Accused of "manipulating" your shot?

Walter Starck of Golden Dolphin just sent this to the uw-photo list on the subject of Digital Photography Contests. One poster accused all digital shooters of "cheating" by touching up their shots:

With consumer level point and shoot digital cameras manufactures tend to

use relatively high levels of preset in-camera image processing for

saturation, contrast, sharpness, etc.  Although this produces a more

pleasing image on the average than does an un-or less adjusted one it

does not yield as good results as does individual image adjustment

appropriate to the subject matter, conditions and intended use.   As all

such adjustments are not fully reversible, higher level cameras tend to

use lower levels of in-camera adjustment and leave final adjustment to

be applied later.  In short, images from professional level digital

cameras are intended to be adjusted and not to be viewed straight from

the camera.

 

Although better cameras permit such adjustments to be made in camera

before exposure they are generally much better and more efficiently made

after exposure as there is often a certain amount of trial and error

involved and as mentioned above such adjustment is easily applied but

not so easily undone.

 

In professional level digital photography some degree of manipulation

after exposure is an inherent part of the process.  Trying to define it

as unethical and prohibiting it is reminiscent of early rules for

automobiles requiring them to be preceded by a runner ringing a bell.

Closer to the point, if digital had come before film might we now be

proposing restrictions on how film is developed?  Trying to impose the

limitations of old technology on a more advanced one is an oft repeated

mistake.  Discovering how best to utilize the new capabilities is the

more intelligent and productive approach.

 

Rules to prohibit misinforming the viewer make sense.  Ones that

restrict the artist in producing the best image do not. If we were to be

completely honest about this matter we might dispense with the ethical

cloak and frankly consider creating a special Luddites category for

those who don't wish to engage the new technology or who find

satisfaction in doing the best possible with limited tools.

So there can be many takes on this one!

One take can be that contest entrants to the digital shootout that used consumer cameras had an advantage over DSLR shooters because the consumer cameras do the sharpening and saturation adjustment in the camera...hmm... (I'm going to catch some flak for that one).

When I developed the rules for my local UPS digital photo contest I thought it would be best to allow any touching up that can be done by a film shooter in a darkroom. Changing the CONTENT of a photo is not allowed.

Personally I don't think it's a good idea to force people to submit shots right out of the camera. If I did that, my Coolpix 990 shots would beat my Fuji S2 shots every time...

Cheers

James Wiseman

Jim,

Some constests want to judge the skill of the photographer, I have been in contests where the slide canister is serial numbered and you pick you entry straight out of that batch to enter, Some allow darkroom/digital manipulation.

IMHO it all depends on the rules of the contest, abide by those. If you don't like them, don't enter.

Also just to let you know, I'm not trying to flame you, but like you I also have strong opions.

Have a great day and enjoy life

Stephen

So many people are hung up on this concept of "being faithful" and they believe they know what that is. Yet, they happily shoot Velvia to take advantage of its pleasing color "infidelity". For that matter, they use strobes and lenses other than 50mm.

I once entered a video contest (never again) that allowed editing. Some of the entrants whined, believing that the only true way to shoot video is to edit in-camera. Of course, they didn't know any other way. They idol-worshipped pro videographers with decade-old opinions on how to shoot and no experience with the big picture. They just didn't get it.

In literally everything we do we alter the "true" appearance of what we shoot. Lenses, film, lighting, focus, white balance, you name it. It's just that some things are taken for granted while others are called cheating. Digital cameras cook their images to a remarkable degree, so who's to say what is right and wrong. Most of our imagers are monochrome for goodness sake!

We all do this for enjoyment and the pleasure our pictures give us. Photography is "painting with light" after all, not capturing a provably accurate image of an event. We should all embrace the strengths of our medium; not be judged by the antiquated standards and limitations of film. Eventually, film will be viewed as a Rube Goldberg technique (sorry, drcraig).

Contests that limit the creative freedom we have are of no interest to me. For that matter, contests seem to be more about validation than enjoyment and, in my limited experience, have a lousy atmosphere. I like the opportunity to learn and grow that they can provide but I don't like the bickering that I've experienced. That said, I promised to help put one on next year...and it's slide film(!)...and I shoot video (and soon a D100). The horror!

I say, anything goes as long as you play nice and treat the environment with respect. Let the best image win! If only there were good judges.

I think the plot has slipped out for a burger...

This topic led by the Walter Starck letter is highlighting a human frailty rather than providing something useful like a compromise...

There will always be a bias - be it chip users or film users they will always tend to slag off each others format. This solves nothing.

Lets look at some facts here:

Both Housed SLR's & DSLR's offer the photographer almost the same pre-shot manipulation.

However, once the shot is captured on slide film that is it, it cannot be manipulated unless scanned into a PC. Then this image becomes digital too.

It is down to the photographers competence in setting up the shot both asthetically and technically to ensure the correct exposure. He can of course bracket, but so can the DSLR photographer, and therefore there should be no reason to post manipulate prior to judging.

If things don't turn out right in any of the frames it's either the photographer or his camera to blame.

In a competition environment extra manipulation would have to be extended to both formats or disallowed altogether.

This is the dilemma faced by judges everywhere, and why I feel it is very difficult to judge film & Digital in the same criteria unless the above has been recognised.

As Walter points out, the P & S cameras are designed for less creative individuals who just want to take home an image without exercising much of their brain whilst doing so. Yet these cameras still manage to concoct some strange exposures and most newcomers to photography have never used manual so wouldn't know how to rectify metering or lighting difficulties for example.

We could create the same arguement with Amphibious & housed cameras too. Nikonos for example relies almost 100% on the ability of the photographer to take home 'that' shot, where as the housed camera takes care of most or all the calculations for you including focusing.

At the end of the day a photographer can turn any subject into something that provides the viewer a vision of beauty or new and interesting information.

Bob

"Straight up gov'ner - I never touched a thing..."

Good feedgback guys! It's perfectly acceptable to have "strong opinions" here in the Galley.

What I think these different Digital Contest Rules that we keep talking about demonstrate is that the contest organizers really don't have a handle on the "technical details" of digital photography. Heck, I can understand that when it's hard to understand how a camera works it's gotta be DARN HARD to make a contest to fairly judge results.

I think the Bonaire digital shootout rules is/are an excellent example. They understand digital pretty well, but I STILL don't think their rules are 100% rock solid.

Now why would I say something like that?

I believe that DSLR shooters have a DISTINCT disadvantage when it comes to handing in their file "straight as it comes off the camera." Let me ask - does that mean they have to turn in a RAW file and the judges turn it into a tiff or jpeg? No? Well if not, then SOMEONE has to "cook" the RAW file and set white balance, sharpening, tone curves, etc. Hmm... Let's look at some specific examples - the Nikon D100 has had complaints of "unsharp" images. But that's just because the D100 doesn't do agressive sharpening in the camera' software. Let's look at the S2 - it has the equivalent of "digital velvia" in the in-camera software. This makes the red and oranges really pop but sometimes they don't look true. The list goes on.

Each camera company and its software engineers design in-camera software to take a sensor's input and turn it into a Jpeg or TIFF, etc. This goes for ANY camera - whether it be a point-and-shoot or a DSLR.

So here's a kicker - did you know that you can set your little digital camera to up contrast, saturation, and sharpness while it writes a JPEG file? Yep, if the camera doesn't "cook" the shot to your liking, you can actually "overcook" it! I can also do this with my Fuji S2pro.

So where does this leave us? Does it look like I am ridiculing any single camera system, contest, or photographer? I hope not, because that's not my intention. My goal is to point out that digital photography is a fast developing field and I invite anyone putting together a contest to engage the folks here at Wetpixel and perhaps we can help make sure that we all understand the constraints and we are all on the same page.

/end editorial

Chees

James

You're exactly right, and I'll add some further.

All digital cameras do a lot of processing already, so it doesn't matter if it's done in the housing or later on a PC. If a distinction is made, some cameras are penalized as James says. Funny thing that it's the better cameras that are most likely to be penalized.

Even further than that, I don't actually care what comes out of the camera since I don't intend it to be my final product. I'm not saying it doesn't matter but that I intend to work with it, so having a contest that judges results half way through the process is not so interesting. Of course, allowing any processing makes judging difficult. Fortunately, the final results of a contest don't have to be so important and maybe all contests will become strictly fun, as they should be.

I think "digital Velvia" is a real hoot, but my film friends just say "huh?".

...the Nikon D100 has had complaints of "unsharp" images.

This always confuses me with digital photography what is a lens for?

Is the problem associated with incorrect positioning of the chip along the focal plain?

Bob,

Do a google search on "Bayer Pattern" to find out more. Basically, CCD imagers do not see in color. They use a mosiac of (3 needed) 4 pixels to read r g and b then "mix them all together" to give the "true" color.

Since the sensor is layed out in some type of array - horizontal and vertical for the Sony CCD and diagonal close packed for the Fuji, some lines will be "fuzzy."

Example, a black diagonal line of 1 pixel width photographed with the D100 will have little tiny jaggies and will have one dark pixel in the middle with a gray pixel on each side of it.

The camera software takes the raw data from the imager and makes a photo out of it. The same software may use a "filter" to remove those gray pixels using some kind of logic. Other times, the photographer himself may do it on his/her computer using Unsharp Mask.

HTH

James Wiseman

I don't see why you can't have both kinds of photography constests. You just enter the type you prefer. I have entered both and they both can be a wonderful learning experience if you let it be.

I'll pass on an observation from my girl friend about this posts. When she was reading these (plus other here and in other forums) She gets the impression that digital photographers are more interested in manipulating the image after the fact, than trying to use ones skills to capture an image. She is not a photographer (I'm working on that) but this impression has not put a good 'taste' for digital in her mind.

James you've taught me some things I didn't know about the sensors with the DSLR's. Some interesting reading there. I have a nephew thats one of Kodaks software engineers and I'll have to get him to go over it in more detail with me.

Have a great day and enjoy life

Stephen

I'll pass on an observation from my girl friend about this posts. When she was reading these (plus other here and in other forums) She gets the impression that digital photographers are more interested in manipulating the image after the fact, than trying to use ones skills to capture an image. She is not a photographer (I'm working on that) but this impression has not put a good 'taste' for digital in her mind.

Hey, Stephen -

That's really too bad (re: her impression). I have not had that impression from the digital divers I know. Perhaps it was because most of them shot film before turning digital?

I started out shooting digital, and I LOVE it when I open up an image to do color correction, etc., and I find that I have to do nothing except sharpen for output.

However, once the shot is captured on slide film that is it, it cannot be manipulated unless scanned into a PC. Then this image becomes digital too.

 

In a competition environment extra manipulation would have to be extended to both formats or disallowed altogether.

I'm certainly no film expert, but aren't unsharp mask, and dodge & burn techniques invented in real darkrooms. One can certainly crop and some color and brightness adjustments can be made in the process of making duplicate slides.

One thing we are missing here guys is

"How long are these competitions going to last?"

From a competition and judging point of view:

I think at the end of the day the simple answer would be to scan slides onto a CDROM at high res. Then all entries are on equal terms.

How practical this is with shoot-outs depends on the facilities at hand for processing slides with a scanner and laptop waiting in the car.

However, this turns a simple photographic competition into something quite different.

99.9% of photographers using slides send them to a commercial lab and can have no say in the original developement and only cropping in the dupes.

if you are in a competition arena and you find a nice subject you would bang off at least a dozen assorted and bracketed shots to allow a better chance of at least one successful exposure.

Ironically whenever I have done this the best shot is always the first!!!

Thanks for the explanation James... It is diabolical really selling the stuff as an imaging device if it doesn't return the goods without manipulation...

It's like selling a lens with the rear element missing. Correct viewing of the image done by looking through a lens after processing.

I think the way to go here is to develope a chip based on the foveon technology but arranged into a cone/hemisphere allowing alternate pairings of colour receptors. The rear of a standard lens adapted by adding a semicircular corrector lens . This would provide almost 3D segregation of at least 4 primary colours working just like the human eye.

Oh shoot, I thought I was back at work for a moment!

Having an interest in "manipulating" your shots does not mean you can't shoot and aren't interested in learning how. Furthermore, you don't have to shoot film in order have good technique nor should you have to swear off editing in order to be considered a real photographer. I know how to operate my camera despite the fact that I edit.

Clearly, garbage-in-garbage-out still applies. Some people will always believe that proper photography means only using the techniques of our forefathers. For them I think film is an excellent choice (and so do they).

For potential new photographers it's more important to understand what you want to accomplish than to fear becoming associated with a bunch of lazy underachievers. Just remember, decades from now there will still be serious, respected photographers only they won't be shooting film.

Hi gang,

Bob, you said:

Thanks for the explanation James... It is diabolical really selling the stuff as an imaging device if it doesn't return the goods without manipulation...

It's like selling a lens with the rear element missing. Correct viewing of the image done by looking through a lens after processing.

I'm sorry I gave that impression. It is NOT the case. Digital SLR camera's DO "return the goods" and better than anything else out there. I just want to make the point that each camera's SOFTWARE processes the shot differently. What they do allow the photographer is MORE CONTROL which is something we all want, right? You have the option of "letting the camera decide" or you can make a jpeg yourself.

That's why we all shoot manual right? We want more control of our tool.

Cheers

James

Thanks for the explanation James... It is diabolical really selling the stuff as an imaging device if it doesn't return the goods without manipulation...

It's like selling a lens with the rear element missing. Correct viewing of the image done by looking through a lens after processing.

The same can be said of slides. You have to send it off for manipulation by the developer, and I've seen enough to know that you can have the exposure pushed a stop or so in either direction or settle for the default. In the digital case, you can choose one of the fixed camera setting, one of the settings in the pc software or choose to intervene.

I think the way to go here is to develope a chip based on the foveon technology but arranged into a cone/hemisphere allowing alternate pairings of colour receptors. The rear of a standard lens adapted by adding a semicircular corrector lens . This would provide almost 3D segregation of at least 4 primary colours working just like the human eye.

Maybe one day someone will invent a good sensor that will do Red, Green, and Blue in the same location, but I suspect that the Foveon technology is not it. Now that there are cameras out in the field, some of the problems that I've expected them to have are beginning to surface, the biggest being a noisy red channel. See the review at imaging-resource.com . I think this is a basic flaw of the technology that can't be easily fixed.

The reason is very technical, so skip this if you don't want to read lots of engineering stuff. The Foveon sensor separates the light input into R, G, B channels by the depth at which the light is detected in the Silicon. It is a property of the material that shorter wavelengths (blue) are absorbed sooner than longer wavelengths (red). However, this relationship is only statistical, that is the shorter wavelength light has a higher chance of being absorbed at any time; therefore, it's more likely to show up at the top of the stack. This is ok if there's lots and lots of light. However, in low light conditions, the variations in the number detected at the bottom (red) will be significantly larger than the number hitting the pixel, leading to a noisier red.

One of the interesting things is that violet light needs to be represented as a combination of red and blue, so the shortest wavelength light needs to produce a signal in the long wavelength channel. I don't know how this works in Foveon or in a 3CCD video imager, but I do understand how it works in the eye. I think the diagrams we see on these things are oversimplifications of how things actually are and I'd love to hear how this actually works. In other words, the simple diagrams we see have no way of differentialing violet from blue (kinda like my dad).

While you may be right, its pretty early to say that the red noise problem can't be overcome with Foveon. 3CCD cameras are knocked for their poor light sensitivity but it doesn't make them uncompetitive. The subtractive nature of the technology puts Foveon at a light sensitivity disadvantage, but that may be moot in the long run. My 3CCD camera has great light sensitivity, after all.

One of the interesting things is that violet light needs to be represented as a combination of red and blue, so the shortest wavelength light needs to produce a signal in the long wavelength channel.  I don't know how this works in Foveon or in a 3CCD video imager, but I do understand how it works in the eye.  I think the diagrams we see on these things are oversimplifications of how things actually are and I'd love to hear how this actually works.  In other words, the simple diagrams we see have no way of differentialing violet from blue (kinda like my dad).

 

While you may be right, its pretty early to say that the red noise problem can't be overcome with Foveon.  3CCD cameras are knocked for their poor light sensitivity but it doesn't make them uncompetitive.  The subtractive nature of the technology puts Foveon at a light sensitivity disadvantage, but that may be moot in the long run.  My 3CCD camera has great light sensitivity, after all.

It's funny you should mention the eye because it works alot like Bayer sensors with its spatially seperated R G B cones. As a raw sensor the eye does not have good resolution and range. The remarkable vision that we have is due to scanning and reconstruction by the firmware in the brain.

I suspect having more noise will make the Foveon uncompetitive at the high end. It's not just the sensitivity, which is less important in still UW photography than most other applications, but the increased noise also has the net effect of reducing the total dynamic range. That's really what people who put up big $$ for cameras with the larger sensors are paying for.

I've never used a video camera, and don't know much about them. I think the 3 CCD's are for Red Green and Blue which gets separated by a prism at the front, like the first camera that Foveon made. If this is the case then video cams should be more sensitive than still cameras because there are no color filter that cut out 2/3 of the light on the sensors. All the light gets to the Silicon.

What I thinking about the eye is that the "red sensor" has very broad response and the blue has very narrow response. To sense violet, or any other color, the brain has to do a lot of processing since the "sensors" aren't orthogonal. I wouldn't describe the eye as using a Bayer pattern at all. Cones have distinct wavelength biases.

I know that 3CCD cameras have 3 imagers and a prism. The classic view is that light is separated into red, green and blue and sent to the 3 sensors. That isn't how it works, though. Yellow light is not a mix of red and green, it is its own frequency but it stimulates both the red and green sensors in roughly equal amounts. When reproduced, yellow will be represented by a mixture of red and green and will appear to be right because of the limitations of the eye.

Violet is interesting because red and blue are not adjacent in the spectrum, so simple bandpass filters that you see described won't work.

I agree that more noise would be a problem for Foveon or any other sensor, I just think its premature to say that Foveon's design is inherently noisy. I also wouldn't say that sensitivity it less important in UW still. Perhaps for macro. Good sensitivity implies low noise.

3CCD video imagers do have a filter only it's called a prism. The color filters in a 1CCD imager pass nearly all the light in the desired band. The 2/3 they stop are the portions they are trying to stop. 1CCD cameras have better light sensitivity that 3CCD ones because their filters are more efficient than the prisms are. Their color performance is not as good, though.

The Bayer CCD's does work in a way similar to the eye. The first plot below is our best estimate of what the eye does and the second is taken from a Kodac CCD spec. Ignor the part above 700nm on the second one, that's inferred and out of visual range as the first plot shows. So the eye and Bayer CCD both use 3 bandpassed filters to separate the colors and each cone or pixel has a specific bandpass response. The reason this works and why we can see violet, both with our eyes and using a Bayer CCD camera is that the filters overlap. Violet is actually pure blue to the sensors and blue is a combination of blue and green. The maping in the brain and a camera is similar with a different offset due to the different responses.

You are right that the 3 chip CCD cannot work by using a prism to split the colors as that will produce no overlap. I should have quit after I wrote I don't know how video cams work I just had a conversation with some here at work that know more about this stuff than I do. She said that the video camera makers don't publish how they do this, but most people in that field thinks that the light splitter in front just produces 3 identical beams and the filtering is done after the spilt with filters responses similar to those of the CCD filters.