Measuring the quality of a recorded disc is a delicate and complicated procedure.
Physical characteristics of a disc and the HF signals which are the output of the photo diode sensor in the play-back head, are specified by the industry in the form of standards. These standards among other things, also set down maximum and minimum levels to ensure correct operation, and it is from these limits that we can then ascertain the condition of a disc.
In order to evaluate the quality of a recorded disc, you need to examine all
these parameters. Any respectable measuring procedure should be consistent, and the
measuring equipment should be accurate enough to maintain measuring consistency as well as reliability.
Furthermore, a measurement
be repeatable over any number of samples and should be confirmed over an adequate amount of samples.
CDRInfo presents the results of a test that started a couple of months ago.
We chose some of the most popular PC burners on the market, and used them
to record a number of blank discs with different Media Identification (MID)
The discs were examined for recording quality with DVD Analyzers provided
by DaTARIUS. The DaTARIUS DaTABANK analyzer, an industry standard, assures
consistent and reliable results.
In the following pages, you will find details regarding the testing procedure,
equipment, and of course, the results themselves. The results have been presented
in two formats. First, we show the behavior of each disc (MID) for each of
the five burners. Then, the same results are presented, but this time the behavior
of each recorder for each of the discs. Although
the results are essentially the same, this dual presentation is provided so
as to easily gauge the performance of a particular disc with all drives, or
a particular drive with all discs.
2. Test Procedure - Definitions
In an effort to ensure the integrity of the results, the disc burning setup and
configuration should be the same for all discs and drives. Our PC testbed consists of a Pentium-4
workstation with the following specifications:
Intel Pentium-4 Processor 3.6 GHz/ 800Mhz FSB,
1GB DDR2-533 MHz dual channel RAM,
Seagate 120 GB 7200 rpm SATA hard drive with 8 MB cache,
Microsoft Windows XP Pro, SP2
The recording drives (burners) used in this test include:
- BenQ DW-1640A vBSHB
- Plextor PX-716A v1.08
- Pioneer DVR-110 v1.22
- LiteOn SHW-1635S vY.S0N
The PC is equipped with Nero Burning Rom 6, version v188.8.131.52 . The same
image was burned onto DVD media, at the maximum allowed speed. In most cases,
the speed was 16x. In cases where the burning speed was lower,
this is noted in the corresponding tables.
Philips 16x DVD+R
Prodisc unbranded 16x
Prodisc unbranded 16x
Prodisc unbranded 16x
BenQ DVD-R 16x
BenQ DVD+R 16x
CMC MAG AM3
Philips DVD-R 16x
CMC MAG M01
Ridisc DVD+R 16x
Traxdata DVD+R 16x
Traxdata DVD-R 16x
Optodisc DVD-R 16x
Optodisc DVD+R 16x
RICOH JP NR03
Ricoh DVD+R 16x
Verbatim DVD+R 16x
Verbatim DVD-R 16x
PI and PO:
Stands for Parity Inner and Parity Outer. Reed-Solomon Product Code
(RS-PC) is a method of
error correction employing a number of groups of Reed-Solomon
computed from rows, columns, or diagonals of a rectangular data array.
DVD uses two-group
method of RS-PC for error correction. The redundant codes of rows and
columns are called
Parity of the Inner code (PI) and of the Outer code (PO) respectively.
DVD discs correct small
read errors using PI, that operates on rows, and corrects large read
errors using PO, that operates
on columns of the data array.
Stands for Parity Inner Error. The number of error corrections made on
incoming rows of data in
the first pass of the decoder using the inner parity correction code.
A row of an ECC Block that
has at least 1 byte in error constitutes a PI error. PIE is measured
over 8 ECC blocks. In any 8
consecutive ECC Blocks the total number of PI errors before correction
shall not exceed 280, as defined by the DVD specification,
16448:2002 or ecma-267, third edition, April 2001. PIE are also called as PI
Stands for Parity Outer Errors. Parity outer fails. The decoder was unable
to correct the data
using the outer parity codes. It is measured over 1 ECC block.
This test uses the measured maximum inner parity (PI Sum8) errors for recordable
DVD as a high-level estimate of the quality of the media.
However, the correlation between actual loss of information and the measured
maximum PIE actually depends on several factors including the playback system
as well as the media
quality. The PI errors of 280 are conservative levels chosen
by the recordable DVD specifications as a predictor of the onset of uncorrectable
JC: Data to Clock Jitter
Jitter measurements are used to indicate the occurrence
of deviations from an ideal duration. The data on an optical disc is carried
by the length of
the pits and the distance between them. This data is extracted with a clock
signal that is synchronized with the pit edges. Jitter measurements are done
by monitoring the length of the pits and the distance between them (Data
to Data Jitter), and by monitoring the length between the edges of the pits
the edge of the clock (Data to Clock Jitter).
Its value is given by the standard deviation of the length of the recorded
marks. It mustn't exceed 8% for DVD-R and 9% for DVD+R of the 1T reference
clock signal (38.2 ns).
3. DVD Analyzer - The DaTARIUS DaTABANK
The DaTABANK quality assurance tool from DaTARIUS was used
to analyze the DVD media in this test. These systems measure recorded
and unrecorded DVD discs and include a Sony Reference Drive, which correlates
with the Pulstec SDP-1000 optical drive, capable of measuring at speeds of up
error (POF) and jitter, are measured using this system.
The DaTABANK was developed with the highest demands on measurement
flexibility of hardware and software. The system’s highly modular design,
allows easy upgrades of both software and hardware.
The DaTABANK has been an industry standard media testing system, installed
in the majority of optical media production plants. The system offers full
control of the production process for all DVD, BD and HD DVD formats, including
dual-layer recordable discs.
One of the main advantages offered by DaTARIUS for the DaTABANK system, is
its easy to use user interface software, called DaTAVIEW. The software is
used to operate the
system, user administration, hardware monitoring, remote control and maintenance,
as well as and view and
results (result display & export & statistical analysis. trend charts).
DaTARIUS has developed the DaTAVIEW software, having in mind that a system
should always present information in a user-friendly way.
The DaTAVIEW software is responsible for the generation of accurate
charts illustrating the parameters specified for each quality measurement,
filters, organizes and illustrates the data imported by the measuring system,
according to the operators needs. Critical quality parameters can be isolated
from the full list of measurements and be separately presented, offering easy
yet stringent control
over the production process. All the measurement/graph
parameters, can be easily set with a few mouse clicks, in a full Windows-based
environment menu interface. The operator is able to configure signals, their
Below is a FE (Focus Error) sample graph for a DVD disc, as produced
by the DaTAVIEW software:
A very interesting and useful tool of the DaTAVIEW software is QualityWeb.
It is a dynamic viewing feature, that offers a clear and complete visual presentation
of disc quality. The operator is able to configure the multiple parameters included
in the display, and thus have an idea of the quality of a disc at a glance. In our example,
both JC (jitter) and POF values are not graphically included within the inner
area of the the red polygon and so show up as exceeding the quality limits
set by the operator:
In addition, DaTAVIEW presents the disc surface diagram, which displays
from one signal (quality parameter), in a unique visual representation:
Below you can see a 3D View of an HD DVD-ROM as illustrated through the DaTABANK
software (Click for larger image).
The results can be printed, saved as an archive for future reference, or exported
to one of several popular formats including .csv and .xls.
DaTARIUS offers remote desktop support through the web for all the DaTABANK
systems. For more information, refer to the following:
DaTARIUS Technologies GmbH
Anton Maria Schyrle-Str. 7
Service Hotline: +43 5672 206 206
DaTARIUS Technologies Inc.
15-A Marconi Avenue
Irvine, CA 92618
Service Hotline: +1 866 383 8378
DaTARIUS Technologies GmbH
Unit 801-5, 8th Floor, Delta House, No. 3 On Yiu Street, Shatin,
New Territories, Hong Kong
Service Hotline: +852 2561 8078
In our tests, we used the DaTAVIEW software to monitor JC, PI Sum8 and
POF measurements. Measurements were taken at nine specific points on the disc,
at a reference speed of 1x, starting from the inner tracks and progressing towards the outer, in an effort to cover
the whole disc's area.
The results were exported
to Microsoft Excel where
then generated. Keeping in mind that the data presented in the graphs is taken at
discrete points, the graphs most likely indicate the trend of how quality changes
from the inner to the outer area of a recorded disc.
The test results are presented in the ensuing pages.
4. Disc Quality Per Drive - BenQ DW-1640A
BenQ DW-1640A vBSHB
The BenQ DW-1640A performed well with the first batch of discs, as we can
see in the first chart above. Note that with a DVD disc, the number of PI errors
occurring in eight consecutive ECC blocks is defined as "PI-Sum8," and
the maximum value for PI-Sum8 is specified as 280 or less. The average PI
Sum 8 values are lower than the 280 limit and
so the disc is in
a good overall condition.
Things aren't the same with the next set of discs, as we can see in the following
chart. More than the half of the discs showed a high PI sum 8 count,
peaking as the measurements move towards the outer areas of the disc's surface.
This was expected, since the DW-1640A reaches its maximum recording speed
(16x using CAV) there, so the possibility of data being written close to the limits is more likely. In the following graph, the red dotted line indicates
the PI Sum8 upper limit (280).
The MID codes that produced high error rates are Ritek R04, Ritek F1,
MCC 03RG20, Optodisc OR16 and CMC MAG M01, as shown on the graph
above. Note here, that the PI Sum8 errors are handled by the first level of
error correction employed on DVDs. Their count normally represents
their severity, which means that high PIE counts normally trigger Parity Output Errors which can in turn become Parity
Outer Fail (POF) errors. The occurrence of POF is not acceptable for
DVDs, since these are uncorrectable errors. So measuring POF for the "problem" MIDs in the previous
scans is mandatory.
The results above confirm the correlation between high PI Sum 8 and the likelihood of POF. A
pleasant exception is the Daxon AZ3 DVD+R disc, which returned zero POF,
despite the high PI Sum 8 counts.
5. Disc Quality Per Drive - Plextor PX-716A
The Plextor PX-716A with firmware revision v1.08, showed decent performance, as we can see in the chart
above. Something to watch out for later, is whether the high PI Sum8 counts, for
the third, fourth and perhaps the seventh measurements, will lead to POF errors, although
the peak value is less than the acceptable 280.
There is an issue here with the Ricoh JP NR03 DVD+R disc, with a PI Sum8 count
that reaches 880. The Ritek R04 disc was also difficult to read/measure.
We did not manage to get measurements for the first three points,
as well as the last. The Daxon 016s DVD-R as well as
the MCC 03RG20 DVD-R discs, also reported high PI Sum8 counts at different areas.
In case you are wondering,
the CMC MAG M01 DVD+R and
CMC MAG AM3 DVD-R discs are missing from these graphs, but with a promise to be
included in the near future.
It is surprising to see POF in the case of the Ritek F1 DVD-R disc, since
while the POF is located in the same area where we had previously measured
high PI Sum 8, the PI Sum 8 count was well within limits (see first graph on this page).
The Ritek R04 DVD+R disc's
condition could be considered as bad, since we were not able to
take measurements at all points on the disc.
The Daxon 016s and the MCC 03RG20 discs also reported POFs, with the errors occurring at the same point where we had maximum PI sum 8 counts in the corresponding graph previously.
6. Disc Quality Per Drive - Pioneer DVR-110
Pioneer DVR-110 v1.22
Very low PI Sum 8 counts for the Pioneer DVR-110 v1.22 drive, shown in the graph
above. However, the Prodisc S05 DVD-R as well as the R05 DVD+R discs, were burned at 4x, being limited by the drive's firmware.
The good results continue for Ritek R04 and the Optodisc R016 MIDs. The
CMC MAG M01 DVD+R disc was burned at 8x speed (maximum allowed by the drive),
but we see a PI sum 8 error count, close
to the 280 limit, for the first measurement point:
In the following graph, we see that the MCC 03RG20 DVD-R disc produced the
highest PI Sum 8 error count, far exceeding the 280 limit, while the Ritek F1 DVD-R disc hovered near the limit for the duration of the scan.
Let's take a look to the POF error counts for the last three MIDs,
shown in the previous chart. MCC 03RG20 DVD-R media produced POF errors, and it was to be expected considering the high PI Sum8 levels. Interestingly enough, the
Ritek F1 DVD-R disc did not generate any POF errors, despite the fact that its
PISum 8 was continually near the limit.
7. Disc Quality Per Drive - LiteOn SHW-1635S
LiteOn SHW-1635S vY.S0N
On this page, we look at the LiteOn SHW-1635S vYS0N burner. The first batch of discs
proved to be a good combination with the specific burner, at least judging from the
PISum8 measurements shown in the graph above.
Half of the media in the graph above reported high PISum8 error rates, climbing well above the 280 limit.
This increases the possibility of producing POF errors for the same media. The
following graph confirms this:
Surprisingly, Ritek F1 media also produced POF errors despite having PI Sum8 values within limits. What is becoming evident however, is that there is a direct relationship between PI Sum8 and POF, and where there are peaks in the PI Sum8 graphs (especially those exceeding the 280 limit), there is a likelihood of POF errors occurring at the same point or points. Note also, that the specific graphs are essentially trends, due to the limited amount of samples (measuring positions)
taken per disc.
8. Disc Quality Per Drive - NEC ND-4551A
NEC ND-4551A v1-8T
The next burner in line is the NEC ND-4551A. Good results with the first batch of discs,
as we can see above.
PiSum8 errors above the limit with the Prodisc F02 DVD-R so this is one disc that could potentially produce problems (i.e. POF).
In the graph below, all media exceeded the limit. What will be interesting is to see if they all produce POF errors.
Fortunately, not all discs that exceeded the 280 limit, produced POF errors. In the chart above, we can now see the "problem" MIDs for the NEC-ND4551A, with the most prominent being Prodisc S05 DVD-R media.
9. Disc Quality Per Disc - Page 1
In the following pages we check the behavior of each disc when burned with the various burners used in this test. On top of the PI Sum8 and
POF error rates, we also present the jitter charts for each disc, measured at
the same area on each disc.
In general, jitter is a good indication of the condition of a recordable
disc, and high values could lead to PI or even more severe errors. A jitter
signal can be introduced when the spots on the recording
layer of the disc (data), are not perfectly shaped. For example, they may
be longer or not steep enough. High jitter compromises the integrity
of the recorded data and thus,
may influence the playability or lifetime, over long term. Jitter
is a low-level quality measurement for optical storage media.
Note that no disc is jitter-free. The maximum allowed jitter value for the
DVD+R discs is 9%, and slightly lower for DVD-R media (8%). Exceeding
these values, however, does not always mean that a disc is problematic. It
means that it tends to be more vulnerable to possible data corruption.
Optodisc R016 DVD-R
The first measurements with Optodisc R016 DVD-R media, show jitter above the limits,
no matter the burner. The dotted red line on the graph above, indicates the 8% jitter
limit for DVD-R media.
In the following graph, we can see the PI Sum8 error rates for the same
MID disc, after being burned with our test drives. Highest error were rates produced by
the LiteOn and the Plextor drives. The LiteOn drive also produced the highest Jitter in the graph above.
Despite having high PI values and Jitter, the discs recorded with the
Plextor and LiteOn burners did not produce POF errors. POF errors were reported with the NEC ND-4551A drive which did not have a high PI Sum8 count, in fact it has very good levels there, did produce high Jitter. Perhaps what is most surprising is that the LiteOn drive, which produced high Jitter and PI Sum8, did not generate POF errors.
MCC 03RG20 DVD-R
Again, high jitter was reported for MCC 03RG20 DVD-R media burned
with all the DVD burners.
The specific disc reported high PI Sum8 error counts with the Pioneer
DVR-110 and the BenQ DW-1640A drives. What we then see below, is that the same drives also produced POF errors. It is also becoming evident that as the PI Sum8 count exceeds 700, it is very likely that POF errors will be produced. If the level is above 1000, then it is almost certain to do so.
10. Disc Quality Per Disc - Page 2
CMC MAG AM3 DVD-R
With CMC MAG AM3 media, we observe very good and the lowest Jitter measurements so far, with the Pioneer DVR-110, while the NEC ND-4551A produced the highest Jitter.
In the following graph, we can see the PIsum8 error rates for the same drives.
The NEC ND-4551A - CMC MAG AM3 DVD-R combination produced unexpected POF errors,
since we didn't see high PI errors. The behavior cannot be easily explained
and the specific disc will be measured again, very soon.
Ritek F1 DVD-R
The Plextor PX-716A with Ritek F1 media created problems and it took 3 attempts before we could take a reading. Unfortunately, we could not measure the Jitter for this combination and hence the absence in the above graph.
It is difficult to find a direct connection between the jitter values measured
and the PI and PO error rates. The reason for this is that the existence of both PI and
PO errors is a result of many parameters. In our case, Ritek F1
DVD-R media produced POF errors when burned with the Plextor, LiteOn and the BenQ drives.
11. Disc Quality Per Disc - Page 3
Daxon 016s DVD-R
Looking at both the above graphs, we note that the Plextor PX-716A is the only drive that exceeds the limit for both Jitter and PI Sum8.
As it turns out, with Daxon 016s DVD-R media, the Plextor PX-716A
and Pioneer DVR-110 drives produced POF errors.
Prodisc F02 DVD-R
Good results with the Prodisc F02 DVD-R disc. There were no POF errors, even with the NEC ND-4551A drive, which had high Jitter and PI Sum8 errors.
Prodisc S05 DVD-R
As it happens, the high jitter reported with the LiteOn and the NEC drives, corresponds to
the PI errors reported with Prodisc S05 DVD-R media.
In the case of the LiteOn SHW-1635S, the
POF is indicative of the high PI Sum8 reported in the graph above.
12. Disc Quality Per Disc - Page 4
Daxon AZ3 DVD+R
The Daxon AZ3 DVD+R disc produced quite good results with the drives in this test. Despite
some PI Sum8 errors above the limit with the LiteOn and the BenQ drives,
none of the discs produced POF errors.
Prodisc R05 DVD+R
Prodisc R05 DVD+R media also produced good results, with all the burners.
Although jitter was higher than 9%, especially the Pioneer and NEC drives, the
PI sum 8 error rate was very respectable, below the 280 limit, and there were no POF errors.
Ricoh JP NR03 DVD+R
High jitter values in most cases with Ricoh JP NR03 DVD+R media, with the BenQ
drive being a pleasant exception.
POF errors were produced with the LiteOn drive in the outer tracks,
while the Plextor and the Pioneer drives also produced some
13. Disc Quality Per Disc - Page 5
Infomer 30 DVD+R
Strangely enough, the Informer 30 DVD+R disc did not produce any POF errors with the Pioneer drive, despite the high Jitter. The PI Sum8 values are well within specs for all drives, and as we are starting to see, Jitter probably isn't as much of a factor in generating severe errors as PI. Only the LiteOn drive produced POF errors although it was well within specs for both Jitter and PI Sum8.
Note that the discs recorded by the Plextor PX-716A could not be measured.
Prodisc R04 DVD+R
Another MID from Prodisc in this test. The Prodisc R04 DVD+R is a good choice with all the drives in our test, giving a low PI error rate and
no POF errors.
Ritek R04 DVD+R
Ritek R04 DVD+R media does not seem to be the best MID for the Plextor and BenQ drives which produced POF errors. The only anomaly here is the NEC ND-4551A which had both high Jitter and PI Sum8 count above the limit, which however did not result in POF errors.
14. Disc Quality Per Disc - Page 6
CMC MAG M01 DVD+R
High jitter in all cases as illustrated above for CMC MAG M01 DVD+R media.
Possible problems with the discs burned with the LiteOn and NEC drives, with PI Sum8 error rates well above the limit.
The LiteOn SHW-1635S drive produced a very high POF error rate. The disc burned with the BenQ drive, also
produced POF errors.
MCC 004 DVD+R
Again we see that any combination of drive and media which produces high PI Sum8, especially above 700, is very likely to result in POF errors. The LiteOn SHW-1635S and MCC 004 DVD+R media is just such a case. The rest of the drive/media combinations seem to be OK.
Optodisc OR16 DVD+R
The last disc in this test is Optodisc's OR16 DVD+R. The disc produced out of specs
PI Sum8 with the NEC, BenQ and the LiteOn drives, while all except for the Pioneer drive, resulted in greater than 9% Jitter:
The BenQ and the
LiteOn drives produced POF errors with this media. A little surprising since the PI Sum8 is not all that much above the limit although the LiteOn drive also produced a high percentage of Jitter.
In the matrix below you can see a summary of the results. The red color indicates
that at least one measuring point showed PI Sum8 higher than 280,
or POFs have occurred for each disc. The green color means that the
drive/disc combination is good or adequate.
16. Final Thoughts
In this article, we tried to address the critical and highly anticipated in the
optical industry issue of quality of a recordable DVD disc. The test procedure
was simple enough and with selected measurements for each disc,
chosen so as to indicate the quality of a recorded disc, hopefully
in a way understandable and familiar to our readers.
The jitter to clock measurement, was chosen in order to provide a low-level
examination of a recorded disc. The PI Sum8 error measurements, indicate the
the recorded data, and reveal potential problems that can manifest themselves which can make a disc harder
to read or even unreadable over the course of time. POF errors, are undoubtedly the most serious errors, since they
represent uncorrectable errors on the recorded media.
However, have in mind that the correlation between actual loss of information
and the measured maximum PIE actually depends on several factors including
as well as the media quality.
A future update of this test would require additional measurements for each
disc. Measuring HF Signals such as I14/ I14H, resolution (I3/I14), I14 Modulation
or Asymmetry. Servo and tracking errors such as the DPD Amplitude, DPD Asymmetry
would also help determine the condition of a recorded disc.
a last note, we should say that is is not wise to directly blame either a disc
or a burner when a measurement is out of specs. Along with the various
parts, the hardware makers develop the software that controls the recording
and the media makers develop recording materials (organic/inorganic dyes, phase
change, etc) and manufacturing technologies that conform to the specifications
and the requirements of hardware. Of course, hardware is usually built and tweaked
by using the disc as a reference, or vice versa. As a result, the combination/interaction
of burners and media is the factor that decides the final quality.
In this test, we tried to give an idea of which is the most suitable burner/disc
combination. Of course, not all MIDs
have the same characteristics, which are dependant on the manufacturing
process and the materials used. Let's skip this factor for now, and assume that
the MIDs show consistent results.
In the end,, all we
really want from a disc is that it should be readable, not only now but also in the long term.
CDRInfo will continue to provide detailed measurements of optical media
in the future.