Canon 5D MK3 “Hot Red Pixels” on long exposure

[neuroanatomist]

Neuro - The above is how I did it. Is there a better procedure... one that may be more successful?

That's basically what I did, except I had a lens mounted with the lens cap on it.

[Busted Knuckles]

Might I suggest a ramping of exposure - start with 1/2 second and see at what point the pixels fail. There will be some noise - but that isn't the same as a pixel going "hot."

I may be off a bit on this next part, but as I understand it, the ISO setting for this effort really can be thought of similarly to turning up the volume on a radio, whatever is out of tune will just be heard/seen louder. Keeping the ISO low will isolate the actual pixel faster, once identified by position, speed up the exposure and the ISO and see if it is still "hot"

Having said all of this (and having a couple of hot pixels in both my cameras) I simply don't do all that much 30 second exposure work and at anything shorter than 5 second the pixels pretty much function as desired.

If you have hot pixels at some fraction of a second, then that is a different issue.

[UmiKaibutsu]

I had the ISO very low anyway. I guess it is not that important. I don't use long exposure that often, I have used it though. Doing candle light panoramic of a restaurants large dining area was 10+ seconds each across 3 images wide. The area was prepped specifically for this image, taken with a 60D some years ago though. The pixels could be troublesome but I guess if it is that important they can be fixed in post. I'm surprised to see it happening in a Canon product.

Regarding the remote, I have a cheapy RC6 but it will not do bulb. I don't have a wired remote and I couldn't do bulb if I am running around a dark room with an LED keychain. The one is ordered is Yongnuo or Hangdong or DingDong something, garbage electronics from China. It's the 603, wait lemme google it... Yongnuo 603 ordered through Newegg. It wont do bulb though.

I have had issues with stuck and dead pixels in monitors, but it decreases as the quality of the monitor increases. Again, I am surprised at this being an issue in a 5D MK3. It is probably the Rising Sun and is a symbol of Japanese exceptional quality. I guess I mistakenly thought it was an error when all along it was meant to be there.

日本　すごい！

[conropl]

...Regarding the remote, I have a cheapy RC6 but it will not do bulb...

This is from the RC-6 remote "Overview" discription from B&H:

"May be used to activate mirror-lock and bulb-shutter functions"

http://www.bhphotovideo.com/c/produc...e_Control.html

I did not look up the others you referenced, but according to B&H the RC-6 should work in bulb mode.

[UmiKaibutsu]

ありがとうございます。　あなたが正しいです。

Thanks, you are correct.　I had to press the RC6 a second time to close the shutter.

[9VIII]

I've had two different experiences with this. On my T3 the sensor cleaning trick didn't do anything, so I mapped out where all the hot pixels were and sent it to canon, it came back producing nice clean black long exposures (I test at 8 seconds because that's the longest my HDR images usually go). They also updated the firmware so once hot pixels start popping up again I'll see if the update enabled in camera sensor mapping.
I also just picked up a used 5d2 this week and man was that sensor a mess! I did the cleaning trick and it improved the image a lot, it wasn't perfect like the T3 was (the T3 was extremely clean to begin with though) but it made a big improvement.

[PhotobyMayo]

Have you tried the free program "PixelFixer"? It has been around for a few years and it helps on long exposures. Basically you take a long exposure w/ the lens cap on, run it through the software, which maps the hot pixels, then you choose the level of repair, then run your long exposures through the program & it subtracts out the hot pixels.

[UmiKaibutsu]

@PhotobyMay: I will avoid the PixelFixer since I worry about the camera firmware, I could hurt it. I use custom firmwares on things only when the item is low cost or has a flash rom that is user repairable or modular. I guess the raw conversion does most of it, but if I keep it under 10 seconds, I don't think I will have too much trouble with it. When I revisit the long exposure fun, I'll try everyones suggestions. For now, I have the images I took already and I would say bringing from raw has toned down the red dots to small grey specks.

Meanwhile, before I do more long exposure, I have to redo the shot I am going to submit to the monthly photo challenge. It will be my first


@9VIII the sensor cleaner activates every time I turn off the camera, so it's already doing that. Maybe I misunderstood. I think that it vibrates for dust particles and that the cleaner could not resolve a stuck or sensitive pixel, which I believe is more of a chemical issue as the semiconductor is malfunctioning and the semi conductor is composed of chemical compounds designed to perform in a particular way. Microvibrations should not affect this. Perhaps the compound in that particular pixel has an irregularity. Perhaps the signal it generates is not being modulated properly, so it could be programmed accordingly and resolve the issue. My guess is that any correction of the problem is going to simply eliminate the pixel from the overall picture (e.g. PixelFixer), in the event that it does, I will then have a dead pixel, as opposed to a over sensitive pixel. The latter being preferable as the effect is exponential to the shutter speed. If I keep the shutter speeds normal, the pixel will not have sufficient time for its heightened sensitivity to be obvious.