Red blotches on every image with new 6D

[Carl]

Attached is a cropped section of a night shot just done with my new Canon 6D.
Problem: 22 images taken and on every image there are red blotches all over the image and they are all in the exact same position on every frame, no variation and not random like noise. There are also quite a few hot and dead pixels to add to the mess.
Images were all RAW, 4 min exposure and only at 800iso. Both RAW and converted JPGs have the same pattern.
I bought this camera as it is supposed to have a very sensitive chip suitable for astro imaging. Does it look like i have a dud or is this mess par for the course? I have a Canon 500d and never seen anything like it.

Can anyone shed any light on the issue please.

Regards,
a very worried Carl

[killswitch]

Run the in camera sensor cleaning about three to four times consecutively, this should reset the pixel mapping and get rid of stuck pixels. Not sure about the red blotches though.. maybe a defective sensor?

Do you have in camera high iso/long exposure noise reduction on?

[Terry B]

Have you used a dark frame?

[Octane]

What software did you use to do the conversion?

H

[Carl]

Answers to all questions
Q1. No noise reduction function on. Shouldn't need it. These red marks are consistent in one location every exposure, not random like noise. have done the same imaging with a canon 500d (no in camera noise reduction and perfect result).

Q2. Yes have done a test with dark frames as well. Same result.

Q3. Have tested conversion with 3 software programs (a) eos software, image browser that comes with the camera (b) Adobe DNG Converter and (c) Adobe Lightroom latest version

Regards
Carl

[iceman]

That doesn't look normal for a 6D. It certainly doesn't match up with the clean, noiseless images I've done with much longer exposures.

[iceman]

Was it taken in RAW?

Can you upload the RAW somewhere?

[Carl]

Thank's Mike
Below is the link to the RAW file; it's about 21mb.

http://filez.bordermail.com.au/io7veh9

If anyone has a 6D can they just do a test for me.
I shot that image in RAW, 4min exposure iso 800 using 16-35mm EF L Canon lens. Not that the lens would mean anything. RAW or processed all images have the same pattern.
Canon rep just tried to tell me its noise from having the shutter open for 4 minutes. Come on guys.

Regards
Carl

[Carl]

Interesting response from the camera shop i bought it from (Authorised Canon dealer not Grey import) . Just rang them to say that Canon suggested i send it to them via the retailer so they can have a look at it.
The guy at the store said no worries, if you think there is a problem just send it back with copy of Canon warranty and receipt and we'll just send you a new one, no questions asked. Unless you want to wait 6 weeks for Canon to look at it.
Not bad service, but if anyone has any ideas i'd like to get to the bottom of this i'd hate to think that there is an inherent fault in the sensor.

Carl

[Octane]

Have you tried opening the file with DPP and not Imagebrowser? IB just displayed the embedded JPG thumbnail from memory.

Once in DPP, open the image and ensure peripheral illumination correction and auto lighting optimizer are disabled.

H

[Carl]

Hi Humayan
Tried all that, no difference i'm afraid

regards
Carl

[CapturingTheNight]

Wow. That's rather disturbing Carl. Any chance you can get a hold of J. H's to do a test with same settings? I had a play with the RAW. My DPP couldn't read the 6D file because of it being an older version for my cameras to see if it automatically detected and removed the hot pixels (like it does on my files). Opened fine in Adobe Camera Raw. I was able to easily tune out the red blotches/pixels/noise , but as Mike says, the 6D should have bugger all of that sort of thing full stop. If they are prepared to replace with new at no cost, I would say go for it by the looks.

[Terry B]

I'm still a bit confused. Are you worried about the individual red pixel that are all over the image or red patches that are bigger than individual pixels ?

[Carl]

Hi Terry
Hot pixels are easily defined and removable. My image is just covered with red and blue dots tat don't look like they have a sharp edge. These dots are in the exact same position on all 20 images, surely it can't be random noise. I have no idea what they are. The guy from Canon suggested that it was acceptable noise, i find that hard to swallow on a camera that has been given a lot of hype about low light sensitivity.

I just did the same exposure test tonight shooting both RAW and JPG. The JPG was not too bad, but he RAW was just as bad as the above sample.

I'm in a pickle here.

Regards
Carl

[Davi5678]

I just got a new 6D last week and it doesn't do that, I'd get it replaced with a new one.

[iceman]

I've done 18 minutes and 53 minute exposures and don't get any of that sort of noise.

[coldknights]

Hi Mike ,
Do you have the camera's long exposure noise reduction on when taking night shots ?

[iceman]

No, never.

[coldknights]

I have taken 30sec shots with the mark 2 and mark 3 and no red dots at all as long exposure noise reduction is off on the camera's settings as well.

[Carl]

Hi Mike
Did you get a chance to download the image from the link is sent you so you can see the RAW image?
Also when you took the 18min and 53min exposure was it in RAW or JPG. Is there a noticeable difference between your result and my image

Regards
Carl

At this stage it's looking like a new camera for me

[Terry B]

I would have thought that if the noise is in exactly the same place on every image then it should easily subtract off with a dark frame. As you say it isn't random noise. Dark current is not "random noise". It is predictable.
I do not have the camera you have and only have a 40D but do have various CCD cameras.
Even in my cold climate a 5 min exposure will have lots of dark current that causes thousands of hot pixels across an image. If you are taking pics above ~5deg C then they will be very obvious.
Try taking a series of dark frames with the camera of the same duration. You need to cover both the lens and the viewfinder as light can leak in via the viewfinder. Average them and then subtract them from the raw image and see if the problem disappears.

[alpal]

Good post Terry.

Ed hasn't said whether he stacked the results in Deep Sky Stacker along with
a proper set of dark frames -
which would subtract out all those spots.

[iceman]

The point is he shouldn't need to.

Those frames do not look normal for a 6D.

http://www.mikesalway.com.au/dishes-in-long-exposure/
http://www.mikesalway.com.au/long-exposure-noise-performance-of-the-canon-6d/
http://www.mikesalway.com.au/np101-high-iso-noise-performance-of-the-canon-6d/

[LightningNZ]

Exactly, there's something really wrong with his camera - it should be returned and replaced.
-Cam

[Terry B]

Interesting. What that seems to show is that there must be something done in camera to reduce the thermal noise before outputing the file as a raw image. There is the physical way that the CMOS or CCD detector can have no thermal noise with such long exposures. Makes the camera not very useable for science images but this is not it's purpose.
I agree there must be something wrong with the camera.
Time to get it looked at by the manufacturer.

[gregbradley]

You got a dud. Send it in for a replacement.

Greg.

[Carl]

Thanks to every ones input on the issue.
It's going back to the shop for replacement.

Regards
Carl

[iceman]

Hi Carl

Just following up on this - did you get a replacement?
Is it better?

[gazza83]

My new 6D bought in the new year had the same problem as the OP's example. It was Australian stock and was replaced no questions asked. The new one is OK.

Cheers