Does anyone know the answer to my question? If so, I would really appreciate if somone answered it
I'm not sure if this is the
reason for the change, but I looked into this a while ago. I didn't get around to posting it till now. The main point seems to be
2.Changed the algorithm to increase the sampling rate
To get an idea of what the improved scanning strategy is up to, I checked the sample lengths (difference in LBA values at the PI Sum 8 samples from the RAW or CSV files) for several versions of KProbe with a "good" disc (FUJI branded ProdiscS03 DVD-R), all scans done with my SOHW 1213S at 4x. After plotting this information with Excel, it's very apparent that the samples fall into 2 bands with Version 2.2.3: ca. 128 and 144 sectors.
128 sectors correspond to 8 ECC blocks (@ 16 sectors).
My guess is that the 144 sector samples result either from having to poll every ECC sector (16 sectors) to acquire the data for PIF (Sum 1) or from delays caused by averaging and/or displaying the results.
The distribution is very similar if PI and PIF are both summed over 8 ECC blocks in this version, so I didn't plot it.
The new version 2.2.5 updates the counts on screen less often than before and shows only one
band- so the sampling is more uniform. However, most samples are actually shorter
than 128 sectors (the most common length is 126
), so that the values may not accurately represent the "sum over 8 ECC blocks".
Maybe Karr can comment on that.
KProbe 1.1.29 also had a common summing region for PI and PIF, so I checked that as well- it also shows 2 bands (like 2.2.3) so 2.2.5 is definitely more uniform.
The version 2.3.2 with the reinstated 8/1 ECC sum shows the same sampling as 2.2.5, so my guess may not be accurate or Karr has managed to reduce the overhead. The following diagram summarizes the sample sizes for these 4 versions:
During this analysis, I ran across a further question regarding the "Sampling count" shown in the saved diagram. This seems to be 7x the number of samples in the CSV or RAW file and I don't understand this factor. I'd expect either the same number or a factor of 8 (or whatever is used for the ECC count).
Hope this helps.