CD-DVD Speed Disc Quality (PI/PIF) Test

Please note that there is a newer guide for CD-DVD Speed available in this forum.
You will find it in this thread:

CD-DVD Speed - A user guide

Latest version of Nero CD-DVD Speed
You will ususally (but not always) find the latest version of Nero CD-DVD Speed available for download at the official CDSpeed site.

You can also download the latest release of CDSpeed from cdfreaks.com:
Download CDSpeed version 4.60 made available on 31 august 2006.

Thanks to Erik Deppe and Nero for making this version available for download.

- DrageMester

---

Greetings, and welcome to the CDSpeed discussion and tutorial thread.

This thread is for the discussion of CDSpeed Disc Quality testing. Any CDSpeed related topic is open for discussion here, but the focus is on the Disc Quality Test. I'm also including a brief tutorial on using the Quality Test.

To begin, visit the CDSpeed web site and download the latest version of CDSpeed. Many thanks to Eric Deppe for his excellent work on this wonderful tool. CDSpeed has long been the reliable standard for a wide variety of CD/DVD testing tools. CDSpeed contains a bunch of tools for testing a variety of disc and drive functions, but here I will focus on the Disc Quality Test only.

CDSpeed now supports a range of drives for error scanning. To see the list of known drives that are supported, look at the current list of supported drives. Each of the supported drives may have some different features supported in the scans, and each drive may report errors a bit differently from the others. It is not suggested that you compare scans of the same disc in different drives, or that you compare scan results from CDSpeed to results from different programs. Each has it's own way of reporting things, and so comparisons are meaningless.

The main purpose of scanning is to compare one disc to the next in the same drive, and to evaluate different media types and burn speeds. It's also strongly recommended that you only compare scans done at the same scanning speed. Also only compare scans done with the same read strategy, comparing a scan done in CAV mode with one done in CLV mode is pointless.
You will find that different drives read in different strategies, and some may not allow the selection of read speeds at all.

To begin, make sure that you have the latest version of CDSpeed, version 3.00 or later. Just click the help menu and "about" to see your version:



To update an older version, just download and place the new "CDSpeed.exe". file in your Ahead program folder in the "Nero Toolkit" directory. The new file should over-write the old one. Also check to be sure that in that directory is a copy of "gdiplus.dll". This file allows the saving of images of your scans.

Now that you have the program installed or updated, start by checking the "options" for the quality test. Do this by opening the file menu on the main CDSpeed screen, then select "options" and the "Disc Quality" options:



At the bottom of the pane you can select the colors for displaying your test results. Choose what pleases you, there's no preferred colors. Above that you can set how your graphs will be displayed. I recommend using the "automatic" option, which will allow the program to auto-scale your graphs. If you prefer to have all your tests come out with the same graph scale, you can manually set that scale here.

Note, if you want to save all your tests in a particular image format, go to the "capture" options and set that format there. CDSpeed will give you the option of changing the format when you do a save, but setting it here will make that format automatic.

OK, now you're ready to do a test. You need a burned disc. It's always best to use a disc that's completely full, so you can evaluate the entire surface of your media and maximum speeds. If you are simply testing a new media, I suggest using the "Data Disc Test" to create a full disc for scanning.You will find that test in the "Run Test" menu on the main screen.

You will find the Quality Test in the "Extra" menu.



CLICK!



If you have a disc loaded, you will see the disc info displayed in the upper right of the window. Below that is the "settings" area. Here you select the speed for scanning (CDSpeed will remember this the next time you use it). As I mentioned above, not all drives will support setting a read speed, and only the max speed is available. You may also see some nonsense speed displayed, like "48x" for a DVD disc, etc. You will have to try the settings on your particular drive and see how it functions in the test. As long as you have enabled the speed graph, the actual speed that is used will be obvious.

You can also set the area of the disc for scanning. Most times you should leave this at the default, which will be the total burned area of the disc.

Below that you will find the settings for the graph displays. Note that there are 2 graphs. The upper graph displays the read speed and the C1 or PI errors. The scale for the speed is on the right of the graph and the error scale is on the left. De-selecting one of these options will remove the appropriate scale from the graph. If your drive does not support that particular function, it's best to deselect that option so as to avoid confusion on the graph scales.
The lower graph is for the C2 or PI-Failures., and jitter. Again, if your drive does not support a feature, deselecting that option is advised.

Most people will be familiar with "PO" errors, rather than "PI-Failures". These are essentially the same type errors, so don't be confused by references to the 2 designations. While there may be disagreement as to what the 2 designations mean, you can generally assume that they are comparable. Again, different drives may report slightly different things. For further explanations of interpreting these errors, refer to the interpreting PI/PO errors thread.

Below the settings area, you will ses the displays for speed, progress, and also the "Quality Score". The "Quality Score" seems to be somewhat arbitrary, and reliance on this score for an indicator of your discs's over-all quality may not be advisable.

So once you have your settings right, just hit "start" and go have one of your favorite beverages.
When the scan is complete you will see this summary in front of the scan window:



And behind that you will see a completed scan:



In case you didn't notice I used a different drive and disc in this image so as to display jitter.
To help you identify the various graphed results, in the above scan the jitter is pink and the general level is just under 12 (right hand scale). The "PO" is yellow (left hand scale), the PI is blue (left hand scale), and the speed is red, (right hand scale).

- Drives with 1ECC scanning interval: Lite-On, Plextor, NEC
- Drives with 8ECC scanning interval: BenQ, NEC

- Nero CD-DVD Speed's quality score is based on Max PIF.

Max PIF	Quality Score
	8ECC	1ECC
1	99	98
2	99	95
3	99	93
4	98	90
5	97	84
6	97	77
7	96	70
8	95	64
9	95	57
10	94	50
11	94	48
12	93	45
13	92	43
14	92	40
15	91	38
16	90	35
etc	etc	etc

Saving your scan

In the upper right hand corner of the test window you will see a small button for saving to image files.



Click the indicated button and you will get the standard Windows "save dialogue" where you can name your file and format. The file name will be auto-filled according to how you have your options set in the main CDSpeed options, simply change this if you wish. If you're going to be saving a lot of scan images, it's advisable to create a naming scheme to keep them straight.

So there you have it, start testing and enjoy!

Use this thread to report your experiences, and ask questions. Also be sure to report if you find the tool working on any drives that are not listed in the compatable drives list.

---

Last edited by DrageMester; 11-09-2006 at 01:01. Reason: Newer CD-DVD Speed User Guide

Some additional notes: make sure to choose show statistics in the options if you want the statistics dialog box to pop up after a test as shown above. Also choose 4x speed for scanning. And yes this tool will work with liteons as shown above! minus the jitter testing.

also dont worry so much about the quality indicator you really need a almost perfect scan to get quality levels in the 90s.


for the Nu drives, In extensive testing with this tool (and others) ive found generally under 15% jitter is considered a good level. ive had skips etc. when playing back movies when the jitter displayed above 25% (which i would consider the upper limit) this is using the nu082 drive. if you notice the jitter rising up a bit while scanning a 8x burn, this is normal as speed increases causes more jitter.

As RDG said PI failures is roughly what POs mean in the kprobe testing, and as the NU drives returns in 8ecc sum blocks they may be higher than 4 and you will still be ok. Generally from what ive seen under 10 is very acceptable with NUs.

only thing i dont like about the nu is the spikes which can skew your quality rating..the liteons dont seem to spike as much. a way to remove spikes would be a great addition

What's the difference between PI and PIF? Any implications in the test if a burner doesn't support PIF?

How much is too much PI failures 4 or 32? will it be the same for all burners?

Anyone have an idea how much jitter is too much jitter? above 15%?

The "quality score" you say is arbitrary, but isn't some algorithm involving Pi, Pi failures, and jitter? Are these the variables involved?

Thanks and good work.

Ok seems like Jamos answered 2 & 3 while I was typing it up, what about 1 & 4?

I'd like to see some more head to heads with jitter involved.

PI is the same as kprobe correctable errors. PIF is errors that are not correctable by the drive basically the same as POs in kprobe2.

depends on how the drive returns the information..on a liteon you would look for 4 or under..on a NU id say 32 would be absolute max but this has to be determined by concensious.

15% or under would be standard imho..the upper limit still needs to be determinded ive found much above 25% i start getting issues with playback.

Never said it was arbitrary just dont worry that you dont get in the 90s..generally a very well burned disk will be in the 70s to 80s.

Here's a suggestion, it would be nice if you could copy and paste of all the relevant information/statistics for that scan "formated for pasting into a thread" was an option, for those that don't want to take the time to post pics. just a thought.

Just want to point out that Erik has "color coded" the maximums to visually indicate the relation to spec as seen here.
Maybe Erik can elaborate a little on the thresholds...

Finally we have an other nice "scan"-tool to play with...

rdgrimes, I couln´t resist it. Just have to add two scans of same disk.
"Movie", burned at 8X with 411@811 (Z-modV4) on spdhacked FW HS0K, booktype DVD-ROM.

Scanned with CDSpeed3;


Scanned with KProbe 2.1.0


To me they look pretty the same. Ouupps, I know we should not compare...

@Pinto yes it is very comparable to kprobe2 using the same liteon drive. (basically polls the same info). And note how his quality score for a nearly perfect scan is only 93 just shows how hard it is to get in the 90s.

Cdspeed3 also works with Benq drives also.
Two important notes on Benq drives at the moment:
1. PIFs do not get reported.
2. Scan speed is locked to somewhere just a little above 1.5x so scans take a while.

one big advantage at using the benq for scanning is for a pretty accurate measure of the jitter rate that would be close to playback speed.

heres a NEC2500A 8x burn scaned with my Benq 822A..the jump in jitter is normal due to the nec changing write speeds. ignore the 99 quality since there is not feedback for PIFs this score is skewed.

what does he need to elaborate on? at a average polling interval at 1.27ecc your maximum Pis should be under 280 (i beleive he polls these at 8ecc).your maximum pi failures should be around 5(4x1.27) or under ( though you need to look at your graph and see if its just a spike thats causing your 10 as your average is very low).

---

Last edited by Jamos; 22-05-2004 at 15:23.

HEY THANX. REALLY COOL NEW FEATURE!
Saves me from buying a Liteon for Testing quality.

Often if not always when scanning a DVD+RW with my Benq 822(B3BC) I get this error-message at the end of the scan instead of the statistic-window:
Error!
NO ADITIONAL SENSE INFORMATION (04091E)
With DVD+Rs until now it didnt happen.
I always use latest Nero(6313 or so) and write single (closed) sessions only.

Also I want to make a suggestion:
couldn't you place such a colored box on each scale as well, to make it clear which scale is there for which graph...

Have you tried upgrading your firmware? b3HC is out.

I know. No I didn't.
I want to wait to see it at an official source first. (like http://www.benq-eu.com/ServiceAndSup...productline=14
) and some people to confirm that it functions fine and has no issues.

So sorry, I can't try that by now. But if no one else has the problem, than its fine with me. I just close the error-window and can see all the results (without the statistics-window). :O)
Oh. Iforgot to mention: Still using Win98se over here. (I know I could use XP or 2K ;O) )

Also the german translation is a bit odd:
'PI Errors' are translated as 'PI Fehler'
'PI Failure' are translated as 'PI Failure'
The problem is, that in german 'Error' and 'Failure' both mean 'Fehler'.

My suggestion:
Keep the english words as you do it with many other untranslateable words, or make clear that 'failure' is worse than 'error' by something like this:
'PI Errors' translated as 'PI Fehler'
'PI Failure' translated as 'PI Unkorrigierbar' (means: uncorrectable)
or
'PI Failure' translated as 'PI fehlgeschlagen' (means: failed)

---

Last edited by KillMe; 22-05-2004 at 16:33.

Sure, if you don´t like german translation, you can always use English language.

BTW, the swedish translation is also poor...

If BenQs don't report PIFs hopefully he can substitute POs (which BenQs do report) for PIFs.

Great app, highly recommended.
Thanks Erik!

What are the chances that LiteOn will implement jitter measuring?

@rdgrimes: Sorry, didn't find this thread. I must be totally blinded...

>If BenQs don't report PIFs hopefully he can substitute POs (which BenQs do report) for PIFs.

I hope PI Failures can be enabled for the Benq later on....

Look here:
http://club.cdfreaks.com/showthread.php?t=95852

Halcyon asks there: BenQ (with Nexperia) doesn't support PIF, only PI and POF and Jitter. Is this a support issue with the software or limitation of the hardware? Does anybody know?

nicw answers there: firmware limitation from what my Benq contact told me.

Halcyon asks there: Any info as to whether they'll be implementing PIF scanning support in later firmwares?

nicw answers there: ....We discussed adding extras in as you suggest. The next step is trying to get the R&D time to get it implemented. I really hope it will happen, they are quite enthusiastic.....

Why can't the feature of setting the drive speed from CD-Bremse be implemented into cdspeed?
I use my LiteON 166s to test my discs. As it is now, I run the cdspeed first. Then CD-Bremse to set the speed to 4x. Then kprobe to test PI/PO.

Now with cdspeed3 I can't see no need of running kprobe anymore. Would be nice to have no need to run CD-Bremse either.
(If the author of CD-Bremse is reading this, I mean this in a good way. I couldn't do my tests without your program:-)


Also, is there a way to run both the cdspeed and "Disc Quality Test" by pressing one button? This would be the ultimate way of doing things as I see things. Press one button, go to the toilet or whatever, come back and both tests are complete.

I just used CDSpeed for the first time and it is an excellent app. Here are the results for a backed up DVD burned on an NEC2500A @4x as well as Kprobe for comparison.

Does anybody know if this tool will support Pioneer drives at some point (107d in particular)? I tried running a disc quality test and got a message "error initializing test". I know Pioneers are not listed as supported drives, but I figured I try running it anyway...

That's up to Pioneer. The chipsets used in the NEC and Pio drives may or may not support error reporting, we don't really know.

Like DJ_Farid, I also have a 166s (Sony rebadged Lite on) that I have started using to testing purposes. I use CD Bremse to lock the read speed but I'm curious if this is as "accurate" as if I were using it with a burner to read it? I just want to clarify because Kprobe specifically says to use with a Lite On Burner.

NS******** type=text/javascript> vbmenu_register("postmenu_599668", true);

It is not accurate to compare with others, but you can use the 166s to compare with scans you have done earlier yourself.

If (nearly) nothing is accurate, why are we doing these tests and how convincing are
these scans? If I can only compare with my own scans, how will others be able to
benefit from them?