Official K-Probe Discussion (Tool for Scanning C1C2/PIPO)

[cfitz]

Judging from the interface and the 'about' page, it looks to me like an intentional leak...

Who knows? It certainly is much more cleaned up and user-friendly than WSES. And I haven't found (yet, anyway) the infamous "Erase EEPROM" menu item from WSES. So in that sense it is safer to be distributed to the general public. On the other hand, many of the features are beyond the understanding of casual users of CD-RW drives, and in that sense it wouldn't appear to be something meant for the general public. Maybe it was meant to be leaked to the CD-RW "enthusiast" (whatever that is). But this is nothing but idle speculation.

So, keep on going cfitz.....

I am kind of busy with work right now, so I don't really have much to play with it. But there are plenty of others, I think, who grabbed a copy and can report what they find. Take Abacus, for example.

cfitz

[cfitz]

cfitz:
I found a hidden feature !
That is , if you click and drag with the left mouse button to
select a rectangle , the chart will "zoom in"
On the contrary , it will restore.

And if you click and drag with right mouse button , the chart
will be shift.

Right ??

Sure enough. Another neat feature. Thanks for the tip. And, again, thanks for bringing the link to our attention in the first place.

May I ask how you found the link?

cfitz

[Abacus]

cfitz:
I got the link from a friend works in QSI now.
Of course , he got the package from his friend works in LITEON.

So ... I will ask him the new link to KProbe

[cfitz]

cfitz:
I got the link from a friend works in QSI now.
Of course , he got the package from his friend works in LITEON.

So ... I will ask him the new link to KProbe

It is good to have friends in high places...

Please don't get anyone in trouble.

cfitz

[Abacus]

It is good to have friends in high places...

Please don't get anyone in trouble.

cfitz

Don't worry about it .
Cause the link is not fixed.
I got the new link for v1.1.1
http://home.pchome.com.tw/cool/cdtools/kprobe.zip
http://home.pchome.com.tw/cool/cdtools/kpsetup.exe

I already downloaded them , and
kprobe.zip is a real zip file of kprobe.exe
if you have not installed kprobe , you must run kpsetup.exe
if you installed kprobe before , you can only overwrite kprobe.exe.
I extracted kpsetup.exe and I found that kpsetup just contains
kprobe.exe and ASPI driver.
According to my friend , kprobe still has many hidden functions.
So now I still try to learn something unknown.

[cfitz]

Excellent! Thanks Abacus - our hero for today.

cfitz

[MediumRare]

Great, Abacus. I'll try and get the new version too.

In the meantime, I confirmed your mouse results:
- mark area with left MB to zoom in. Careful: the scaling is zoomed as well, so mark entire y-range.
- mark with left MB and pull up out of top of graph to return to 100%
- click and drag right MB to scroll with time
- apparent values < 1 are discretization artifacts from "connecting the dots" (especially noticeable on semilogarithmic representation). Check this by zooming in.

Selecting the entire disc to scan is relatively easy: set the end MSF-value to a large value (e.g. 99:x:x) and the tool substitutes the disk maximum for a C1/C2 scan (NOT for ATER). It is nice to be able to select the scan range and a fixed reading spped (other than 8x and max).

What, incidentally, is ATER/ADER (alternative to C1C2)

And I had a revelation re. CD-Doctor results: the Doctor suppresses counts of 1. In retrospect, this should have been clear from the scale (which goes from 1 to 100): values directly on the axis are not shown. Karr Wang has elected to start at 0.1, which shows the single counts, but makes low error rates appear dramatic when displayed logarithmically.

G

[cfitz]

It's like the ultimate beta-testing going on right here. I bet the developer(s) are reading along with us.

You may well be right. Note that the latest version Abacus kindly linked for us (v1.1.1) is dated 4/5/2003 - today.

First, look at this screen shot:



Then compare it to some of the comments made in this thread so far:

The comparison makes me appreciate several things that CD-Dr. does well:
- add caption to graph
- report on maximum and mean error counts

Unlike CD Doctor which automatically tests to the end of the data on the disc and draws the graph appropriately, K’s Probe, like WSES, requires that you explicitly set the limits for the test.

All of these comments have been acted upon. Now K's (wonderful) Probe allows you to add a caption to the graph by clicking on the configuration button ( - this is also where the log/linear graph choice is made now)

In addition, the maximum and average C1/C2 counts are displayed rather than the current and total. Finally, there is a new "Disc Size" choice below the "MSF" and "LBA" choices. This new choice automatically sets the scan to cover the whole disc.

Wonderful! Coincidence? Maybe, maybe not. But even if it is, it is a happy one.

Now, any chance we will be able to save charts in png format soon??? Let's hope so!

cfitz

[cfitz]

Selecting the entire disc to scan is relatively easy: set the end MSF-value to a large value (e.g. 99:x:x) and the tool substitutes the disk maximum for a C1/C2 scan (NOT for ATER).

As I noted above, this isn't necessary anymore with version 1.1.1. Just select "Disc Size".

What, incidentally, is ATER/ADER (alternative to C1C2)

http://www.cdrlabs.com/phpBB/viewtopic.php?t=9387

And I had a revelation re. CD-Doctor results: the Doctor suppresses counts of 1. In retrospect, this should have been clear from the scale (which goes from 1 to 100): values directly on the axis are not shown. Karr Wang has elected to start at 0.1, which shows the single counts, but makes low error rates appear dramatic when displayed logarithmically.

Sorry. I didn't know that people were unaware of this characteristic of CD Doctor. I have always recognized that you couldn't see error counts of "1" on CD Doctor because the scale starts at 1, but it was evident to me that they were there based on the average values CD Doctor reports. I don't think it is fair to say that CD Doctor is suppressing them, because they do get counted in the average. But you are correct that they aren't visible on the chart. I recognized this as a shortcoming, and allowing the scale limits to be set by the user or to at least start at 0.1 by default as Mr. Wang does is one of the suggestions I sent to Miyuri for improving CD Doctor.

cfitz

[cfitz]

And I had a revelation re. CD-Doctor results: the Doctor suppresses counts of 1.

One more thing: were you aware that you can save the results of your CD Doctor test in a comma separated variable file? This is useful for doing your own enhanced analysis or plotting, and is one way to view the error counts of "1" that CD Doctor isn't really suppressing.

Hey, that's another addition that would be nice for K's Probe - the ability to save the data to a file for post-processing. Has anyone found a way to do this?

cfitz

[MikeTR]

I am kind of busy with work right now, so I don't really have much to play with it. But there are plenty of others, I think, who grabbed a copy and can report what they find.

Wish I could be of any help, but as I said my Litey is now in my girlfriends computer. I'll be sure to give it a run as soon as possible.

[rdgrimes]

The problem I was having with the BMP files in version 1.1.0 have strangly vanished in v 1.1.1.

[MikeTR]

It's like the ultimate beta-testing going on right here. I bet the developer(s) are reading along with us.

You may well be right. Note that the latest version Abacus kindly linked for us (v1.1.1) is dated 4/5/2003 - today.

Looks like Lite-On's 'after-sales-service' should take a look at how their R&D operates. "This could be the beginning of a beautiful friendship.........."

Thanks for the link Abacus.

[cfitz]

The problem I was having with the BMP files in version 1.1.0 have strangly vanished in v 1.1.1.

Yet another apparent response to the feedback in this thread. Great!

cfitz

[rdgrimes]

I would also ask that the saved BMP include the statistical data, not just the graphs.

[cfitz]

Yes. Good point. It would be nice to have the max/avg/etc. data all in one place with the charts (like CD Doctor)

cfitz

[Dartman]

Thanks guy this is another great tool. I couldn't use the 166s drive with CD doctor but it works fine with their program. It does appear to agree pretty closely to CD doctor in the readouts I'm getting. It also now apears
that the 166s has the cleanest error correction of all my Lite ons. It is higher at the begining but still very low then just reads most of the way through with no errors pretty much after that, even on some disks that the other 2 got higher readings on, and one that had a nasty c2 error on the 48x.
Well hopefully they keep developing it and maybe support other drive's just to see who's tops at reading disks.

[MediumRare]

And I had a revelation re. CD-Doctor results: the Doctor suppresses counts of 1.

One more thing: were you aware that you can save the results of your CD Doctor test in a comma separated variable file? This is useful for doing your own enhanced analysis or plotting, and is one way to view the error counts of "1" that CD Doctor isn't really suppressing.

Hey, that's another addition that would be nice for K's Probe - the ability to save the data to a file for post-processing. Has anyone found a way to do this?

cfitz

Sorry maybe I formulated that too strongly. I really should have realized that 1's aren't shown. How do you create the CSV-list with CD-Dr?

I just managed to download the new version- I had kept trying but couldn't access the site with Opera. I switched to IE just now, and that worked.

Will try it a bit, but I have a cold and have to get to bed.

In case Mr. Wang is monitoring this, let me mention one more wish: in case the option has not yet been implemented (I haven't tried 1.1.1 yet), I would like to be able to save my preferences from one run to the next (selected drive, preferred read-speed).

As to logarithmic scale, if the plot routine is flexible enough to show partial decades, it may be useful to set the lower limit to 0.5 to deemphasize the low-error range.

Let me emphasize that these are heady times- a bit like Christmas, a new toy to play with and someone who cares about the suggestions made. Many thanks to all who are making this possible!!!!!

G

[cfitz]

How do you create the CSV-list with CD-Dr?

From the "Result Viewer" window, select File->Save just as you would to save the png graph. Then in the "Save as type" drop-down box, change from "Portable Network Graphics (*.png)" to "Comma Separated Value (*.csv)". Easy (but somewhat hidden)

As to logarithmic scale, if the plot routine is flexible enough to show partial decades, it may be useful to set the lower limit to 0.5 to deemphasize the low-error range.

Or add an option in the new configuation dialog to to fix the scale limits to user defined values.

Will try it a bit, but I have a cold and have to get to bed.

Please take care of yourself. And don't catch that new SARS disease. That is pretty scary stuff.

cfitz

[MediumRare]

How do you create the CSV-list with CD-Dr?

From the "Result Viewer" window, select File->Save just as you would to save the png graph. Then in the "Save as type" drop-down box, change from "Portable Network Graphics (*.png)" to "Comma Separated Value (*.csv)". Easy (but somewhat hidden)

OK I'll have to play around more with options...

Please take care of yourself. And don't catch that new SARS disease. That is pretty scary stuff.

The guys at work are making snide remarks- "a Canadian with a cough- do we have to panic?"

but I'll be fine.

G

[cfitz]

]The guys at work are making snide remarks- "a Canadian with a cough- do we have to panic?"

That's odd. Around here they say "A Canadian - do we have to panic?"

No offense to you, Spazmogen and all our other friends from up north...

cfitz

[jsl]

Do the PI/PO DVD graphs show the number of errors (before correction) in the Parity of the Inner/Outer code per ECC block or...? In that case it would be nice if you could measure the Parity Inner errors in 8 consecutive ECC blocks (PI sum 8 ) too, which according to the DVD specifications should be <280.
Do you think the PI/PO tests are reliable on a Lite-On DVD-ROM or are they only made for Lite-On's coming DVD writer? It would be interesting to see some tests on some known good and bad DVD±R/RW media.

[EddieKM]

Since it is under development, I hope it can offer jitter test in the next release. It is really useful for the one who always burn audio cdr. How do you all think about it?

[Abacus]

cfitz:
The new version of KProbe are released .
The link of KPobe v1.1.2 has no change.
http://home.pchome.com.tw/cool/cdtools/kprobe.zip
http://home.pchome.com.tw/cool/cdtools/kpsetup.exe

KProbe v1.1.2 add csv format and jpeg format
try it ..

[rdgrimes]

1.1.2 also adds the data to the exported image. Very cool.
Since we seem to be getting everything we ask for here: Can I have a couple of new drives?