MTR 6 ANNOUNCEMENTS - READ FIRST!

[geezerbuttz]

I need to clarify the DVDSectorProbe license state.

As it stands , the SectorProbeApp has no protection , no serial numbers , etc.
Anybody can run it.

You cannot do anything with it , as it requires MTR to send it the data it needs , and yes , that data is encrypted.
The license file to 'activate' DVDSectorProbe , actually just allows MTR to send the data to the probe app.

For sure you do not need the SectorProbe app , but , there is nothing to stop MTR from tripping up and defective sectors without it , thus prolonging your extraction time.

You also do not get access to the probe library if you are not licensed.

The RnR replacement extraction engine will be completely dependent on it.

[geezerbuttz]

Cleaning up the bundles.

The App Image was in excess of 50Mb.. to big.
I spent the day optimising all graphic work , as that was where the bloat was.
Got it down to 38(DD)Mb

All I have left , ( after I looked over the code yesterday ) is to disassociate two pieces of code from each other.
Then I can package it.

I will do a standard on-line notification , but the download package will be held online in the forum.
Only 5.0 holders will be able to download it , and only lifetime will be entitled to a free license.
( the lifetime status does not include a license to the SectorProbe application)

I will be back.

[geezerbuttz]

been quiet for a few days , going over the code , doing some obfuscation on the frameworks ..

Just realised , I need to add an addition to DVDSectorprobe output.

I , still need to check , bit I think I need the count number that a sector or file has been hit.
that way We can filter Ifo that have no effect on the disk.

Sorry , but better now than later ..Of course I will have to regenerate my IMF library ( again )

Its tiring

[geezerbuttz]

ok thats done , but tomorrow , I have to redo the IMF library due to the changes in the file structure.

Thats all thats left.

[geezerbuttz]

On a side note.
You must give probe permissions to save files on your system.
As we pass the save location from MTR to it , it cannot generate a permission request as no save panel is presented.

Just a FYI

[geezerbuttz]

The joy of extracting every single disk again...
Somebody get me a beer.

[GrizzlyHippo]

The joy of extracting every single disk again...
Somebody get me a beer.

Can I just ask why you are doing this? Does this mean that the latest MTR will give better quality rips? If so, then we'll all need beer...

[geezerbuttz]

im building the basic library of extraction Maps , so we have a basic library to start with.
Its the up to the users , who support the probe app , to start generating their own maps to get uploaded.

Once uploaded , everybody who has access gets to use them.

Complicated disks who have a map , rip in minutes.
The next advanced rip engine will be completely dependent on thier information.

[geezerbuttz]

The library is growing.

No crashes , 40 disks in...

I made one small change in the probe code , but this does not effect the file format.

When probing regions , 1 was breaking the region down into seperate regions:
i.e ...

Region 134 , became , Region 1 , 3 , 4 ...

This is ok , but I decided to also set the region mask to "134" as an addition to the region individual values.
Of course , if you check for region free ( important in RCE disks ) then it appends A bit setting of 256 ( thats all regions selected )

( A side note: I own the media of every disk Im indexing )


Progressing....

[geezerbuttz]

A small additional change in probe.

The Device Region in the UI was fixed on the initial region setting of the disk.
During investigation the region internally changes , but the UI does not reflect this.

I added the ability to reflect this in the UI.

Currently the current selected language , is not displayed...
I may see to add this.

None of the above effects the output structure , so nothing needs to be recalculated.
G

[geezerbuttz]

A rather obvious fact has risen.

Audio for subs , language , and menus can be set differently on a player.
Currently , im setting them all to the same value.

The fact that they can be set differently , can be used as an exploit in DVD navigation design.

This leads me to think that I need to implement two seperate scanning routines.

One as it is , setting all languages to the same value , and another that iterates over the languages , setting them independent of each other.

This leads to many more passes required to scan a disk , as for 5 languages , we have to do a 5x5x5 == 125 extra passes in the initial phase.
That result propagates further down the chain of navigation paths.

I will get it implemented. Best to be sure , and give the options.

[geezerbuttz]

ok , that is done ,

No change in current state outputs , but its best to have the cases covered.

default condition is OFF.

Can be changed in preferences.

I rewrote the initial scan routines to simplify the initial data creation.

Just need to add on the UI the individual languages currently running.
G

[geezerbuttz]

ok , a look over the code , and I realised that grouping all languages together , and assuming that there is a menu and subtitle or audio for each language is a mistake.

So , i just need to return from the language accessor a dictionary now of all the languages for menu/ spu/audio, and iterate with those.

It will reduce the possible combinations..

Im on it.

[geezerbuttz]

Well , i never thought that would prove to be so complicated to do.

I completely blocked yesterday , I was looking at the objC array sorting methods , and iterators..

No Joy...

Coded my own sort routine ...

Side note: it vastly improves the performance of the Deep Scan routine.
Need to test it a bit...

( so just to be clear , in deep scan mode , it builds a table incorporating for every region , every possible menu , audio and subtitle combination.
It also enables to identify when specifics channels are not set , and we can now bypass them,

we then run that as the initial pass , that sets the initial packets with the required registers to resume on stage 2, reading and changing regions, audio , etc as we go...

Its deep man , very deep.

[geezerbuttz]

Well , running the probe on multiple disks is going well , but.

Having to press start and save for every disk is a pain...

So: i will add two new prefs into MTR today to ease this .
An Auto Start Probe , And AutoSave Probe ... settings...

These will of course be in MTR.

G

[geezerbuttz]

Thats Done:

So to recap. Under general prefs you can set auto start , and auto save for probe.
This will automate the process completely when both are set , except under a read failure , of a file existence check.

Under those conditions , user interaction will be required.
G

[geezerbuttz]

Somebody posted a question:
"Why are you doing all this?"

The answer is as following.

Yesterday I encountered a disk that took 18hrs to image.
Far to long:

A little optimisation got it down to approx 12:
still far too long.

Today , after a small session , i got it down to 3 hrs for a deep image.
Much better.

Thats why Im doing this. Ironing out any nasty suprises. That can only be done by running as much as I have through it.

[geezerbuttz]

First Navigation Error on a disk.

It was getting the correct PGC but getting the cell count wrong:
Thus it was trying to read a cell outside of the table:

Added a check to make sure the cell lays within the table , and a plan of attack to adjust the cell to a valid range , if the out of range condition is met.

A few more box sets left to probe....

[geezerbuttz]

These disks from Columbia all exibit an error in navigation.
They keep trying to access / call a NextCell that doers not exist in the program chain.

I patched a check for it , but I have added a visual warning as well

It a self dismissing , but it will give you an idea of when is going on.

back to building the library.

I hope you all use this to grow the library.

[geezerbuttz]

I also have an idea , for later , to implement a bulk probe run.
I.e you load four disks , you can instruct MTR to run a probe on all disks.
This is good for overnight runs.
Of course auto mode for running and saving must be active.


That will come later.

[geezerbuttz]

An observation:

Im doing a box set that has sat on the shelf for 10 years , never opened.

Every sing disk is sticky , with a clear sticky film across the surface of every disk.
It seems that the oils from the storage case have leached out and built up on the surface over the years .
The only way to safely remove it is 90% alcohol.

Just a heads up , as this makes them unplayable.
G

[geezerbuttz]

have hit a disk that has a 100% mashed file system:

MTR has to build its table before it can be passed to probe.
It makes sense that if a IMF file for this disk exists , we should use the hints in it for the IFO Bup.
If no IM , we do it the hard way.

Its locating the BUP that takes time on these disks , so ...

Thats for later...

[geezerbuttz]

As luck would have it, this last disk is proving to be a problem for both apps.

Im on it, good job i found it now.

[geezerbuttz]

Right ,

Low Level mapping is now running..
Currently BUPs are marked as Generic when they are found.
They will be corrected in a post pass on disk completion ( if possible )

Currently the only missing part is the cell generation for defective reads.
That will be easily completed tomorrow.

These maps are so comprehensive ( low level ) that while they take a while to generate ( depending on how many sectors are defective ) they truly are a 100% mirror of the readable data on the disk.

I am thinking to allow these ( MIF ) files to be placed online with the IMF.
The loader code already can separate what files for each modes.

Of course , MIFs are only good for ISO ( at the moment he says grinning )

More soon.

[geezerbuttz]

Low level scanning is running...

Mapping of defective cells is completed ,
File structure rebuild is complete from the low level blocks on the disk. ( phew )

This disk that is broken , still takes time to mount the initial fileStructure due to missing IFO/Bups.
I will look at that tomorrow , and perhaps limit the search radius when looking for missing data.

Ultimately I think it would be bettor to have a preference to bypss the initial scan all together , presenting an empty panel , that you can then populate with a low level scan.

That can be saved for later use , uploaded ( eventually ).
This will make MTR truly 'dumb' on launch , as it will rely 100% on maps in that mode.

This of course will be added after I kick this out the door.

A few notes about this problematic disk.
For the first time I have found defective sectors in the menu system , not the main feature.

Ifo/Bup 3 and 4 are completely missing , although the VM makes a call to VTS 3.
That crashes all players.

Never come across a disk like this in all my years.

And ironically the disk is 6.66Gb in size.

[rthpjm]

Hey GB, out of idle curiosity, does this problematic disk cause trouble in a hardware player???

[geezerbuttz]

yes... But here are no visible issues on the disk, its clean.

[TheSystemHasFailed]

yes... But here are no visible issues on the disk, its clean.

I've had that happen with both DVDs and CDs. The manufacturing process used in those cases caused the CD (and DVD) to degrade over time.. Apparently the coating used wasn't sealing the disc correctly at the factory and as a result it's no longer reflecting back to the laser diode correctly.

I thought it was something I did UNTIL I saw a documentary on the archival work being done at the library of congress where they had two disc that were the same title but one on closer examination was 'fading' (media wouldn't reflect) and they explained the reason for it (one was made at the factory for the record label, one was subcontracted).

In fact I have a CD audiobook of The Fountainhead that this is happening to right now.. I cannot rip the tracks.. but I could play them.. now.. they're skipping when I play them. Of course the company I bought it from told me "sorry, but tough cookies.."

[geezerbuttz]

i have implemented a faster skip method when an unreadable sector is found.

have to test this though , as i need to make sure my sector alignment code is 100%

Of course , if there is no prior Navigation pack , or we are in an ifo , or dead space , it will fall back to 1 sector niblle mode.


Next i need to look at any method to speed up the initial read:
I think for Ifo i will go with a "if its not where it should be, dont bother looking ' approach.

But for Bup.... thats tricky

G

[geezerbuttz]

ok , i stripped and ceaned all drives and player to rule out dirty optics.
No difference , nothing can play it ,so im going to mark it as broken.

I need to run the error correction skip code a few times on disks that are known good with errorData on the disk , just to cinfirm the c=new method is working correctly.
Tie up a few loose ends because of the new code ( mostly UI related )

thenI can,hopefully package this up...