What do you mean with FD? "Floppy Disk" or "Floppy Drive?" Also if you speak about the Floppy Drive, which backup do you mention in your last topic? Is it the one made with Pasti or the other made with Kryoflux?
List of difficult to copy disks
Moderators: DrCoolZic, Moderator Team
Re: List of difficult to copy disks
Very interesting, thank you DrCoolZic
What do you mean with FD? "Floppy Disk" or "Floppy Drive?" Also if you speak about the Floppy Drive, which backup do you mention in your last topic? Is it the one made with Pasti or the other made with Kryoflux?
What do you mean with FD? "Floppy Disk" or "Floppy Drive?" Also if you speak about the Floppy Drive, which backup do you mention in your last topic? Is it the one made with Pasti or the other made with Kryoflux?
Looking for a CosmosEx unit for Falcon...
-
DrCoolZic
- Fuji Shaped Bastard
- Posts: 2270
- Joined: Mon Oct 03, 2005 7:03 pm
- Location: France
- Contact:
Re: List of difficult to copy disks
By FD I mean floppy disks
and yes of course all informations are from KF stream files as pasti does not provide flux details
and yes of course all informations are from KF stream files as pasti does not provide flux details
Visit *** http://info-coach.fr/atari ***
Re: List of difficult to copy disks
OK, thanks for the information. I firstly tought the problem cames from my internal drive on my STE.
Is there a way to check the integrity on a floppy once the KF made the job? I mean, when you'll release your tool, will it be easy to see if a floppy had a problem?
Also may the problem come from the drive on the PC, too? Or is it only the floppy disk?
Is there a way to check the integrity on a floppy once the KF made the job? I mean, when you'll release your tool, will it be easy to see if a floppy had a problem?
Also may the problem come from the drive on the PC, too? Or is it only the floppy disk?
Looking for a CosmosEx unit for Falcon...
-
DrCoolZic
- Fuji Shaped Bastard
- Posts: 2270
- Joined: Mon Oct 03, 2005 7:03 pm
- Location: France
- Contact:
Re: List of difficult to copy disks
These are very good questions.
When i first started protection analysis I had strange results I could not explain ... Even using the SPS tools did not make sense of some images.
Now that I have a tool that allow to see visually the flux and resulting decoded data it is easy to see that some disk have problems ...
For example look at this example I am not an expert in "magnetism" but obviously the sampled data indicates some problems (notice the strange histogram) - could be a magnet close to FD, flux migration, time deterioration, ??? (just guessing here).
But for sure you wont be able to use this floppy.
Look at the disk layout for this FD You can see that some areas seems almost intact and in some other place (like the upper part) things are pretty bad!
Finding these problem visually is relatively easy but it is much more difficult by software.
So my advice is to visually look at each dump you do. This is what I started to do
Trying floppies on a real machine before dumping helps but unless you spend hours playing the game you might not reach a place where the FD has been destroyed!
When i first started protection analysis I had strange results I could not explain ... Even using the SPS tools did not make sense of some images.
Now that I have a tool that allow to see visually the flux and resulting decoded data it is easy to see that some disk have problems ...
For example look at this example I am not an expert in "magnetism" but obviously the sampled data indicates some problems (notice the strange histogram) - could be a magnet close to FD, flux migration, time deterioration, ??? (just guessing here).
But for sure you wont be able to use this floppy.
Look at the disk layout for this FD You can see that some areas seems almost intact and in some other place (like the upper part) things are pretty bad!
Finding these problem visually is relatively easy but it is much more difficult by software.
So my advice is to visually look at each dump you do. This is what I started to do
Trying floppies on a real machine before dumping helps but unless you spend hours playing the game you might not reach a place where the FD has been destroyed!
You do not have the required permissions to view the files attached to this post.
Visit *** http://info-coach.fr/atari ***
-
DrCoolZic
- Fuji Shaped Bastard
- Posts: 2270
- Joined: Mon Oct 03, 2005 7:03 pm
- Location: France
- Contact:
Re: List of difficult to copy disks
Hard to copy but correctly imaged by SCP !
Here is the game Fire and Forget. Track 79 has 70 sectors like "au nom de l'hermine"
The game has correctly been recorded by the SuperCard Pro It is interesting to see that SCP seems to report more "clean timing"
Below is the layout of this disk on the left is the result using SCP and on the right using KF
Here is the game Fire and Forget. Track 79 has 70 sectors like "au nom de l'hermine"
The game has correctly been recorded by the SuperCard Pro It is interesting to see that SCP seems to report more "clean timing"
Below is the layout of this disk on the left is the result using SCP and on the right using KF
You do not have the required permissions to view the files attached to this post.
Visit *** http://info-coach.fr/atari ***
Re: List of difficult to copy disks
Tracks like this were not too uncommon with the Amiga. Since the Amiga's controller had a very wide spectrum for reading, data was often times varied throughout the track, like the Rob Nothern stuff - but worse. I am not sure what a WD1772 would do in these cases.
Yes, it could also be just dirty, but typically (at least with my analyzer built into the SCP software), the data is really wide all over the place. Pretty easy to clean the disk though, but the 3.5" disks are pretty good about staying clean - unless the disk was stored in a high humidity environment. For some reason, mold likes to grow on iron particles.
The "clean timing" difference is the hardware difference, namely the resolution and interface circuitry. SuperCard Pro uses 25ns capture, which is nearly twice that of Kryoflux's 41.66ns capture. Also, keep in mind that bits were typically written on even multiples of 2us time, which means that if the bit was written at 4us, SuperCard Pro can read it back as exactly 4us. Kryoflux can never do that because 41.66 is not evenly divisible into 4us. The SuperCard Pro's capture buffers are hands down better than the Kryoflux board. I have been meaning to take some scope snapshots showing the difference. Kryoflux's buffered read data input is skewed by hundreds of nanoseconds and rounded, whereas SuperCard Pro's data is not skewed and is square. It's neat that your program can show the difference in quality. From a "preservation" viewpoint, SuperCard Pro is clearly superior.
Yes, it could also be just dirty, but typically (at least with my analyzer built into the SCP software), the data is really wide all over the place. Pretty easy to clean the disk though, but the 3.5" disks are pretty good about staying clean - unless the disk was stored in a high humidity environment. For some reason, mold likes to grow on iron particles.
The "clean timing" difference is the hardware difference, namely the resolution and interface circuitry. SuperCard Pro uses 25ns capture, which is nearly twice that of Kryoflux's 41.66ns capture. Also, keep in mind that bits were typically written on even multiples of 2us time, which means that if the bit was written at 4us, SuperCard Pro can read it back as exactly 4us. Kryoflux can never do that because 41.66 is not evenly divisible into 4us. The SuperCard Pro's capture buffers are hands down better than the Kryoflux board. I have been meaning to take some scope snapshots showing the difference. Kryoflux's buffered read data input is skewed by hundreds of nanoseconds and rounded, whereas SuperCard Pro's data is not skewed and is square. It's neat that your program can show the difference in quality. From a "preservation" viewpoint, SuperCard Pro is clearly superior.
I am the flux ninja
Re: List of difficult to copy disks
Does anyone know if the Hewson games (Zynaps) are heavily protected?
I am the flux ninja
Re: List of difficult to copy disks
They have primitive protections - at least what I examined.
Tried Zynaps STX - there is only 9 sector in track 79 (others have 10), and there are some fuzzy bits. However, I don't see where game checks all it. Made conversion to ST format, and game works. Maybe there is some test in later stages (encrypted), but not much likely - it is from 1988. Possible that disk was not in good shape when image is made. I will tell more after couple days for Zynaps ...
Tried Zynaps STX - there is only 9 sector in track 79 (others have 10), and there are some fuzzy bits. However, I don't see where game checks all it. Made conversion to ST format, and game works. Maybe there is some test in later stages (encrypted), but not much likely - it is from 1988. Possible that disk was not in good shape when image is made. I will tell more after couple days for Zynaps ...
Famous Schrodinger's cat hypothetical experiment says that cat is dead or alive until we open box and see condition of poor animal, which deserved better logic. Cat is always in some certain state - regardless from is observer able or not to see what the state is.
Re: List of difficult to copy disks
OK, thanks... they had so-so protection for the C64, so I didn't expect anything crazy for the ST.
I am the flux ninja
Re: List of difficult to copy disks
I copied/imaged Barbarian, BattleZone, Star Glider, and Forgotten Worlds today. I am not sure if these are heavily protected or not. I got Obliterator but the original won't load. 
I am the flux ninja
-
DrCoolZic
- Fuji Shaped Bastard
- Posts: 2270
- Joined: Mon Oct 03, 2005 7:03 pm
- Location: France
- Contact:
Re: List of difficult to copy disks
If you put the images somewhere I should be able to find out some if not all protections used.JimDrew wrote:I copied/imaged Barbarian, BattleZone, Star Glider, and Forgotten Worlds today. I am not sure if these are heavily protected or not. I got Obliterator but the original won't load.
Visit *** http://info-coach.fr/atari ***
Re: List of difficult to copy disks
Battle Zone is not copy-protected - it is published by Atari, and I did not see any protected title from them.
There is 2 Barbarian for ST: Psygnosys and Palace SW (Barbarian Ultimate Warrior). Looking STX image info, later has 10 sectors/track on tr. 79, while others are with 9. That can be copied with better SW on Atari self. Psygnosis Barbarian has variable sector count too + uses track read, what means usually some pattern in gaps. Medium hard to copy, I would say.
Starglider: sector #247 on tr. 79 - what WD1772 can not create by formatting, used in lot of protections. Not much hard to copy.
Forgotten Worlds: similar to Barb. Psyg. - variable sector count on tracks + track read .
None of mentioned has variable density.
There is 2 Barbarian for ST: Psygnosys and Palace SW (Barbarian Ultimate Warrior). Looking STX image info, later has 10 sectors/track on tr. 79, while others are with 9. That can be copied with better SW on Atari self. Psygnosis Barbarian has variable sector count too + uses track read, what means usually some pattern in gaps. Medium hard to copy, I would say.
Starglider: sector #247 on tr. 79 - what WD1772 can not create by formatting, used in lot of protections. Not much hard to copy.
Forgotten Worlds: similar to Barb. Psyg. - variable sector count on tracks + track read .
None of mentioned has variable density.
Famous Schrodinger's cat hypothetical experiment says that cat is dead or alive until we open box and see condition of poor animal, which deserved better logic. Cat is always in some certain state - regardless from is observer able or not to see what the state is.
Re: List of difficult to copy disks
It is the Pyscnosis version. It's interesting that Pysgnosis was Amiga, Pyscnosis was Atari ST, and Psyclapse was C64 - all from the same company. I never knew that until recently.
I just purchased ALL of the example programs that DrCoolZic mentions in his 2011 Atari ST copy protection manual. So, those have a bunch of the difficult protections.
I just purchased ALL of the example programs that DrCoolZic mentions in his 2011 Atari ST copy protection manual. So, those have a bunch of the difficult protections.
I am the flux ninja
Re: List of difficult to copy disks
Considering Zynaps: there is no any protection check in game code - went thru all 16 levels. That's little strange, but happens.
There are Atari ST titles with PsyClapse logo at start: http://www.atarimania.com/pgelstsoft.aw ... 21&step=25
There are Atari ST titles with PsyClapse logo at start: http://www.atarimania.com/pgelstsoft.aw ... 21&step=25
Famous Schrodinger's cat hypothetical experiment says that cat is dead or alive until we open box and see condition of poor animal, which deserved better logic. Cat is always in some certain state - regardless from is observer able or not to see what the state is.
Re: List of difficult to copy disks
You are wrong. There's drive motor wobble, head friction and quartz difference. Hence something written at 2us or 4us, for sure won't come back as 2us or 4us. It will be 1.95us or 2.03us, or 3.97us or 4.09 us. Hence 25ns or 23ns or 30ns or 40ns or 50ns won't make a difference.JimDrew wrote:Also, keep in mind that bits were typically written on even multiples of 2us time, which means that if the bit was written at 4us, SuperCard Pro can read it back as exactly 4us.
Re: List of difficult to copy disks
No, I am not wrong, re-read what I wrote. Regardless of drive speed, if something was written at 4us it can be read back at exactly 4us with SCP. Kryoflux can never read or write at exactly 4us. Even if the data was returned as 1.9500us or 2.0300us, SCP can read the data as these exact values because they are even multiples of 25ns. So, it does matter. DrCoolZic's tool clearly shows the difference that resolution and better hardware makes.
I am the flux ninja
-
Hippy Dave
- Atari Super Hero
- Posts: 515
- Joined: Sat Jan 10, 2009 5:40 am
Re: List of difficult to copy disks
The Kryroflux board samples at 24MHz (41.66nS).mr.vince wrote:You are wrong. There's drive motor wobble, head friction and quartz difference. Hence something written at 2us or 4us, for sure won't come back as 2us or 4us. It will be 1.95us or 2.03us, or 3.97us or 4.09 us. Hence 25ns or 23ns or 30ns or 40ns or 50ns won't make a difference.JimDrew wrote:Also, keep in mind that bits were typically written on even multiples of 2us time, which means that if the bit was written at 4us, SuperCard Pro can read it back as exactly 4us.
The SuperCard samples at 40 MHZ (25nS)
Both cards can do exactly 1uS and multiples thereof.
Floppy Drives can do 500,000 samples per second (2uS), thus, need at least a 1uS resolution. The Kryoflux board is 24 times oversampled -- more than enough. Now that there is conversion software for SuperCard, KryoFlux and STX you have a choice because both boards are more than adequate.
Code: Select all
1/41.66..nS = 24,000,000 Hz = 24MHz
1/25.00..ns = 40,000,000 Hz = 40MHz
uS Hz KF SC
1 1,000,000 y
2 500,000 y
3 333,333 y
3.33.. 300,000 y n
4 250,000 y
5 200,000 y
6 166,667 y
6.66.. 150,000 y n
7 142,857 y
8 125,000 y
9 111,111 y
10 100,000 y
11 90,909 y
12 83,333 y
13 76,923 y
13.33.. 75,000 y n
14 71,429 y
15 66,666 y
Re: List of difficult to copy disks
No, Jim you are wrong. Drives are immune to your reality distortion field, so you never ever get exact multiples of 25ns. You are doomed when you try to read something that's a pure multiple of 41.66ns. Guess what, we can read that and you can't... 
DrCoolZics pictures show a perfectly well imaged disk without any processing on the right (which we apply later, in software, when we have a choice) and one with filtering and / or post-comping applied on the left (on the fly, which can't be reversed later).
Hippy Dave is spot on.
DrCoolZics pictures show a perfectly well imaged disk without any processing on the right (which we apply later, in software, when we have a choice) and one with filtering and / or post-comping applied on the left (on the fly, which can't be reversed later).Hippy Dave is spot on.
Re: List of difficult to copy disks
If you want a "preservation" quality image, you want the highest sampling possible. You can take a picture with a 1 megapixel camera and see the object in the photo, but a 2 megapixel camera will give you better detail. It's pretty simple really... the higher the base resolution is the less manipulation is required to make it better later. Wow! Wrong again! You need to get your facts straight... the picture on the left is unfiltered (raw flux spooled from the disk) and is the worst case scenario you get, which is better than any 'cleaned up' version would be when starting with the image on the right. This is due to SCP's higher resolution and better hardware design. I have not built in any type of filtering at this point into the SCP software, and I don't know that I ever will. It's clearly not needed.
Does a 25ns vs. 41.66ns resolution make a difference? Absolutely, if you want the most accurate representation of the disk. Would a 10ns capture be better than 25ns? Absolutely! However, in my testing anything less than 20ns captures show no difference in the capture quality because you are at the limit of what the drive electronics can provide. There *is* a difference between 20ns and 25ns. I wanted to use a 200MHz internal clock (instead of the normal 160MHz I am using), but it was not reliable for all CPUs. This would have given me the 20ns capture I wanted. But, I settled for reliable.
A 4000ns (4us) pulse is not evenly divisible by 41.66ns. The closest that you can come is 3999.36ns. Likewise, a 6000ns (6us) pulse is also not evenly divisible by 41.66ns. The closest you come is 5999.04ns. Even 1000ns (1us) pulses are not evenly divisible by 41.66ns - the closest you can come is 999.84ns.
Does a 25ns vs. 41.66ns resolution make a difference? Absolutely, if you want the most accurate representation of the disk. Would a 10ns capture be better than 25ns? Absolutely! However, in my testing anything less than 20ns captures show no difference in the capture quality because you are at the limit of what the drive electronics can provide. There *is* a difference between 20ns and 25ns. I wanted to use a 200MHz internal clock (instead of the normal 160MHz I am using), but it was not reliable for all CPUs. This would have given me the 20ns capture I wanted. But, I settled for reliable.
A 4000ns (4us) pulse is not evenly divisible by 41.66ns. The closest that you can come is 3999.36ns. Likewise, a 6000ns (6us) pulse is also not evenly divisible by 41.66ns. The closest you come is 5999.04ns. Even 1000ns (1us) pulses are not evenly divisible by 41.66ns - the closest you can come is 999.84ns.
I am the flux ninja
Re: List of difficult to copy disks
OK, someone here does not understand how sampling works...
http://en.wikipedia.org/wiki/Nyquist%E2 ... ng_theorem
http://en.wikipedia.org/wiki/Nyquist%E2 ... ng_theorem
-
DrCoolZic
- Fuji Shaped Bastard
- Posts: 2270
- Joined: Mon Oct 03, 2005 7:03 pm
- Location: France
- Contact:
Re: List of difficult to copy disks
It seems that the picture I have published is the source of an interesting debate (personally I think that Mr. Shannon have said all that need to be known on the subject). Therefore, in order to stay as neutral as possible, I have tried to understand why I am getting different results with KF and SCP on my Fire & Forget disk.
FYI I have not yet sampled many disk with SCP because I want to receive my head cleaning kit first. So far I have only done Dungeon Master and Fire & Forget. On DM I could not see any visual difference on the disk layout but on F&F the results were somewhat different. I guess that the main reason is that the DM diskette is in good condition, where the FF seems to have “suffer” from some deterioration. If you look at some tracks (presented below) the raw flux transitions exhibit some sort of “wobulation” and they were certainly not written like that.
The KF sample was done about two years ago so I decided, to be fair, to redo the sampling on the same drive. Resampling the FD gave even worse result to the point that several of the tracks on the end of the disk could not be read without errors after many retries? I have therefore cleaned the disk (my first time experience) using a QTip impregnated with head cleaning solution. I have only cleaned the side on bottom of the disk (looking at the label) and proceeded slowly and not too vigorously on the surface that did not look dirty. Then I have redone the sampling and … yes, to my surprise, cleaning make a big difference. Now not only the disk read without error but it also gives results extremely close to the one with SCP.
Some information on what you see. The Fluxes displayed are NOT processed but the clock and the data recovered from the fluxes are HIGLY processed by a complex DPLL that is supposed to remove as much as possible the fluctuation from the drive read channel (both slow and fast variation). I give some explanation about my DPLL here http://www.cbmstuff.com/forum/showthrea ... =16&page=4 and here http://info-coach.fr/atari/hardware/FD-Hard.php#WD1772. The disk layout graphic is obviously using decoded data (highly processed data).
Now some pictures. Track 76 with KF before cleaning: here you can see the "wobulation" and the very dirty reading of flux that results in the last three sectors read incorrectly.
Track 76 with KF after cleaning: Here you can still see the "wobulation" but less pronounced and a more "clean" reading resulting in 3 last sector read correctly
Track 76 with SCP: The same sector sampled with SuperCard Pro
To come the famous disc layout diagrams ...
FYI I have not yet sampled many disk with SCP because I want to receive my head cleaning kit first. So far I have only done Dungeon Master and Fire & Forget. On DM I could not see any visual difference on the disk layout but on F&F the results were somewhat different. I guess that the main reason is that the DM diskette is in good condition, where the FF seems to have “suffer” from some deterioration. If you look at some tracks (presented below) the raw flux transitions exhibit some sort of “wobulation” and they were certainly not written like that.
The KF sample was done about two years ago so I decided, to be fair, to redo the sampling on the same drive. Resampling the FD gave even worse result to the point that several of the tracks on the end of the disk could not be read without errors after many retries? I have therefore cleaned the disk (my first time experience) using a QTip impregnated with head cleaning solution. I have only cleaned the side on bottom of the disk (looking at the label) and proceeded slowly and not too vigorously on the surface that did not look dirty. Then I have redone the sampling and … yes, to my surprise, cleaning make a big difference. Now not only the disk read without error but it also gives results extremely close to the one with SCP.
Some information on what you see. The Fluxes displayed are NOT processed but the clock and the data recovered from the fluxes are HIGLY processed by a complex DPLL that is supposed to remove as much as possible the fluctuation from the drive read channel (both slow and fast variation). I give some explanation about my DPLL here http://www.cbmstuff.com/forum/showthrea ... =16&page=4 and here http://info-coach.fr/atari/hardware/FD-Hard.php#WD1772. The disk layout graphic is obviously using decoded data (highly processed data).
Now some pictures. Track 76 with KF before cleaning: here you can see the "wobulation" and the very dirty reading of flux that results in the last three sectors read incorrectly.
Track 76 with KF after cleaning: Here you can still see the "wobulation" but less pronounced and a more "clean" reading resulting in 3 last sector read correctly
Track 76 with SCP: The same sector sampled with SuperCard Pro
To come the famous disc layout diagrams ...
You do not have the required permissions to view the files attached to this post.
Visit *** http://info-coach.fr/atari ***
-
DrCoolZic
- Fuji Shaped Bastard
- Posts: 2270
- Joined: Mon Oct 03, 2005 7:03 pm
- Location: France
- Contact:
Re: List of difficult to copy disks
Not much comment to me they look the same
Therefore my advice cleanup the disk before sampling
Therefore my advice cleanup the disk before sampling
You do not have the required permissions to view the files attached to this post.
Visit *** http://info-coach.fr/atari ***
-
DrCoolZic
- Fuji Shaped Bastard
- Posts: 2270
- Joined: Mon Oct 03, 2005 7:03 pm
- Location: France
- Contact:
Re: List of difficult to copy disks
Seems like we have the same source and understandingrittwage wrote:OK, someone here does not understand how sampling works...
http://en.wikipedia.org/wiki/Nyquist%E2 ... ng_theorem
You were a bit faster than me on the subject Visit *** http://info-coach.fr/atari ***
Re: List of difficult to copy disks
I (almost) always check disks if they need to be cleaned and have been using head cleaning kits for the last 20 years so i am used to clean disks/drives
I use Blu-Tack (Tack-It) to hold the metal protector in place and also use blu-tack so i can spin the disk easy and i always use the type of cleaning cloth that is ment for camera lenses and glasses since they will not scratch the surface.
DrCoolZic about those last images... is the SCP also from the cleaned disk?
I use Blu-Tack (Tack-It) to hold the metal protector in place and also use blu-tack so i can spin the disk easy and i always use the type of cleaning cloth that is ment for camera lenses and glasses since they will not scratch the surface.
DrCoolZic about those last images... is the SCP also from the cleaned disk?
