When does a photo cease to be a photo and turn into CGI?

Gary...LFC said:
https://en.m.wikipedia.org/wiki/Computer-generated_imagery

If you're going to be the grammar police on this thread please don't post on it.
Click to expand...

Gary...LFC said:
I mean photos with depth that "pop" off the page, not they just pop out of the camera with ease.
Click to expand...

You've really got to slow down and read a thread.

There's no point in having 'an idea' for a discussion and then just looking for aggro where non exists. You missed 'the point' on both those occasions.

Back to your original question:

As a few people said, it's a recurring theme, and you wouldn't believe how often it's someone's first post :thinking:

I can understand the confusion, but the underlying fact is what you're seeing as 'CGI'* and we'd call digital art, is when it's done badly or overdone. The example you gave of Dave's is proof in point, he probably spent as long in PP on that image as the other 'photographer' with the composited sky. The difference is one is done well, the other quite crudely.

*CGI is computer generated, as opposed to being generated by humans with computers, and again that's not 'Grammar police' but it's helpful to be specific with terminology when discussing the details. we understand what you mean, but only after you added further detail.
 
Phil sums it up very well (as usual); CGI is generated by a computer and it is, generally speaking, a 3D model of the world, it is not just some 2D images. That is why you can move around in a video game for example and see things like the backs of buildings. Yes someone defined the shape of the building and the materials it is made from but the surface textures, lighting, point-of-view, etc. are all rendered by a computer. The thing you see on screen does not exist as an image until it is rendered.

If the question is “what amount of editing changes a photograph into something that is not a photo” then I think that is in the eye of the beholder. With the current state of technology I would say if it looks to you like a convincing, realistic scene then you can regard it as a photo. The problem with this definition is that I don’t think we are too far away from the point where technology can render photo-realistic scenes so would that make CGI images photos?

Also it seems to me that the “painting with light” definition breaks down at that point. Years (decades) ago there was a TV programme where David Hockney did some paintings with Quantel Paintbox (an early computer paint application) and he described it as “painting with light. Computer monitors are light boxes and we can paint the light they produce, bring that together with photo-realistic rendering and I think we may be left scratching around for a new definition for what we do with our cameras.
 
Phil V said:
You've really got to slow down and read a thread.

There's no point in having 'an idea' for a discussion and then just looking for aggro where non exists. You missed 'the point' on both those occasions.

Back to your original question:

As a few people said, it's a recurring theme, and you wouldn't believe how often it's someone's first post :thinking:

I can understand the confusion, but the underlying fact is what you're seeing as 'CGI'* and we'd call digital art, is when it's done badly or overdone. The example you gave of Dave's is proof in point, he probably spent as long in PP on that image as the other 'photographer' with the composited sky. The difference is one is done well, the other quite crudely.

*CGI is computer generated, as opposed to being generated by humans with computers, and again that's not 'Grammar police' but it's helpful to be specific with terminology when discussing the details. we understand what you mean, but only after you added further detail.
Click to expand...

It probably took me longer as I am a bit of a fiddler :)

If I had the tripod with me I would have been tempted to get a long exposure on the sky as well (separate image, but same position for camera) I would then have looked to add the sky in to give me multiple layers of still / movement.

Photography for me is about trying different things and keeping doing the parts I enjoy :)

I know from experience that time blended landscapes take a lot more effort / time than the single shots I take. Just figuring out a composition takes me ages.

Dave.
 
sirch said:
If the question is “what amount of editing changes a photograph into something that is not a photo” then I think that is in the eye of the beholder. With the current state of technology I would say if it looks to you like a convincing, realistic scene then you can regard it as a photo. The problem with this definition is that I don’t think we are too far away from the point where technology can render photo-realistic scenes so would that make CGI images photos?
Click to expand...

Oh, we are already well beyond that point, especially if you exclude people from the equation.

For skilled architectural renderings, the remaining give away these days is that the CGI is probably better lit than any real photograph. :)

e2a: For example

http://www.lunas.pro/3d-rendering/interior-visualization-concept-loft-concrete-kitchen
 
Last edited:
Musicman said:
Oh, we are already well beyond that point, especially if you exclude people from the equation.

For skilled architectural renderings, the remaining give away these days is that the CGI is probably better lit than any real photograph. :)

e2a: For example

http://www.lunas.pro/3d-rendering/interior-visualization-concept-loft-concrete-kitchen
Click to expand...

Fair enough, so playing devil’s advocate are those examples photographs? After all if it walks like a duck and quacks like a duck…
 
chris malcolm said:
I can give myself more latitude in processing the image afterwards if I deliberately shoot with settings which increase the processing latitude. Doing that results in images which straight out of the camera don't look quite as good.
Click to expand...

I'd be very interested to learn what you mean by this - please!
 
sirch said:
Fair enough, so playing devil’s advocate are those examples photographs? After all if it walks like a duck and quacks like a duck…
Click to expand...

Quick take: CGI imagery like that is a synthetic photograph, much as we have synthetic motor oils or diamonds. They may be indistinguishable from the 'real' thing and serve many of their purposes (for which they may be better in many cases) but we treat them differently at a conceptual level.
 
Last edited:
Musicman said:
Quick take: CGI imagery like that is a synthetic photograph, much as we have synthetic motor oils or diamonds. They may be indistinguishable from the 'real' thing and serve many of their purposes (for which they may be better in many cases) but we treat them differently at a conceptual level.
Click to expand...
Indeed!

That CGI image looks more 'real' than many 'photographs' I see here, including the double exposure with the mountains in this thread.

Turning the OP's assumption on its head I suppose o_O
 
Midland Red said:
I'd be very interested to learn what you mean by this - please!
Click to expand...

I usually shoot RAW + JPEG, and only go back to process from the RAW when the changes required are too much for the JPEG. JPEGs have settings like white balance and sharpening already "cooked in" which gives limited latitude to changing those, plus they have a lot less dynamic range. Turning the sharpness down, the colour saturation down, and the contrast down, effectively increases the amount of the original RAW dynamic range represented in the JPEG. That increases the amount of change of white balance, shadow lifting, etc. which can be done to the JPEG, the final changes of course being to restore the contrast, saturation, and sharpening to the usual levels. Without any processing, however, the JPEG images look dull, dim, and flat. Since processing from RAW (in my particular workflow) takes me a lot longer, then anything which can reduce the number of times I have to go back to the RAW saves me image processing time.

In cases where composing a shot carefully and getting the verticals really vertical etc. takes too much time, it's easier to shoot a little wide to leave enough latitude for straightening up and carefull cropping.
 
Phil V said:
Indeed!

That CGI image looks more 'real' than many 'photographs' I see here, including the double exposure with the mountains in this thread.

Turning the OP's assumption on its head I suppose o_O
Click to expand...

Should I throw a(nother) spanner in the works and mention architectural visualisations which are a composite of real photography and CGI elements, usually the new building? These are particularly common in external / aerial views.

[emoji48]
 
chris malcolm said:
I usually shoot RAW + JPEG, and only go back to process from the RAW when the changes required are too much for the JPEG. JPEGs have settings like white balance and sharpening already "cooked in" which gives limited latitude to changing those, plus they have a lot less dynamic range. Turning the sharpness down, the colour saturation down, and the contrast down, effectively increases the amount of the original RAW dynamic range represented in the JPEG. That increases the amount of change of white balance, shadow lifting, etc. which can be done to the JPEG, the final changes of course being to restore the contrast, saturation, and sharpening to the usual levels. Without any processing, however, the JPEG images look dull, dim, and flat. Since processing from RAW (in my particular workflow) takes me a lot longer, then anything which can reduce the number of times I have to go back to the RAW saves me image processing time.

In cases where composing a shot carefully and getting the verticals really vertical etc. takes too much time, it's easier to shoot a little wide to leave enough latitude for straightening up and carefull cropping.
Click to expand...

Thanks, Chris - very informative
I shall change camera settings for sharpness and saturation, and see how that affects processing jpeg images
Cheers!
 
This might be slightly off topic but there is also the philosophical question as to when does a photo become not your photo. I am thinking about presets where someone else designs a preset / filter algorythm, you apply it which changes the image. Of course it can always be replicated with sliders in software however if you simply rely on someone else slider settings are they putting their style to make a finished product...

And yes there is also the debate with digital that if you select say the image setting in camera to vivid or something else its using the camera manufacturers algorithm of image correction....

Discuss....
 
Are you saying all my ektachrome slides belongs to Kodak and Fuji owns the Velvia ones ? ;)
I would say its not so much about the tools, more the idea
 
Last edited by a moderator:
jamesev said:
This might be slightly off topic but there is also the philosophical question as to when does a photo become not your photo. I am thinking about presets where someone else designs a preset / filter algorythm, you apply it which changes the image. Of course it can always be replicated with sliders in software however if you simply rely on someone else slider settings are they putting their style to make a finished product...

And yes there is also the debate with digital that if you select say the image setting in camera to vivid or something else its using the camera manufacturers algorithm of image correction....

Discuss....
Click to expand...

I don't think it matters who did what. In the days of wet processing different labs may well produce different looking results from the same negatives and different film types would produce different “looks” to the finished print.


I am beginning to think that a reasonable definition of what we mean by photograph is a rendition based on light captured from a scene where the composition largely* represents the original physical scene.

I.e. As long as all or nearly all of the components of the photograph were captured rather than drawn, painted or generated then it is a photograph. I think we have to allow for “nearly all” to accommodate enhancements like spot removal where the spot overlay is computer generated. This also allows for things like replacing the sky with a different sky (becasue the sky is still in the same place compositionally) but not say moving a mountain from the centre to the edge or editing in a flock of flamingos.


*Allowance needs to be made for techniques like IR, false colour astro and ICM; all of which may not look realistically like the original but are renditions of the captured light.
 
Gary...LFC said:
... rather than a moment in time captured on a camera.
Click to expand...

Regarding retouching...

"Capturing a moment in time" is fine if you're taking casual shots, maybe even street photography. Those shots don't need retouching. For commercial work where the photographer isn't trying to capture a moment in time, images need to be retouched. They need to be faultless and eye popping to create perfection which, 99% of the time, simply cannot be achieved in camera.

I shoot in RAW, so by default, the resulting data files I capture are flat, desaturated and bland in appearance. They need to be enhanced to either match the scene I was photographing, or to take them a level beyond.

I'm an interior and architectural photographer. If I present two images to my clients, one straight from camera (perfectly executed) and a second which has been carefully retouched (often blending multiple ambient and flash exposures), they will always take the second image.

Regarding over-processing...

There are two sides to over-processing an image. Handling it correctly and handling it badly.

This guy I follow on Instagram technically over-processes his images, but does it very well:
https://www.instagram.com/jr_ffm/

Over processing an image poorly is a trait commonplace with a lot of beginner or amateur photographers - much like putting a sock down your pants, they're covering up shortfallings or are lacking confidence in their photos. Grunge HDR, over processing of landscapes, ridiculous levels of saturation etc etc. I was guilty of this in my early years. Most of us are.

Bottom line is, if executed correctly, any processing or 'over-processing' will look great. If handled badly, it'll look sh*t. If you find the processing of an image frustrating, distracting or over the top, then it's been executed badly. You can't tarnish all processing with the same brush.


Craigus said:
HDR in general tends to look 'false' in my opinion.
Click to expand...

If HDR is executed correctly, you can't tell it's HDR. I presume your referring to the over processed grunge look that many amateurs use.
 
Last edited:
Oliver Pohlmann said:
Regarding retouching...

"Capturing a moment in time" is fine if you're taking casual shots, maybe even street photography. Those shots don't need retouching. For commercial work where the photographer isn't trying to capture a moment in time, images need to be retouched. They need to be faultless and eye popping to create perfection which, 99% of the time, simply cannot be achieved in camera.

I shoot in RAW, so by default, the resulting data files I capture are flat, desaturated and bland in appearance. They need to be enhanced to either match the scene I was photographing, or to take them a level beyond.

I'm an interior and architectural photographer. If I present two images to my clients, one straight from camera (perfectly executed) and a second which has been carefully retouched (often blending multiple ambient and flash exposures), they will always take the second image.

.
Click to expand...
Very well explained!
 
It's not how much post processing is done, but what is done. If the processing inserts or removes objects from the scene, or inserts or removes implied objects outside the scene, such as things casting shadows or shade into the scene, or light sources, then the image is no longer a record of what could have been seen at the time by an observer. That means a suitably equipped observer, of course, because long lenses can see details beyond the resolution of unaided human vision, and long exposures or flash can show details too dim for unassisted eyesight.

Is the observer of a scene allowed to make temporary changes, such as moving a traffic cone out of the scene, and then putting it back afterwards? Is the photographer allowed to remove the traffic cone from the image by computer, because he hadn't the time to move it when taking the photograph? That's just a question of what someone decides are the rules, or decides how the rules are to be interpreted. I don't think anyone would object to someone telling a sketch artist to do a realistic sketch of their newly installed kitchen sink and cooker layout, but to ignore the debris. Therefore I see no reason for someone to object to a photographer removing unimportant debris from an image by computer processing.
 
I'm going to refer to something I wrote a few days ago in the latest Colour vs Monochrome thread.

Me said:
I think we must draw a distinction between the subject and the purposes of a photograph; that is the intent in taking it and the uses to which it will be put.
Click to expand...

Any debate about what makes a photograph a photograph and what is and what is not acceptable must be framed within these terms of reference. Why was the photograph taken what are your other people going to be doing with it?

If you're a photojournalist, intending your exclusive photo of Donald Trump being entertained by two blondes in a hotel room in Moscow to be published on the front page of the New York Times Daily News, then the authenticity of the image is probably paramount. Adding a few novelty items on the bedside table is not going to do you any favours if anyone finds out they exist only in an extra layer of a Photoshop file.

Heck, just paying a random guy in an orange wig to be in the hotel room because Donald was actually playing golf in Florida at the time is going to get you into trouble, even if the NY Daily News publishes your unedited original JPEG and no digital jiggery pokery occurred.

Equally, if your spiffy photograph of a Komodo Dragon from a very low angle is going to be used for the front cover of a new edition of a 1950s pulp SciFi novel, dropping a 3D model of a nuclear power station into the background, turning the sky a sickly yellow-green and giving the lizard red and black stripes isn't likely to be that much of a problem.

Each purpose has its own rules. Some of them are clearly codified in Codes of Ethics or the Terms and Conditions of a competition. Outside of those more formal settings, it shades toward the individual photographer to draw their own lines in the sand: "I'll edit my digital originals within an inch of their lives, but the scans of my Kodachromes stay exactly as they came out of the Coolscan" may be a perfectly reasonable position to take.

There are 1024 shades of grey between what's acceptable in one context and what's not in another and there are probably more opinions than that, which is why questions like this keep coming up again and again.
 
Last edited:
You must log in or register to reply here.
Back
Top