Review of a very simple dead pixel test for my e500

chensin · 15th November 2006

Just carried out a simple test in order to find out whether CCD has any dead(bad) pixel.

- Designed condition -
aperture: F4.2
shutter speed: 30 second
lens cap: always covered
viewfinder: Covered with soft cloth in the process (consulted with Olympus staff and he said light leakage may occur if the viewfinder is not covered)
Format: Tif
Software: Dead pixel test (http://www.starzen.com/imaging/deadpixeltest.htm)
Threshold for hot pixels: 60
Threshold for dead pixels: 250
Others: the whole camera body was also covered by a cloth in the process

I did the test twice. The only difference is on/off Noise reduction.

- Result -

1st time
hot pixels:2915, max luminance: 123
dead pixels: 0

2nd time (with Noise reduction on)
hot pixels: 0
dead pixels: 0

- Conclusion -
No dead pixel was found in this camera. The noise reduction function is very effective.

- Question -
The only strange thing is that there is a very small patch of 'light-leakage' sign at the upper left coner of the picture although I tried with differnt shutter speed (more than 1 second) and aperture. It supposes that pic taken under my designed condtion should be totally black except for hot pixels (noise). But the brightness of the patch is not considered to be dead pixel at least based on the judgement of the software. After noise reduction, this patch can be totally removed. I'd like to ask experts here:

1. Why can such 'light-leakage' sign be seen all the time?
2. The hot pixels of 2915 is considered to be normal?

Thanks a lot!

OlyFlyer · 16th November 2006

chensin,
This is actually posted on the wrong forum. Ask admin to move your message to 4/3 section where all 4/3 dSLRs have a "Common Home" now. I am just visiting now and then here, not many other people who have a dSLR comes here any more so chances that you'll get answers here are very limited.

I have no idea about #2. As for #1 I think it is simply a third problem, long exposures result in the CCD heating up generating noise. The CCD must be allowed to cool down between shots. The heating up normally turns up in one corner and the area gets bigger the hotter the CCD gets. That means it is probably not light leakage that you see but noise due to the heated up CCD.

chensin · 16th November 2006

Originally Posted by OlyFlyer

chensin,
This is actually posted on the wrong forum. Ask admin to move your message to 4/3 section where all 4/3 dSLRs have a "Common Home" now. I am just visiting now and then here, not many other people who have a dSLR comes here any more so chances that you'll get answers here are very limited.

I have no idea about #2. As for #1 I think it is simply a third problem, long exposures result in the CCD heating up generating noise. The CCD must be allowed to cool down between shots. The heating up normally turns up in one corner and the area gets bigger the hotter the CCD gets. That means it is probably not light leakage that you see but noise due to the heated up CCD.

Thank you very much for the anwer to my question. It sounds very reasonable. And I will post in 4/3 section in the future. Thanks again for your reminding.

microcosm · 17th November 2006

Thread was moved from Olympus section as requested. Cheers.

17th November 2006	#4
microcosm Moderator Join Date: Sep 2006 Posts: 8,002	Re: Review of a very simple dead pixel test for my e500 Thread was moved from Olympus section as requested. Cheers.

15th November 2006	#1
chensin Member Join Date: Nov 2006 Posts: 496	Review of a very simple dead pixel test for my e500 Just carried out a simple test in order to find out whether CCD has any dead(bad) pixel. - Designed condition - aperture: F4.2 shutter speed: 30 second lens cap: always covered viewfinder: Covered with soft cloth in the process (consulted with Olympus staff and he said light leakage may occur if the viewfinder is not covered) Format: Tif Software: Dead pixel test (http://www.starzen.com/imaging/deadpixeltest.htm) Threshold for hot pixels: 60 Threshold for dead pixels: 250 Others: the whole camera body was also covered by a cloth in the process I did the test twice. The only difference is on/off Noise reduction. - Result - 1st time hot pixels:2915, max luminance: 123 dead pixels: 0 2nd time (with Noise reduction on) hot pixels: 0 dead pixels: 0 - Conclusion - No dead pixel was found in this camera. The noise reduction function is very effective. - Question - The only strange thing is that there is a very small patch of 'light-leakage' sign at the upper left coner of the picture although I tried with differnt shutter speed (more than 1 second) and aperture. It supposes that pic taken under my designed condtion should be totally black except for hot pixels (noise). But the brightness of the patch is not considered to be dead pixel at least based on the judgement of the software. After noise reduction, this patch can be totally removed. I'd like to ask experts here: 1. Why can such 'light-leakage' sign be seen all the time? 2. The hot pixels of 2915 is considered to be normal? Thanks a lot!

16th November 2006	#2
OlyFlyer Senior Member Join Date: Mar 2006 Location: Europe, Sweden, outside Lund Posts: 2,038	Re: review of a very simple dead pixel test for my e500 chensin, This is actually posted on the wrong forum. Ask admin to move your message to 4/3 section where all 4/3 dSLRs have a "Common Home" now. I am just visiting now and then here, not many other people who have a dSLR comes here any more so chances that you'll get answers here are very limited. I have no idea about #2. As for #1 I think it is simply a third problem, long exposures result in the CCD heating up generating noise. The CCD must be allowed to cool down between shots. The heating up normally turns up in one corner and the area gets bigger the hotter the CCD gets. That means it is probably not light leakage that you see but noise due to the heated up CCD.