My 14 day return period runs out next Wednesday. If Canon doesn't have something interesting to say about it by then, I'm going ask my brick and mortar (pro camera store) for an extension of the return period until Canon's announcement. If the store can't do that, it's going back.
I just can't believe that Canon has screwed up yet another major product release.
CMOS wrote:
I just can't believe that Canon has screwed up yet another major product release.
Like someone else asked, do you happen to have a picture worth printing that shows the problem?
I'm just trying to figure out how big of an issue this really is. If it only comes up in shots that less than 1% of all users take, I'd hardly call that a screw up.
Mattbtn wrote:
Like someone else asked, do you happen to have a picture worth printing that shows the problem?
I'm just trying to figure out how big of an issue this really is. If it only comes up in shots that less than 1% of all users take, I'd hardly call that a screw up.
Actually, I am seeing the problem in photos that I'd consider printing. There's an amusement park not far from where I live. It's got lots of neon accent pieces. I was out there a few days ago with the wide angle getting some fun shots of the signage. Plenty of those shots have black dots bordering the neon lights.
Those weren't test shots, I was actually trying to compose something interesting. They might not be masterpieces, but they were legitimate shots.
Hey.Underpants wrote:
Its a flattering soft focus.
Seriously though, does a "real" photo that has this problem exist?
Indeed several of us are curious about this. I'm honestly beginning to wonder if Canon did this as a move to aggresively fight ESO Syndrome in it's users. If the pictures are bad they are purposely made unuseable.
For those that don't know, ESO = Equipment Superior to Operator.
CMOS wrote:
Actually, I am seeing the problem in photos that I'd consider printing. There's an amusement park not far from where I live. It's got lots of neon accent pieces. I was out there a few days ago with the wide angle getting some fun shots of the signage. Plenty of those shots have black dots bordering the neon lights.
Those weren't test shots, I was actually trying to compose something interesting. They might not be masterpieces, but they were legitimate shots.
Mattbtn wrote:
Like someone else asked, do you happen to have a picture worth printing that shows the problem?
I'm just trying to figure out how big of an issue this really is. If it only comes up in shots that less than 1% of all users take, I'd hardly call that a screw up.
Wow, your statement is pretty similar to how many were reacting to the posts 1D3 and AF problems.
Doesn't have to be worth printing to see that there is a problem here, so much that Canon has acknowledge it, with comments coming soon.
Curator wrote:
Do you have any normal shots, something that you would actually want to print, that shows this problem?
I'm done with spot hunting for now, but . . . .
The same day I received the camera - I found out about the spot issue. My first decision is whether or not to keep the camera - so I tired to 'provoke' the dots, and see what reduces the impact. ISO 50 & HTP<off> seem to minimize the issue with my camera. At this point I'm probably going to keep the camera - since I have a good idea of when the dots occur, and how they affect image quality.
The grid picture - shows lots of dots at the center/bottom. The dots are at grid highlights - right side of each reflection. These are not noticeable at 24x36 print size - and most (sane) people do not inspect large prints under magnification. However, those dots are annoying when post-processing, and will litter any night shot of city lights.
But if you are cropping a high ISO shot - the dots could become an issue.
I don't have a shoot scheduled until next Tuesday - but I know in fashion shots the dots will rarely even appear - but that does not get Canon off the hook.
- I had to fight to get them to appear
- HTP on, ALO on, NR off
- ISO 3200
- Evaluative metering
- Shooting RAW + .jpg/large/fine
- 35mm/1.4L and 200/2.8L primes
- Lightroom 2.1 on Vista Home Premium, 64-bit Intel
I first tried some outdoor Christmas lights at night (handheld), then a test arrangement of lights spanning a matte black and a white background indoors with lights off at night (tripod), both with the 35mm at 1.4, and finally another night shot with the 200mm at 2.8 on a tripod. I shot -1, 0, and +1 EC. I didn't test with HTP or ALO off although I haven't used either one other than for testing.
- Nothing obvious with the handheld shots.
- A handful of dots with the 200mm.
- Almost all of the dots were on green lights
- More obvious at longer exposures (max exposure was 1/2 second)
- I didn't see any on white lights.
- Dots were 1 or 2 pixels
- Cranking up exposure/curves, and they appear to not actually be black, just dark
- I had to look at 300% to find them, then I could see them back at 100%.
I actually had a more noticeable (but still minor) issue with lens/UV filter reflection. CA and fringing were much more noticeable than black dots.
While I'd still like a "perfect" camera/sensor, I'm not going to break down Canon's door for a refund or replacement.
Dots Bad?
Maybe we should look at the dots as a mark of distinction. I can hear an art director . . .
"What? No black dots? Go reshoot this with a mkII and get back to me"
At the Gallery: "and here we have a lovely city skyline taken by a wonderful photographer. His attention to detail is amazing - he actually signs every light in microdot form, which really transcends his work to a unique plateau above other local artists . . . ."
Wonder how long Nikon will take to rush a BD model to market?
Anyone know if Canon is still shipping the cameras out to dealers? If they stopped, that would be a possble indication that Canon is thinking of a recall for a hardware problem. If Canon is still shipping out, maybe they think it's a firmware problem.
I guess it's also possible that Canon will keep shipping and then just have to recall more cameras...
Great find! This has got to be it. Samsung recently got a patent for a "Column analog-to-digital converter of a CMOS image sensor for preventing a sun black effect "
Bad news is that both of these solutions look like hardware. You guys have identified a real structural problem with CMOS sensor. I can't wait to see how Canon is going to fix this one.
Sorry if this was posted but I cracked up when reading this: