Precision, accuracy, and reliability of disc quality (PI/PO/jitter) tests
| Disc Testing Methods and Software Discuss, Precision, accuracy, and reliability of disc quality (PI/PO/jitter) tests at Blank Optical Media forum; Quote: |
-
Posted:
13-02-2007
- # 126
If you were familiar with scientific investigations, you would not have said so. Any measurement is subject to variations, though in different extents. If the drive is not defective and there are no physical irregularities in the disc, the differences between test runs will be small and even negligible in many cases. You need to do more homework, IMO.Quote:Originally Posted by monohousecall me a offtopic 1-poster, but I don't believe ANY test on ANY drive can give accurate nor precise result, simply because if you repeat the same test you will get different results, even on the same drive under the same conditions.
the ONLY thing that really matters and that really makes a difference is whether or not you can safely read the data, that is ALL, regardless of whatever may be underneath the drive, tests will NOT tell you anything that you don't probably already know (you would probably test when a file is not readable - which is not very useful anyway) - and in case you think that using the tests can help you figure out if the disc needs to be replaced before it needs to be replaced is also inaccurate for the same reason that tests are inaccurate - different time - different results - different situation - not once it happened that yesterday the disc didn't work and today it does.
You have misunderstood the first reply, which wrongly assumes that no scan is inaccurate.Quote:I agree with the first reply - tests are unuseable.
By using some reliable drives to test the writing quality of discs, you will be better informed in several aspects.Quote:....unless you think quantum physics in which case the only way to be sure is to measure using an objective measurement equipment....and if you're going that far just to know if it is time to replace the disc - maybe it is time to do some tests on your brain ?
If you use a consumer drive that does not give false-negative results in C1/C2/PI/PO errors, disc quality test will be at least as sensitive as TRT and will be indicative of the readability of the disc by that drive at the same speed setting.Quote:Originally Posted by scoobiedoobiescanning in consumer drives isn't really about true 'accuracy' in comparison to a pro testing device anyway, they are for comparing trends in burn quality with different drives/speeds/firmware, judging disc stability, etc. A quality scan in a consumer drive shouldn't be considered a real substitute for testing in something like a CATS device, and it shouldn't be a substitute for a Scandisc or Transfer Rate Test to confirm with certainty disc readability.
- Today (MyCE Staff)
- Posts: 15,596
- Posted:
13-02-2007
- # 127
I am very pleased with my Lite0n as a testing/scanning drive, as over time I have learned to understand the results that it gives me and what it adds into the mix. I am also willing to believe in the results that it gives me because I can (and often do) re run the test at the same and different speeds and get a very close match with 99% of the tests that I do.
Where as my NEC it could 'almost' be another disk each time I re-run the same DVD, its a bit better with CDR's, but I dont trust it anywhere near as much as the Lite0n, but it is ok in the TRT area where it is better at showing read problems.__________________
ASUS A7N8X nForce 2, AMD Tbred, 2GB G.Skill EL Plats, Radeon 9550 Seasonic S12, Maxtor 200GB x4 WD 74 GB, Win2000/Slackware
LG GSA-4167B - LiteOn SHM-165P6S - Optiarc DVDRW AD-5170A Sony CPD-E500 also E4500 C2D & nF4 SLI 185 opty via KVM
- - - - -
DFI LP RDX200 CF-DR Athlon 4200x2 (LDBFE) Zalman CNPS9500 AM2 2 GB G.SKILL HZ's, x1650XT 24Pipe, Raptor 150GB, 1TB Hitachi DS x4, 320GB Hitachi DS x2 Corsair HX 620 in Eclipse 62, LG GSA H42N - Lite0n SOHW-832S - NEC ND-3550A - Asus CRW-5232A Mitsubishi Diamond Plus 200 - Posted:
07-03-2007
- # 128
Verbatim MCC 004 16X DVD+R burned at 8X default
Burner: Lite-On LH-20A1P 487C KL0A(stock) Burn Size: 4482 MB (CD-DVD Speed Data Disc)
First burn after flashing new firmware KL0A AND Restore EEPROM using original EEPROM backup created just after purchasing this drive.
Nothing checked in the SmartBurn Utility.
CD-DVD Speed's 4X jitter curve appears to indicate that jitter is highest(9.3%) at the start of the burn and pretty much declines gradually out to the edge of the disc(about 8.0%--there is a small rise at 0.7 GB indicated.) CD-DVD Speed's Disc Quality 4X graph pleased me very much, in PIE, PIF and Jitter.
DVDScan Speed 2 Jitter test curve tells a slightly different story.
I have cut and pasted the write curve, CD-DVD Speed's jitter curve and DVDScan's jitter curve in to composite below for comparison. DVDScan's curve starts relatively constant(just above 10%), then begins to increase at about 0.4 GB--reaching 12.0% at about 1.1 GB or very near the point where the drive switches from CAV-writing to CLV-writing. After the switch to CLV-writing, DVDScan's jitter curve pretty much declines gradually out to the edge of the disc(10.4%).
IMO, DVDScan's jitter curve makes more sense. Precisely at the writing strategy switch there is a small drop in the jitter curve and the reported jitter declines thereafter.
Although I would prefer a disc with low jitter, CD-DVD Speed's jitter curve seems to display what I want to see, while DVDScan's curve seems to fit the facts of the burn more closely. The fact that DVDScan was created by Lite-On people for use with Lite-On drives also nudges me towared the DVDScan result.
But it's just my opinion.
__________________
Useful software with my burner:
DVDFab+++++DVDShrink+++++ImgBurn+++++CD-DVD Speed - Posted:
07-03-2007
- # 129
Why don't you scan jitter at the same speed with the two programs? In my experience reported jitter can vary quite a lot depending on scanning speed, so you need to use the same speed to make a fair comparison.Quote:Originally Posted by uSerKeyCD-DVD Speed's 4X jitter curve appears to indicate that jitter is highest(9.3%) at the start of the burn and pretty much declines gradually out to the edge of the disc(about 8.0%--there is a small rise at 0.7 GB indicated.) CD-DVD Speed's Disc Quality 4X graph pleased me very much, in PIE, PIF and Jitter.
DVDScan Speed 2 Jitter test curve tells a slightly different story. - Posted:
12-03-2007
- # 130
Quote DrageMester:
You are absolutely correct.Quote:Why don't you scan jitter at the same speed with the two programs? In my experience reported jitter can vary quite a lot depending on scanning speed, so you need to use the same speed to make a fair comparison. 
I followed up with two more jitter tests,
DVDScan at Speed 4
CD-DVD Speed at 2X
I am still uncertain on interpreting the results.
The re-test at Speed 4 with DVDScan produced a curve contour very similar to CD-DVD Speed's jitter curve at 4X.
But the re-test at 2X with CD-DVD Speed produced a curve contour similar to its curve at 4X, but still different from DVDScan's curve at Speed 2 during the CAV portion of the burn.
__________________
Useful software with my burner:
DVDFab+++++DVDShrink+++++ImgBurn+++++CD-DVD Speed - Posted:
12-03-2007
- # 131
Quote from Erik Deppe:
I have frequently found when using my LH-20A1P 487C, DVDScan Jitter Test curves resulting from Speed 1 and Speed 2 tests are quite a bit different from Speed 4 test results. Not only in levels and averages, but also in shape.Quote:"Jitter with LiteOn drives
The scanning speed seems to affect the way jitter is reported. At the moment jitter can only be reported at 2x and 4x but I should be able to make it work for all speed settings."
Does it mean the Speed 1 and Speed 2 curves are wrong? OR That the Speed 4 result is contradicting the lower speed test results? Not necessarily. Correct me if I am wrong with the following statements.
1. It has been stated repeatedly in numerous threads that all disc quality testing by consumer drives are producing results based not only on the disc being scanned but also based on the scanning drive and the software used. Different drives and different software will produce different results on the same disc.
2. Such test results not only reflect media quality and burn quality but also reflect scanning drive characteristics at various speeds(with each particular firmware version) and software calculation & orientation (and possibly PC configuration.) I think everyone agrees on these 2 points.
What follows from 1. & 2. as far as jitter testing on DVDs has never been adequately explained as far as my searching has gone. I do think optical industry firmware developers consider jitter an important disc/drive characteristic and therefore it interests me in my enquiry into improving burn quality with my drives. The feature in the scanning drive's electronics which provides for jitter reporting suggests that jitter data are not something which can be completely ignored in DVD disc quality testing.
One confounding problem in interpreting jitter test graphs and numbers is ascertaining the portion of the reported data which reflect characteristics of the disc from the portion of the data which reflect characteristics of the scanning drive.
If the portion of the reported jitter data due to the scanning optical drive system is relatively small compared to the the portion due to disc characteristics, then the resulting data and graph should mostly reflect the resulting jitter due to the media and burn attributes. However, if the portion due to the scanning drive is relatively large, then the test results may very well be useless in ascertaining disc quality as far as jitter is concerned.__________________
Useful software with my burner:
DVDFab+++++DVDShrink+++++ImgBurn+++++CD-DVD Speed - Posted:
16-03-2007
- # 133
For Nero CD-DVD Speed I suggest you scan at 4x. The 1x and 2x scanning speeds are awfully slow, and IIRC you might even have problems getting the 1x scanning speed to work with CDSpeed 4.7.0.2. If you have plenty of time, I believe that 1x jitter scanning is the most "correct".Quote:Originally Posted by DJouneytso what are your software (Nero CD-DVD Speed 4.7.0.2 / 4.7.5.0, DVDScan 004) and speed (1x, 2x, 4x, 8x) suggestions to a Lite-On SHW-1635S user for jitter scanning
For DVDScan you can scan at the slower speeds and choose a longer sampling interval. Since I think most other people (although there aren't that many) use 4x Jitter scanning, you might consider using that as well.
I prefer the 1x scanning speed for TA Jitter, because it's the right scanning speed according to ECMA standards, and it's just as fast as the higher scanning speeds (the difference being only fractions of a second).__________________
Do not meddle in the affairs of dragons, for you are crunchy and taste good with ketchup. - Posted:
17-03-2007
- # 134
Quote TL0:
My Lite-On LH-20A1P is reporting pretty much in line with TL0's Lite-On 1635S as far as variation between scans.Quote:"Until enough repeat testing has been done with your particular Lite-ON drive which supports the TA test, I would be wary of relying on the reported peak shift values as I noticed my 1635S shows disturbing large variations between scans of the same zone (+/- 2%). DC Jitter varied much less between scans (+/- 0.2%)."
Two tables below show results of tests on 16 discs burned at various speeds with various firmwares. Observations are prelimary based on only 16 discs representing only 3 MIDs, only two burners and only one scanning drive (LH-20A1P 487C KL0A fb-eos.)
All MCC004 media are Verbatim, made in Taiwan(CMC) from the same spindle.
TA jitter tests were run 5 successive times at each scan speed for each disc--with the lowest and highest of the 5 values of reported DC jitter and Peak shift included in the tables.
All tests were done at position "Middle Zone."
A. Two prelimary observations on the values shown in the DC Jitter Table are:
(1) most of the reported DC jitter values reported at 2X speed are less than or about equal to the values reported at 1X,
(2) most of the reported DC jitter values reported at 4X speed are greater than or about equal to the values at 2X.
Disc 3 burned with KL0A is a notable exception to (2).
B. A CDS screen capture from tests on the disc showing the lowest DC jitter values(disc 8) in the table has also been included. LG's JE07 firmware burned discs providing some of the best looking histograms of the discs tested.
C. Peak shift values are less easy to understand. The Peak Shift table backs up what TL0 posted previously: the reported values vary more widely than the DC jitter values.
I have highlighted Disc 3, which exemplifies a characteristic found with several of the discs:
Peak shift values reported were lowest at the 4X scanning speed. (Disc 3 was certainly the odd-ball of the all the discs tested.)
D. The 2nd CDS screen capture is from tests on the disc reporting the lowest Peak Shift values in the table(Disc 12).
E. The 3rd CDS screen capture is from tests on the disc reporting the highest Peak Shift values in the table (Disc 3).__________________
Useful software with my burner:
DVDFab+++++DVDShrink+++++ImgBurn+++++CD-DVD Speed - Posted:
19-03-2007
- # 135
Quote scoobiedoobie:
"P.S. - Regarding TA Jitter testing, IMO its weakness is only testing at one location, though certainly more thorough at that one location. So I have a request already - particularly when CD-DVD Speed is maximized, it would be great if we could have multiple TA Jitter tests done and display all of them at once, similar to the look of the advanced quality scan. They could be at preset spots such as 0mb, 500mb, 1000mb, 1500mb, 2000mb, 2500mb, 3000mb, 3500mb, 4000mb, and at the edge (or something to that effect). Having the actual stats listed for each of the tests might be the biggest hurdle there. While having a fairly small number of samples still could miss a potentially poor jitter area on the disc, it would decrease the chances and give users performing the test somewhat of a 'standard' for doing multiple scans and posting them on CDFreaks, for example. While testing at one spot is obviously informative at that location, it may be a different story at another area of the disc."
After using the TA jitter feature a few times, I thought scoobiedoobie's suggestion made a lot of sense.
So I created a composite as a follow-up suggestion, included below. The composite consists of sections of TA jitter test displays run on a disc at regular intervals. Although the small individual histograms are not as easy to view as the current one spot enlarged display, the composite gives a broader view of the TA jitter qualities of the disc as a whole--so I think would be useful as I believe scoobiedoobie intended in the above statement.
1. Another possible addition is the ability for the user to select a testing interval for the compound TA jitter test. I used 400 MB in my example, but maybe 200 MB and 100 MB intervals could also be selections. I think with even 48 test points, the time for the complete test would be dominated by the spin-up time(hopefully only once.)
2. I included a CD-DVD Speed screen capture of the Disc Quality Test on the same disc. I have a suggestion here as well:
Since the jitter test is now performed after the PIE/PIF test, is it possible to add an additional Speed Selection Box under settings in this tab? One speed selection box for the PIE/PIF portion of the test and a separate speed selection box for the jitter test. So a user could set the PIE/PIF test to 8X or whatever and the jitter test to 1X or 2X or whatever? Each user could set the testing speeds to suit their own requirements or the particular characteristics of their drive model.
3. Last suggestion for this post, in the Statistics screen capture is it possible to add the scanning interval for the Jitter portion of the test (and the speed if #2 is added)?
I am sure other members can make additional suggestions for Erik Deppe to consider.
Thanks again, Erik for the many improvements in this version.
P.S. - I think I have read most of the posts about this version of CD-DVD Speed and have noticed no one has asked this question. So I guess I will be the dummy who asks. The TA jitter displays shows two parallel histograms, one above the other--I assume one is for lands, the other for pits: Which is which?
__________________
Useful software with my burner:
DVDFab+++++DVDShrink+++++ImgBurn+++++CD-DVD Speed - Posted:
16-04-2007
- # 136
hi,
My english isnt that good to go all that technicall like u all. But in my experience the best reader is a picky stand alone dvd player wich playes any originals just fine if not perfect in the first place ofcourse. Seeing is believing to me. Ive made many burns and scans whit my latest burner,it occurs more and more to me when i compare the results whit eachoter i noticed great similaritys between the scans made throug cd/speed and dvdinfo pro aswell.
urs kabster, - Posted:
16-04-2007
- # 137
Hi, kabukijo
I think I understand what you are saying.Quote:But in my experience the best reader is a picky stand alone dvd player wich playes any originals just fine if not perfect in the first place ofcourse.
I assume you mean the "best reader" for testing the quality of burned discs, not the best reader for actually watching the image on the disc.
That would make sense if your purchase of a standalone player is to test disc quality as a primary priority.
I looked for a standalone player to watch the movies on my disc, so I tried to choose one which was not picky.
I guess decisions in buying "things" depends on what you want to use them for primarily.
__________________
Useful software with my burner:
DVDFab+++++DVDShrink+++++ImgBurn+++++CD-DVD Speed - Posted:
16-04-2007
- # 138
Well yeah, infact an picky reader if it comes down to playing backups to determine the quality burn of it u see. An stand alone wich only accepts close to original quality burned discs u see. And ive tested an few stand alones trust me. Like, one time i had an Cyberhome 402 model. And u coud threw just about anything in that machine from cheap media to very cheap media regardless on wot speed u wrote them. Realy, and ive got by now a couple of hundreds dvd movie backups by now, written on 10 - 15 different brands and speed certified disks. So, wot i mean is. Those are the type of stand alones i rather avoid because they read toOO good, and that way u never know exactly how well the quality is of fresh burned disc in my opinion u see. U are probebly wondering, why keeping an picky reader. Well, ive have had a few movies wich played just fine throughout the whole movie. After a few years i still coud play them flawless on that same outstanding stand alone player BUT.... As u may expect, once i tried to re-copy that same disc through my computer whit the same burner whit wich i wrote that movie disc back then. I get ONe unrecoverable Read error after another. And there is very litle u can do to backup such an problematic disk. Eventualy, i managed to backup that very same disk just to see if it woudnt affect the content. And, ofcourse through out the movie i missed some data resulting in big gaps wich resulted in chapter skipping , pauses etc during playback u see.Quote:Originally Posted by uSerKeyHi, kabukijo
I assume you mean the "best reader" for testing the quality of burned discs, not the best reader for actually watching the image on the disc.
Urs kabster, - Posted:
16-04-2007
- # 139
Also, nowdays once u found the write combo u better stick to it. by that i mean, if it works using the write media , burn speed, etc. Neva, step down from it just stick to it. My moto is , if it aint broke dont fix it Dont try out to many different brands of discs. Or u can start all over in ur quest to the holy grail. That wos my problem at first. Because u get easily paranoid over this Its been an half year a year perhaps since i use verbatims because of the many mixed result whit other media. And poeple trust me when i say its ONE of the best if not the best media one can buy for any burner. And i use the brand verbatims for CD aswell now. Only, now iam trying to figure out wot brand i shoud use for dual layer as the quality of many different brands even verbatims doenst seem to be much different quality based from those Maxell ive tried so far!!! But, its to early to say. I just began to use Dual Layers 3 months ago.
Urs kabster, - Posted:
11-01-2009
- # 140
Wizzu asked me to comment on my own methodology here. I'm afraid I don't have much to add to the fine comments here already, but here goes.
I view my test setup like this:
1) Drive needs to report (i.e. measure) validly what is sees. If it reports really bad data, the disc should perform badly, and the opposite for good reported discs. This is the basic measure, but hard to test for.
2) Drive needs to report consistently the valid measures it gives. This of course means statistical reliability, so scan-to-scan measure on the exact same disc is very low (even if the drive heats up a bit). This is much easier to measure for, but it doesn't guarantee validity, of course.
3) Drive needs to report meaningful data OR alternatively, it's reported data output must be able to be converted to some meaningful range. Example: if drive measures PIE or PIF, those must remain within the DVD standards (for 1ECC or consecutive 8ECC blocks). If it doesn't report like this, then it's measures must be able to be converted to such a range. Again, it must pass requirements 1 and 2 first, or it's useless.
4) Drive will hopefully be consistent over long periods of time. This should really be requirement 2.1, as it's just a special case of reliability over a longer period of time. If a drive starts to drift, it becomes very hard to start compensating for the drift statistically. This usually means one just has to discard the drive. Of course, testing for drifting means testing with a high quality pressed test discs that do not degrade so easily themselves. That is, I don't believe in measurement 'calibration' with DVD-R discs.
5) I prefer CDR drives for CD-R testing and DVD-R drives for DVD-testing. In my opinion DVD-R drives have a tendency to under-report CD-R error rates (either in absolute or in relative CD-drive performance terms), which makes their use somewhat problematic. I don't use ROM drives due to many problems with requirements 1-4. Haven't tried BD-drives yet.
6) I try to use several drives using different manufacturer chipsets, pick-up heads and read strategies so that I get more real world correlation with the kind of variance that may exist amongst reading drives. Currently my semi-retired setup consists of BenQ, Optiarc, LiteOn, Plextor and Pioneer drives.
7) Software: use the one you know and have found reliable, suits your plotting needs. I'm a great believer in plotting ALL results to the same scale (manual maximum PIE/PIF values in the scale), so that side-to-side comparisons are easier. I don't believe in automatically adjusting scales, it takes too long to deduce them and they expose too much micro-detail at the expense of macro-scale, for me.
8) Statistics of measurement. A book could be written about this. Several hundred have been. I'm not an expert, but I can recommend some. My current ref is: Measurement Errors and Uncertainties: Theory and Practice, 3rd ed by Semyon Rabinovich and I'm still learning a lot. Far from being an expert. Perhaps a neophyte. The point is that to really understand measurements and what you can and cannot deduce from them, especially once you start calculating figures from raw data, you need to understand statistical reasoning. A good ref is a must. I can also recommend Dealing with Uncertainties: A Guide to Error Analysis by Manfred Drosg, although not as complete.
From my POV, there's not much else to add. I've had to retire lot of my scanning due to other commitments. Still, I keep checking back every now and then and learning from you guys and gals. So thank you for keeping up the community!
PS I once thought about TA jitter related measurements as well, but it turned out to be too difficult/time consuming for me. As such, I only use it as a 2ndary confirmatory measure - if at all. There are too many issues for me: sampling representativeness (thus reliability and validity, both), jitter measure specification used, reading strategy optimization and jitter correlation (i.e. what does jitter actually imply, aka validity of the measure) and many other smaller issues. I just gave up. Maybe you guys can crack it.__________________
Scan with a minimum of 3 diff. drives (reason: PUH/chipset/strategy variance). Post only 1 result if all scans agree - Posted:
14-01-2009
- # 142
Thanks a lot, Halcyon! __________________
François "Wizzu" Marchand, the artist formerly known as Francksoy
I won't reply to technical help requests by PM, but ask you question in the forum and notify me, I'll be glad to help if I can!
People who found this also searched for
- accuracy and reliability definitions in microbiology
- accuracy of disc test
- accuracy precision and reliability
- accuracy vs jitter
- defination of disc realiabilty
- disc quality test inaccurate
- highest peak cd software
- imgburn unrecoverable read error
- interpreting jitter
- juistheid vs betrouwbaarheid
- manfred drosg -unsicherheiten -uncertainties -neutron -nuklearelektronik -winterthur
- measurement errors and uncertainties theory and practice
- pie accuracy
- ta jitter
- verschil tussen juistheid en betrouwbaarheid
- a person who can record data with great precision also ensures great accuracy
- a person who can record data with great precision also ensures great accuracy.
- accuarcy vs precisison vs reliability
- accuracy and precision in software testing
- accuracy and precision of pi
- accuracy and precision when using a word
- accuracy and reliability
- accuracy and reliability definition
- accuracy and reliability definitions
- accuracy and reliability homework help
- accuracy and reliabilty
- accuracy and reliabilty of call bells
- accuracy discs
- accuracy of benq 1640 as scanner