AUFIT Program

[dlfrsilver]

crackdown disk 1 dumped with kryoflux as raw stream files :

https://app.box.com/s/zj6hf5gyil6epn2dekbr

EDIT : even better, i have redumped the 2 disks on my atari ST, and guess what, it works under steem !

https://app.box.com/s/3qjzgfzbyxd5hms7y2md

[DrCoolZic]

crackdown disk 1 dumped with kryoflux as raw stream files :

https://app.box.com/s/zj6hf5gyil6epn2dekbr

EDIT : even better, i have redumped the 2 disks on my atari ST, and guess what, it works under steem !

https://app.box.com/s/3qjzgfzbyxd5hms7y2md

Thanks very much just one last request do you have the KF stream files for disk ?

and this is perfect to have the stx file so I can compare
I will look at the images tomorow

[dlfrsilver]

crackdown disk 1 dumped with kryoflux as raw stream files :

https://app.box.com/s/zj6hf5gyil6epn2dekbr

EDIT : even better, i have redumped the 2 disks on my atari ST, and guess what, it works under steem !

https://app.box.com/s/3qjzgfzbyxd5hms7y2md

Thanks very much just one last request do you have the KF stream files for disk ?

and this is perfect to have the stx file so I can compare
I will look at the images tomorow

The floppy drive i use with KF and SCP is not handling correctly the floppy disks if they are too old.....

I guess i will have to either clean the head or change the drive.

[DrCoolZic]

I have looked at the disk 1 done with KF
It seems like the FD is in bad shape the tracks 1-4 have a lot of garbage and it used to crash Aufit (no all fixed).
But track of interest 08.0 ID over index

And track 79.0 this track has non formated area at end of track that generate fuzzy bits but not inside a sector so it is ignored but sector 6 has a lot of protection that I need to anlyze I will let you know

[DrCoolZic]

The floppy drive i use with KF and SCP is not handling correctly the floppy disks if they are too old.....

I guess i will have to either clean the head or change the drive.

Never mind. Wont be able to generate Pasti from the files but still interesting protections to look at
Thanks

[JimDrew]

You need to get a drive head cleaner kit!!

[DrCoolZic]

New version 0.2a of the program.

Major change from 0.1b is the capability to write Pasti files. This is still experimental but should work.

- Lot of bug fixes so the program should be able to read most images even if input is not good quality.
- The Track buffer and Sector Buffer reports have been improved.
- Now Buffer display uses the currently displayed revolution. This allow to compare the decoded data for the different revolutions.
- This version includes a reasonable documentation with added information about usage of the program, about protection analysis, and Writing Pasti files. Added a new complete section about analysis of the quality of the imaged floppy disks.

This new version contains most of the features planned for first round.
You should be able to read SCP file (including the coming support for No Flux Area / Strongbits) and Kryoflux stream file. Full analysis of protections, and now you should be able to write Pasti STX file directly from flux transition input

I have only tested on few test cases and this seems to work. Currently the files generated are very “conservative” and therefore a lot of information is written and making bigger than normal file but it should work in all cases (apart from bugs! )

If you can try and let me know if you were successful I would appreciate.

One thing important if you try to write Past file is to make sure of the quality of the image.
I have added a chapter on the subject so please read it before creating Pasti file. Obviously if you are starting with a bad image you won’t be able to get a working Pasti file.
In coming version I will make more checking to help determine if the image has problem but for now it is visual.

Aufit02a.rar

[Brume]

Great, thank you very much!
I was waiting for it, you made it!
Now, it's time to test it for me

[DrCoolZic]

Great, thank you very much!
I was waiting for it, you made it!
Now, it's time to test it for me

Let me know if it works.

But again be careful using good images. On the images you have sent me about 50% are not good.

As mentioned in the doc even if you have diskettes that works on Atari that does not mean you can create good images. Some reason are using different floppy drives and the fact that the Atari performs automatically retries when read fails.

I will continue to add information about quality of images started in this thread http://www.atari-forum.com/viewtopic.php?f=95&t=25934

[Brume]

Let me know if it works.

Yes, it works fine with the disks I tried! Thank you very much for this very nice addition

But again be careful using good images. On the images you have sent me about 50% are not good.

True, but all the disks were imaged before I start to clean up my floppy drive (and the disks when necessary). I'm the process to re-image the disks with KF and SCP

[DrCoolZic]

I have found that several images created with Aufit do not work! I am working on the problem

[DrCoolZic]

I need help!

I have converted the Back to future Game with the Aufit program. This game is apparently protected by a Rob Northen protection.
It seems that this is the same protection as described by Sarnau at http://www.sarnau.info/atari:protection_mighty_bombjack

• First sector number is function of track number
• Sector 6 slower
• First 16 bytes equal to specific value

The clock used for the bit cell changes inside sector 6 from period 2000 to 2100. There are almost always one flux transition in the region where the clock changes which is out of normal range.
On a real Atari this results sometimes (but not always) in fuzzy bits because the bit after the glitch are shifted.

I have tested the fuzzy bytes for this game on a real Atari using my Panzer v0.22 program http://www.atari-forum.com/viewtopic.ph ... 25#p237714
setting the test to 5 revolutions and it is about fifty-fifty to get fuzzy bytes.

The DPLL in Aufit is apparently better than the WD1772 as I always read the same result (never fuzzy bytes)!
I therefore simulate fuzzy bytes in this specific case. Usually the first 30 bytes are read correctly and after the bytes are changed.

Now I am generating a stx file that looks exactly similar to the stx file generated by the Pasti imager. (I have remove all optimizations so the file contains all the sector data …)
To compare the two files I am using my Pasti Reader http://www.atari-forum.com/viewtopic.ph ... 25#p244419

But when I load the file under steem it does not work!
I am a complete ignorant in debugging code under steem or Hatari so if someone can have a look and find why the program do not load I would appreciate.

The files are at https://mega.co.nz/#!90hTHaxA!Vz9RdDHG0 ... a3SFJfZfjU

[Jeff_HxC2001]

Here both files seems to work on the real machine (with an usb hxc).

But the track viewer confirm that there is no fuzzy bits in your file.

[DrCoolZic]

Here both files seems to work on the real machine (with an usb hxc).

But the track viewer confirm that there is no fuzzy bits in your file.

Thanks Jeff this is very interesting. I wonder why it does not work under Steem? May be a hidden check in the the Pasti DLL
All information seems good especially the bit position and sector timings. They are very close and should not create problems.
I have asked Steven Seagal to look at the problem under Steem.

Regarding Fuzzy bits this is very strange. On a real Atari machine most of the time the sector is read the same (no fuzzy bits) but Pasti imager detect it as fuzzy sector.
With scp and kryoflux images of the game I always read the sector the same. As I said I have artificially created fuzzy bytes for this sector (to be as close as possible as what is generated by Pasti imager) but according to Sarnau this does not seems to be part of the protection. Only the 24 bytes or so are checked and these ones are never fuzzy? So I might remove this patch.

[DrCoolZic]

I found the problem bit 5 of trackFlags has be be set to 1 so readTime in sector descriptor is used
Explanation in http://www.atari-forum.com/viewtopic.ph ... 75#p244675
Now it works new version of Aufit to come soon that fix this problem.

[DrCoolZic]

FYI I have also removed the patch that was adding Fuzzy byutes and rogram works perfectly.
So it seems that Aufit in this case does a bit better than Pasti Imaging

Also In Aufit I have implemented the file optimization described in Pasti documentation: when a sector reads the same as what has been saved in the track data then the sector is not saved in the sector data. This is a semi nightmare to implement and in fact recent version of Pasti imaging program does not use it anymore So for example Back to future saved by Pasti is 1818 KB and Saved by Aufit 1393 KB (this is with 3 extra tracks!)

So a bit more tests and new version of Aufit to come with hopefully much more successful conversions

Back to future III converted
https://mega.co.nz/#!5ppQ2ApS!a53KWy9g9 ... 3F2Mpy480E

[DrCoolZic]

New release of Aufit V0.2b bug fix - just replace the executable

- fix trackFlag bug see above
- CRC stored in reverse order to be same as Pasti
- optimize flag not set for safety
- fixed bug in sector position (startBit)

Still some stx files not working I am investigating ....

Aufit02b.rar

[Brume]

Thank you
I'll test it tonight.

[DrCoolZic]

With version 0.2b almost all games I have tried seems to convert correctly.
The only two problems are "Au nom de l'Hermine" and Blood Money
But for blood money even the Pasti generated version does not seems to work? this seems to be a Steem problem from what I understand.
Au nom de l'hermine uses 70 sectors on track 79.0. I think I have another game that uses the same protection I have to find it!

EDIT
Actually I have found it it is Fire and Forget and it seems to convert correctly?

[Brume]

Au nom de l'hermine uses 70 sectors on track 79.0. I think I have another game that uses the same protection I have to find it!

I managed to produce a STX version of Au Nom de l'hermine with Aufit (with the previous version).
It worked fine with the RAW stream done with KF.
With SPC, I had to redo the dump: with 2 revolutions, the game stopped at the first screen. It works now with 4 revolutions.

[DrCoolZic]

can you put the scp version on your ftp site

EDIT
In fact the conversion seems to be ok. Apparently the game does not start automatically.
You have to open the disk A and click the loader.prg

So now I only have one game not working and It look like it is not a problem with Aufit

[Brume]

Yes, no problem. You'll get it in a few hours (as soon as I'm back at home).

[dlfrsilver]

With version 0.2b almost all games I have tried seems to convert correctly.
The only two problems are "Au nom de l'Hermine" and Blood Money
But for blood money even the Pasti generated version does not seems to work? this seems to be a Steem problem from what I understand.
Au nom de l'hermine uses 70 sectors on track 79.0. I think I have another game that uses the same protection I have to find it!

EDIT
Actually I have found it it is Fire and Forget and it seems to convert correctly?

Au nom de L'hermine use the STARTER protection (it's the grand sister of the amstrad CPC KBI-19 sectors of the same duplicator).

Here is the story : KBI (Klein Boussuge Informatique), was a french duplicator based in the Saint-Cloud in the West of Paris. At a moment the company was renamed KB STARTER. STARTER was the new name of their long track protection, duplicated on 'Trace Mountain' Unix duplication machines.

Au nom de l'hermine on amstrad CPC use a KBI-19 (for 19 sectors), while 16-bits machines like atari and amiga use one called 'STARTER'. This one was used by Loriciel, Silmarils, titus, ere informatique and maybe infogrames. The starter protection use 70 sectors.

look at Teenage Queen from Ere, it uses the starter-70 sectors protection. Pasti is unable to image it, while Super CARD pro can image it. And aufit can make STX files out of SCP and KF files for games using this protection, which is great

[Brume]

look at Teenage Queen from Ere, it uses the starter-70 sectors protection. Pasti is unable to image it, while Super CARD pro can image it. And aufit can make STX files out of SCP and KF files for games using this protection, which is great

Are you sure? On my side, Pasti can make a STX file of Teenage Queen from Ere. At least it did it two weeks ago (before DrCoolZic released Aufit). Which version do you have? I have the two disks version, and it has the STARTER protection on it and Pasti made STX w/o problem.

[dlfrsilver]

look at Teenage Queen from Ere, it uses the starter-70 sectors protection. Pasti is unable to image it, while Super CARD pro can image it. And aufit can make STX files out of SCP and KF files for games using this protection, which is great

Are you sure? On my side, Pasti can make a STX file of Teenage Queen from Ere. At least it did it two weeks ago (before DrCoolZic released Aufit). Which version do you have? I have the two disks version, and it has the STARTER protection on it and Pasti made STX w/o problem.

When i got Teenage Queen in 2 disk original release before my father burned the house, pasti was unable to image the Starter protection track on each of my atari (STF and STE). It exited with an error message like "too many sectors on this track" or something like that.

With Aufit, and my KF stream file dump, i made STX files two hands tied in the back