5d mark II strange black dots?

SwingBopper wrote in post #6916749
I totally agree with Tonylong here. In that spirit I'd like to point out that my testing shows the BlackDots to be sensitive to shutter speed not to ISO. Holding ISO steady and varying shutter speed shows dots at 1/250 sec and getting progressively worse with longer times. At 1/500 there are no black dots; but there are groups of dark pixels (sometimes red or blueish) where the black dots manifest at longer times. This makes me wonder if the dots are just an extreme case of a larger problem in a "sensor bloom" situation caused by sensor circuitry pre-processing RAW data. In this image (1/250 ISO 6400) the dots (now gray) have almost disappeared. At 1/500 the red bloom remains but no gray or black dots. Also notice the red bloom around the vertical lattice at the right. Has anyone else looked into this?

Is the red just CA? How can we know?

tekkie wrote in post #6917167
I think his point is the majority of the people making this out to be some huge deal doesn't even own the camera, whilethe majority of the people that have the camera accept that the issue is there and is waiting for a fix

There have been a wide variety of responses to this, most of what I'd consider measured attempts to examine, analyze and discuss the problem. Of course, this was intitially brought to our attention by real users and real users continue to buy the camera. Many of those real users have found this issue on their cameras by simply shooting Christmas lights or city lights. But they are still overall happy with their cameras.

Of course, there are those who don't own the camera who are looking at this and hoping for a "fix" before taking the plunge. Is that some kind of bad thing? I don't think so -- after all it is a real issue that affects real images. I'd certainly want this hashed out if I was getting ready to buy into a body like this.

There are a minority of people who rant, as if this is some kind of Canon Armageddon, but that is a minority of people. I and others try to be a voice of balance and reason.

This thread is to discuss the problem. If the problem is not relevant to a particular photogs shooting, then all is good. For some of us it's more of a concern for some less.

One thing that would be useful to all of us when discussing photographic issues is to preface our statements, either implicitly or explicitly, with a simple phrase: "For the type of photography I do (YMMV), I have this view:".

Just thinking of that before we express an opinion could keep us from sounding as if our view was the only view.

radiohead wrote in post #6915364
Really?

So if something is known to be wrong I have to actually own it to be able to say so.

Sounds like complete nonsense to me.

I've been through the same argument over @ dpreview. The people who think Canon has done a grand job will usually hit out at anything that dares even remotely criticise Canon, even if it's well founded. They'll use any excuse to make our thoughts and opinions invalid (at least in their minds).

Dave

RDKirk wrote in post #6916493
You sold your grandmother to buy a camera? Dude.

I've said this before and I'll say it again. If it's a major financial burden to buy a new camera, don't buy it. It's totally foolish to incur a major financial burden for either a hobby or a job.

And I would counter argue that it's up to the individual as to why and how they want to buy a camera (or any other consumer product), not you.

Dave

dpastern wrote in post #6917338
And I would counter argue that it's up to the individual as to why and how they want to buy a camera (or any other consumer product), not you.

Dave

And I would add to the counter argument, why should a person's financial position in buying the camera have anything to do with the price, performance, quality opinions, concerns of a camera.

Rock Photo Star wrote in post #6917641
And I would add to the counter argument, why should a person's financial position in buying the camera have anything to do with the price, performance, quality opinions, concerns of a camera.

Agreed.

Dave

SwingBopper wrote in post #6916749
I totally agree with Tonylong here. In that spirit I'd like to point out that my testing shows the BlackDots to be sensitive to shutter speed not to ISO. Holding ISO steady and varying shutter speed shows dots at 1/250 sec and getting progressively worse with longer times. At 1/500 there are no black dots; but there are groups of dark pixels (sometimes red or blueish) where the black dots manifest at longer times. This makes me wonder if the dots are just an extreme case of a larger problem in a "sensor bloom" situation caused by sensor circuitry pre-processing RAW data. In this image (1/250 ISO 6400) the dots (now gray) have almost disappeared. At 1/500 the red bloom remains but no gray or black dots. Also notice the red bloom around the vertical lattice at the right. Has anyone else looked into this?

I will start with I DON'T own a 5d2 but I am on the long look list. That is I would like to get one.

I have secretly thought that the Black Dots were caused by over loading the pixels at these locations and that it would be fixable BUT at a performance cost. What I mean is the fix would make the chip LESS sensitive and thus we would lose some low light performance. Just a guess but to it makes a measure of sense.

Since the thread is going no where until Canon releases a fix or something, I want to know if we would be willing to get rid of the spots at the cost of reduced low light performance?

If this were the case would the PP fixes be good enough?

I have held back thinking that if this one slipped what else did Canon miss? For now at least it looks like nothing. Sure hope it stays this way.

I'm sure that canon is already testing several fixes. Evaluation/testing takes time. Be patient if this is an issue in your shooting.

dslrvids wrote in post #6917827
no strange black dots as far as I can tell. Shot this around 1000-6400 ISO at night.

eeeeenough!

syntrix wrote in post #6917204
Is the red just CA? How can we know?

Good question. I'm not a CA expert by any means; but would you expect CA in the lattice on the right of the photo where there are no blown out highlights?

News Alert:

I don't want to muddy up the waters here, but I've been curious as to whether artifacts like these "black holes" show up in other bodies.

So, I don't have many Christmas light shots, but today I took a couple flashlit night snow flurries, and I thought of a family fireworks shoot I did last Fourth of July, all done with the 5D Classic. At a 100% view, I'd be hard pressed to find such artifacts, but guess what? When I zoomed in at 400%, there were some pretty noticeable ones.

Now, the immediate reaction is that this proves that the 5D2 blackholes are not really a problem, but one thing I will note is that many of the black holes I've seen from the 5D2 are considerably more pronounced than the ones I've seen from the 5D, which is what bothers me about the 5D2 shots: they are not just noticeable at 400%, but at 100%.

So, this was educational for me, because it is evidence that points to the possibility that we are looking at a sensor architecture issue that just is more pronounced with the 5D2.

Anyway, if your curious, I've posted a few in a pbase gallery:

www.pbase.com/tonylong​/dec_20_08_cmos_artifa​cts

Hope this is educational and entertaining!

Speaking as a product photographer who shoot watches and jewelry a fair bit, the black dots problems will be a major drawback if its not resolved completely.

I have been waiting for the camera to be available here but this problem has reinforced my belief that

1. only the desperate buy the first batch/generation of new technology and
   
2. manufacturers - no matter how good historically they have been will eventually have someone at the top who is more a marketing type of guy rather than an engineer and push products ahead before its ready - or I suspect - sit back and not push R&D until competition makes it necessary to speedtrack products that are not completely resolved of all issues.
   
3. not enough testing is done to make these products fault free. What is critical failure on their part is that they forget that many professionals rely on the gear to be working properly out of the box. 1D3 and 5D2 are body types that pros use and 50L infamous back focussing are just some recent examples.

SwingBopper wrote in post #6916749
I totally agree with Tonylong here. In that spirit I'd like to point out that my testing shows the BlackDots to be sensitive to shutter speed not to ISO. Holding ISO steady and varying shutter speed shows dots at 1/250 sec and getting progressively worse with longer times. At 1/500 there are no black dots; but there are groups of dark pixels (sometimes red or blueish) where the black dots manifest at longer times. This makes me wonder if the dots are just an extreme case of a larger problem in a "sensor bloom" situation caused by sensor circuitry pre-processing RAW data. In this image (1/250 ISO 6400) the dots (now gray) have almost disappeared. At 1/500 the red bloom remains but no gray or black dots. Also notice the red bloom around the vertical lattice at the right. Has anyone else looked into this?

That's just flare and perhaps a bit of CA and does not appear to have any connection to black dots. I see that on my XTi when I have a image with very high contrast edges when the highlight is blown or nearly blown. I saw the same effect when using a demo 50D that a Canon rep allowed me to evaluate. Using my own lenses, I took some shots where there were plenty of specular highlights to see what worst-case situations would look like. Bloom is a phenomenon that is mainly confined to CCD type image sensors because of their physical construction which doesn't allow for very good isolation between photosites to minimize charge coupling. The effect is not very likely in CMOS sensors except for extreme overexposure situations where the image would be worthless anyway.

syntrix wrote in post #6917204
Is the red just CA? How can we know?

It could be CA, but in this case it looks more like flare ... or even a little of both.

AllenF wrote in post #6917738
Since the thread is going no where until Canon releases a fix or something....

Most sensible thing said in this whole thread.

syntrix wrote in post #6917198
You never directly answered my question if you own one. But I can extrapolate that you don't have one. But you are concerned about these things as a possible future owner:

1. black dots
2. pricing drop
3. Bright day IQ
4. seeing more samples (there's tons out there already), did you mean maybe more samples on the next firmware update?

I Simonius wrote in post #6914255
Im not buying it until I see Canons answer

- sorry, I thought that was unequivocal.

Re: samples, the ones Im interested in are those that cover the sort of shooting I would do, processed the way I would process. As it stands I dont have enough to go on, but Im sure there will be enough once the camera production is up to speed

I have seen shots along the above lines and have not been impressed so far. I would like to be impressed, as i was with the 5D samples before I bought that Camera

I am keeping a keen eye on the BD issue asmore info and new samples arrive, it is very much of interest. I speculate that if the fix is unsatisfactory the price will drop sooner rather than later and that will be to my advantage. Im in no hurry