Which test is more accurate? CD-DVD Speed or PlexTools Professional?

I’ve ran some quality scan with both CD-DVD Speed and PlexTools Professional 2.20.
The result doesn’t match so well.
Especially the average PI/PO is with great differences.
So which software is more accurate?
Or which result I should follow?

I would think Plextools is the way to go. Plextor,Plextools same company. Yheir product. :wink:

Their product.

I would say it’s your choice, both don’t have to agree with each other. But both will tell you the quality of your burn

I assume you burn it with Plextor drive. Then Plextools scan on the SAME drive will show a more accurate figures. If you scan with CDSpeed with BenQ drive for example, it is just showing how the BenQ drive reads the burnt disk.

However, if your Plextor and BenQ drive are both good, then they shouldn’t be much different figures. Little difference is normal but not much, ie an excellent scan VS a coaster.

It is known that the BenQ 1620A is a little too optimistic. As said before in this forum, the error measurement tests are to be taken with a grain of salt since it will only show you how that particular drive can read that particular disc. It’s a good indication though as PI/PO scanning can reveal problem areas on the disc but comparing two different drives (using two different scanning software tools) can show a very different result. Neither will be more accurate, it all depends on the drive you’re using to read the disc with.

I think the issue here is a fundamental difference in the way the two programs reports average PIE.

CDSpeed always reports PIE average for an excellent scan in the 5-15 range, while PTP reports values under 1.0.

I always wondered, for example, how PTP could report an average PIE under 1.0 when the graph showed a solid blue with no values under 5. The only answer I could come up with was that the graph simply didn’t show results for every ECC block (only PIEs per second) but that the average PIE it gives is for each ECC block.

CDSpeed, on the other hand, shows averages for what is graphed (per-second PIE counts).

I may have the exact specifics wrong, but I think the difference is to do with different interpretations of “average”.

only PIEs per second
what is a PIE per second?

IMHO it shows the maximum value of all values represented by one vertical line

Maximum values of what … 8 ECC (Sum-8) blocks? Or maximum values of all ECC blocks read in 1.0 seconds?

BTW, I’m just guessing here … it makes no sense that one program reports 20,000 total PIE with an average of 0.13, while another program also reports 20,000 PIE and an average of 2.9.

PS. Alex, does your PXSCAN calculate the average or does the drive supply it? I assume the former.

This is not true for my 1620. Really excellent scans give me averages of 0.6 or below, just like Plextools (based on firmware B7T9 for Benq).

I think the main reason for a discrepancy between CDSpeed and Plextools has more to do with the speed of the scan.

Scanned at a different rotational speed, the same disc will produce different results. Plextools scans at 2X CLV to achieve a truer result (and speed is closer to the speed on a set-top dvd player, for example). On the Benq 1620 with CD/DVD speed, scans are set by default at 8X. If you reduce the scanning speed to 4X (the minimum for the Benq), your results should be closer, but not exactly identical. PIFs, for example, under CD/DVD speed use a larger block size, so sample-by-sample, the peaks are allowed to be four times higher–which is why the PIF graphs may look different. Total PIF count, however, should be roughly similar.

My homepage describes what you see in PlexTools:

Lets say the picture is 500 pixels width, and your disc is a fully written dvd-r with 2,298,496 sectors. Then it has 143656 ECC sectors, i.e. 17957 PIsum8 values.

One line then represents 36 values. PlexTools will show the highest of those 36 values.

PS. Alex, does your PXSCAN calculate the average or does the drive supply it? I assume the former.
The drive supplies those 17957 values. Making some stats from it is up to the software

If I had to vote for ‘scanning accuracy’ comparing BenQ 1620 & CD/DVD Speed and 716A & Plextools Pro, I’d vote for Plextools (or Alex’ implementation, of course) …

PTP does count the PIF over 1 ECC block (if Sum-1 is selected) as stated in ECMA specs whereas CD/DVD Speed does scan 8ECC for both PIE and PIF (but faster and both at a time which is far more convenient) … therefore, PIF exceeding the allowed value of 4 per 1ECC cannot be determined with 1620 & CD/DVD Speed.

As for accuracy … there is no real accuracy. Even a CATS analyzer’s PI/PO scan will tell you only one thing: how good a Pulsetec drive can read your DVD at a given speed (apart from the option of measuring the more interesting geometrical disc specs).

OK, I’m probably not making myself clear…

Attached below are two scans taken from the Quality Scans thread. Each shows a roughly similar burn - about 20,000 PIE across the entire disc.

One program (Plextools) reports a PIE average of 0.15, while the other program (CDSpeed) reports an average of 2.5, which is over 16x higher.

And neither bears much resemblance to the graph, although the CDSpeed one seems closer.

The original question in this thread asked about the discrepancy between averages, which is something I’d also wondered about. Clearly, one of these programs is calculating averages incorrectly (or based on a totally different criteria) and I’d appreciate any input on which one.

By PlexTools: Avg PIE = Total PIE/quantity of ECC blocks. But not of SUM8 intervals (why???).
Full DVD-R - 143656 ECC blocks.
Full DVD+R - 143444 ECC blocks.

Again Avg PIF = Total PIF/quantity of ECC blocks, SUM1 - that’s right.

I can’t to understand what is divider which CD-DVD Speed uses for average PI/PIF calculation.

PlexTools always scans at 2x, your CDSpeed scan was at 8x? I would guess this would account for at least some of the difference.

Nope … ~20,000 total PIE across the entire disc for both scans … speed doesn’t come into it at all.

The calculation should be: Total PIE ÷ Interval. One of the programs is using a different interval or …?

a little tip plextool can scan at 2x 5x 8x 12x once you start sum8/1 go read speed setting page on plextool and bump it up to the one that ends in 5x or 8x or 12x I found that 5x has bit higher errors but still gives good scan. and can do scan in about 15 minutes for each part ie pi/pif takes two passes and about 30 minutes much better then over an 1hr at 2x. these times are for scans in high acc. setting

One program (Plextools) reports a PIE average of 0.15, while the other program (CDSpeed) reports an average of 2.5, which is over 16x higher.
Sounds like one is doing avg per ECC block, and the other is doing avg per 16 ECC blocks