Ultimately, it was seen that scanning in the iHBS312 generally produced lower errors than scanning in the iHBS212, and that scanning at 8x shows similar trends to 4x however exaggerate errors somewhat.
But this wasn’t a very complete examination. Nor was it that scientific. I spent the time to burn some fresh discs – both LTH (TYG-BD-Y05) and HTL (PRODIS-CR0) burnt on the LG GGW-H20L and verified. These will be immediately scanned at all supported rates to do a comparison between the iHBS212 and iHBS312. It was a marathon effort that took a whole weekend to achieve. This allows us to compare scan speeds for both drives.
Then we look at scanning results across drives at a fixed speed of 4x, but instead using 4 LTH (of which 2 are part of the above set), and 4 HTL (of which 2 are part of the above set).
Finally we look at scan repeatability where one LTH and one HTL disc is subjected to 5 repeat scans at a fixed speed of 4x and the numbers are presented.
Unfortunately, the sample number is still a bit small – but I think it is still better than most other examinations.
Drive to Drive
Speed: LTH Disc 1 on iHBS212
Note that selecting an 8x scan results only in a 6x scan on both drives for LTH media. Thus there are only 1x, 2x, 4x and 6x speeds.
Surprisingly, it would seem that scanning speeds make little difference at least on quality media. The drive seems to handle the discs fairly well, although errors do increase slightly at higher rates and more spikes appear. This may suggest that these spikes are not high error rates present on the disc – but maybe a tracking error of the drive at higher speeds.
Speed: LTH Disc 2 on iHBS212
Speed: HTL Disc 1 on iHBS212
This particular disc seemed to have a pretty whacky jitter pattern – maybe the LG isn’t that good with these Prodisc discs.
Speed: HTL Disc 2 on iHBS212
Speed: LTH Disc 1 on iHBS312
The same issue is occured with the iHBS312 with LTH scanning only possible at rates of 1x, 2x, 4x, 6x only.
Speed: LTH Disc 2 on iHBS312
Scanning Speeds: HTL Disc 1 on iHBS312
Scanning Speeds: HTL Disc 2 on iHBS312
Drive: Scans at 4x – LTH Media
Here are the results from the four LTH discs scanned across both drives at 4x. The differences between the drives are not too significant. Left is iHBS212, Right is iHBS312. Looks like Disc 1 may have gotten some dust on it – hence the slight anomaly. The drives are pretty much in agreement trend-wise.
Drive: Scans at 4x – HTL Media
Here are the scans from the four HTL discs scanned across both drives at 4x. Again, the differences are not too significant.
Repeatability: iHBS212 LTH Disc 3
Repeatability: iHBS212 HTL Disc 3
Repeatability: iHBS312 LTH Disc 4
Repeatability: iHBS312 HTL Disc 4
From these results, I feel that the 2x scanning speed may have merits in the fact it is at CLV. The number of spikes is less, and back in the DVD days, we did see that the 4x standard scanning speed is CLV as well, thus meaning that the timing of the pits is consistent but the disc itself gets slower. In general, I prefer CLV for burning too as it seems to be more consistent in quality in my experience.
The 4x scanning speed does result in occasional spikes, with a CAV strategy, but the results are very close to the 2x result but requiring less time. In some cases, especially with HTL, it seems the 4x speed results in lower errors. The 6x and 8x speed results in even more spikes and a higher error rate, but on quality discs it seems that the difference in scan speeds can be quite small and the time savings significant from scanning at higher speeds. Unfortunately 8x scanning is not possible with LTH discs. Speeds above 4x seem to cause a reduction in sample numbers which will probably affect accuracy.
Interestingly, scans at 1x seem to produce higher readings than at 2x or 4x which is rather surprising. It is anticipated that at slower speeds, the requirements of disc rigidity and mechanism tracking will be lowest and should result in the lowest error rate, but alas, this is not the case. 6x scanning increases the error rate quite markedly – and is likely to underestimate the quality of a disc by over-reporting the error.
Results from both drives follow similar trends, thus legitimizing the scanning of discs with these models, however, the absolute numbers reported are sometimes significantly different. It would appear that the iHBS212 reports higher levels of error than the iHBS312 (noting that LTH disc 1 seems to have picked up some dust skewing the results on a summary figure basis), so while results are not comparable on the “micro” level, the trend and any serious issues will still be represented fairly well by both drives and at virtually all speeds. It also seems that the difference is higher for the HTL media tested, and less for the LTH media tested.
Therefore, both drives are good scanning candidates, although I do need to note the bug with scanning on iHBS312 needing for a speed selection after the spin-up phase to prevent the scan from hanging. The other thing is that when scans are started, sometimes there is an erroneous spike on the first few samples which can skew the result – these scans were discarded.
In some cases, dust may have settled on the disc despite best efforts to ensure a clean environment. This may lead to isolated thick-spikes in the results which aren’t adequately represented in the summary tables.
In general, similar trends were seen in repeat scanning, although the figures did differ somewhat, making the difference between scan speeds seem less significant. This doesn’t tell us anything about the accuracy of the figures however.
From the experiments, it seems that scanning LTH discs works just as well as scanning HTL discs, and the speed trend is similar, although it is not possible to perform a scan at 8x on LTH media.
I would say it is clear that 4x scanning has its place, and 2x does as well. Scans at higher than 4x exaggerate errors by varying amounts and reduces sample numbers and are probably less comparable. 8x scanning should be avoided (my mistake for thinking this was not the case). 6x scanning can be acceptable on quality media. 1x scanning results in unexpected increased errors over 2x which may be drive and disc related, and so proves to be inconclusive, but I don’t think anyone wants to spend one and a half hours scanning one disc.
The method was employed with care to improve confidence in the results. Make what you will of these results.