One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

All questions about how to use KryoFlux go here.
User avatar
IFW
Posts: 3079
Joined: Mon Nov 08, 2010 2:42 pm

Re: One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

Post by IFW »

I was just commenting on that:
1, shifted tracks can easily happen with disks written with a 1541 - accidentally
2, shifted tracks can happen with duplicated disks too - but only when the duplicator does that on purpose, e.g. to compensate for the track capability of the target drive
3, The majority of commercially duplicated disks are not track shifted, think say 99%. So if you want to check whether your drive truly can see to track -8 use a few commercial disks, e.g. games to see if side 1 is readable without the *T warnings.

ErikFromCalifornia
Posts: 12
Joined: Sun Oct 28, 2012 6:24 am

Re: One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

Post by ErikFromCalifornia »

Now that I'm a bit more familiar with the command line, I'm getting familiar with the KryoFlux Application to see what it can do to dump flippy disks.

Question #1:

One of the things that I find I can't seem to modify is the drive density through the GUI. However, I can add this to the advanced settings, but I'm not sure if this would conflict with the previous command that seems cooked into the application. There's doesn't seem to be another way to do it.

Here's where I set the drive density:
KryoFlux Advanced Settings
KryoFlux Advanced Settings
However, the result is that you get double parameters (see red below):

DTC.exe -l15 -fC:\kryoflux\dtc\testing\Dump1\track -g2 -y -i0 -fC:\kryoflux\dtc\testing\Dump1.d64 -g2 -y -i6 -r5 -a0 -b-8 -d0 -p -e83 -dd0 -dd1

Question #2:

I'm dumping a STREAM as well as a guided format (I believe) which produces a D64 file too. In the GUI, I check the FLIPPY MODE check box in both the STREAM image profile and the "CBM DOS sector image" image profile. I also set double sided for both the STREAM image profile and the "CBM DOS sector image" image profile too.

Is it necessary to double specify this in both image profiles? For example, if I only specify this in the STREAM profile, does this affect error checking at all?

DTC.exe -l15 -fC:\kryoflux\dtc\testing\Dump1\track -g2 -y -i0 -fC:\kryoflux\dtc\testing\Dump1.d64 -g2 -y -i6 -r5 -a0 -b-8 -d0 -p -e83 -dd0 -dd1

Question #3:

Being that I'm dumping Flippy Disks, how do I know if I SHOULD use the LOW density or HIGH density? What guidlines should I follow when making that consideration? Is there a recommendation as to which drive density I should typically stick to for C64 Flippy Disks?

Question #4:

It would seem 5 revolutions are the standard that seem to be always captured. Is there any benefit to boosting the revolution count (for example to 10 revolutions)? More the better right? Or maybe not so? If so, again, what guidelines should I follow when making that consideration as to when I should bump up the revolution count?

Question #5:

@ mr. vice: you posted in the past the following:
For dumping C64 disks, use the following command line:
dtc -p -b-8 -fdumpdir/<NameOfGame_> -i -i2 -y -g2 -i6 -l8 -t10
Which you mentioned here:

viewtopic.php?f=10&t=306

I don't understand the use of "-i -i2". Is this possibly a deprecated command for older KryoFlux software? If not, can you explain the benefit of using "-i i2"? Obviously, the KryoFlux manaul recommends a different command for typically dumping C64 Flippy Disks.

The KryoFlux manual recommends:
To dump a C64 flippy disk, use the following command line:
dtc -p -b-8 -f<dumpdir/dumpfile> -i0 -y -g2 -i6 -l8 -t10
Can you please provide some clarity?

User avatar
mr.vince
Posts: 2127
Joined: Tue Oct 05, 2010 5:48 pm

Re: One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

Post by mr.vince »

#1 & #3: yes you can. It's in the menu for drives. The density setting is for the DRIVE to force it into DD or HD mode. DD disks should be read DD, HD disk should be read HD. 5.25" drives can't sense media.

#5: is explained in manual. -i is -i0 and -i2 is for preservation raw files. Newer dtc versions produce the same output even without -i2. It's all exporters passing their requirements onto -i0.

ErikFromCalifornia
Posts: 12
Joined: Sun Oct 28, 2012 6:24 am

Re: One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

Post by ErikFromCalifornia »

In regards to #1 & #2:

For those who overlook this, Drive density can be set here in the GUI as mr.vice kindly reported:
DriveDensity.png
DriveDensity.png (4.96 KiB) Viewed 1077 times
Sorry, I overlooked the option being available under Drive. I kept going through the settings and couldn't find it there yet saw it set in the outputs (always at -dd0). I didn't realize there were further options under DRIVE. But that's clear now! ;)

As for #5:

It states in the manaual:
-i0 Stream file and preservation mode enabled. Zero is not mandatory so giving only -i defaults to 0.
So this explains the lack of 0 in -i being the same as -i0. It doesn't mention anything in regards to -i2. But as you said:
Newer dtc versions produce the same output even without -i2.
I'm therefore of the understanding that -i2 is no longer needed. SO, when I see people using "-i -i2" I could just as well use either "-i" or "-i0" to accomplish the same thing. This is good to understand!

@mr.vice, what about Question #4?

Earlier stated:
  • Question #4:

    It would seem 5 revolutions are the standard that seem to be always captured. Is there any benefit to boosting the revolution count (for example to 10 revolutions)? More the better right? Or maybe not so? If so, again, what guidelines should I follow when making that consideration as to when I should bump up the revolution count?
I ask because I'm wondering if higher revolutions would be beneficial for later analysis down the road if I have to return to the STREAM files without having to read from the original disk AGAIN?

It sort of brings me to my additional question ...

I'm of the understanding that the slower the drive speed, the higher the samples that can be achieved per revolution of the disk media. If I have commodore disks that were originally written on DD disks by say the 1571 that rotates at 300 RPM, isn't it better to capture STREAM files with a drive at a lower RPM too? If capturing a STREAM at 300 RPM's, isn't that a truer representation of the original disk medium than having captured the STREAM at 360 RPM's?

Typically the PC drives run at 360 RPM. And the streams I have captured were also at 360 RPM (earlier in my test captures) despite the original drive that wrote some these disks were at 300 RPM.

I'm thinking, if I can, see if I can jumper the drive to where it will run a 300RPM? Or maybe this would not make much of a difference?

For preservation and with the goal in mind of never having to re-read the original disks, which route should I take? I rather not copy all my disks at 360RPM while I find out later down the road it would have been better to transfer them at 300RPM, in DD (Density Line Low mode) with also -r10 instead of -r5? Advice on this is appreciated! :)

User avatar
IFW
Posts: 3079
Joined: Mon Nov 08, 2010 2:42 pm

Re: One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

Post by IFW »

Usually 5 revolutions is as good as it gets for a disk that is in an acceptable state - 5 revolutions is enough for any kind of analysis and it's also less strain on the disk surface than spending more time with it.
Yes, you might have to redump some of the disks, but that can happen regardless of whether you took 5 revolutions or 50 :)
It happens because of various contributing factors, such as head not settling properly, stepper failure, dirt on disk, dirt on drive head, damaged disk surface, weak signal etc. - anything that can possibly go wrong with a device like a disk drive usually does go wrong sooner or later...
So as far as preservation goes there is nothing to be gained from more revolutions - it's a good compromise for the vast majority of the disks.

As for sampling: the number of samples is exactly the same whether you sample a track at 300RPM or 360RPM, or any other RPM :)
What you sample is the time elapsed between subsequent flux reversals - that's the information carrier on a magnetic disk.
So what happens when you sample at different RPMs is that the flux reversals get shorter or longer depending on the RPM.
The flux reversals need to be interpreted for later processing and treated at the speed of the target platform, regardless of the RPM they were sampled with. In order to do that, you need to know the source (sampling) RPM and the target (platform) RPM. The former is derived from the index pulses detected during sampling, the latter is what is the default for the target platform/format or the manually overridden value given by the user.

User avatar
mr.vince
Posts: 2127
Joined: Tue Oct 05, 2010 5:48 pm

Re: One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

Post by mr.vince »

If, on a weekend, I give you a finger, don't grab for the hand. No need to repeat questions.

Dump as many revs as you like. The more the better... 5 is OK. 10 is great. 100 might have, for a damaged track, the one read that has the difference. But realistically 5 are OK.

If you can dump at the correct sped. Correct speed gives best reads. Might work with HD/DD switch alone or separate jumper configuration.

User avatar
IFW
Posts: 3079
Joined: Mon Nov 08, 2010 2:42 pm

Re: One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

Post by IFW »

As for density: that's a tricky thing for various reasons.
Usually DD disks read fine with DD as well as HD mode.
Drives that actually react to density changes during reading (they always do for writing) change their amplification settings, so for borderline flux reversals or some magnetic noise they may become more receptive.
This is a blessing and a curse at the same time :)
Some difficult or impossible to read DD disks may become readable in HD mode, but tracks that contain longer than usual flux reversals may become unreadable, or very difficult to read as the drive may see a flux reversal in a more sensitive mode, when it's just a glitch ignored by the filtering in DD mode. It is also common to "detect" a flux reversal in HD mode, when it just takes too long for HD, but it's more or less acceptable for DD. This is especially affecting some custom c64 disk formats as well as Apple disks.

For the above reasons, when you read a DD disk in HD mode, the samples will compress less, especially for tracks containing unformatted data - the filtering gain will turn them into lots of small flux reversals, while in DD mode you can normally read fairly long unformatted sequences without any problem.

I'd recommend reading DD disks in DD mode, and for tracks that fail to read in DD (weak signal etc.) try HD mode.
This will have the benefit of significantly smaller archive (zip etc.) files for your stream files (since the unformatted/noise areas will compress a lot better), and the peace of mind that fake flux reversals caused by HD gains/filtering are a lot less likely to happen.

For HD disks you really don't have a choice: they must be sampled in HD mode, otherwise what you read is often illegible random noise, as it is out of spec for DD.

User avatar
IFW
Posts: 3079
Joined: Mon Nov 08, 2010 2:42 pm

Re: One-pass "flippy", KyroFlux, 1571, and ZoomFloppy Exploration

Post by IFW »

You will also have to establish whether your drive:
1, actually changes amplification in DD mode
2, which density line setting means DD mode

Due to the behavior described above, there is a simple test for this:
Sample a disk with only 1 side formatted with both possible density line settings, then compress the stream files with e.g. zip.
If the archives are about the same size, your drive is not reacting to DD mode during reading.
If one archive is significantly smaller, that's the one created in DD mode reading - so use the density line setting used for those stream files as DD mode in the future.

Sadly, there is no other way to verify as amplification change is not standard, and some drives may be hardwired, others jumpered and others won't do it at all... those drives should be avoided, if you want to archive DD disks.
This means that setting the density line low or high has no further meaning, until you establish it: either setting could be DD for your specific drive, or actually none of them... as described before, you can create a working drive hardware with HD mode only for reading, as it will work most of the time.

Post Reply