Home › forums › Archives › Discussion Archive

Shuffle play

Fri, 2005-01-14 08:07 — Archive

Posted Jan 14,2005 8:07 AM ianshepherd

Hello Chaps,

Trai described a cool technique here:

http://www.tfdvdedit.com/members/forum/openthread.cfm?forum=1&ThreadID=5...

Did anyone ( LP ? ) test it, and if so how did they get on ? And, how does it compare with Scenarist's Shuffle feature - is it the same thing ?

Ian

Posted Jan 14,2005 2:45 PM benchmarque

I recall working with it a bit, but found inconsistencies with certain parameters in v2 & v3. Values would appear differently and there was some other weird stuff going on. I need to check my system for the builds; thanks for reminding me!

-Ben

Posted Jan 14,2005 3:18 PM Trai

Hi Ian and Ben,

Yes. This capability needs some more attention. Verison 2.05 would be best to test this, I think.

Interestingly, another requirement for this to work (add this to the list of requirements in that post Ian linked to), is that the PGC set up in Shuffle mode (or Multi PGC/Shuffle Mode; i.e. not one-sequential ), cannot have a Call System or Jump command in the pre/post or Cell commands! So the only way "out" of this PGC is to link to another PGC in the Title, which means to set this up properly for most project types, is you have to have a multi-PGC title. The second PGC (set up for sequential playback in the PGC settings), a dummy PGC, essentially, would receive the link command, and from there, navigation can be sent out of the domain with a jump or call system command.

I've very keen on this capability getting verified. After John finishes this next version of Seamless Multi-Story sometime next week, hopefully, this and other neat capabilities (as well as finishing up pre-mastering), will be high on the list.

Many thanks for your awesome testing!

Trai

Posted Jan 14,2005 4:02 PM Trai

Hi Ian,

Yes. This is Scenarist Shuffle mode capability. But I don't think they've considered all the spec requirements to make it work! We'll see, eh (I love "DVD Suspense"! :-)

Take care,

Trai

Posted Jan 14,2005 5:16 PM benchmarque

In looking at my project - though it's not even close to being spec. compliant we now know - it's clear there's definitely some funky stuff going on. The build is just a modified version of one of my article projects, and has one :20 track with 4 PGs. Playback starts off promisingly, skipping around the PGs randomly without a hitch, but then pauses at the end of the 6th PG.

In addition, the Apple DVD Player's info bar shows some weird stuff: when the 4th randomly selected PG plays, the "Remaining" time changes to 1193:02, while "Elapsed" continues counting. "Chapter" displays chapter 3 no matter which PG is playing. I took the picture below after the DVD has been parked on the 6th PG for 11 minutes; notice how "Remaining" has decreased (but not by 11 minutes).

Sorry for the poor image quality - I had to use my phone to get the shot!

-Ben

center

Posted Jan 14,2005 8:48 PM Jake Russell

Ok when I run a 12 PG shuffle project through Sonic Verification in Fusion/Creator 2.1.7 I get the warnings below. So Sonic want 8 max PGs but I haven't tested Scenarist yet.

The 2nd error is not totally accurate. It is not One-Sequential-PGC of course but there is also only one PGC in the Title.

When Shuffle mode is set in the Pres Info options the option to set the PG to PTT also is greyed out. But the seamless check box is still active! So I can set the PG/Cell to seamless or non-seamless in Shuffle mode.

With regards to the funky time display, isn't that to be expected due to the fact that it is not One-Sequential PGC? I know the LCD usually has dashes but I'm unsure how the Apple DVD Player reacts.

Cheers, and I'll see what else i can find out tomorrow,

Jake

center

Posted Jan 14,2005 10:01 PM Jake Russell

Well from my Tests with the Sonic authored projects I'm finding that things are random and a PG is not repeated. I'm only testing in the OSX 10.3.x player at the moment.

With regards to the Time on the Player being messed-up, it seems to me that all works well until the last PG in the PGC is played. So if there are 6 PGs and the order is: 5, 2, 3, 4, 1, 6 all would be ok. If PG 6 is accessed earlier in the list the timecode goes funky. So 5, 2, 1, 6, 3, 4 would cause timecode issues as the 4th PG to play is the last in the PGC. Runtime is fine but the post commands are never reached.

That's what I'm finding anyway,

Jake

center

Posted Jan 14,2005 10:38 PM Jake Russell

Ok I'm also using projects with 12 PGs. I'm finding that things are still random and a PG isn't repeated. I haven't tested things in settop boxes yet though!

The timecode issues seems to be to do with erratic start times from the Apple DVD Player! I'm using PGs that are 5 sec each and if the player plays from launch it will sometimes be fine all the way through the 12 PGs but it will then play a 13th. Random of the 12 PGs in the PGC and hang.

Other times when using File > Open VIDEO_TS folder the start time is not 00:00:00:00. Some times it's 5 secs, 30 secs and so on. If it is 5 secs then the time is fine until the last PG, 5 secs from the end. If it is 30 secs then it if fine till 30 secs in as the time remaining reaches 00:00:00:00 30 secs too early.

Jake

Posted Jan 14,2005 11:46 PM Trai

Excellent work Jake!

Are you setting a loop count number? I think to get through the 12 PG's differently each time, say twice, I think you would enter "24". Then make sure a PGC still time is set, and then link over to a dummy PGC's pre command are (in the same Title) to jump back to the menu.

Many thanks. Please let us know all that transpires (when you get a chance to continue, of course),

Trai

Posted Jan 14,2005 11:48 PM Trai

Hi Jake,

I've checked the spec, and I don't see any maximum number of PG's mentioned (maybe Joe might have dug deeper? :-)

I'll keep looking.

Thanks again,

Trai

Posted Jan 14,2005 11:50 PM Trai

Hi Jake,

The spec is very clear that the cells (as programs) are to be set as non seamless in shuffle mode. Interesting stuff you and Ben are working on!

Thanks,

Trai

Posted Jan 15,2005 12:02 AM Trai

Oh yeah Ben. The Nav Packs are on this latest challenge like a duck on a bug!

A true "DVD Mystery"; that I'm sure we'll solve.

If you, or Jake or Ian, or anyone has any questions. Let me know.

Thanks again,

Trai

Posted Jan 15,2005 11:34 AM Jake Russell

Ok now it's not 3 or 4 in the morning here in the uk I am seeing things clearer!

The time starts at whatever the first PG is in the list. So in my 12 PG project, 5 secs per PG, 1 min for the PGC:

PG 1, 00:00:00:00
PG 2, 00:00:05:00
PG 3, 00:00:10:00
PG 4, 00:00:15:00
and so on....
PG 11, 00:00:50:00
PG 12, 00:00:55:00

So if PG 9 was the first PG in the Shuffle Play list, the Elapsed Time would start at 00:00:40:00 and Time Remaining would be 00:00:20:00. After 30 secs the Elapsed Time would be 00:01:10:00 and Remaining would be showing something similar to the posted screen shots. 1932:02:

I still haven't been able to get to the post-commands but I'll see what I can do,

Jake

Posted Jan 15,2005 12:34 PM Jake Russell

Ok I hadn't set the still time for every PG in the Sonic. But now it works with or without two PGCs. So just the Shuffle PGC and Jump works on the software player. But as we now know, it's best to link PGC then Jump etc.

The time now displays as you'd expect with no funky business so I'll test the projects in various generation players asap. Thus far it seems if things are set correctly that it can be done in the authoring app or Edit,

Jake

Posted Jan 15,2005 1:40 PM Jake Russell

So far so good. Old Toshibas, Philips Recorder, various Sony's all play random and don't repeat.

Except if I set the Shuffle to 12 with 12 PGs then it will play 13 PGs. So one PG is repeated after the 12. It seems if you have 12 PGs you need to set it to 11 not 12. So number of PGs - 1.

Jake

Posted Jan 15,2005 2:06 PM Trai

Lovin' it! Thanks Jake,

Trai

Posted Jan 15,2005 2:43 PM Jake Russell

Ok it's official, as if we didn't know! You MUST Link PGC then Jump. Lots of the players will be fine Jumping from a shuffle PGC but some wont. I've got two projects I'm testing and the one with the Link PGC then Jump is 100% so far. The straight Jump from the shuffle PGC post commands is not working in some players. All Sony so far. The 12 random PGs play fine but when it's time to Jump to the VMGM it stops.

So great work Trai. The Link is a very important part.

Also to get the Playback count to be 12 out of 12 it needs to be set to 11.

Cheers,

Jake

Posted Jan 15,2005 3:05 PM Jake Russell

Xbox has no probs with both versions. I've had to call it a day but I'll test the PS2 when I get a chance,

Jake

Posted Jan 15,2005 3:31 PM Trai

Thanks Jake,

You've done a great deed for the DVD industry (if you didn't already have 2 DVD producers hats, I send you another! :-)

Now, when we're able, we need to ramp up the number of PG's in the Shuffle. According to spec, we should be able to do up to 99.

Many thanks again,

Trai

Posted Jan 15,2005 4:21 PM benchmarque

I've got this working ok as far as I can tell - even in the notorious Pioneer player.

Jake is correct - the number of PGs to be playback is the desired # minus 1. My build has only 4 PGs, and setting "PG Playback Count" to 5 will display 6 PGs before exiting.

But do we need to set Cell still times to 1? I just tested my build on 2 players with the "still time" parameters (PGC & Cell) set to 0, and everything works ok.

-Ben

Posted Jan 15,2005 5:09 PM Trai

Excellent work to you too, Ben!

Set the PGC still time for 1 second. No cell still times needed. The cells just need to be set as non seamless.

Thanks again,

Trai

Posted Jan 15,2005 8:29 PM Jake Russell

Ok I went ahead and authored a 99 PG PGC project. Finished things in Edit. Well I only ever got 1 of my now 3 hats :-) and it sits right on one of my monitors...

I only had a chance to test 3 players, cheap sony, old Toshiba and xbox.

Sony DVP-NS330. Played all 99 PGs randomly not repeating any PGs in that 99 List. But instead of Linking and Jumping it continued randomly playing PGs 1-12 randomly. Then one PG played for a 3rd time before navigation continued. So a total of 112 played PGs. PG 2 was the PG that played 3 times post the random 99. As soon as PG 2 played for the 3rd time things continued as programmed.

Toshiba SD-100E. Exactly the same as the Sony above except the PG that played 3 times was PG 11. So a total of 112 PGs played. PGs 1-12 repeated randomly and when PG 11 played for the 3rd time things continued as programmed.

Xbox. Played all 99 PGs randomly not repeating any PG but after the 99th it froze. No post commands read and no extra PGs played.

Hmmmm..... to be continued....

Jake
---------------
navpack two

Posted Jan 15,2005 11:32 PM Trai

Hi Jake,

what was your loop count setting?

Thanks,

Trai

Posted Jan 16,2005 4:08 AM Jake Russell

Hi Trai,

It's 98,

Jake

Posted Jan 16,2005 6:36 AM Jake Russell

Trai & Ben, (Ian & Everyone:-)

>No cell still times needed. The cells just need to be set as non seamless.

In my earlier tests with regards to the time being displayed correctly I found that this is not the case. I found each Cell needs to be 1 sec. If it was 0 sec then at some point you would get 1193:02: etc. If it was set to 1 sec the elapsed and remaining time was always correct for the given program. Ben are you saying you can get the Apple DVD Player time to display fine for the complete cycle with a Cell Still time of 0 secs?

Of course it could just be a Apple DVD Player issue but none the less if we can get that working too then cool.

The 99 PG test I had issues with was with the Cell Still time at 0 sec! I just tested it with the time set to 1 sec for all 99 PGs. Only 98 PGs were random. The 99th was a repeat, PG 83. PG 77 was not played at all. Then it continued as above to cycle through PGs 1 - 12 before repeating PG 6 for a 3rd time and running the post commands. This test was on the Toshiba above,

I really really am going to have to stop and get back to it when I have a chance... I might quickly drop the number of programs or the count below 98 just to see if it makes a difference!

Jake

Posted Jan 16,2005 9:04 AM Jake Russell

It works! So yep I have a project the shuffles 99 PGs with no repeats at all and then goes on it's merry way. I actually have multiple shuffle PGC's on the same disc :-)

All players play without a problem so far. Xbox is also fine. I've only been able to do it in Edit so far as when I set the Cell Still Time to 1 sec in Sonic it is Still 0 sec when I open the build in Edit! I'm testing that too though...

Jake
Oh and what was the problem before I hear you cry? Lets say you can PM me as it was on the level of, is it plugged in!

Posted Jan 16,2005 9:47 AM Jake Russell

Ok I've done it with Fusion and Edit now. No probs and the time is fine when the Cell Still Time is set to 1 sec.

Jake

Posted Jan 16,2005 10:33 AM Jake Russell

Hi Trai,

I've done as you've said with regards to PGC layout. Shuffle PGC, post commands Link to PGC2 then Call VMG... but I'm looking at the Sonic site and it says that there can be no pre/post commands at all.

http://www.sonic.com/products/dvdcreator/faq.htm#pgcs

------
IMPORTANT RESTRICTIONS FOR RANDOM AND SHUFFLE

PGCs that are assigned Random or Shuffle mode cannot have Pre or Post commands associated to them.
------

Jake

Posted Jan 16,2005 4:20 PM Trai

Hi Jake,

Sonic is incorrect. They also got wrong their warning message in Creator's verifier, that you can only have "8 PG's" in a shuffled PGC. :-)

The DVD-Video Spec is very clear in several locations that certain pre and post commands for a PGC in Shuffle mode can be present; just not particular ones.

But there is another way to get navigation over to the dummy PGC if you don't want to use PGC post commands, just set the next PGC # for the Shuffled PGC, over to the dummy PGC (in the same title). If there's no post commands, this next PGC # setting for the PGC will move navigation over to the next PGC, or the one designated.

So you got things to work with 99 PG's with a cell still time of 1 second? The spec doesn't say that's necessary, but for sure, anything to help the video buffer while the player first decides on, and then seeks the next PG, couldn't hurt (maybe as the number of PG's in the project increase, the cell still time could provide that extra margine of safety?)

Great work, and thanks again Jake,

Trai

Posted Jan 16,2005 4:54 PM benchmarque

No - Apple DVD player time display is still weird. I'll go ahead and change the still times and verify....

-Ben

Posted Jan 16,2005 5:12 PM benchmarque

So I changed the Cell still times to 1, but left the PGC still time at 0 and the Apple DVD player displays the time remaining relative to the PGs linear position. I can't really tell what difference setting the PGC still time to 1 makes with this project, but if it's supposed to be that way, so be it.

I also tried deleting the post-command linking to PGC 2 and changing the Next PGC value to 2 instead, and it all seems to work the same as before.

-Ben

Posted Jan 16,2005 5:58 PM Jake Russell

Well it's working with and without the 1 sec setting on the Settop boxes and there is no LCD time of course. I was just trying to sort out the apple display too.

Yep Ben I found that if it was set to 0 secs the time remaining would cound down from which ever the first PG in the Cycle was but if it was set to 1 sec the remaining would change with each PG but never get messed up.

Jake

Posted Jan 17,2005 4:28 AM Trai

Hi Guys,

One thing I found in the spec that could derail the setting of a Cell Still time for these programs, is this sentence:

"When a PGC still is set, any Cell Still of the Cell which has the possibility to be presented lastly in the PGC shall not be set".

Since we're playing these programs in shuffle mode, and if we've set Still times for their Cells, it's very possible, and really likely that we will have a program/cell being played last with a Cell Still time, which is a no no.

SO...after reading the spec for the thirtieth time (various requirements for proper set up of a Shuffle mode PGC is spread througout the entire book!), I'm beginning to question the need for a PGC still time to be set.

The schematics shown for the Shuffle and Random PGC (3.3 System Model, Pag VI3-29 ), has a very prominent graphic that describe playback being "dumped" into the PGC still time setting. But, you know, it doesn't say that a value must be present; only that if a PGC is in One-Sequential mode the value must be "0" (no PGC still time to be set. This is where DVD SP messes up when you set a Track to pause after playback, it uses the PGC Still Time, even with it's one sequential PGCs it creates! And players will lock up, we've found through experience over the years!).

So, I think you guys are on to something with the Cell Stills. But if we set the Cell Still times for the programs (along with setting each cell to be non seamless), we really shouldn't be setting the PGC Still time with any value, to stay within spec (if my interpretation of the above quote is accurate. If it's not, then just not setting the last cell in the PGC with a still time, would do the trick, and we could then set a PGC Still time). And, like I said, after many re-readings, the spec doesn't actually say a value must be in there for the PGC Still, only that navigation "goes" into the PGC Still (before ending up in the PGCs post command area).

For sure, I don't think we need be too concerned about Apple's player making a mess of the time readouts, as Random or Shuffled, or multi-PGC Titles won't allow the players to display times. The main thing, is that we get 100% of players playing the programs properly, and then exiting the PGC back to the menu (or to a dummy PGC that uses a GPRM to keep track of the number of times we want the whole shebang to loop, if we don't want to place a large value in the loop count setting), on cue.

Please let me know if you have questions.

Thanks again,

Trai

Posted Jan 17,2005 4:49 AM Jake Russell

It works for me either way I set the Cell Still Times. Players only have a problem if the post commands don't link then jump.

FWIW, the last PG in the PGC, PG 99, has it's Still Time setting greyed out when authoring in Sonic. Isn't that due to the setting for the PGC?

Cheers,

Jake

Posted Jan 18,2005 4:58 AM ianshepherd

Great work guys - I got a little lost in the details there somewhere, maybe one of you would like to make a new thread with a summary of the technique, the final word on cell/pgc still times, and any outstanding issues we need to test ?

Thanks !

Ian

‹ Moving Button Hotspot Build uploaded! Moving buttons success ›