Page 1 of 1

Japanese site tests effects of direct sunlight exposure

PostPosted: Sat Apr 14, 2007 7:08 am
by MediumRare
The Japanese site dvd-r.jpn.org has a test of the effects of direct sunlight exposure on DVD media. They looked at 26 media types (hello dolphinius_rex :wink:) over several weeks, and it looks like it's still continuing.

Although this isn't under scientifically controlled conditions, it gives a good idea on how UV affects these discs. And it is/was obviously a lot of work (26 discs tested 8x each to date). I couldn't make out which drive/program they use for the tests though.

G

PostPosted: Sat Apr 14, 2007 2:00 pm
by dolphinius_rex
Nice!!

I'm jealous though... if I had had a roof to use for such a test, I would have done one also. I still have enough media to do one with if I can find an appropriate place to use :wink:

PostPosted: Sat Apr 14, 2007 4:57 pm
by RJW
Did these type of cd-r's more then once.
Only problem is that in Netherlands we don't see that much sun.
However these days it is sunny and 25 Degrees C.

Also I can say that there is a significant difference in putting disc's with the dye site or the printed site up in the sun. DUH !!

PostPosted: Sat Apr 14, 2007 6:33 pm
by Scour
Nice test with surprising results :)

But I miss Ritek DL-media

PostPosted: Sun Apr 15, 2007 12:09 pm
by MediumRare
RJW wrote:Did these type of cd-r's more then once.

Have you reported on the results? I think it was Halc and/or some guy in Iceland who were trying to organize a systematic test like this here a few years ago.
RJW wrote:Also I can say that there is a significant difference in putting disc's with the dye site or the printed site up in the sun. DUH !!

Those were my thoughts too when I read the report. I'll see if I can find an address for the author- maybe he'll be able to add some information.

G
(who first saw this here.)

PostPosted: Mon Apr 16, 2007 10:38 am
by Wesociety
I suppose my location (Phoenix, Arizona) would be a great place to perform sunlight tests as it gets up to 120 degrees Fahrenheit here in the summer.

PostPosted: Tue Apr 17, 2007 5:54 am
by Halc
I did my own test like that (yet to be published). Indirect + direct sunlight expose (window still, behind double glazed window, indoors).

My results also indicate that there is a BIG difference in UV tolerance among discs.

However, I don't think my tests (60 days) agrees with all the results the Japanese testers had.

What could explain the results? Exposure time? Batch variance? Initial burn quality?

One thing I know differs is that I did scans with 4 drives (all different maker/PUH/chipset) and calculated averages.

Some drives seemed to be much more sensitive to the changes on the disc by UV (e.g. Nec ND-4551).

Sorry to be a tease, I will publish my results here later when I'm free to do so (still a month or so I guess).

PostPosted: Tue Apr 17, 2007 11:00 am
by Scour
In c´t-magazine test they found out that media burned in a NEC was more reliable than media burned in Benq

PostPosted: Tue Apr 17, 2007 11:25 am
by MediumRare
Halc wrote:Sorry to be a tease, I will publish my results here later when I'm free to do so (still a month or so I guess).

I'm looking forward to that!

G

PostPosted: Wed Apr 18, 2007 4:49 am
by RJW
Halc wrote:I did my own test like that (yet to be published). Indirect + direct sunlight expose (window still, behind double glazed window, indoors).

My results also indicate that there is a BIG difference in UV tolerance among discs.

However, I don't think my tests (60 days) agrees with all the results the Japanese testers had.

What could explain the results? Exposure time? Batch variance? Initial burn quality?


Exposure time - Shouldn't have significant inpact except for failures over longtime. In other words. What can degrade in your test in 6 days can decay in the japanese test in 16 days or the other way arround. Because of different exposure time and intensity of the sun.

Batch variation -YES

Other test conditions. How are the disc's put in the sun. I wonder if there will be a difference between clear and non clear jewelcases. Maybe should do some tests now that we have sun in holland about that one.

Initial burn quality (Based on a Philips article in Japanese Journal of Applied physics. It shows that the burn power incombination with the dye can have significant influence on the burn quality.
Specially overheating has serious impact- ThermoBallanced writing was created to overcome these problems. Now some folks might call it off as pure propaganda for techniques as Power REC /WOPC. However the source of the article says enough I think.).

PostPosted: Wed Apr 18, 2007 12:09 pm
by dolphinius_rex
RJW wrote:Initial burn quality (Based on a Philips article in Japanese Journal of Applied physics. It shows that the burn power incombination with the dye can have significant influence on the burn quality.
Specially overheating has serious impact- ThermoBallanced writing was created to overcome these problems. Now some folks might call it off as pure propaganda for techniques as Power REC /WOPC. However the source of the article says enough I think.).


I don't mind giving a bit of a hint on my own longevity study results to say that my findings support the Philips article you mentioned :wink:

PostPosted: Sat Apr 21, 2007 8:04 pm
by evilboy
Scour wrote:In c´t-magazine test they found out that media burned in a NEC was more reliable than media burned in Benq
Which issue exactly?

PostPosted: Sat Apr 21, 2007 9:39 pm
by Scour

PostPosted: Mon May 14, 2007 4:51 pm
by MediumRare
I think mailed the author a while ago via his feedback option (I wish cfitz were still active- not only because of his knowledge of Japanese).

In any case, there's now a picture of the test setup present: the dye layer is exposed on top!

The test is still continuing with results up to 7 May now available.

I don't know if it's new or if I missed it the first time, but clicking on the brand entry in the summary table brings up a new page with the detailed scans for each media up to "15 days". I hope the author will add the remaining images.

He used 3 LiteOn's with CD-Speed for the scans: SHM-165P6S MS0R, LH-18A1P GL0F and LH-20A1S 9L02, always the same drive for each disc. The DL discs were tested with a Plextor PX-755A 1.05 and PxScan.

Very nicely done! =D>

G

PostPosted: Thu May 17, 2007 6:33 am
by RJW
Not so sure about c't fast conclusion that NEC burn quality is better as Benq.
Based on other test results that I've seen I think the conclusion can not be made and that there are to limite test results to make any better then stability concerns in this case. (No clear winner.)

PostPosted: Fri May 25, 2007 5:14 am
by evilboy
MediumRare wrote:clicking on the brand entry in the summary table brings up a new page with the detailed scans for each media up to "15 days". I hope the author will add the remaining images.
The remaining images are available but the author chose not to display them.

Example:
RITEKF1 15 days later
http://dvd-r.jpn.org/beam/4/5/B4_RITEKF1.png

RITEKF1 18 days later
http://dvd-r.jpn.org/beam/4/6/B4_RITEKF1.png

:roll:

PostPosted: Fri May 25, 2007 1:57 pm
by MediumRare
That's a good find- thanks! :D It looks like he just needs to update the cgi script.

G

PostPosted: Mon Jul 09, 2007 8:56 am
by Halc
Here's a short summary of my own UV exposure test results.

Do note a couple of things:

0) First measurement is before any UV exposure (other than out of the jewel case and into the burner). Second measurement is exposure after 60 days on a windowsill in Finnish summer, but inside of course. No high RH or high (surrounding) temperature exposure.

1) PIF cannot be counted as absolute. Those who you know my stance, know why. The numbers for the disc (reads) are comparable with each other within this test, but NOT across other tests. So, consider them 'relative PI failure averages across four drives'.

2) It's important to look at delta of change, not the final amount of errors only.

3) Couple of oddities: Philips -R 16x and esp. Gigatain +R 8x (TY fake) had a falling PI failure rate (that is, less PIF after exposed to UV). I have my own theory as to why, but don't want to spoil it for you yet :)

4) Why PIF? Well, it doesn't really matter if I used PIE or PIF. Averages or total with some statistical variance measure thrown in. They all correlate so strongly that the results are almost identical (corr >0.9).

5) Each disc was scanned with four drives, using same fw and same settings for each disc. Averages are across the four drives.

My apologies for the slightly unreadable graph. My excel graphing skills are not that good.

Image

Code: Select all
Budget DVD-R 4x       Princo
MAM-E DVD-R 4x       MAM4XG02
Fujifilm DVD+R 8x   YUDEN000-T02-000
Gigatain DVD+R 8x   YUDEN000-T02-000 (fake)
LG DVD+R 8x           PRODISC-R03-003
Memorex DVD-R 8x   CMC MAG.AE1
Ricoh DVD+R 8x       RICOHJPN-R02-003
Samsung DVD-R 8x   OPTODISCR008
Sony DVD-R 8x       SONY0D81
TDK DVD+R 8x       TDK-002-00
Maxell DVD-R 16x   MXL RG04
Philips DVD+R 16x   MBIPG101-R05-001
Philips DVD-R 16x   RITEKF1
TDK DVD-R 16x       TTH02
TY bulk DVD+r 16x   YUDEN000-T03-00
Verbatim DVD-R 16x   MCC 03RG20
Ricoh DVD+RW 8x       RICOHJPN-W21-001
Verbatim DVD-RW 4x   MCC 01RW4X
Verb.DVD+R DL 2.4x   MKM-001-000


To me there are roughly three groups of discs in the test:

1) discs that withstand UV exposure very well (60 days during Finnish summer, when sun almost doesn't set is very cruel test)
2) discs that do stand some exposure, but clearly deteriorate
3) discs that do not stand exposure to UV well

Yes, the discs are old, many of them not sold anymore (at least where _you_ live) . This is the downside of doing a fairly big test like this (UV was just a small part of it) on top of one's normal day job.

In fact, I think this is the last test of this kind I'll ever do, unless I really find out a way to streamline the whole process to 1/5th the time or even less :)

I wish I had the possibility to test several disc samples of each type, from different batches and with several different exposure times, but time and real life seriously cut into what I could muster.