Page 5 of 7

Re: AUFIT Program

Posted: Wed Jan 21, 2015 11:31 am
by Mug UK
Cheers DrCoolZic. Will do some tests over the next week.

Sent by my Note III

Re: AUFIT Program

Posted: Sat Jan 24, 2015 8:50 pm
by greatguy
DrCoolZic wrote:
greatguy wrote:I tried to make an STX from my fully working Chambers Of Shaolin SCP image without success. First I thought it was due to some emulation problem in Steem, but the Atari Mania version boots just fine.
The SCP images are sampled with four revolutions and disks written from them works on real hardware.

Files attached below.

Chambers.Of.Shaolin.zip

Can you provide test ccase (scp files)?
How many revolutions where sampled?


The scp images provided below are sampled with four revolutions.

https://www.dropbox.com/s/mv4p5aac5p5jssb/Chambers%20of%20Shaolin%20SCP.7z?dl=0

Re: AUFIT Program

Posted: Sun Jan 25, 2015 8:57 am
by DrCoolZic
Thanks for the dump of chambers of Shaolin here is my analysis

Disk 1. First all tracks are sampled at around 1950 µs - do not know if this is on purpose or nat (could be because of your drive) and therefore have about 6400+ bytes
Track 00.0 has classic 9 x 512 bytes sectors all others have 5 x 1024 + 1 x 512
All tracks have a strange A1 sync at end of id and and of data fields
cos-d1-t00.0.png

lets zoom to see more clearly
cos-d1-t00.0-zoom.png


To understand exactly we need to look at the track content
cos-d1-t00.0-track-start.PNG

Here we can see that at end of id field we have 14 0B and as we know (see viewtopic.php?f=104&t=27448) this is obtained by writing 00 29 A1 (hidden sync)

We have the same sequence at end of data field
cos-d1-t00.0-track-middle.PNG


and the end of the track is filled with F7
cos-d1-t00.0-track-end.PNG



Disk 2
Same trick as d1 but something more. All tracks on side 1 are using a bit width 5% below normal (around 1900 µs) so they are long track with about 6550+ bytes
Here is a picture with clock diagram zoomed
cos-d2-t00.0-zoomclock.png




STX files have been generated with new Aufit viewtopic.php?f=104&t=25906&p=266034#p266034
Seems to work OK on Steem 3.6.4

Re: AUFIT Program

Posted: Sun Jan 25, 2015 9:07 am
by DrCoolZic
greatguy wrote:
DrCoolZic wrote:
greatguy wrote:I tried to make an STX from my fully working Chambers Of Shaolin SCP image without success. First I thought it was due to some emulation problem in Steem, but the Atari Mania version boots just fine.
The SCP images are sampled with four revolutions and disks written from them works on real hardware.

Files attached below.

Chambers.Of.Shaolin.zip

Can you provide test ccase (scp files)?
How many revolutions where sampled?


The scp images provided below are sampled with four revolutions.

https://www.dropbox.com/s/mv4p5aac5p5jssb/Chambers%20of%20Shaolin%20SCP.7z?dl=0

By the way I did not test but I bet that Aufit 0.4c cannot create a working stx file as the 00 29 A1 sequence was not decoded correctly

Re: AUFIT Program

Posted: Fri Jan 30, 2015 12:56 pm
by Brume
DrCoolZic wrote:Aufit 0.9.5489.31220
Here it is :)


Thanks for sharing it :)
Well, I wasn't able to play with it until yesterday night (Stonish project needs more time than expected...).
The new version is very cool. I like very much the "one-window-only" GUI, instead of the previous version that required both windows (one for the main interface, the other one for the display of the disk).
Also as stated in your previous messages, .STX files generated with Aufit are now smaller. That's great. And I like the feature that allows to store a disk in .ST format.
Just one thing: I don't like very much the colors when a disk is displayed. I found the old colors more clean and it was easier to spot a damaged sector. Of course it's a matter of taste ;)

Thanks again for this great release, DrCoolZic.

Re: AUFIT Program

Posted: Fri Jan 30, 2015 3:01 pm
by DrCoolZic
Brume thanks for feedback.
For colors I can revert to old colors. I just changed so that colors are consistent across different views. So track view and disk view use medium blue for normal track and same green for gap intra sector. The gaps between sectors stay black so it is easier to see complete sector.

PS: I am currently working on fixing Panzer so Aufit dev is on hold.

Re: AUFIT Program

Posted: Mon Feb 02, 2015 2:27 pm
by DrCoolZic
New release 0.9.5511.xxxx de Aufit

This release allow to save non protected files in ST and MSA format.
If the optimize check boxis checked (default) in the Write Image view, the STX ans MSA files will be much smaller ...
Also changed colors in Disk view (Brume let me know if this is more what you were expecting?).
Plus minor fixes.

To get this new version you just need to run Aufit (probably twice - checked on first run). You will be propose to update just accept and the new version will be installed.

New documentation v0.9a http://info-coach.fr/atari/software/pc- ... /Aufit.pdf

If you do not have Aufit yet goto viewtopic.php?f=104&t=25906&p=266034#p266034

Re: AUFIT Program

Posted: Mon Feb 02, 2015 4:40 pm
by JimDrew
Can we have an option for English vs. French - where the check boxes are the correct direction? :)

Re: AUFIT Program

Posted: Mon Feb 02, 2015 5:03 pm
by Hippy Dave
JimDrew wrote:Can we have an option for English vs. French - where the check boxes are the correct direction? :)

What you really want is the "right handed" version. :D

Re: AUFIT Program

Posted: Mon Feb 02, 2015 9:02 pm
by Brume
DrCoolZic wrote:To get this new version you just need to run Aufit (probably twice - checked on first run). You will be propose to update just accept and the new version will be installed.


Sorry, this doesn't work for me...
I closed and reloaded several times Aufit, it's still 0.9.5498.31220 and no new version is offered.
I even deactivated the firewall+antivirus for a short time: same result.
Is there no way to install it manually?
Thanks DrCoolZic.

Re: AUFIT Program

Posted: Tue Feb 03, 2015 6:01 am
by JimDrew
Hippy Dave wrote:
JimDrew wrote:Can we have an option for English vs. French - where the check boxes are the correct direction? :)

What you really want is the "right handed" version. :D


Yeah, I guess. I have never seen a left handed check mark before.

Re: AUFIT Program

Posted: Tue Feb 03, 2015 8:18 am
by DrCoolZic
In fact this has nothing to see with English versus French (also it could have the case for English vs. Arabic).

Indeed Hippy Dave has found the root of the problem: I am left handed and when I started to design the GUI it looked to me much more aesthetic to have it this way. I was surprised to find out that the default was the other way around and that I had to add a "Rigth"ToLeft property to get it "Right"! It appears to me more natural natural, as we are writing from left to right, to first have the name of the option then the check box. But to be honest, as most of the other GUIs use it the “wrong way”, I must admit that it gets confusing when you have several check boxes back to back.
Therefore I publish today a “LeftToRight” version rather than my “right” version for the right handed people ....

I also need to confess another idiosyncrasy to Brume who complains about colors: I am what we call "Daltonien" in French or "Color deficient" in English. For example the two first pictures here http://daltonien.free.fr/daltonien/arti ... _article=9 look EXACTLY the same for me but I have been told they look different to most people. This is why I use for example Green and Orange in the button grid instead of the hard to distinguish Green and Red. I also use in the Disk view Blue for normal sectors and Red for errors ….
But I guess I may also need to implement a way so that user can change the color scheme?

Now for the auto-update I do not understand. I have tested on two machines and it works fine. Do others get updated? I have now changed the checking so that it test on Internet before starting. If this slow down the startup too much let me know I will change back to checking after start (requires two starts to update).
If you update does not work for you here is the new version 0.9.5512.16391

Re: AUFIT Program

Posted: Tue Feb 03, 2015 8:38 pm
by JimDrew
The auto-update worked fine for me - and the checkmark is now correct! Thanks!

Re: AUFIT Program

Posted: Wed Feb 04, 2015 10:29 pm
by Maartau
Auto-update also worked fine, thanks :thumbs: .

Re: AUFIT Program

Posted: Fri Feb 06, 2015 9:56 am
by DrCoolZic
This post is a reply to the question is it possible to find if a "sector has been overwritten" with Aufit
see viewtopic.php?f=102&t=26976&p=267397#p267397

The new version of Aufit has new features that can help finding if a disk is an original or not.

When a disk is produced on a mastering machine the formatting and sector content data of a complete track are written in one pass. The write gate is turned on at a specific point (usually the index pulse) and turn off at the same point one revolution later. Due to speed variation it is not possible to close the write gate at exactly the same point as when the gate was open. This result in some transitions being shorter or longer than normal and these transitions are called the track write splices.

When a disk is produced on an Atari machine the formatting data of a complete track are written in one pass. The write gate can only be turned on at the index pulse and turn off at the next index pulse one revolution later. For same reasons explained before we also have track write splices located at the index pulse. In order to write data in a sector a write sector command need to be used. In that case the FDC read the address header and when the right header is found it switches from reading to writing the sector data and when it finishes it turn off the writing gate. Due to speed variation the start and end of the sector write is not aligned with the formatting data. This results in two sector write splices at the beginning and the end of the sector. If we find these sector write splices (as well as possible sector speed variation) this is a good indication that the sector has been overwritten.
It is normally difficult to detect the write splices on Aufit track view. However if you switch the disk view the write splice are seen as white marks on the track. The white marks are out of band transitions and in most cases (unless done on purpose) they correspond to write splices.

For example let’s format a floppy disk on an Atari. First we can see the write splices (in white) for the tracks that are aligned close to the index line (in blue).
track-write splice.PNG

At the end of the format command several information is written on the diskette. For example the boot sector, the FAT, the directory information, etc. As we have explained this information is written using a write sector command and result in two write splices on a specific sector (one at beginning, one at end of the sector). So here we see the sync marks in black followed by the address header in yellow followed by the sector gap in green. Inside this gap we see a write splice in White. After we have the sync marks in black followed by the data segment in blue. At the end of the sector we have the gap between sectors and at the beginning of this gap we see the second write splice. So we can be almost assure that this sector has been overwritten on an Atari machine.
sector-write-splice.PNG

Note that detecting the duplication of a diskette with a board like SCP is possible but more difficult.

Re: AUFIT Program

Posted: Fri Feb 06, 2015 1:36 pm
by DrCoolZic
Analysis of AMC provided here viewtopic.php?f=102&t=26976&p=267397#p267362

Disk 1
T00.0-T77.0 => 1 sector with no data + 10 x 512 sectors
AMC-T00.0.png

T79 => Track with 58 sectors (truncated sectors within sectors) with CRC/Fuzzy bits errors
AMC-T79.0.png

t79-disk.PNG


Disk 2
T00.0-T07.0 => 10 x 512 sectors + 1 sector with no data
T08.0-T73.0 => 10 x 512 sectors + 1 truncated sector

In the sample track 10.0 miss one sector header and therefore image is not good.

Attached is an image of disk 2 layout
Two things to note:
- lots of sectors have write splice (white dots at start-end sector data). This make me thing that the disk has been written on an Atari?
- Second side is not Atari track. However there are some data and sync marks (black lines on side 1). I am not knowledgeable about Amiga but might be Amiga track? These track cant be read reliably on Atari as only two sync?
disk2.png


PS New version of Aufit with better display of Out of band transitions and NFA transitions + Display of sync mark in black
This new version will be released today or tomorrow.

Re: AUFIT Program

Posted: Fri Feb 06, 2015 1:59 pm
by DrCoolZic
New version 0.9.5515.26942
- Improved display of OOB (white) and NFA (Red) in Disk View
- Added display of Sync Marks (Black) in Disk view
- Aufit Update: check done after program started (requires two run to update)

Run onece or twice to get updated

@Brume: still problem with the update?

Re: AUFIT Program

Posted: Fri Feb 06, 2015 2:12 pm
by DrCoolZic
As you know you can do file association for example with SCP and RAW file. Double clicking on associated file starts the program and loads the file(s) clicked.

But after making test I found out that in this case the auto-update does NOTwork. So you have to start the program with the start menu to get the update.

Re: AUFIT Program

Posted: Fri Feb 06, 2015 3:53 pm
by JimDrew
Nice... I like the write splices. This same routine should also be able to detect weakbits.

Re: AUFIT Program

Posted: Fri Feb 06, 2015 3:56 pm
by DrCoolZic
Yes but only for weakbits obtained by violating MFM bands

Re: AUFIT Program

Posted: Fri Feb 06, 2015 3:59 pm
by DrCoolZic
oops even this is not always true :(

Re: AUFIT Program

Posted: Fri Feb 06, 2015 8:24 pm
by JimDrew
That's not actually the case. You can determine exactly where weakbits are, even those that are within the normal MFM bands, and without reading multiple revolutions. :)

Re: AUFIT Program

Posted: Thu Mar 05, 2015 10:19 pm
by Brume
DrCoolZic wrote:@Brume: still problem with the update?

Yes, there is still no auto-update. I'm just coming to see you've updated the program last month...

Re: AUFIT Program

Posted: Thu Mar 05, 2015 10:59 pm
by DrCoolZic
this is really strange. Works for me and apparently for most people?

Re: AUFIT Program

Posted: Fri Mar 06, 2015 6:18 am
by JimDrew
It works for me, but this is why I have my own auto-update routine in the SCP software that just pulls the latest version from my server, renames the original version as it's version number, and then renames the downloaded file.