PI/PO scanning

vbimport

#1

Looks like the read speed has a big effect on the result. I guess it’s one contributory cause to why the errors seem to increase by the position on the disc when reading in faster speed using CAV technique too?
The DVD specification of 280 PI errors in 8 consecutive ECC blocks is when reading in 1x speed right?

Edit by rdgrimes:

I’m making this the PI/PO discussion thread. Let’s use it for discussing PI/PO scanning topics.


#2

Thank you jsl!

Good point! Don’t know how I could have overlooked that. Yes those specs is most likely for 1X reading.

Now my brain got some more thinking and testing to do :stuck_out_tongue:


#3

Guys, I created a new thread cause I think we need one for these discussions of PI/PO scanning. It’s a topic that needs a lot of discussion!

I too have been thinking about the read strategy, not only speed. The discs are recorded in CLV 4x (2.4x), so why not read them at the same speed (CLV). It’s a constant speed. Also doing some testing regarding the ECC sum setting, it’s at 1 by default.

By way of a more practical test, I’ve have playback problems on a couple movies that had PO spikes near the end (as little as 60 max when scanned at 6x (PCAV)). These same movies also show red blocks in DVDInfo near the end.
Reading them at lower speed eliminated the spike, but not the playback problems. :frowning:


#4

Just playing around here. Here’s a 4.4GB movie scanned in the 401S:

with ECC sum set at 1:

with ECC sum set at 8:

As you can see, the total count is not significantly changed, but the max count is more or less changed by a factor of 2x.

Here’s the same disc scanned in the 166S. Keep in mind that 166 scans at 8x CAV.

Scanned with ECC sum at 1;

now scanned with ECC sum at 8;


#5

Originally posted by rdgrimes

I too have been thinking about the read strategy, not only speed. The discs are recorded in CLV 4x (2.4x), so why not read them at the same speed (CLV). It’s a constant speed.
This is what I’ve been doing with my CDR stuff for a pretty long while, i.e. burnt @CLV, tested @CLV.
As simple as that. I mean, I don’t see why I should do otherwise.
But thank God, the others’ racing mentality slowly but surely (?) seemed to start dying out, lol.


#6

Originally posted by rdgrimes
Just playing around here. Here’s a 4.4GB movie scanned in the 401S:

I can see that it’s on a Ricoh media, but what brand does these discs have? I have tried the media that came with my DVD-writer (also a Ricoh) and it reads fine. But on Ritek 3 of 4 discs are unreadable.

RITEK

Media from LITE-ON


#7

That must have been a bad batch as my ritek is way better than Arita, TDK and memorex…

All these discs is identified as ricohjpnr01. Not tested with Lite-On though, only with plextor PX-708A and memorex dual-x


#8

Right now, the best media I have is the Fuji/Ricoh DVD+ burned at 2.4x, and the +RW is better than the +R. The Memorex +R Ricoh has been “variable” at best.


#9

Since the standard is for 8 ECC blocks, should we be scanning with Kprobe set for 8 ECC? Seems like that would give the magic number of 280 or less, right?
Is there a standard for PO?


#10

I have received word from Karr that it is best to set the ECC sum at “8” to conform to the DVD standards.

For DVD spec., it is better to set ECC sum to 8.
The another benefit is increasing the accuracy of PI/PO.
It is because if you set ECC sum to 8 , KProbe can reduce
the burden to update screen data.

Also, in a future version of Kprobe, he will include that setting in the saved scan image.


#11

The following is a translated excerpt from the German ct’ magazine:
“With Kprobe you can visualize the remaining errors after the first (PI) and the second (PO) error correction phase. To get comparable results (of Kprobe) with our laboratory measurement (Audiodev), you should set Kprobes PI/PO Sum Error to 8. …
Kprobes PO-results are corresponding to (our laboratory measurement) Parity Inner Failures (which should not exceed 4). Parity Outer Failures (PO), the final reading errors are not reported by KProbe.”

“According to the specification the error margins of PISum<=280 and PIF<=4 are valid if the disc is analysed using 1x speed”.

According to an older article in the ct’ on Kprobe, this program was designed to work with Liteon DVD-Writers, not DVD-Roms. Does anybody really know whether the results Kprobe delivers on both drive types are RELIABLE?

My thoughts:

  • Kprobe’s results should be read with caution! This program gives only an indication of the burn-state of the DVD±R(W).
  • As long as Kprobe’s PO results are within (lets say) 15-20 max, this disc should be considered as clearly readable. Keep in mind that we are testing with a speed of 8x or higher instead of 1x according to the specification.
  • No burned disc is error free. As long as the drive’s ECC can easily cope with the errors in a way that transfer rate is not affected I would consider this disc as within specs.
  • You should test a disc with important data always on two different drives, to make sure it is readable (if you want generate CRCs first and then recheck them on the disc).

Just my two cents,
Moonstar


#12

My experience has been that it’s not too difficult to get PO values under 10 at 8x sumECC, when scanning at 4x or lower. Scanning at 2.4x gets them considerably lower.
I still see an increasing curve of errors near the end at 4x, which seems odd to me because the linier speed is constant at 4x. this curve seems to flatten at 2.4x.
I do not recommend scanning in DVD-ROM, partly because of the locked speed at 8x. Scanning at 1x is just TOO boring! :frowning: But it sure does eliminate the errors.


#13

The statement that DVD-ROM drives are not very accurate in reading error rates (as opposed to Lite-ON DVD burners) seems very plausible. I did a new test yesterday with an Arita DVD+R 2.4x disc. I’m sorry I do not have the graphs anymore, but here were the basic results done with three different Lite-On DVD-ROM drives, and performed at different times during the day.

Read at 4x speed, 8 ECC blocks
AVG PI/PO-values:

Lite-On 165 @ 12.00: 6.8/4.6

Lite-On 166 @ 12.00: 13.6/4.4
Lite-On 166 @ 14.00: 12.5/4.4
Lite-On 166 @ 16.00: 9.7/4.4

Lite-On 166 @ 12.00: 123.0/4.5
Lite-On 166 @ 14.00: 194.4/4.4
Lite-On 166 @ 16.00: 118.6/4.4

So the PI-values are completely random. The PO-values, however, seem very consistent (4.4-4.6) and as they correspond to the official Parity Inner Failures, the value might be useful for judging media quality.

Can anybody tell me how to calculate the PiSum value that is mentioned before (which needs to be <= 280 if read at 1-speed)?


#14

Originally posted by rdgrimes
I do not recommend scanning in DVD-ROM, partly because of the locked speed at 8x.

Is there any way, hardware or software, to get the Lite-On DVD-ROM drives running at 1-speed instead of 8x?


#15

KProbe is designed to work with the DVD burners. Either the DVD-ROMs do not respond to the read speed commands, or they are unable to respond and are locked at 8x on recordable media.


#16

Originally posted by rdgrimes
KProbe is designed to work with the DVD burners. Either the DVD-ROMs do not respond to the read speed commands, or they are unable to respond and are locked at 8x on recordable media.

Are you willing to do a ‘repeatability’-test for me? I mean, test the same burnt DVD three times during the day and see if the results match or vary as much as the once I posted above…?

That way I can decide whether or not it’s worth buying the Lite-On burner to do testing… :slight_smile:


#17

Are you willing to do a ‘repeatability’-test for me?

What is it you are looking for? Any 3 scans will vary. I’m not sure I understand your scan results completely, but the variation in PI is not unusual. Those discs that are more “marginal” in quality will generally exhibit a lot of variation in error scans.


#18

OK, here’s the same DVD+R scanned at 3 times during the day, about 3 hours apart, the earliest scan is first.

As you can see, some variation is normal. Experience has shown me that the higher the error rate on a disc, the more variation you will see from one scan to the next. This disc is pretty high quality, so the variation is less. How a drive handles error correction and how it “likes” the media is also a factor.


#19

Anyone here heard of a new program made by Sanyo that can do a Pi/Po scan as well? It was posted on the cdfreaks news page a while ago. Funny how there’s no thread about the program at all…


#20

Yeah. Maybe since the Sanyo UM Doctor II program only works with two optorite drives and costs shitloads of money to buy?