10D image defect

This light-arc artifact or fault abruptly appeared in all my 10D

images beginning last weekend. See below (full frame):

At first I thought it would be a lens problem, but this light-arc

fault persists, regardless which Canon EF lens is attached.

Camera/ambient temperatures are not a factor.

I�ve examined the 10D sensor via Zeiss stereomicroscope and saw

nothing suspicious.

I�ve searched Photo.net and the web, but can�t find anything out

there describing this particular problem.

Has anyone seen this problem with the 10D or offer any advise..?

The 10D is still under warranty, but I�d be very grateful for any

info before I go down that �Customer Service� road.

Many Thanks in advance!

Scott<div></div>

Hi RSS,

Bummer, could be an amp or memory clock problem. Take it in to Canon for repair. Good luck!

I had a band of light almost exactly like that on an 8x10 B&W negative the other day.

I thought it was being caused by flare from the front and rear surface of a red filter I was using.

From the picture the arc looks like exactly that: an arc, part of a circle. I can't imagine why electronic system using a row column matrix would fail and reproduce a perfect arc? The odds would seem to be a billion to one.

Therefore I suspect it has to be something optical.

Did you just buy and install a bunch of UV filters to protect your lenses?

I bet it goes away if you use a lens shade.

Thanks Folks for your comments so far!

Here's a little more background...

This 10D I purchased (B&H) in early June 2003, so perhaps it's one of Canon's first shipments of this model.

The light-arc occurs regardless of the mounted lens, so far this includes: 70-200mm f2.8L IS, 300mm f2.8L IS, 28-135mm IS and 85mm f1.8 lenses. I always use factory hoods, but usually no filter, unless specific need or effect, but this light-arc still there regardless, even with a polarizer on the lens.

The light-arc is always in the exact same centered location in the image frame.

The light-arc is less apparent when image is overexposed, but it is still there.

As I mentioned earlier, stereomicroscope examination showed no dirt, hair or anything physically "funky" with the sensor.

This 10D camera has never pointed at the sun (or other dumb tricks). All-in-all I'm very very careful with my gear...

Please keep the ideas coming..!

Go through the sensor clean operation and see if something isn't lurking aorund inside your camera body. This happened to me about 3 days ago, it was a hair in front of the sensor.

Rick H.

I still think that it is flair.

Try this: put a lens cap on it, and cover the eyepiece and take a picture. If it is still there we have pretty much ruled out light leaks.

If it goes away try no lens cap viewfinder covered. Maybe the foam gasket on the mirror went away.

Otherwise look for something shiny inside the camera. The lenses you are using throw a 35mm image circle. That is a lot of extra light to bounce around inside the camera.

I have a digitial Rebel that has this exect same problem. If I shoot at a shutter speed above 1/200 sec it becomes obvious and becomes less obvious with slower shutter speeds. I have this problem both with my 70-200 F2.8 lens and the kit lens that comes with the camera. The lens shade on the 70-200 lens has no effect.

I have emailed Canon tech support about this problem and gotten no response. My next step is to call them about it.

Keep in mind, dust, hairs, etc. should cause a dark defect. Good trick Neal, try no lens and cap.

Not first shipment of model, I got both of mine In March of 2003.

In my film days, this type of defect was caused by hairline cracks in the back of the coating on the mirror causing a reflection during the exposure. (Say that three times real fast!)

Send it to Canon. Good Luck!

Just waiting for the 1D MkII...

I have seen the identical problem with my own Canon 10D and I returned it to Canon last week for servicing. The band (or light-arc) appears in exactly the same place. The people in the shop that I had bought it from had never seen anything like it. Thankfully I still have a month left on the warranty. I'm very curious to know if you managed to solve this on your camera. Also, in some ways I'm glad to see someone else has experienced a similar problem and I'm not going crazy.

My own experimentation with this problem ruled out the lens, as it occurred with several lenses. In fact, I could actually get the same bands to occur without a lens at all. There was certainly nothing on the sensor and nothing looked unusual inside the camera. I also tried turning on the mirror lock to see if that had any effect, but to no avail. The best way to see the bands clearly was to photograph a blue sky.

What I did find was that the bands were much more prominent with a faster shutter speed. They were barely noticeable at 1/60 of a second but very noticeable with any speeds greater than 1/250 of a second. Perhaps this is something wrong with the shutter mechanism? It seems unlikely that it could be dirt on the sensor as then the band would be darker than the surrounding image.

I suspect that this is a mirror problem. The mirror may not be rising all the way and sealing the image chamber.

You could test for this by blocking the viewfinder window to stop light leaks coming in there. Another test might be to put it on manual with the same setting and circumstances as a picture known to exhibit the problem, but take a picture with the lens cap on and leave the viewfinder exposed. If you get a streak in that one, I would suspect the mirror.

A poster in another thread elsewhere suggested there might be a leak between the lens and camera, but I find this unlikely. You could test for this by wrapping a black cloth around the lens mount in a situation known to produce the problem.