#1  
01-10-2019, 07:33 PM
jwillis84 jwillis84 is offline
Free Member
 
Join Date: Aug 2017
Location: College Station, TX
Posts: 800
Thanked 217 Times in 174 Posts
Fumbling around I found the HDD from my Toshiba RD-XS32 DVD recorder got "put in my computer" one evening.

It turned out all the recordings were stored in a single UDF 2.0 volume, which could be mounted as a readable file system under Linux or Windows.

Inside were several folders:

TSB0003
-TS_HDDAV
-TS_LBBRY
-TS_LOG
-TS_MANGR
-TS_PFDAT
-TS_THMNL

Inside the TS_HDDAV were four files:

TMAPDATA.IFO
TS_HDDMG.BUP
TS_HDDMG.IFO
TS_HDDMV.DAT

Peering at the IFO files

The TS_HDDMG.IFO seems to be a Video Managment VRO Information File with a list of Three recording titles.

Aiming Womble at the TS_HDDMV.DAT file and running Tools > MPEG MBS Scanner finds Three separate timebase segments.. which it quickly extracts to its timeline as three segments without re-encoding. Playback is fine, segments can be individually selected and deleted or moved around.

So I can pull the recordings off.. at least if they are not fragmented. I know DVD_VR allows for fragmentation and a DVD recorder probably will at some point.

Are there better tools for extracting VRO volumes? Or extracting DVD recorder videos?



I am brand new to reading VRO files.. I don't know what tools there are, or what I should be looking at.

I've heard of the Panasonic DVD Movie Album software that came with their DVD-RAM camcorders but it seems to have vanished from the Internet or is no longer available.

One of the interesting things about using an old Toshiba RD-XS is apparently the HDD could be swapped from model to model and they could playback each others files.. so I don't think the format changed throughout the series. A method for extracting recordings on one should work on all the others.

.. I guess I could reformat the recorder HDD to run in non-VR mode, which I guess would be DVD-Video mode.. and "maybe" it would produce only non-fragmented DVD-Video video_ts folders? (I am totally speculating here.. DVD storage formats is "way" outside my wheelhouse..).. but if they're all smashed end to end in a video file, and Womble breaks them out as segments by timebase resets.. maybe that is a better way? Other than "frame accurate" edits that VR mode promised.. I think they were about the same.

Last edited by jwillis84; 01-10-2019 at 07:54 PM.
Reply With Quote
Someday, 12:01 PM
admin's Avatar
Ads / Sponsors
 
Join Date: ∞
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
  #2  
01-10-2019, 08:59 PM
jwillis84 jwillis84 is offline
Free Member
 
Join Date: Aug 2017
Location: College Station, TX
Posts: 800
Thanked 217 Times in 174 Posts
Womble MBS details here > VRO REC

and two pics

interestingly it preserves the Close Caption stream from PBS and plays back on VLC when switched on


Attached Images
File Type: jpg wom-pic1.jpg (66.0 KB, 11 downloads)
File Type: jpg wom-pic2.jpg (38.1 KB, 9 downloads)
Attached Files
File Type: txt Untitled-Dat.txt (3.6 KB, 6 downloads)

Last edited by jwillis84; 01-10-2019 at 09:24 PM.
Reply With Quote
  #3  
01-12-2019, 10:34 PM
jwillis84 jwillis84 is offline
Free Member
 
Join Date: Aug 2017
Location: College Station, TX
Posts: 800
Thanked 217 Times in 174 Posts
What a challenging headache. I've learned much about the DVD-Forum and DVD+/-VR in the last 24 hrs, and the tools that until only recently ceased being available.. one outstanding tool is still available and an opensource project that is still underway in 2018.

Getting over the UDF format and the file system was only part of the problem. The +/-VR video storage files contain the segments and metadata that identifies the individual recordings stored within them. Things like Line Input, Time and Date recorded, User or EPG assigned Program Title, ect.. In the recent past Camcorders by the same companies that produced the DVD recorders tended to "shared" similar or exact same formats. To "Edit" the VR video on those Camcorders they often included Matsushita developed and rebranded software, because they were a member of the DVD-Forum which defined the "VR" file storage method. DVR/Camcorder equipment would be hooked up to a PC and then either edit the files on Internal HDD on the Camcorder, or pull them off and to the PC for editing.

To make things more of a headache, there are two formats +VR and-VR, +VR being defined and developed by Philips and Sony and -VR being developed by Toshiba and all the rest. So for exampled, Toshiba DVR recorders tended to use DVD-VR principles when storing video in the .IFO and .VRO files and Philips or Magnavox tended to use +VR.

They did not exactly follow the same Spec blueprints when making their DVR recorders. For one thing they were sold "while" these specifications were still be worked out in technical working groups... so the Specs didn't really exist, they just developed them side by side and never expected anyone would look at them and compare.

This is all very old news.. and I am not the first to come down this road.. many people in the US and Japan have made attempts at deciphering the file system (often UDF with -VR) and the storage format in the files.

It "looks" like to me that the +/-VR styles differ mostly in how they implement "editing" or "fragmentation" support on existing files. +VR actually removes and reassigns blocks to different files at the time of Edit, where -VR creates an "Edit List" with pointers to different segment blocks on the same disk and doesn't actually rewrite the blocks until "finalized".

This would be why you can high speed Dub a -VR recording back to a HDD from DVD media and get back your edits and be able to resume a project.

The difference also comes during "finalization". DVD+VR has "already" made the changes to the DVD disc so that it can be played in a DVD player.. so there is "no" real finalization step.. its said to be "always play ready" or more "DVD player compatible"

But DVD-VR needs a "real" finalization step, to close out the tracks.. essentially "commit" the edits to tracks by re-sequencing the blocks into a final form that a DVD player can actually read.. and this is why it takes time. -- it needs to make the track more "DVD Player compatible"

However DVD-VR can accidentally end up unreadable to both a DVD-VR machine and a DVD-R machine because during this transformation from a DVD-VR disc into a DVD-R disc it is neither "fish nor fowl" if the recorder looses power during "finalization" it becomes a mess of half finished tracks and segments and blocks in the middle of moving around.. and data loss can occur.

There are advantages to both formats +VR and -VR, one is perfect for fast edits and quick ready to play discs, but not as flexible for editors.. there is no "undo" with +VR, there is infinite "undo" with -VR -- at least that's how I naively understand it at the moment.

The actual specs are hard to come by for DVD-VR as the DVD forum sells them for a lot of money. The DVD+VR specs I think would be easier to come by.. but they are long out of business.. or ceased to be an organization. Even the DVD forum is in "reduced" activity mode.

+VR is also basically used on expensive editors but cheap playback devices, where -VR is used on expensive editors (more so than +VR editors) and potentially cheap or expensive playback devices.. since until its finalized, only an expensive -VR edit machines of the exact same type that understands that flavor of -VR can playback the recordings.

Compared to a +VR disc that is almost always playback ready for a +VR editor or +R playback device.

-VR playback requires more CPU horsepower and closer to a full power PC level machine, essentially an edit station.

+VR playback requires less CPU horsepower and closer to a cheap bargain basement microprocessor, essentially a walmart special

this might explain a little of the difference in price for a Toshiba, Panasonic -VR recorder as opposed to a Funai or Magnavox +VR recorder (and) the subsequent "loss" of editing flexibility and features

JVC joined the -VR empire and placed premium features in their products.. and there was an actual organization like the DVD-forum to join. The +VR "rebel" alliance was cheaper, anyone could join, there were no membership dues.. but less organized and less strict.. a few manufacturers could dominate by fiat.. and their products were ultimately much cheaper to produce and sell. --- hence cheap overseas companies tended to adopt +VR first, and adopt "All" formats later as the specifications became better known and were reverse engineered.

DVD-RAM to my knowledge was always -VR, although it could in theory be treated as raw HDD and randomly accessed, but the +VR alliance never took up the format judging it too expensive a media type.

BluRay has similar features, but I have not dove into trying to understand it.

Of course this means -VR tracks underneath should always be contiguous, hence why Womble could separate them out using its MBS scanner by timebase resets without understanding the IFO files. I'm not so sure this would work on a Maganavox +VR format recorder however.. but it might, since they are always supposed to be "play ready".. the losses would mostly be in the area of Metadata to identify the files.

The real problem with +VR recorders is that they were not beholden to use the UDF volume format (since they were like "rebels" ya'kno?) and could make up their own file system to hold the files if they wanted. HDDFs 0.7 often pops up.. and I've never heard of that volume format. This is the layer somewhat like formatting a hard disk with FAT32. On top of that they would place the IFO and VOB type files to hold recordings.


CPRM handling is interesting as well, but I haven't deep dived on that either.. it seems to implemented not by encryption.. but by operating system policy enforcement. That would make sense so that a wholely different storage format with different rules and calculation for file size or time remaining would be required.. and actually hidden by the compression ratio, if any of the encryption.. encryption actually mostly "expands" and does not compress.. or "bloats?" recording size. A CSS for example would need to comingle or otherwise transform a DVD-Video recording on HDD if they went with that method.

Last edited by jwillis84; 01-12-2019 at 11:27 PM.
Reply With Quote
The following users thank jwillis84 for this useful post: sanlyn (01-12-2019), Tester (01-14-2019)
  #4  
01-18-2019, 11:14 AM
jwillis84 jwillis84 is offline
Free Member
 
Join Date: Aug 2017
Location: College Station, TX
Posts: 800
Thanked 217 Times in 174 Posts
I've made some really good progress on this.

I burned a DVD-RAM disc of the same content and could compare the DVD-RAM -VR files to the HDD -VR files.

It looks disarmingly similiar and deceptively simple.

The TMAPDATA.IFO file is a TMAP (or Time to VOBU map). Each recording has a TMAP (section or list), which indicates which VOBU with snippets of video to be played, one after another until done. VOBUs are stored in VOBs but in this case the huge .VRO file on a DVD-RAM disc serves as the one stop shopping cart for "all" VOBUs. (In a normal DVD-Video disc recordings are broken out in to "Title Sets" or TS_VideoXX directories with separate VOB files and VOBU snippets stored inside those..)

The HDD drive has a .DAT file that serves as the .VRO and one (huge single) .IFO file for (all recordings) without their TMAP data, instead TMAP data is kept in the one TMAPDATA.DAT file.. so it all kind of makes sense. - the One (TS_HDDMG.IFO file) lists recordings, which point to their TMAP data section in TMAPDATA.IFO which point to "VOBUs" or blocks of video in TS_HDDMV.DAT.

TS_HDDMG.IFO is kind of a recording "metadata" directory with human readable information
TMAPDATA.IFO is kind of a "File Allocation Table" which points to blocks called "VOBUs" stored in
TS_HDDMV.DAT which is the Hard Disk Drive "Movie" storage file

This (is Not) the simple approach I was originally interested in.. but since the DVD-Forum proved a barrier to understanding.. it got my curiosity up and I kept looking.

You can throw away the TS_HDDMG.IFO and ignore the TMAPDATA.IFO if you don't care about discontiguous or "edited" recordings.. an MPEG player will mostly handle any hiccups reading the TS_HDDMV.DAT and Womble can even slice it up nicely when there are time base "jumps" in the playback.. it just assumes they're recording splits.

So why didn't they just go with a flat file and separate MPG files for recordings?

1. Metadata - mpg files don't really have a standard place for it other than the file name
2. VOBUs - as a dvd recorder the eventual target is to produce DVDs which are based on VOBUs "rendering" at the time of burning would take a lot of CPU power and time, storing it in the format required to make a DVD saves a lot of time. (and) These DVD recorders could "record direct to DVD media" bypassing the HDD.. so working in VOBU space was especially important.

Emulating the whole -VR storage method makes converting that into DVD-RAM -VR or other -VR media a lot easier, its just a few changes or fixups and its ready to burn, practically something that could be juggled in a very small RAM memory space.

The downside of course is people are - kind of - forced into learning more about DVD-video and the limitations that enforces on their work flow.. more than they would like.

A cheap PC doesn't have software that stores in the -VR style format, and the dark arts of mystical hidden -VR formats due to DVD-forum secrecy.. more or less "ensured" the death of the platform once businesses with proprietary interests left the market.. a fairly short sighted point of view in my opinon.

Of course there are costly PC software "DVD Authoring" packages which can do it for you.

Last edited by jwillis84; 01-18-2019 at 11:39 AM.
Reply With Quote
  #5  
01-18-2019, 11:24 AM
lordsmurf's Avatar
lordsmurf lordsmurf is online now
Site Staff | Video
 
Join Date: Dec 2002
Posts: 13,510
Thanked 2,449 Times in 2,081 Posts
Just replying so you know others are reading.

- Did my advice help you? Then become a Premium Member and support this site.
- For sale in the marketplace: TBCs, workflows, capture cards, VCRs
Reply With Quote
The following users thank lordsmurf for this useful post: jwillis84 (01-18-2019)
  #6  
01-18-2019, 05:46 PM
jwillis84 jwillis84 is offline
Free Member
 
Join Date: Aug 2017
Location: College Station, TX
Posts: 800
Thanked 217 Times in 174 Posts
This website has a very good diagram of how IFO and VRO files are related.

www.pixelbeat.org

BUP files are simply duplicated IFO files designed to be recorded after the VRO so that they wind up in a different location on a DVD disc in case the first copy is damaged. They serve the same purpose on a HDD as a backup or duplicate copy of the file allocation table.

On the Toshiba HDD the "VOB Map" (which is the TMAP) is not located "inside" the IFO file, but stored in a file outside called TMAPDATA.IFO

An often cited problem with "file" management is "file size" is not recorded for each program, but has to be calculated or estimated from the "Metadata" for a recording.. but that is not a direct measure of the space consumed by all the VOBUs associated with one recording (nor is it accurate).

A calculation from Metadata would go - like minutes x MB/min for that recording speed, which would only be a worst case scenario and maybe not accurate after editing.

After "editing" and "splits or merges" (any) VOBU may end up associated with "multiple" recordings. If so the "actual" HDD space consumed by a recording may be inaccurate. It could be the HDD is "full" with both "used" and "unused" VOBUs that have not been deleted, or recycled (or) the HDD may still have unused space because multiple recordings depend on a few of the same VOBUs and so did not have to duplicate them.

This "disassociation" from traditional "file space" measurement people are use to.. adds confusion and misconception to the ideas of HDD "cleanup" and "defragmentation". Its really more akin to "shuffling the deck chairs on the Titanic" because the "real" filesystem, the VOBUs and the TMAP behave very differently.

Toshiba was not the only DVD recorder manufacturer, its hard to believe others who were members of the DVD forum would deviate too far from the -VR model for their recorders since it would put them at a competitive disadvantage, and they would have to invent a totally different way to store the same things because ultimately their devices purpose would be to create compatible DVD-video discs.

This has not been proven yet.. but it is suspected.

"Calculated file space consumed" could be used for HDD space management , but could be not the same as "Burned file space consumed" when a recording is output to DVD media.. so yeah.. this whole question of resource management was messed up. "High Speed Dubbing" would be different as well since individual IFO files would have to be created for each recording in order to remain complaint with DVD-VR, since "that" is not the same thing as HDD-VR...

what a headache.

it all sort of worked a lot better if you took the quoted "speeds and free space" measurements as "guidelines" and not real numbers to be relied upon in every situation.

Last edited by jwillis84; 01-18-2019 at 06:08 PM.
Reply With Quote
Reply




Similar Threads
Thread Thread Starter Forum Replies Last Post
Is Toshiba D-VR7-K-TC2 a quality machine for VHS to DVD? Flipper Restore, Filter, Improve Quality 11 12-08-2020 02:52 PM
Telecined program with interlaced titles processing? ELinder Restore, Filter, Improve Quality 0 12-22-2018 06:44 PM
WTB: Toshiba SV-970 S-VHS VCR Tom Monahan Marketplace 6 10-11-2015 05:16 PM
For Sale: Toshiba RD-XS34SU Nightshiver Marketplace 0 08-17-2015 09:07 PM
Titles vs. Chapters in Ulead DVD Workshop NJRoadfan Author, Make Menus, Slideshows, Burn 3 12-17-2010 12:33 AM

Thread Tools



 
All times are GMT -5. The time now is 08:49 AM