DVDPlanner

[Amnon82]

I am going away for a couple of days, so will investigate when i'm back.

... still on vacation ?!?

[Zeul]

nope
I have been rewriting the muxer and making loads of changes/fixes. Actually as it happens i hope to have the last fix done tonight.

[Zeul]

0.1.4
Fix:
Selecting an Audio or Subtitle track before any plays had been added and attempting to alter the Language/Extension would cause a crash
The mux order for audio/subs was not reflected in the actual VOB if loaded from a scenarist script
If muxing is halted then an error would show
Field Based Video muxing was broken in previous release
Complete rewrite of the muxing engine to correct an issue with consecutive I frames in the video - speed has also been increased by ~40%
Some parts of the GUI were still accessible during mux
If an asset has pan-scan data embedded, this is now confirmed by the presence of pic_display_ext OR if the horizontal size in the seq_display_ext <> actual horizontal
Some VMI commands (if imported via scp) were not being coded correctly
Button commands were only corrected during mux, so importing an scp and then exporting as .dpp would give incorrect button commands if reimported
Several PGC corrections had to be made including a rewrite of the tmapti table for a slide show
AutoAction flag was not being read from an scp file
Chapter position was not being forced if the preceeding VOBU was < 0.4s (the preceeding VOBU should have been extended)
Cutting a track would not remove it fom the available track list
Removing tracks / pgcs/pgs etc didn't always work correctly.

Correction:
LetterBoxed source can now only be set if the menu/title source has aspect ratio of 4:3

Addition:
You can now set the final slide to carry through to the audio end instead of truncating
Each Audio or Subtitle stream can have the muxing ID specified. Obviously all ID details (language, extension, displaymode, audio type etc) must match within the current domain; this is verified
When a track has been added to a PGC the current audio / sub IDs are 'locked' for that track
Display Mode of the video track can now be set. If the mode is 'Both PanScan and LetterBox' then all 3 substreams are now supported
A warning is now shown in the subtitle panel if a displaymode is missing from this language/extension. If a stream has multiple modes like wide/pan then this is counted as an individual.

Please check your e-mails

[bigotti5]

There seems to be a problem with the VTS_VOBU_ADMAP

two tracks, video and audio, - two titles - one track in each title - navigation commands - create

first navpack of vobid2 appears two times in VTS_VOBU_ADMAP

e.g

[0000002c] VOBU_11: start sector 2517 [000009d5]
[00000030] VOBU_12: start sector 2720 [00000aa0]
[00000034] VOBU_13: start sector 2922 [00000b6a]
[00000038] VOBU_14: start sector 3084 [00000c0c]
[0000003c] VOBU_15: start sector 3160 [00000c58]
[00000040] VOBU_16: start sector 3178 [00000c6a]
[00000044] VOBU_17: start sector 3196 [00000c7c]
[00000048] VOBU_18: start sector 3213 [00000c8d]
[0000004c] VOBU_19: start sector 3213 [00000c8d]
[00000050] VOBU_20: start sector 3225 [00000c99]
[00000054] VOBU_21: start sector 3240 [00000ca8]
[00000058] VOBU_22: start sector 3372 [00000d2c]
[0000005c] VOBU_23: start sector 3519 [00000dbf]
[00000060] VOBU_24: start sector 3653 [00000e45]

Edit:

If you remove a program from a pgc and add a new program the PTT_SRPTI table will be wrong (more chapters than programs)


PGC_1 (program chain): [Title(TTN): 1] [00:00:09.23 / 25 fps] (Programs: 2) (Cells: 2) (uses VOB-IDs: 1)
[Ch 01] [Pg 01] [Cell 01] [V/C Id: 1/ 1] : time: 00:00:04.04 / 25 fps [Pos: 00:00:04.04] [Frames: 104] SP/ILVU/DISC/SA:[ no/ no/yes/ no]
[Ch 02] [Pg 02] [Cell 02] [V/C Id: 1/ 2] : time: 00:00:05.19 / 25 fps [Pos: 00:00:09.23] [Frames: 248] SP/ILVU/DISC/SA:[yes/ no/ no/ no]

PGC_2 (program chain): [Title(TTN): 2] [00:00:09.23 / 25 fps] (Programs: 2) (Cells: 2) (uses VOB-IDs: 2)
[Ch 01] [Pg 01] [Cell 01] [V/C Id: 2/ 1] : time: 00:00:04.04 / 25 fps [Pos: 00:00:04.04] [Frames: 104] SP/ILVU/DISC/SA:[ no/ no/yes/ no]
[Ch 02] [Pg 02] [Cell 02] [V/C Id: 2/ 2] Layer Br.: time: 00:00:05.19 / 25 fps [Pos: 00:00:09.23] [Frames: 248] SP/ILVU/DISC/SA:[ no/ no/yes/ no]

TTU_1 PTTs:
[00000010] PTT_1: Programm Chain number (PGCN) 1 [0001]
[00000012] PTT_1: Programm number (PGN) 1 [0001]
[00000014] PTT_2: Programm Chain number (PGCN) 1 [0001]
[00000016] PTT_2: Programm number (PGN) 2 [0002]
TTU_2 PTTs:
[00000018] PTT_1: Programm Chain number PGCN 2 [0002]
[0000001a] PTT_1: Programm number PG 1 [0001]
[0000001c] PTT_2: Programm Chain number PGCN 2 [0002]
[0000001e] PTT_2: Programm number PG 2 [0002]
[00000020] PTT_3: Programm Chain number PGCN 2 [0002]
[00000022] PTT_3: Programm number PG 3 [0003]
[00000024] PTT_4: Programm Chain number PGCN 2 [0002]
[00000026] PTT_4: Programm number PG 4 [0004]

[Amnon82]

Yabadabadoo. A working version !!! I'll post you the bugs I find on my system using DVDPlanner 0.1.4 with PAL sources.

I started my first mux ... so be tuned ;)

Here the first bugs:

http://img377.imageshack.us/img377/9685/runtimeerror3836yb.png

1] If I click on 'Video Play List' I got the runtimeerror 383. If I click on the cross it is ok.
2] Also if MCAffe InternetSecurity 2006 is running the muxroutine goes crazy. I must disable the antivirus for this. Take your time ;)

http://img48.imageshack.us/img48/8233/runtimeerror139vs.png

3] I wanted to add a sst subtitle created by SubRip 1.30 beta and got the runtime error 13

Some questions:

1] The 'BOV Track Creator'-tab is empty, right? When yes, why not hide?
2] Is it right, that I only can add 1 sst atm.?
3] How do you convert frame numbers to timestamps 00:00:00:00?
4] What language do you use to code your program?
5] Is there a need to analyse video for chapter placement?
6] What does the analysis of the video does? Check the framecount?

[bigotti5]

Some minor things

In a title, if there are jump, link or call commands in pre-, post- and/or cellcommands title_playback_type in VMG_PTT_SRPT should not be '0'

Creating a PAL project without any menu vob, menu related attributes should contain 0 and not 4096
VMGM_MAT, VTSI_MAT offset 0100
VMGM_VTS_ATRT offset 0014

[DVD] ERROR 4041 (ref. DVD-3 4.1.1) :
ERR_DVD_DATA_FOUND_WITHOUT_VOBS
When no VOBS are associated with the VMGI, indicated by a zero value of VMGM_VOBS_SA, the Video, Audio and Sub-picture stream attributes should specify ‘0’ in every bit. This error indicates some bits from an attribute field were non-zero.

[DVD] ERROR 4224 (ref. DVD-3 4.2.1) :
ERR_DVD_VTS_DATA_FOUND_WITHOUT_VOBS
When no Menu-VOBS exist, all Menu-related attributes should contain ‘0’. (VTSM_V_ATR, VTSM_AST_ATR, VTSM_SPST_ATR.)

[mpucoder]

That's one of the wonderful things about the Philips verifier - if you make the video attributes all zero (which really is what they should be) you will get
>>> [DVD] ERROR 5607 (ref. DVD-3 Annex N, Table N-2) :
VTSI: The VTS_ATRT VTS_V_ATR specifies a TV system 1 (PAL) different
from the TV system 0 (NTSC) specified by the VMGI_MAT VMGM_V_ATR.

No matter what you do, you will get an error with a PAL DVD without menu video.

[bigotti5]

other verifiers (interra) will complain this too if it is non zero (without getting an error as philips 5607)

Imho assets with sequence_display_extension are misinterpreted sometimes

'PanScanData present' should only appear if 'display size horizontal' in 'sequence_display_extension' is smaller than 'horizontal size' in 'sequence header'

[Zeul]

1] If I click on 'Video Play List' I got the runtimeerror 383. If I click on the cross it is ok.

Dumb error. Fixed


2] Also if MCAffe InternetSecurity 2006 is running the muxroutine goes crazy. I must disable the antivirus for this. Take your time ;)

All i am doing in reality is opening a file, writing to it and then closing it (for the VOB files), or open a file and read it (for the video, audio, subs). Is mcaffe scanning the file everytime it is written too?


3] I wanted to add a sst subtitle created by SubRip 1.30 beta and got the runtime error 13

I haven't used the latest beta, so i will give that a run and fix.


Some questions:
1] The 'BOV Track Creator'-tab is empty, right? When yes, why not hide?
2] Is it right, that I only can add 1 sst atm.?
3] How do you convert frame numbers to timestamps 00:00:00:00?
4] What language do you use to code your program?
5] Is there a need to analyse video for chapter placement?
6] What does the analysis of the video does? Check the framecount?

1) Correct. The next major release i hope will have some bov management
2) Incorrect. All 32 streams can be used. When you have added a sst, if you click on the 'Subtitle Streams' the add subtitle stream icon will be enabled. DVDPlanner will not allow you to open multiple streams without actually putting any subs in them. When the stream has subs then the next stream can be used (as shown by the enbeling of the subicon)
3) a simple frame to timestamp routine that requires the framecount and the formatstandard passed to it.
4) VB
5+6) Chapters must be placed at an I frame GOP header. You can specify a chapter place and DVDPlanner will put the cell change on the next available I frame. If you perform an analysis first the I frame positions are noted, so when you add the chapter it will positioned correctly on the chapter list. It will also ensure that you do not add chapter positions that would create a VOBU < 0.4s or > video end. It basically does the same as Scenarist does when you import a video file.

[Zeul]

Imho assets with sequence_display_extension are misinterpreted sometimes

'PanScanData present' should only appear if 'display size horizontal' in 'sequence_display_extension' is smaller than 'horizontal size' in 'sequence header'

Interesting. My understanding was that the fact that the header was there implied panscan. Perhaps mpucoder would comment :D

[bigotti5]

Interesting. My understanding was that the fact that the header was there implied panscan. Perhaps mpucoder would comment

My suggestion is only the way scenarist handles such files

TmpegEnc add a sequence_display_extension by default in a 4:3 stream and so it is not handled properly

[mpucoder]

I test for either a Picture_Display_Extension (any value) or Sequence_Display_Extension with width == 540. This is the only legal width for DVD panscan.

[Zeul]

I test for either a Picture_Display_Extension (any value) or Sequence_Display_Extension with width == 540. This is the only legal width for DVD panscan.

Thanks

[Zeul]

0.1.5
Fix:
Pressing the Video Play List caused a runtime error.
Duplicated entries in VOBU_ADMAP/VTS_C_ADT tables
Mulitple errors in IFOs caused by removing vts/title/languages/pgcs/programs/cells
Removing a title caused an error
Adding a Language and PGC (under certain circumstances) caused an error
PanScan is now only recognised if sde=540 or by the presence of pde
Trying to view a subpicture without clicking a subtitle in the viewer caused an error
Because subrip sst files count from 0 in the color palette (instead of 1-16), a crash was caused when trying to view.

Correction:
If no menu vobs exist then the IFO now reflects this (all entries are 0); even though the Phillips Verifier throws an error on PAL

Addition:
Better support for subrip - in the options you can default to subrip mode or as with muxman (for uniformity) add the line 'Generator Subrip' to the sst
An ini file for some simple default settings

Change:
Removed the Add Cell button as it isn't needed

[bigotti5]

Now I get an error if I remove a cell from a program and add a new cell
"The max number of Cells (255) will be exceeded for this PGC if this scene is added"

[Amnon82]

The problem with McAfee was that I set it up wrong. It was set up to scan all files ...

[bigotti5]

If you want to add two tracks (same attributes and audio is long enough) to the same title (same PGC), it is not possible to set the joint seamless - imo it should be possible.

[weetabang]

RE McAfee Virus Checker Problem while muxing:

If you close and re-open a file while muxing, McAfee checks it every time you open it if set to scan all files. If you leave the file open throughout the writing process and just flush to disk, McAfee doesn't go crazy. It may also improve overall performance. I'm a VC++ programmer so I don't know how to do that in VB.

Hope this helps.

[Zeul]

My thinking was on the same lines - keep the file open as each vobid is written, until all sectors are written. thanks for the heads up.

[Oldeman]

Zeul,
When do you plan to make DVD Planner available for the rest of us blathering idots?

[Zeul]

@Oldeman

A free version will be available soon.

[bigotti5]

Some more minor things

Menu subpicture 720x576 in script

Item=SubPicture
{
Name=Subtitle Stream 1
DisplayMode=Normal
DropFrame=0
SubLanguage=English
SubExtension=Unspecified
StreamNumber=0
Item=SubPicture Play
{
File=D:\path\menu.bmp
Origin=0,0
DisplayArea=0,2,719,574
Forced=Yes
Start=00:00:00:00
......
after muxing
Display Area: x=0 x1=719 y=0 y1=575
imo should be
Display Area: x=0 x1=719 y=2 y1=574



Open a scenarist script with menu causes strange values for "Highlight end time" and "Button selection end time" in menu vob.
Without modifying the script I got -1 (ffffffff) but imo this is only valid for stills, it should hold the same value as vobu presentation end time of the last vobu from this cell.
Shortening the script values by one frame gives a value 3600 ticks shorter than last vobu presentation end time.
So how to set up the script to achieve correct values?


Button group type has incorrect values for a 4:3 stream
Offset 009B shows 4352 (1100) but for 4:3 it should be 4096 (1000)

[Zeul]

-1 should appear if the cell still time > 0 (the type of cell - still, slide,motion is unimportant).

[bigotti5]

cell still time == 0 in my case

[Zeul]

it seems i am setting the endtime to = -1 if the highlight end time = end_ptm; which is incorrect. I am changing that now.

[bigotti5]

Highlights in Menus need to be switched on and off on cell boundaries so highlight start/end time must always be equal to cell start/end_ptm if cell still time == 0

[Amnon82]

What are the quickfixes in 0.1.5.1?

[Zeul]

0.1.5.1
Fix:
Viewing a subpic a 2nd time could cause a crash
Left a debug line in the code stopping the addition of cells
and something else i can't remember :D

[Amnon82]

@Oldeman, I think maybe next weekend. Some stuff is to test. As you can read also the GUI is unstable atm. If you want to test the same releases I do atm write Zeul an email.

[Amnon82]

@Zeul: Can you add CLI commands? Check your email ...

[Zeul]

@Annon82
Check you pm

[Amnon82]

@Zeul I did. THX it helps me. keep on going.

[Zeul]

0.1.5.2
Fix:
Incorrect highlight button modes were set for 4:3 sources
A 4:3 PS/Letter source did not have the correct highlight button mode set
If bmps were greater than the max resolution then the spu details were incorrect
Audio Delay was not imported from a dpp file correctly

Change:
Altered the I/O routines. This has improved speed further and should stop the McAfee problems

Addition:
Improved the way subtitle streams are handled. When added to a title/language, DVDPlanner will now try to associate corresponding language/extension display modes to the same track (eg English Wide and English LetterBox will be grouped together creating 1 track. Selecting Maximum Streams would now show 1 subtitle; whereas Custom streams would now show 2 streams available - only 1is available for menus)
Added C/L support:
-i InputFile (.dpp or .scp)
-d Destination Folder -> Only the root folder is needed, DVDPlanner will create the VIDEO_TS folder itself
-create Starts the DVD Creation
-exit DVDPlanner will close after compile
-analyse Each video file will be analysed on import for accurate chapter adjustment if the script is incorrect

[Amnon82]

Yes you did it !!! Also the analyse. I'm now adding it to AutoQ2 Internal Releases and test a little. Hope the freeware version will be released soon.

[bigotti5]

Because subrip sst files count from 0 in the color palette (instead of 1-16), a crash was caused when trying to view.

Played around with subrip (ver 1.17.1, 1.30b10, 1.40b3, 1.50b3)
palette count is always from 1-16 (palette count is always from 0-15 if you use son format which is correct for maestro)

but Pixel_Area and Display_Area are calculated wrong

Sst claims

Max image size
NTSC:720x480
PAL:720x576

Pixel_Area (Py0 Py1)

Py0 >= 0
Py1 <=image height –1
Py0 < Py1
Py1 – Py0 <= 477(NTSC)/ 572(PAL)

Display_Area (Dx0 Dy0 Dx1 Dy1)

Dx0 >= 0, Dx1 <= 719
Dx0 < Dx1
Dy0 >=2
Dy1 <= 479(NTSC)/574(PAL) (not 575)
Dx1 – Dx0 = image width –1
Dy1 – Dy0 = Py1 – Py0 <= 477(NTSC)/ 572(PAL)

Subrip calculates Py1 as image height and not image height -1
and Py1 – Py0 is not <= 572 (testedPAL only)

[drob]

Regarding the xpstyle rendering, i really think we can do without it if it means quicker loading and rendering time.

[Zeul]

I have already added an option to disable XP themes

[bigotti5]

Now Highlight end PTM is equal to cell presentation end time in a still (tested still time 255 and 10) and not -1.

In motionvideo with still time specified, it works as it should, value -1.

Start new project by button wont let you choose PAL - NTSC, so change requires restart?

[Amnon82]

AutoQ2 displays now the progressbars of DVDPlanner 0.1.5.2 in the GUI. For sources of GetProgress v2 modified click here (http://forum.doom9.org/showthread.php?goto=newpost&t=110413)

@Zeul: Check your PM for the latest AutoQ2 NON PUBLIC release.

[Amnon82]

@Zeul: Drob found a bug in the CLI-commands of DVDPlanner. Please add the "-signs in front and in the back of the input/output pathes. Read your PM.

[Zeul]

what is the bug. a c/l like:

DVDPlanner.exe -i c:\muxtest\command_line_test.dpp -d c:\muxtest -create -exit

works for me no problem

[Amnon82]

I know that your testline works, but if you have a line like

DVDPlanner.exe -i c:\program files\work folder\project files\dvd test\test dvd.dpp -d c:\program files\work folder\project files\dvd test\ -create -exit

... creates a problem like drob reported me. I'll look tonight into it and post you my results.

Why not do it like the others do:

DVDPlanner.exe -i "c:\program files\work folder\project files\dvd test\test dvd.dpp" -d "c:\program files\work folder\project files\dvd test\" -create -exit

[Zeul]

@Amnon82
The c/l has been changed as requested, the dpp file and destination folder should now be enclosed in quotes.
The dpp format has been changed very slightly. Please alter your dpp creation in AutoQ2 to as follows:
The video standard of the video should now show in the track information NOT the settings.

Item=Track
{
Name=Track_1
DisplayMode=Only PanScan
Resolution=NTSC
Item=Video Play List
...

[Amnon82]

@Zeul: Thx. The download links are still 0.1.5.2. What version number do the new build have? Can you PM me the link or mail it to me.

[Zeul]

The new release isn't quite ready yet - perhaps today

[Amnon82]

No problem. I code AutoQ3 today ...

Great news: The freeware will come tonight or tomorrow ...

@zeul: I can't wait :p

[Zeul]

0.1.6
Fix:
The highlight end time for stills was broken in previous build.
Recording the button commands was flawed for multiple highlights per cell or when the same track scene was used across several domains
LinkCN was searching for the scene being used instead of the actual cell name
When analysing video/stills for pan-scan data, if the source had color description info, the analysis was off by 1 bit, resulting in incorrect display modes
The audio offset in the SYNCI tables could reference the incorrect audio packet if the previous audio packet spanned 2 vobus
The audio buffers were incorrectly set on loadup resulting in potential buffer overflows
If a link command was specified in a dpp file - it would be incorrectly processed
The DTS timestamps for field based sources of the 'B' frames were incorrect - resulting in buffer overflows
It was possible to try and remove the Audio Streams header - caused a crash

Change:
Removed the built in resource file for XP themes (an external manifest is used instead). There is now an option to enable/disable XP themes. If enabled loadup is slower.
--If using c/l i suggest renaming the manifest file to disable XP theme, and then resetting the filename afterwards
Altered the way the track list is handled so that a combination of PAL and NTSC tracks can be made - of course only one standard can be used in the DVD structure
--Slight alteration to the dpp file to reflect this change - the video standard is now required for each track
--The global standard is set by the first cell that is added to the structure (which is reset in the event of a clear)
--This will prevent any errors thrown by importing a scp/dpp that has PAL and NTSC tracks
The startup screen has been removed as a consequence of above
Altered the c/l slightly; the input file and the destination folder should now be enclosed, like -i "c:\test.dpp" -d "c:\test_out"

Addition:
DTS audio can now be used (I don't have any DTS sources of 1536kbps to test - but the code should be sound) - at least 1 AC3 stream must be present in a menu (if audio is being used)
PCM audio has also been added (I have no multichannel (ie > 2 channels) pcm to test, so although the muxer should be correct the IFO won't show the multichannel extensions
When a PCM stream is added the Emphasis and Mute flags can be set
Improved detection of exceesive bitrate errors and underflows
--If underflows are detected then an option is now given to continue mux or abort (the total number of underflows are reported in one of the system panels)
A check on the number of VMI commands, to ensure that the limit of 128 per PGC is not broken
The Video/Audio/Sub panels are hidden from view during import, reducing track building time
If required the same audio file can be used for several streams/plays (previously it was only allowed once per track)
Added a new button to clear the current structure (this will also re-enable the audio/subtitle track details if they had been disabled due to being used)
The VBV delay is now corrected in the mpeg stream if it is not set to 65535
A more robust pre-check is now under taken to ensure no spec violations
Random play and shuffle play can now be used
If there are script errors pertaining to pre/post/cell/button commands then a list will be shown at the end of the script import
--with an option to dump the list. If there are errors then this will abort any c/l options

I am working on the freeware now :D

[Zeul]

see this thread from now on:

http://forum.doom9.org/showthread.php?t=110943