I can't imagine shooting or doing music or painting or living for any end product.  The real end product is that eventually you die.  For me life is about the moment, the creativity of the process, the inspiration and natural high of looking for and finding those visual treasures in nature, not just the product.  I have to think about product too, especially for marketing, but if it's only about product and not the actual process of creation, I'm not interested.  If the shooting part of wilderness photograhy is nothing but being somewhere pressing a button for you, we must not even be the same species.  But, to each his own.  No problemo.

I'm more interested in the sensor blooming thing than philosophy rants, thank you.


Well, that's a big opinion.  I haven't seen too much evidence that the more planned out and "paradigm" oriented shooters get such superior results.  Anyway, I do get an awful lot of unsolicited advice related to all this and if admiring your words of wisdom displayed this way gives you some sense of satisfaction, hey, no harm.

In that case I am also interested in what type of polarizer to use (circular or linear) and how to use it! I can imagine that a circular polarizer has little effect, while a linear polarizer allows you to play with the incedent light!

Do you mean a raw file?  I don't have enough space on my website for anything like the necessary 11 MB.  Full size tiff is 33 MB.  If you don't need to do a raw conversion, wouldn't a crop with a particularly bad area in the middle work?  I could upload that and provide a link or email it to you, but that won't work for anything like raw or full size tiff or even jpeg.

Would you mind sharing your idea?  The experts here can't even agree about what the problem is (sensor blooming, CA, or raw conversion artifacts) and no one has come up with any suggestions for a fix that works in the worst areas.


This information has not been previously posted in either thread; thank you.  However, the worst areas in my case could not conceivably be painted out because they're so bad you can't even tell just what was supposed to be there.  I'm disinclined to buy and try assorted expensive converters, since the Canon and Photoshop converters produce identical failures to mitigate this.  I absolutely don't believe that this is conversion artifacts and I doubt that any converter could give acceptable results in the worst areas, even given the possibility of some mitigation.  Lightly dispersed noise like this is fixable in several ways, including tedious use of the cloning tool.

I suspect that if there's a solution here it's in the realm of prevention rather than cure.  I'm particularly hopeful that greatly underexposing such scenes will reduce or eliminate the effect.  Of course that will only be feasible if the scene has no deep shadow areas.

This posted image is at least 1/2 stop overexposed if you look at the rocks in the background.  This would overexpose the specular highlights by 2 stops or more.  Slide film cannot handle that range, why would you expect a CCD or CMOS sensor which is more sensitive to overexposure than film to do better.  If you were using negative film, this image would not have been a problem as the exposure latitude of the film would have covered the difference in exposure between the specular highlights and the background.  
A histogram of the image indicates overexposure in the individual channels especially the blue channed which is consistent with the altitude in the Sierras.
There is a great deal to be said about learning the limitations and behavior of media, be it digital or silver based.  You then take that information and integrate it into what the camera is seeing and make adjustments accordingly.
That is why cameras have manual modes and spot meters.
Galen Rowell said it takes 8-10,000 hours to become proficient in a skill whether it be photography or music.  There is a certain amount of truth in what he said.  This is where experience comes in.
A polarizing filter would have helped immensly in the example you posted, as would have metering off the highlights as well as the rocks in the background using a spot meter, then choosing an exposure that would result in an approximate =1 reading on the highlights.

Here is a quote from http://www.northlight-images.co.uk/Canon_1DS_MkII.html .....

"The 1Ds MarkII sensor shares with the 1D MarkII several substantial improvements made possible by the Canon-developed, on-chip RGB primary color filter together with larger micro lenses that have much smaller gaps between them than those on the 1Ds. These narrow gaps greatly increase the efficiency of light convergence while greatly reducing birefringence. Canon has also improved the design ofthe photo diodes in the CMOS sensor by eliminating transistors in each pixel, making a greater portion ofthe surface area of each pixel sensitive to light. Normally, it's nearly impossible to provide both high resolution and minimal false color, but the EOS-1Ds MarkII has a 3-layer optical low pass filter that works with the DIGICII chip to reduce false colors while having minimal impact on resolution."

I gather you are working with old shots, but it might be interesting to try to recreate the problem and to assess what (if any) differences in effect can be made by using different lenses. As I understand it, the problem of birefringence is a result of an optical interaction between the lens and the microlens array (try goolging birefringence microlens eos ). This may be simplisitc (or wrong) but it appears that light incident at a particular angle on one microlens ends up contaminating ajacent microlenses - I seem to recall reading that the effect is most pronounced when the light is convergin (or diverging) at the plane of the microlens and hence would be sensitive to lense design.

Have a look at the picture of the lenses at the link above and you can imagine how incident light could be scattered at various wavelengths to nearby microlenses. One can also imaging how this effect would be most notable at highlights and edges where the scattered light can overpower the intended light

- just a thought

m

In going through my zillion or so Sierra images, I've just run into a whole series of shots that I really like and found that nearly all of them are irretrievably ruined by sensor blooming.  With a lot of work I was able to salvage a couple, but when a whole major area is nothing but this weird colored noise garbage, where you can't even tell what was supposed to be there, there's no hope.  

These images were also shot into the sun with a lot of bright specular highlights on water and also some from ice.  It occurs to me now that a polarizer ought to be able to totally prevent sensor bloom.  Anybody with any experience to verify that?

In any case, now I've had a really convincing episode to show me that shooting and shooting and shooting without ever looking carefully at anything is not such a hot idea.  Well, I was in the grip of "Sierra fever" and I don't really have any regrets, but I do plan not to have such huge lags between shooting and looking at what I have in the future.

I can't even imagine shooting like that.  I shoot for the end product, not to be somewhere pressing a button

Didger, I know I have said this before, but I will repeat it...this is not sensor blooming but artifacting from the raw convertor.

Why do I say that it is not sensor blooming - whilst CCD sensor do suffer from blooming, CMOS sensor do not. I'll give you two references from the net to back up this claim as no-one is apt to believe me without them

Also, since CMOS technology has no blooming or smearing effect, it opens the market to many outdoor applications", says Dr Joachim Linkemann, Basler Vision Components Product Manager.

CCD vs CMOS comparison
Drop down to additional CMOS advantages.

Well, whatever it is, it's not just greenish.  It's various (green and purple, the purple being more obvious)  and it only happens when I'm shooting into the sun and there's bright specular highlights on water or ice.  If it's CA, it's nothing like normal lens imperfection type edge CA and it also absolutely doesn't respond to the CA fixing you can do with the Photoshop converter.

I just tried comparing a Photoshop conversion with Canon conversion of the worst image and there's no difference.  The Canon conversion has much more intense yellow, like Velvia, but the weird color noise grunge is the same, so apparently the Canon 1ds converter doesn't work right for 1ds images.

I don't really care what it is.  I still hope that a polarizer will help, in case Jonathan is right and the problem (whatever you call it) is because of absolute light levels being too high for those particular sort of specular highlights (shooting into the sun).  In that case you can use a polarizer to cut them back to whatever extent you want and your problem is solved.  Next time I'm in such a situation, I'll do some testing and let ya know.  I hope maybe Jonathan already knows.

Nothing of the sort.  I've gone through way over half of my huge collection and this is only the second episode, and from exactly the same sort of circumstances.  There's no way I would get this very often because there's so rarely any need to shoot straight into the sun.  I have also not run into any other disappointments that I could have been spared with a more proper and intellegent approach of looking at stuff before shooting so much.

As for the many many things I'm doing totally wrong if I'm ever to even have a remote hope for success as a photographer; hey this is a total win/win situation for all concerned.  If I succeed, I'll be happy and you'll be happy too that someone doing everything wrong beat the odds and succeeded anyway.  We'll all get together and have a big celebration, right?  

If I fail you'll of course all be very happy and nod sagely and have the "I told you so" glow of satisfaction.  I'll still be happy because I'm doing all this and enjoying it a whole lot (wilderness and Photoshop) and not making money with it eventually won't subtract anything from that joy and inspiration because it's NOT about "product".  I'll just have to continue to do other things for money.  Some combination of my creative endeavors has kept me from starving all this time just fine and I've never once had a day job and I've still got both my ears too.  :D

Umm, do some of you ultra-competent and "correct procedures" folks already have definite arrangements that your hard disks will all be enshrined and mined for luscious gems to inspire future generations?

Just out of more than idle curiosity... while you are trying filters, how about trying a neutral density filter or split density or something like that? Can't do any more harm than the polarizer would.  Just maybe, the specular light is too much for the cameras sensors and the density filter could pull that back a notch.  I have absolutely no way of knowing if it will work but just brain storming.  :p

An ordinary circular polarizer can dramatically reduce or totally cut reflections from water or ice (but not metallic surfaces), but I'm not actually sure about specular highlights.  Maybe not.
ND filter would be useless for sure.  You could just as well use a faster shutter speed to reduce the light without the inconvenience of an ND filter.  In fact, underexposing a couple stops or so just might work for a scene that's fairly evenly lit like these were so that you could pull up a couple of stops in the conversion and still maybe have a good image; better than all that weird garbage.  I'll try underexposing and polarizer at the next opportunity.  There's plenty of such opportunities, but not many worth photographing, except for a test.

Could you post a link to one of the bad ones - preferably full size - there "may" be a solution....

Lin

A jpg would be fine - just something which could be used to analyze the condition. Sensor blooming doesn't happen with CMOS based cameras, but various color fringe issues can and do occur (transverse or lateral chromatic aberration - micro lens exacerbated purple fringe, etc.). There are excellent post processing remedial solutions for most of these, but I can't make a suggestion until I actually see the problem, and the solutions sometimes won't work on a crop, but do require the entire image.

Lin

What - you didn't notice all the cotton candy moss growing there when you were shooting?

Yeah, I know, but this has rarely been a problem.  Generally if I'm even suspicious there could be a problem I bracket and then blend the over and under shots.  Shooting into the sun where there's very bright specular highlights is a special situation and I can now easily enough deal with it, now that I'm aware of the problem.

I'm still a little confused about just what it is that you are seeing. What is the substance in the pond and is the small amount of excess color in it (CA) the problem?

The excess color (slight fringe) at the pixel level can be easily removed and the slight overexposure of that portion of the image can be resolved, but I'm still not clear on whether that's the problem which you refer to?

The small amount of misconvergence on the white substance is fairly normal for any digial instrument and would not normally show up in prints unless they were tremendous enlargements. On the other hand, if the issue is the substance itself (it doesn't appear to be reflection, but indeed some type of pond scum) I assume it was actually there when you took the shot. This doesn't appear to be an artifact of any type, just an inability of the sensor to resolve any further detail at this level.

Your explanation makes the most sense of anything offered so far.  However, the practical problem is not explanation, but prevention or cure.  Only mild occurrences can be cured at all and even these are annoyingly labor intensive to cure.  I'm pretty confident, however, that now that I know exactly the circumstances when this happens I can prevent it, perhaps with a polarizer, but almost certainly by underexposing a stop or more.  If latitude considerations don't allow that, I'll bracket shots to be sure those extreme specular highlights don't cause this problem in the future.

Geez, that's what we've been telling you all along! Doesn't work technically, compositionally, nor artistically..., but go ahead, be stubborn, be a loner, be the grossly misunderstood artist that makes avant-la-lettre art that the world is just not ready for yet! By the time you're dead, there is nothing easier than reformating a harddisk!
::        

Well, sorry, that one was too easy, and you probably were hoping someone took the bait.

I don't think a polarizer makes a difference in sensorblooming as it will not necessarily diminish the intensity difference between adjacent pixels.

btw are you actually saying that you see the same type of artifacts as in the "WHAT IS THIS???" thread? And that you know find that it occurs far more often than you initially thought/hoped?

In that particular case I would seriously suggest to look again at JF's assesment of CA as the possible source. The fact that the overall color in that image appears as greenish in the downsampled version does suggest that there is a particular pattern to it, and not entirely random. The color (greenish) suggests CA, where blooming on most RGB type bayer chips would be bright purplish...

I hope I'm not repeating something that has already been posted but I noticed this link to a review by Keith Cooper on the DXO RAW conversion program.  In the review he cites the same phenomenon you have been reporting with water highlights exhibiting color artifacts.  In this particular case he implies that DXO is superior to ACR at limiting the effect.  He also references a previous post of his where he shows the effect and how he had to "paint" it out.  Not necessarily the solution you may want for batch processing but perhaps you'll find it interesting.  Again, I apologize if this information has previously been cited.

Joel

Keith Cooper

OK, I've cleared some space on my website and put up a high quality hi res jpeg showing the worst of this garbage.  If you think that this could conceivably be fixed, but maybe not provable with this jpeg, I could send you a CD with the raw file and a full size tiff, or if you have a location I could FTP the material to, that would work also.  In any case, maybe seeing the picture gives you some idea of exactly what's going on.

Yeah, but normally 1/2 stop overexposed is no big deal and conventional wisdom says shoot for the right end of the histogram to reduce shadow noise and I can't see just blue over on my 1ds field histograms either.  I do leave my camera permanently in saturated pixels flashing mode, so it's a little strange that I got neither any histogram or flashing blown pixels warning with this.  I've only seen this weird color noise problem twice now, both times shooting into the sun where there were very bright highlights.  Now I know that in the future I'll underexpose at least a stop and check to make sure that the histogram isn't then falling off the left side on those very rare occasions where conditions are such that this is liable to happen.  I'll also experiment with a polarizing filter.

Yeah, me too.  It just looks like a totally weird wrong colored unidentifiable layer of some sort of garbage.  In some of the other frames this time and on the other occasion it was more clear that it was very screwed up areas of densely clustered highlights on flowing water.  In any case the actual scene had nothing that looked like this whatever it is.  I tried all sorts of things removing these green and purple colors (successfully), but even converting the image to greyscale leaves that area looking weird, pixillated and unidentifiable as to what it might actually be.

It's definitely looking like this is a prevention and not a cure situation.  Something I had to go through in the process or racking up my 10,000 hours on the way to technical mastery.  :p

The histogram on the 1Ds is "colorblind" it shows the green channel only. Hence if shooting bright blue or red you will blow those channels without warning in the info view. I recently documented a large stained-glass window in a new church and the blue was 5 stops brighter than the reds. The green channel was below that. My luminosity histogram and exposure went nuts, but when enabling the RGB histogram on the MarkII's it becomes a perfect aid to the correct exposure - although heavy post processing was needed to reproduce what I subjectivly saw. The reds and greens had to be brought several stops up from the shadows.
The contract proofs showed me that trying to reproduce the vivid glow of dense blue glass is almost impossible, and we had to settle for a rather boring interpretation of a great piece of art. The artist was very satisfied though and I couldn't really disagree with that. But it is very important to be aware of the primitive luminosity or green channel histograms limitations, at least when in high altitude or in extreme warm lighted rooms like theaters or night clubs.
MHO