Thorough review with exhaustive media tests, as usual, D_R.
Unfortunately, you used DVDInfoPro for the BenQ and LiteOn scans. I can understand that the dual plots save you a lot of fiddling with the IDs of the diagrams when you set up your review. However, the results that DIP displays differ drastically from what other diagnostic tools show (at least for LiteOn drives- I can't speak for BenQ).
I don't normally use DIP, but installed the newest version to check into this after the numbers shown seemed "funny" to me.
I'm attaching 4 scans of the same DVD I did today with my LiteOn SOHW-1213S@1653S (CS0P). The first 2 are done with CD-Speed (v4.10) and KProbe 2.4.3. Here the sampling is 8 ECC for PI, giving PI Sum 8, and 1 ECC for PIF.
As has been shown many times before, the information given is essentially the same, except for the usual scan-to-scan variations and known "idiosyncrasies" (CD-Speed calculates averages based on the displayed values, not on all samples and the KProbe PIF average is ca. 8x too high, which is unnoticeable here).
The next scan is from DVDInfoPro 4.5.0.3.
The first thing noticeable is that the PI values are ca. 1/2 what the other tools show and that the sums are
drastically smaller! According to DIP, the scans are done for 8 ECC blocks for both PI and PIF.
This is not true. 8 ECC for PI should look like the CD-Speed scan. The PIF scan looks like what KProbe and CD-Speed show and is definitely for 1 ECC.
The last scan was done with KProbe using 1 ECC for PI and 8 ECC for PIF (this is
not the standard setting). The PI plot looks very much like what DIP shows, PIF values are higher.
So now it's clear that DIP uses
1 ECC not 8 ECC (as indicated) for both PI and PIF.
But the sums and averages still differ drastically from what the other programs report. I tried out a few ideas on RAW data from KProbe before it dawned on me what DIP does: it calculates sums and averages of the
values displayed in the graphs (like CD Speed for averages) and not of all the measured values! You can check this for the PIF values: 26 1-values and 11 2-values on the graph give 48. That this can falsify information is evident when comparing averages in CD-Speed and KProbe.
I suppose the the shape of the graphs still give some information, but I prefer having proper numbers as well and using 1 ECC for PI is definitely a no-no.
One of the strong points of CDR-Labs reviews is the constant format which allows comparisons of various tests done at different times, and it's unfotunate that this isn't fully possible with this latest test. (I realize that the current format has definite limitations, particularly with the performance section!).
I sure hope you take this into account in your upcoming tests!
G