5dIII hot pixel question

[HoneyBadger]

Hi all. I just upgraded from my 20d to a 5dIII. I love it but today I noticed a bright red dot in a picture without enlarging the picture. I went back and enlarged others around the same spot and could find the dot again at all ISO's.. even at 100. I just did a 30 second exposure at 800 ISO with the lens cap on and I got this.
http://i.imgur.com/GRrg0.jpg
Now I expect a couple bad pixels but I can count about 12 spots and some spots have way more than just 1 bad pixel I believe. I tried the manual cleaning trick multiple times for up to 4 minutes and I still have atleast 10 spots and it seems to not be improving so I stopped. Is this average or did I get a lemon? Also adobe camera raw fixes about 2/3rds of them but definitely not all now that I know where to look. I have 23 more days to return it to amazon. Thanks guys for your help.

 

[MattBicePhotography]

Whoa, That should be on its way to warranty service. That is not normal, or acceptable at all.

 

[HoneyBadger]

Thanks. That is what I thought. I hate to exchange things but in this case because of the money I want it to be almost perfect. 1 or 2 speckled in different places wouldnt be too bad. Anyone else think this is not acceptable? I hope the one I exchange for is not worse.

 

[Drizzt321]

Hmm...makes me want to do the same thing to double check for hot pixels. *whoosh* Off to test!

 

[HoneyBadger]

Lol I wish you luck. Let me know what you find so I can make a comparison to what is standard on these cameras.

 

[RunAndGun]

This early in and still being within your exchange/return window, send it back. I have a MKIII and it's been good, so far(got it "opening day"). BUT my first MKII(Dec '08) had hot pixel clusters start to show up within a few months and I ended up sending it in to Canon under warranty. I have another MKII body I bought in Oct/Nov '11 and it has lit pixel clusters that I started to notice in March/April, about the same length of time, I believe, that the pixels started appearing on my first body.

 

[Mt Spokane Photography]

I did not find any hot pixels on my 5D MK III, in fact it was amazingly devoid of them. But then, I did not try to create them with a super long exposure.
If you actually need 30 sec exposures at ISO 800, take a dark frame and subtract it. Thats pretty much the standard for long exposures.

 

[victorwol]

It is sort of normal for 30 seconds exposure if you don't let the camera do its noise reduction by taking the second pic automatically. But be aware that if you are using any resolution other than full size RAW you will see even more rd pixels since the substraction of noise and hot pixels seems to work only on full size raw, at least on my experience..l.

 

[fallenflowers]

i also have a 5d mark iii, i never knew about this testing until i read your post. so i decided to give it a try with cap on at iso 100 30s i saw only 1 ho pixetl, iso400 30s this time i saw 8, iso 800 30s 14 hot pixels and iso 2000 at 30s around 20 hot pixels. when i turned on noise reduction for iso 800 at 30s i think i saw only 1, so if you are concerned about hot pixels, just turn on nr and you should be fine. i know your hot pixels are consistent at any given iso, i think it's normal (12), if you decided to send to canon they would probably say the same, unless you have dead pixels, than that would be different story.

 

[fallenflowers]

looking at the pics u provided at iso 800, you are getting almost the exact number of hot pixels as mine, so i think this is normal for 5d mark iii. how about you guys out there who has 5d3 can you try this testing and post your results?

 

[Chewy734]

But, I thought HoneyBadger don't care?

I'm gonna try your long exposure technique on my 1D X and report back my findings.

 

[mrmarks]

You may need to heat up the sensor via live view for 3-4 minutes before the manual sensor cleaning http://photography-on-the.net/forum/showthread.php?t=1170332

 

[ed24]

I don't really know what to say....

http://imgur.com/TVESJ

 

[HoneyBadger]

oh wow. Do any show up in pictures you take?

 

[HoneyBadger]

Also do you guys noticed different colors of black when you zoom into your test pictures. Zoom into mine and there are a lot of black and lighter black patches.

 

[HoneyBadger]

I thought this was pretty interesting. I used the up arrow next to the histogram in ACR and it showed this at 66.7%. The larger dots all correlate with a hot pixel when I turn off that setting. Are the other ones pixels that have been mapped out?

http://i.imgur.com/04VZe.jpg

 

[ed24]

This is weird. I'm going to do some test shots this week to see if I can replicate the so called dead pixels in normal photos...

 

[smithy]

Much to my horror, my first 5D3 had a hot pixel towards the centre of the frame which was visible at all ISOs, and even on the LCD display. Fortunately I noticed it immediately (hard not to, really) and contacted the authorised dealer from whom it was purchased, and they provided me with a replacement which has been flawless.

It's definitely worth getting a replacement, if possible.

 

[Chewy734]

Have any 1D X owners looked at this phenomenon? I took some photos at varying ISOs and I was able to see 1 hot pixel at ISO 100, and 3 hot pixel regions at ISO 1600. This was shot with the body cap on at a shutter speed of 30s.

I was able to see all those hot pixels in my photos when I use such long shutter speeds.

 

[ereka]

This is very interesting. I've had my 5DMkIII since March. The first time I used it, I shot RAW + jpeg as I was only able to use the jpegs because I was travelling abroad and also didn't have any RAW conversion software with me that was capable of converting the RAW files. The in camera jpegs showed a few seemingly randomly placed black dots surrounded by lighter pixels - they looked a little like 'X' marks the spot! I'd never seen anything like it on my 1DMkII. They were quite noticeable at 1:1. When I returned to the UK, I upgraded from LR3 to LR4 and developed the RAW files in LR4. I haven't noticed the dots in any of the converted files, not that I've looked very hard but there's nothing that immediately stands out. As the in-camera jpegs are too mushy due to overly aggressive noise reduction (even with noise reduction turned off in camera), since I came back to the UK I've only shot RAW and converted with LR4. I've been reasonably happy*** with IQ and haven't noticed any hot pixels/dead pixels at all - not that I've been actively looking for them. When doing beauty retouching I often work at 1:1 or even up to 4:1 on occasion, so I'd have thought I would have noticed any hot pixels if they were there. To be honest, I'm not going to go looking because even if there are a few hot pixels/dead pixels they obviously aren't affecting overall IQ without a lot of pixel peeping (even with). I am intrigued to know though what my particular type of hot/dead pixel is/was - can anyone enlighten me? Until reading this thread, I'd assumed they were some sort of artefact produced by the in-camera jpeg conversion, particularly as I haven't noticed any (yet) in the RAW files.

*** I don't want to go off topic, but I say only 'reasonably happy' because IQ from RAW files converted with LR4 are quite good but the RAW files are noisier than I'd like even at low ISO (400 or less) and need quite heavy NR settings in LR4 (e.g. up to luminosity 70, detail 60) which results in loss of some fine detail. Also, although I don't use in-camera jpegs it would be nice to be able to do so e.g. for client proofing. I'm of the mind that this is not quite good enough for a camera at this level and I'm hoping for a firmware update that will improve the in-camera NR. I'm also hoping for a LR4 update as although I like the results, it is very sluggish when doing local adjustments.