, that's very good, as it's the official way to do things. But I encourage you to try to set the LB on a cell where the pause is hardly noticeable, to avoid deceptions with most players. PgcEdit has a special preview of the LB (in the burn dialogue, when you have to select the LB cell) with an intentional 2 seconds pause at the LB position, to be able to select the best LB point.

[Ghitulescu]

Originally Posted by Emulgator

You may still ignore Eclipse's error report in that case...

No!

You may ignore it, as an individual, but not someone that presses DVDs. It might be that the replicators refuse your image, because it doesn't fulfil the DVD requirements. End of story.

You can ignore it for your DVDs at home ... because you're responsible only for these