New BenQ 1ECC error scanning thread
| BenQ / Philips Writer Discuss, New BenQ 1ECC error scanning thread at CD, DVD and Blu-ray Writers forum; Off topic Quote: |
-
Posted:
22-03-2007
- # 1
Off topic My bad, only tested BSLB scan at 4x CLV, while comparing with Liteys.Quote:Originally Posted by scoobiedoobieBSLB and 4x 1 ECC works fine in my drive.
Heck, even with CD/DVD Speed v.4.7.0.2 we can scan at 4x (CAV) 1ECC.
Nice comparision to Lite-On drives, because sampling rate is bigger (~+18-20k) and all my QS96-97 discs now show up with QS99, hehe.
[Edit. Just found out DW1620 fw B7W9 also is able scanning at 4x (CAV) 1ECC.
]
Someone has to start a new thread about these findings, you'r welcome guys.
.
- Today (MyCE Staff)
- Posts: 15,596
- Posted:
22-03-2007
- # 2
The discussion of 4x CAV/ 1 ECC testing in the Benqs already exists in one of the CD-DVD Speed threads if you didn't know. - Posted:
22-03-2007
- # 4
Well like Erik said:
Quote:Originally Posted by ErikDeppeThe nice thing about 1 ECC scanning with BenQ drives is that you will get about 20000 samples more than with LiteOn drives
- Posted:
26-03-2007
- # 5
Good suggestion.Quote:Originally Posted by pinto2Someone has to start a new thread about these findings, you'r welcome guys.
[thread splitted]
Everyone is welcome to post findings here, including comparing with other 1ECC drives ie Liteon and Plextor.__________________
Click here to join CDFreaks.com
Join CDFreaks Distributed Computing Team, have fun while helping to save some lives!
My review PC & Specs & Media
Blu-ray to DVD guide
Forum Rules | Blu-ray UDF Driver
BenQ/Philips FAQ and tools: BQFlasher, MCSE, WOPC Tool, WinDWFlash
LG BE06LU11 Blu-ray | LG GGW-H10N Blu-ray/HD-DVD | LG GBW-H10N Blu-ray | Panasonic SW-5582 Blu-ray - Posted:
26-03-2007
- # 6
Thank you guys! I was just about to start a BenQ 1ECC scanning thread myself! 
I wanted to join the discussion over at the CD/DVD Speed thread, but IMO page 67 is no place to start such an interesting topic!
Now this is just a suggestion, but I think it'd be mighty cool if all the findings were located in this thread from now on.. as it will be much easier to locate for others!
Here's MY thoughts on the subject now..
While comparing two scans of the same disc, one at 1ECC and one at 8ECC, I find it quite relieving that I can finally look back at all my 8ECC scans and know how freaking good they were!
Before this little quirk existed, I always had trouble interpreting what an 8ECC scan meant, when compared to a higher-sampled 1ECC scan. I believe I finally have my answer!
To the best of my initial interpretation:
2-4 PIFs @ 8ECC = 1 PIF @ 1ECC
4-8 PIFs @ 8ECC = 2 PIFs @ 1ECC
8-16 PIFs @ 8ECC = 4 PIFs @ 1ECC
Now this would mean that a max of 32 PIFs @ 8ECC is equal to 8 PIFs @ 1ECC, which is off compared to the generally accepted (although worst-case scenario) rule of 32~4.
I have yet to scan some discs with higher PIF counts to confirm this information.. so be patient as I have more work to do!
Hopefully after a few other people shed some light on the subject, we can all [finally] know what to really expect with our 8ECC scans, and not even have to spend the extra time running the 1ECC scans!
Note: Jitter and PIEs are almost identical between the two scanning intervals. This means that both scans are pretty accurate when compared to each other, right? - Posted:
26-03-2007
- # 7
Please keep in mind that regardless of which interpretation you come up with, it will only be an approximation and in some cases it will be completely wrong.Quote:Originally Posted by zoinbergsBefore this little quirk existed, I always had trouble interpreting what an 8ECC scan meant, when compared to a higher-sampled 1ECC scan. I believe I finally have my answer!
To the best of my initial interpretation:
2-4 PIFs @ 8ECC = 1 PIF @ 1ECC
4-8 PIFs @ 8ECC = 2 PIFs @ 1ECC
8-16 PIFs @ 8ECC = 4 PIFs @ 1ECC
Now this would mean that a max of 32 PIFs @ 8ECC is equal to 8 PIFs @ 1ECC, which is off compared to the generally accepted (although worst-case scenario) rule of 32~4.
Have a look in this post for an in-depth explanation.
My rule of thumb for max PIF in 8 ECC scans is that 12 PIF is acceptable and corresponds to 4 PIF in a 1 ECC scan. This is also just an empirical value and will in some cases be completely wrong (it could also be 12 PIF at 1 ECC). - Posted:
26-03-2007
- # 8
Ooh! Thanks for that post you referred me to DM! That was EXTREMELY helpful! I can't believe I never found it until now.. as I've been over that thread before.
I agree with you now.. it is a total approximation when going from 8ECC to 1ECC.
I also like your rule of thumb. - Posted:
26-03-2007
- # 10
Oh! And before I forget.. just so all the noobies out there know (and so I can remind myself as well!) -- disc quality scans are a SMALL piece of the pie when researching the overall quality of a written DVD.
Transfer Rate Tests (TRTs) as well as ScanDiscs tell just as much, if not more about how well data is written.
And don't forget the all powerful standalone player test either, to see if your disc will work outside of a high-speed DVD writer setup!
Who cares if a disc scans well, but doesn't play! Right? - Posted:
27-03-2007
- # 11
A higher scanning speed is a very likely cause for the rise in PIE towards the edge.Quote:Originally Posted by zoinbergs[I wonder why the PIEs rose up toward the end of my 8ECC scan] - Posted:
29-03-2007
- # 12
First I would like to thank zevia for starting a new thread (for us) and for his understanding about the importance and inpact that 1ECC BenQ drives scanning capability has for BenQ users.
For years we have had discussions about Lite-On's being much better scanning drives then BenQ's because they (Liteys) were said to be more reliable as scanners, simly because they scan at 1ECC.
And I can see even today that some of our senior members are still quite pessimistic about this new BenQ 1ECC ability...
I never doubted BenQ 8ECC scans, not even when some said they were unreliable or useless simply because they had problem to read (badly burned) Lite-On discs.
In this and next post I'll show you some scans and then maybe you'll have posibility to make up your own mind (about BenQ's 8ECC unreability).
Keep in mind though; there is no perfect drive, no absolute truth and no perfect world.
Test disc; CMC MAG AM3 (DVD-R) burned at 18x speed with BenQ DW1800 aka Lite-On LH-18A1P (MediaTek chipset). Disc burned as ISO image with CD DVD Speed containing a movie.
- pic1, scan BenQ DW1655 at 4x 1ECC
- pic2, statistics of DW1655 scan
- pic3, scan Litey LH-18A1P at 4x 1ECC
- pic4, statistics of LH-18 scan
. - Posted:
29-03-2007
- # 13
All that Liteys have over BenQ now is TA Jitter test for testing. Maybe Erik can fix that also.Quote:Originally Posted by pinto2For years we have had discussions about Lite-On's being much better scanning drives then BenQ's because they (Liteys) were said to be more reliable as scanners, simly because they scan at 1ECC.
Now i'm sure comes the doubt with posts saying it to new to tell bla bla bla.
I don't think Erik would release this testing if he doubted it.
I'm going to buy 1 more BenQ 1650 series unless a great deal comes along.
*ducks head* - Posted:
29-03-2007
- # 14
Looking at above scans we can see they are very similar in both PIE/PIF and jitter graphs. Lite-On drive shoving a little less of everything but that might be the "better reader" impact.
Note, BenQ drive sampled almost 20k more test samples. Can the difference in total values (Litey less errors) have something to do with it. I don't know, and I think nobody can say this for sure.
Edit. I see rolling56 managed to sqeeze in while I was posting, but I hope you know what I mean by "above scans".
Now to the more interesting part of the story; how will a BenQ scan at 8x speed and 8ECC look like. And how will Litey react to a rised 8x speed.
Check it out by yourself and remember my statement above; are BenQ's really that bad 8x speed (8ECC) scanners...?
- pic1, scan BenQ DW1655 at 8x 8ECC
- pic2, statistics of DW1655 scan
- pic3, scan Litey LH-18A1P at 8x 1ECC
- pic4, statistics of LH-18 scan
. - Posted:
29-03-2007
- # 15
IMO the higher sample rate is probably related to the slower scan speed of the Benq, CAV maxing at 4x while the Liteon is 4x CLV. Technically this makes for an imperfect comparison between the two drives, although I think that the speed difference in the early part of a 4x CAV scan is not a signficant variable considering the relatively slow speed of even 4x (though I prefer CLV scan speeds). I tend not to get into the 'are Benq drives reliable/accurate/etc' conversations much anymore. No consumer drive can be directly compared with a professional testing device, however I believe that Benq drives are just as useful at testing media as any other drive, whether at 1 ECC or 8 ECC. Getting familiar with the behaviour of your drive(s) and how they test can make useful testing devices of most scanning drives. 1 ECC is a useful addition when scanning with Benq drives for things such as unusually problematic discs that don't necessarily show problems in other tests, but I'll continue to be comfortable using 8 ECC scanning in my Benqs for the majority of my scanning, as well as scanning with my Liteon often. - Posted:
29-03-2007
- # 16
Congrats to your 3k posts scoobiedoobie your posts are (as always) a great contribution to this forum.
Yes, I agree with you, our drives are far too primitive to be considered as anything near the perfect scanners. This goes for all drive makers (saying this although I don't own or ever owned a Plextor).
I have no intention to tell everybody that this drive is better scanner then that one. My idea was only to show how unreliable all this scanning is and how little it can differ amongs these drives.
Repeating my line from above; "Keep in mind though; there is no perfect drive, no absolute truth and no perfect world."
Ofcourse 1ECC scanning capable drive will always be more "accurate" then the one scanning at 8ECC. But with all these more or less primitive drives, the difference is not that big as some members here want to teach us (and has been teaching us) with in deep theory posts that are just theories, although they don't understand that as it seems.
Just adding a scan of disc posted above, this time side by side, the right scan taken two hours laters. All settings the same.
. - Posted:
29-03-2007
- # 17
Yeah i agree it's only scanning. My real test is watching or using the data i burned. I scan a few after i open up a new spindle and when buying new drives but it's nice to have BenQ 1ECC. - Posted:
29-03-2007
- # 18
Thanks Pinto2, I'm not sure how I've managed to accumulate this many posts.
3000 Posts: A CDFreaks spamming odyssey
- Posted:
29-03-2007
- # 19
Congrats scoobiedoobie.Quote:Originally Posted by scoobiedoobieThanks Pinto2, I'm not sure how I've managed to accumulate this many posts.
3000 Posts: A CDFreaks spamming odyssey
I don't think i'll ever catch pinto2 to get that title he has. - Posted:
29-03-2007
- # 20
I've a feeling there are many members out there having difficulties to understand all this talk about 8ECC, 1ECC, PIE, PIF, POS, samples, aso.
[Note aimed at all "expertise" on this subject; I'll try as long as possible to simplify ECMA terms, wording, aso in my expressions and explanations to keep this post handable and maybe also better understandable.
If and when you find simplier explanation/s, please post them.]
Cont. There isn't any easy answer to all that I posted at start (and I have no intention to do this for you here), most of the time senior members just refer to ECMA without any further (simple) explanation. When visiting that site, I and I suppose also many other members just drawn in information, an information very complex, highly techical and hard to understand.
To somehow ease this feeling and make this more understandable I can recommend you this document. Not that easy either but much more simple/understandable then the very same information at ECMA.
Samples.
We have learned that at 1ECC drive collects much more samples then at 8ECC. You may ask; why is it like this, what are samples, how is disc structured to provide samples, how many samples can be taken on disc?
For in-depth studies, definitions like (DVD) bytes, sectors, pockets aso can be found with google and at ECMA, so I'm not going into details here.
I just like to explain and show how samples on a DVD are calculated. You can open the document I was linking to right above now.
We know that a fully burned DVD disc contains approx. 4.38GB (gigabytes) of data. These bytes are aranged in sectors (or pockets). Every data sector contains 2048 bytes of (user) data and is arranged in 16 rows making in total 32768 bytes of raw data for every ECC block (see example in document).
When we say 1ECC this means drive takes a sample of every ECC sector on drive. 8ECC means equally only one sample is taken in/of any of these 8 ECC sectors in a row, thus also less samples taken by drive (compared to 1ECC).
How many samples can there then be sampled on a DVD disc? To get this value we have to go back and check discs total amount of data and devide this value with the value of a ECC block.
Example with 1ECC on a DVD-R, total data devided with ECC block makes; ~4707100000/32768 = 143649 samples. A drive only able to sample at 8ECC can only collect or sample 1/8 of this value, or 17956 (randomly) samples.
Using calculator you can also calculate the values for DVD+R media.
I know all this might be difficult and hard to understand but I hope some of you out there know a little more now what we are talking about here. Thanks.
Off topic
@rolling56, wanna trade? - Posted:
30-03-2007
- # 21
Naaa I'll just battle it out with TaxQuote:Originally Posted by pinto2Off topic
@rolling56, wanna trade? - Posted:
30-03-2007
- # 23
First a little correction to my post above. Bolded "on drive" should ofcourse be read as "on disc". (Overlooked this when posting, while busy at my work.) Sorry.Quote:..."When we say 1ECC this means drive takes a sample of every ECC sector on drive. 8ECC means equally only one sample is taken in/of any of these 8 ECC sectors in a row, thus also less samples taken by drive (compared to 1ECC)"...
@seeker010, can't see I'm discussing PIF's in my above post*.
What I said is that scanning samples in a row of 8 ECC blocks are taken randomly amongs those 8 blocks.
Example, lining up 8 ECC blocks from a to h. Scanning samples are taken on bolded blue block.
a + b + c + d + e + f + g + h
a2 + b2 + c2 + d2 + e2 + f2 + g2 + h2
a3 + b3 + c3 + d3 + e3 + f3 + g3 + h3
a4 + b4 + c4 + d4 + e4 + f4 + g4 + h4
...
....
.....
*BTW, DrageMester posted here how trivial a 8ECC /1ECC sum PIF calculation can be. - Posted:
30-03-2007
- # 24
all 8ecc scans are taken over 8 blocks, not 1 block in 8 blocks. this is true for pisum8 (pie) or pif over 8eccQuote:Originally Posted by pinto2First a little correction to my post above. Bolded "on drive" should ofcourse be read as "on disc". (Overlooked this when posting, while busy at my work.) Sorry.
@seeker010, can't see I'm discussing PIF's in my above post*.
What I said is that scanning samples in a row of 8 ECC blocks are taken randomly amongs those 8 blocks.
Example, lining up 8 ECC blocks from a to h. Scanning samples are taken on bolded blue block.
a + b + c + d + e + f + g + h
a2 + b2 + c2 + d2 + e2 + f2 + g2 + h2
a3 + b3 + c3 + d3 + e3 + f3 + g3 + h3
a4 + b4 + c4 + d4 + e4 + f4 + g4 + h4
...
....
.....
*BTW, DrageMester posted here how trivial a 8ECC /1ECC sum PIF calculation can be.
Quote:Originally Posted by dragemasterCalculating the Sum8 value from 8 separate Sum1 values is trivial, the reverse however is impossible, because there are many ways the reported number of PIF could have been distributed among the 8 ECC blocks.
Examples:
7 = 1 + 0 + 1 + 2 + 0 + 1 + 0 + 2 (Maximum is 2)
7 = 0 + 0 + 1 + 3 + 0 + 1 + 1 + 1 (Maximum is 3)
7 = 0 + 0 + 1 + 6 + 0 + 0 + 0 + 0 (Maximum is 6)
So how does one go about coming up with a reasonable calculated value when it is in fact impossible to calculate it correctly? - Posted:
30-03-2007
- # 25
Ofcourse they are, I never said anything else.Quote:Originally Posted by seeker010all 8ecc scans are taken over 8 blocks,
..."Example with 1ECC on a DVD-R, total data devided with ECC block makes; ~4707100000/32768 = 143649 samples. A drive only able to sample at 8ECC can only collect or sample 1/8 of this value, or 17956 (randomly) samples."Edit. I see my simplified example can be confusing indeed. Not my intention though.Quote:Originally Posted by seeker010...not 1 block in 8 blocks...
BTW, we have had this discussion (what is and what is not sampled over those 8 ECC blocks) years back so I hope we don't need to make that all over again.
