DW1620 OEM firmware discrepancies?

vbimport

#1

Newly purchased OEM benq DW1620 attached to the slave channel on my secondary IDE, master drive on secondary is a Pioneer 106s DVD-ROM. According to Nero Info tool DMA mode is on for both drives, device manager shows Ultra DMA mode 5 transfer mode, Windows XP SP2, Intel P4 2.8, using Intel 82801EB Ultra ATA controller embedded into DFI Lanparty motherboard.

The drive came with a few blank media from various manufactures, I tested the drive with a couple of Ridata 8x DVD+R discs using the G7M9 firmware, factory default for this particular drive. One was a DVD movie and the other a CD-Speed generated data disc, both burns turned out just fine with both my Pioneer 106s being able to read both discs and my DVD set-top player being able to play the DVD movie.

Satisfied that the drive worked okay I cross-flashed to the retail B7P9 firmware, I read a post on here that the benq qscan program only worked with the retail firmware so I was anxious to try that out (I’ve sense figured out that it works with OEM firmware too)

I purchased some “Great Quality” GQ DVD+R 8x media from Fry’s Electronics, they come up as Prodisc R03’s as you can see from the scans below. I’m still trying to research the best DVD media for the drive. I know there a lot of people on here that are “cross flashing” from the OEM firmware to the retail firmware, I haven’t read any posts that say whether or not this is a good thing or something that shouldn’t be done. My guess is that it may depend on the media with which works the best with the drive/firmware, newest version not always being the best.

Using Nero CD-DVD Speed 3.55, I decided to test out the various firmwares with these apparent cheapo discs. I didn’t see a way of controlling the burn speed of the “create data disc” routine, under qscan these “GQ” discs show 12x being the max. For each version of firmware I burned a seperate disc and ran the Quality Test on them at 4x scan speed:

First up, B7P9: Very ugly towards the end, using the ScanDisc function of CD-DVD Speed the end of the disc has nothing but red squares using the Pioneer 106s, surprisingly the 1620 reports no problems being able to read the end of the disc.

How about that B7S9: Quality score pretty much says it all :Z

Lets crossflash back to OEM firmware G7M9, maybe retail firmware isn’t so great of an idea on an OEM drive: Rather high PI errors, but a quality score of 96, DiscScan reports no unreadable errors with either drive.

I thought I might as well try out the unofficial OEM G7Z9 firmware: Most definitely a step backwards.

So it seems obvious, flash the G7M9 and problem solved right? I did that and used Nero 6.6.0.0 to burn six large .avi files to one of these “GQ” Prodisc R03’s at 4x speed (thinking lowering the speed might equal a better burn for these discs). The 1620 won’t copy the files back onto the hard drive, CRC error (I can play them from the 1620 just fine though) – yet the Pioneer can read the files just fine, both playing right off the disc and copying them to the HD, after copying them back to the HD I did a file compare with FC command in dos and they compare 100% okay. :confused:

I then burn several DVD .iso files again at 4x, 1620 can read them just fine, Pioneer has problems just loading the disc (takes a long time) and doing a DiscScan shows problems reading towards the end of the disc, 1620 doesn’t have a problem. :confused:

I’m not sure if my problem is from Nero, the burner or the chepo discs.

Cliffnotes: Just because you get a good looking scan from CD-DVD Speed’s “Disc Quality Test” doesn’t mean that it’s a “reliable and multidrive redable” disc, and using retail firmware on an OEM drive is not necessarily a good thing, I guess.

Thanks to ImageShack for Free Image Hosting


#2

Testing PRODISCR03 at 12x now. First with B7S9 firmware.

To make better scanning, I’ll use one SOHW-1613S@1653S CS02 drive.


#3

Inertia pointed out in this thread http://club.cdfreaks.com/showthread.php?t=118419 that the GZ79 is actually older than G7M9 by a few days, looking on the unofficial firmware page http://dvdpro.club.st/firmware/BENQ_FW.htm sort of shows it as newer than G7P9 (above it in the list anyways), if you look at the file dates inside the archives it is not.

I flashed my OEM drive to the G7P9 firmware and now the drive is working rather well with these “GQ” discs.

AVI file compilation Burned at 12x with Nero 6.6.0.0

iso burned at 8x with Nero 6.6.0.0, now we’re cooking. Disc reads just fine in the Pioneer 106s, DiscScan shows no unreadable blocks.

So, if you’re having problems with an OEM drive that you’ve crossflashed to retail you might want to go back to OEM firmware and re-test. Something must be different between Retail and OEM drives, atleast with this one it appears to be the case.


#4

In my opinion, most (or all) of the quality variation you have experienced is due to inconsistent media performance. You would have to test multiple media samples per firmware version to get a valid sampling for judging media. Most people don’t do this and take the result from one burn as the standard for a given firmware. If the tests happen to be performed using poor quality inconsistent media, false conclusions can be drawn regarding firmware.

Fry’s “Great Quality” media is one brand that I would personally never buy. I was warned by a Fry’s employee that returns for this house brand were greater than all others put together. It may have a PRODISC R03 media code, but this in itself does not mean that it is a top quality version. It could well be lower grade manufacturing seconds or rejects. It could also be some noname small media manufacturer that can’t get their discs tested and entered into firmware under their own code, so they use a competitors media code (from the stamper) whose writing strategy may work (hopefully) with their media.

There is no significant difference in the hardware of OEM and retail versions of the DW 1620 drive. Large numbers of people (including myself) have flashed from the OEM to the retail firmware versions with excellent results. There is no known functional difference between the binary programming content of OEM and retail firmware with the same code designation.

If you can show a reproducible difference in the results from firmware G7P9 vs. B7P9, it would be very interesting. I am inclined to believe that your latest success with G7P9 would have shown the same result with B7P9. Inconsisent quality media can have results that vary all over the place. This can lead to false conclusions if the unreliable media is considered to be of fixed quality and firmware or software the only variables to be considered. :slight_smile:


#5

B7S9 vs. B7M9

Test Disc #001 CD-Speed Create Data Disc

PRODISCR03 SmartBuy 8x DVD+R at 12x
DW1620 B7S9 firmware

6:35

Scanned in Lite-On SOHW-1613S (firmware CS02 for SOHW-1653S, OmniPatched) at 8x

B7M9 PRODISCR03 12x + SOHW-1613S CS02

General Information
Drive: LITE-ON DVDRW SOHW-1653S
Firmware: CS02
Disc: DVD+R (PRODISC R03)
Selected speed: 8 X
PI errors
Maximum: 416
Average: 2.37
Total: 21014
PI failures
Maximum: 2
Average: 0.04
Total: 679
PO failures: n/a
Jitter: n/a
Scanning statistics
Number of samples: 124349
Average scanning interval: 1.15 ECC
Glitches removed: 6

Scanned in BenQ DW1620 B7M9 firmware at 8x

B7M9 PRODISCR03 12x + DW1620 B7M9

General Information
Drive: BENQ DVD DD DW1620
Firmware: B7M9
Disc: DVD+R (PRODISC R03)
Selected speed: 8 X
PI errors
Maximum: 8
Average: 2.38
Total: 20953
PI failures
Maximum: 5
Average: 0.06
Total: 711
PO failures: 0
Jitter
Maximum: 10.5 %
Average: 8.14 %
Scanning statistics
Number of samples: 16571
Average scanning interval: 8.65 ECC
Glitches removed: 0

B7M9 PRODISCR03 12x + DW1620 B7M9

General Information
Drive: BENQ DVD DD DW1620
Firmware: B7M9
Disc: DVD+R (PRODISC R03)
Selected speed: 8 X
PI errors
Maximum: 13
Average: 2.74
Total: 26281
PI failures
Maximum: 5
Average: 0.06
Total: 747
PO failures: 0
Jitter
Maximum: 10.5 %
Average: 8.28 %
Scanning statistics
Number of samples: 17908
Average scanning interval: 8.01 ECC
Glitches removed: 0

Test Disc #002 CD-Speed Create Data Disc

PRODISCR03 SmartBuy 8x DVD+R at 12x
DW1620 B7S9 firmware

6:38

Scanned in Lite-On SOHW-1613S (firmware CS02 for SOHW-1653S, OmniPatched) at 8x

B7S9 PRODISCR03 12x + SOHW-1613S CS02

General Information
Drive: LITE-ON DVDRW SOHW-1653S
Firmware: CS02
Disc: DVD+R (PRODISC R03)
Selected speed: 8 X
PI errors
Maximum: 14
Average: 2.58
Total: 25317
PI failures
Maximum: 4
Average: 0.26
Total: 5275
PO failures: n/a
Jitter: n/a
Scanning statistics
Number of samples: 124360
Average scanning interval: 1.15 ECC
Glitches removed: 6

Scanned in BenQ DW1620 B7M9 firmware at 8x

B7S9 PRODISCR03 12x + DW1620 B7M9

General Information
Drive: BENQ DVD DD DW1620
Firmware: B7M9
Disc: DVD+R (PRODISC R03)
Selected speed: 8 X
PI errors
Maximum: 14
Average: 3.08
Total: 31171
PI failures
Maximum: 9
Average: 0.36
Total: 4831
PO failures: 0
Jitter
Maximum: 10.6 %
Average: 8.27 %
Scanning statistics
Number of samples: 17919
Average scanning interval: 8.00 ECC
Glitches removed: 0

Test Disc #003 Nero Burning Rom v6.6.0.3 Data DVD

PRODISCR03 SmartBuy 8x DVD+R at 12x
DW1620 B7M9 firmware

6:26

Scanned in Lite-On SOHW-1613S (firmware CS02 for SOHW-1653S, OmniPatched) at 4x

B7M9 PRODISCR03 12x + SOHW-1613S CS02

General Information
Drive: LITE-ON DVDRW SOHW-1653S
Firmware: CS02
Disc: DVD+R (PRODISC R03)
Selected speed: 4 X
PI errors
Maximum: 9
Average: 2.70
Total: 22778
PI failures
Maximum: 2
Average: 0.02
Total: 197
PO failures: n/a
Jitter: n/a
Scanning statistics
Number of samples: 117612
Average scanning interval: 1.09 ECC
Glitches removed: 0

Scanned in BenQ DW1620 B7M9 firmware at 8x

B7M9 PRODISCR03 12x + DW1620 B7M9

General Information
Drive: BENQ DVD DD DW1620
Firmware: B7M9
Disc: DVD+R (PRODISC R03)
Selected speed: 8 X
PI errors
Maximum: 14
Average: 3.03
Total: 26034
PI failures
Maximum: 6
Average: 0.03
Total: 268
PO failures: 0
Jitter
Maximum: 10.6 %
Average: 8.52 %
Scanning statistics
Number of samples: 15460
Average scanning interval: 8.31 ECC
Glitches removed: 0


Test Disc #004 Nero Burning Rom v6.6.0.3 Data DVD

YUDEN000T02 That’s 8x DVD+R at 12x
DW1620 B7M9 firmware

6:44

Scanned in Lite-On SOHW-1613S (firmware CS02 for SOHW-1653S, OmniPatched) at 4x

B7M9 YUDEN000T02 12x + SOHW-1613S CS02

General Information
Drive: LITE-ON DVDRW SOHW-1653S
Firmware: CS02
Disc: DVD+R (YUDEN000 T02)
Selected speed: 4 X
PI errors
Maximum: 9
Average: 1.23
Total: 8524
PI failures
Maximum: 3
Average: 0.06
Total: 934
PO failures: n/a
Jitter: n/a
Scanning statistics
Number of samples: 107179
Average scanning interval: 1.20 ECC
Glitches removed: 0


Scanned in BenQ DW1620 B7M9 firmware at 8x

B7M9 YUDEN000T02 12x + DW1620 B7M9

General Information
Drive: BENQ DVD DD DW1620
Firmware: B7M9
Disc: DVD+R (YUDEN000 T02)
Selected speed: 8 X
PI errors
Maximum: 7
Average: 1.34
Total: 8987
PI failures
Maximum: 6
Average: 0.10
Total: 939
PO failures: 0
Jitter
Maximum: 10.1 %
Average: 8.08 %
Scanning statistics
Number of samples: 15495
Average scanning interval: 8.29 ECC
Glitches removed: 0



#6

Thanks for the input Inertia, I can see why Fry’s sees a lot of returns for these “GQ” discs, these probably are some sort of factory seconds from Prodisc remarked for Fry’s, still it’s nice to have a source of cheapo media around when you’re not concerned about longevity. I did burn a few others of these GQ discs with B7P9 with odd results, and high PI Failures towards the end of the disc resulting in errors reading from the Pioneer. The two “GQ” discs I burned today with the G7P9 read just fine in my brothers external usb Benq DVD-ROM drive.

Thanks for the scans Kenshin, are you using a retail DW1620 or an OEM crossflashed to retail firmware?

At any rate, I’m going to leave the G7P9 on there for awhile and see how it goes.


#7

Here is the first scan again of the .iso I made above using G7P9

same .iso 8x burn using G7P9, even better.

I keep reading how DVD burners “learn” the media to which they are burning to but I have yet to see any explanation as to how this works, does the 1620 have this feature too?


#8

Ahhh, the old optical anthropomorphism. :stuck_out_tongue:

The worst part is that after they “learn” the media, they start having optical sex with other drives and create hybrid firmware in preparation of The Matrix. :eek:

This “feature” is limited to the minds of those who experience the mystical intentions of opto-mechanical devices, to which ordinary mortals are oblivious. :cool:

This “learning” is not to be confused with WOPC II (Walking Optimal Power Control), which is simply a real-time adjustment of the laser power to enable optimal writing quality throughout the disc. Unfortunately, the dumb WOPC “forgets” the power strategy every time it does it and has to do it all over again without “learning” anything. :sad:


#9

OEM DW1620, but it is DW1620. Unless BenQ has two different ways to make DW1620, one for DW1620 for BenQ and the other for DW1620 for everyone else.

It’s just like Sony U12, U14, U16, and so on. And Lite-On SOHW-1213S and other “retail” models have their OEM counterparts. Also, Pioneer with their DVR-A08 and DVR-108. Samsung has a different method to name their drives. The only difference in the Samsung case is about firmware which is written to the “produced” drives at the final stage in the assembly lines. It must be the same for BenQ.


#10

BenQ’s (or IO magic’s) “Z” firmware does not look like the best and the latest firmware development from BenQ. B7P9 and B7S9 are probably just “Pro” firmwares that were created by adding a few new features to some firmwares older than K and M.

Hm. I need to re-assemble my GSA-5163D enclosure. :sad: I’ll use the enclosure to test DW1620 as well (externally.)


#11

Are you limiting your comments to the 1620 (maybe you know something about it that we don’t) or perhaps to all DVD recorders? Certainly the WOPC II (Walking Optimal Power Control) is a wonderful technology, but do we know for sure that it’s not supplemented by other more traditional approaches? A lot of people in here are getting some strange results that tend to clear up over time, and that smells of some sort of a learning capability.

I remember when derisive comments similar to above were being made in the LiteOn forum, and then CodeKing actually tracked down the technique. He also went so far as to produce a tool to erase the learned data from some of the LiteOn drives, since it occasionally goes wrong and has caused problems for some users as per the following link:

http://club.cdfreaks.com/showthread.php?t=112103&highlight=eprom


#12

Dangerdog, Did you notice that the BenQ Booktype application does not work with this version of firmware?


#13

It doesn’t? I’m not sure why I would want to change it though since it defaults to DVD-ROM.


#14

Your right, not to worry (except for a few) because of default BenQ booktype DVD-ROM settings.

Whoops, Now not sure if my statement about the BenQ Booktype app not working, is absolutely true. I just flashed back to B7xx firmware (trying to find best for SONY08D1) and realized that I did not ask drive for info when I was in the G7P9 firmware…So apoligize if I mislead anyone.
Just for info, 47L9 seems to burn my Sony08D1 best.


#15

Thanks for the link, Spartane. :slight_smile: I hadn’t seen it and this certainly explains the talk about drives “learning”.

The dictionary defines learning as:

  1. The cognitive process of acquiring skill or knowledge.
  2. Profound knowledge.

Of course, definition no. 1 would apply here. Problematically, though, cognition is involved in learning. Cognition meaning: The psychological result of perception and learning and reasoning.

My point in this hair-splitting is that “learning” seems a misnomer for what is happening in the EEPROM. If the EEPROM was actually “learning”, even if the calibration was incorrect it should recognize this and correct it without having to be cleared. Recording (not learning) a small series of burns and recalibrating from the results is what it seems to be about.

I applaud the efforts of the guys experimenting with this theoretical tweak. On the other hand, I frankly don’t see from that thread at this time that the results are consistent or particularly useful. In fact, the outcomes go all over the map. It may be fun to play with, but if testing is done with a particularly good or bad media sample wrong conclusions will be derived. And in fact, most people seem to be doing their testing with the poorest quality media hoping that it may improve by “learning” (recalibration).

In any event, as C0deKing stated, this experiment should only be performed on a drive with major problems, not on one that is performing properly. In the former case, there is little to lose and a backup will recover the original data.

And as for drives “learning” with cognitive ability (implying consciousness), I stand by my comments poking fun at it. :bigsmile:


#16

Along the same lines, it would appear that Plextor has something new that they refer to as AutoStrategy. Although I personally have no experience with it, it would seem that the drive can “learn” how to properly burn some of the more marginal media types.

Here’s the link:
http://club.cdfreaks.com/showpost.php?p=786342&postcount=12

from this thread:

http://club.cdfreaks.com/showthread.php?t=109469

To see the list of AutoStrategy compatible media, go here:

http://www.plextor.be/technicalservices/technology/cdrmedia.asp?choice=Supported%20media

and click on the “DVD+R” link over to the right from the PX-716A label.


#17

Thanks again for the links, Spartane :slight_smile:

From Plextor U.S.A.

AUTOSTRATEGY - A technology that determines the standard deviation of any blank disc and automatically optimizes the write strategy for unknown media, enabling high-quality disc recording. AUTOSTRATEGY technology is the culmination of five years of research and development by Taiyo Yuden, a leading Japanese supplier of quality CD and DVD recording media.

From Taiyo Yuden (Japanese Translation)

…AUTOSTRATEGY is the technology which applies laser light to a disk, discriminates the characteristic of a disk, and can write in on the optimal conditions for every disk. Compared with the system which specifies a disk by the conventional media ID, the influence of the characteristic variation at the time of disk manufacture is reduced, and the optimal writing can be performed. Moreover, it writes in also about the disk with which Media ID are not registered, and it is said that conditions can be optimized.

When optical energy is specifically added to a disk by laser, the heat distribution of a record side is analyzed in an instant. Based on the result, a laser output and irradiation time are controlled the optimal and it is said by forming a precise pit that write-in grace is raised.

Since crude media are written in and it becomes discriminable in front by adopting AUTOSTRATEGY technology, it is said that the trouble “writing is not made” and “it is unreproducible although written in” is mitigable. In the company, it is supposed that activity which is considering as a chance that the drive which adopted this technology in a few days will be commercialized, will cooperate with a drive maker from now on, and makes AUTOSTRATEGY an industry standard is performed…

AutoStrategy seems very promising, and it seems to work as specified from reports in the threads. I notice that nowhere in the Plextor and Taiyo Yuden descriptions do they use the “learn” word. Apparently many forum members like “learn”, as they have fun anthropomorphizing their hardware and ascribing mysterious qualities to it. :wink:


#18

OK, how about this Plextor link:

http://www.plextor.be/products/px-716uf.asp?choice=PX-716UF

From that link, here’s one of the points listed under the “PX-716UF Benefits” section:

Intelligent Recording

‘AutoStrategy’ is a :bigsmile: self-learning :bigsmile: writing technology to optimise the writing quality on any type of media


#19

Your’re right, Plextor is not immune to this hype. :cool:

I think that this is an example of intelligent marketing to mesmerize unsophisticated viewers. "Self-learning, intelligent, and intelligence are all contained in this blurb. Plextor has no been shy in exaggerating earlier technologies, such as VariRec, and are continuing that tradition here with their terminology.

Personally, I prefer the traditional dictionary meaning that learning and intelligence are characteristic of higher life forms, not machines. Language does change and can lose important distinctions in the process. This is an example of how an original meaning might be adulterated. The exaggerated notions of “learning” or “intelligence” to describe the functions of a machine are anthropomorphisms that are gaining common currency in this age of burgeoning technology.

I still consider it to be hype. :bigsmile: