What makes digital identifiably digital?

fraley

fraley

Beware of Claws
Local time
10:13 AM
Joined
Sep 6, 2004
Messages
516
What makes a digital photo recognizably digital? Can you articulate the identifying characteristics of a digital photo as opposed to a film photo that has been scanned? Arthury has an excellent posting in the gallery:

http://www.rangefinderforum.com/photopost/showphoto.php?photo=56072&cpage=1#poststart

I admired it and played my usual guessing game, is it digital/is it film? I can usually identify a digital posting. But I can't put in words the quality that I'm responding to. Anyone else wonder about this?
 
It's an M8 and therefore no comment. :) I think these people are very sensitive. :D

No... just kidding Arthur Y is cool. I know him at LUF.

You need to own and seriously shoot a digital camera and likewise seriously shoot film and develop your own. Then you can tell one from the other.

I'm sure Arthur can tell one from the other. He's been shootimg M for a while.

Best,

-Ron
 
it's tough to compare scanned film images vs. digital capture, but depending on the speed of the film, can make it easy to spot the film scan. Grain is the number one thing I look for. In areas of constant tone or smooth gradients, especially lighter areas, the total lack of any grain at all tells me it wasn't shot with 400ISO film :)

Another way, using your posted link as an example, is in the total lack of a gradient once a certain level of black is reached.

See this shot for an example of the smooth gradient from light to dark, and the wide range of "blacks":
http://www.rangefinderforum.com/photopost/showphoto.php?photo=46398
 
The single biggest factor in being able to identify a digital image as digital is bad post processing. Oversharpening being the most common problem.
 
I don't know how to describe it, BUT I know it when I see it :D

ask an easier question like the difference between records and CDs now that one I CAN answer :cool:
 
i find it pretty easy to be honest..

digital cameras have shorter dinamic ranges > like someone said, the gradiants are much pure ( lack of grain ) but also very short - in a black and white photograph its easy to spot, there no so many amount of blacks or even whites. the range is shorter.

in colour its even easier > the total absence of grain, or the brutal amount of sharpening, over saturated colours, over contrasty images and the overuse of layers to boost dinamic range > sometimes its nicer to see a HDR image but 99% of it just looks unatural.. the human eye have a dynamic range as well and can spot where a scene wouldnt be interpreted as it sees > pick black shadows full of details and overblown areas with lots of details. ahn ahn. not going to happen.

usually the graduations are rough - they end abrubptly. not in a smooth manner like film. and to end, if the image is not oversharp ( which usualyl is ), its over smooth to hide high iso noise, which its another tip - or sometimes the faces have baby skin face where pepople usually dont have, or long colours have super smooth gradients.

easy to spot, not easy to describe though. So far only one digital camera ( I've used more than a dozen professionnal SLRs ) could fool me from time to time > A Canon 1Ds, version 1. beast digital sensor Ive ever seen. Overexpose a bit because the shadow noise will bit ya hard. But it renders film like colours, contrast and quality.
 
I find it very difficult to tell a film vs digital image on the net. Both images are digital and the only way I notice is when I see a printed image of reasonable size. It starts to show on printed media around 5x7 and by 8x10 i'm seeing a difference.

Also I think most of us are comparing digital cameras to puny 135 mm film cameras.... when the 120 is pulled out and printed.. I can tell without a close inspection. Digital still has a long way to go. But would note that people are accepting lesser quality in images; so the detail in film may be of little importance any longer.

I think I'm repeating myself from earlier posts here.
 
>i find it pretty easy to be honest..

Poppycock. I could put particular 11x17" prints from an Epson R-D1 RAW file in front of you and tell you it was from a 35mm film scan and you'd never know I was lying to you.

Depends on the print and person making it...
 
AusDLK said:
>i find it pretty easy to be honest..

Poppycock. I could put particular 11x17" prints from an Epson R-D1 RAW file in front of you and tell you it was from a 35mm film scan and you'd never know I was lying to you.

Depends on the print and person making it...
Click to expand...

Quite so.

Digital has a shorter dynamic range? Depends entirely on how the shot was taken and the post processing. I can give you a print from a film neg that has only a 4 stop range, simply because I shrunk the dynamic range to achieve a certain effect.

It isn't always as easy to distinguish between a pro print from a digital file or a film neg. When it does become easy is when you compare a 1-hr lab print from a 3.2 MPix digicam to a wet print of a properly scanned neg. But then we're comparing apples and oranges, aren't we?
 
It's always easy to tell the difference between my digital shots and my scanned film images because the digital ones haven't got any dust or hairs on them :)

But seriously, I was looking at some of my digital shots recently and it struck me that the colours are just not right - too vivid and saturated sometimes, at other times looking like artificially coloured black and white photos. However, my digital camera is a long way from the leading edge of the technology (I only use it for record shots). With nice digital images shot on good cameras I can't tell for sure whether or not they're film.
 
AusDLK said:
>i find it pretty easy to be honest..

Poppycock. I could put particular 11x17" prints from an Epson R-D1 RAW file in front of you and tell you it was from a 35mm film scan and you'd never know I was lying to you.

Depends on the print and person making it...
Click to expand...

poppycock yourself. You are talking about two digitally produced prints, which is, besides being an asinine example, not the comparison being discussed by the poster you are belittling :/

Let's be fair - there are certain films that have an unmistakable fingerprint. Scans from them are rarely confused for digital captures. On the other hand, there are also films that can produce scans that could easily be confused for digital origin because the grain is close to non-existent. By the same token, there are images taken digitally that are clearly a result of digital imaging, and also images that don't show their digital heritage so clearly. But arguing that there is no way a person could tell the difference simply because you can't is just rude.

Once you know what to look for, it can be rather easy to identify digital captures vs. film scans when comparing images on the internet. But obviously there can be cases that are not so easy, because the film scan was manipulated or the digital capture was processed with that intent, or because of the image itself. I'm not sure what that proves.
 
Last edited:
Hm, a JPEG file is RGB with 256 values for each colour, the source does not make a difference as long as it was capable of recording this. Somethimes I can pick out certain characteristics In shots I took myself.

What proenca wrote about wrinkles makes sense to me, I think I could find differences if I would be looking for something like that, but not in a small JPEG file on a webpage :)

And I'm pretty sure I can fool proenca with a 8x10 Frontier print from XP2 and a FB print from a 6Mpixel dSLR printed on real RB paper with a digital enlarger.
http://www.polycolor.de/Fotolabor/digital%20fiber%20base%20prints.html
 
Hm, a JPEG file is RGB with 256 values for each colour, the source does not make a difference as long as it was capable of recording this.
Click to expand...
It's not just the number of levels that a medium is capable of producing, because the distribution of values within each 256-level colour channel also makes a difference. For example, an image that contains a linear spread of colour values in relation to the original source will look quite different to an image that has most of its colour values compressed within the middle of the range with much poorer linear resolution at the ends of the scale (ie one that reproduces a linear scale accurately will look different to one that reproduces a linear scale as an S-shaped curve) .

Best,
Alan
 
A printed digital photo and a normal film photo, side-by-side, are easily distinguished. The digital photo will lack detail in either the shadows or the highlights, it won't have both (unless the photographer decided to dabble).

Online, it's not as easy due to the film photo having been converted into a digital version of itself.
 
Socke said:
And I'm pretty sure I can fool proenca with a 8x10 Frontier print from XP2 and a FB print from a 6Mpixel dSLR printed on real RB paper with a digital enlarger.
http://www.polycolor.de/Fotolabor/digital%20fiber%20base%20prints.html
Click to expand...

No, because while you can probably make a print that is hard to tell if it is film or not, you can't make a digital print that is "unmistakably film" without making an effort to make it look like film and not digital.

The bottom line is you can use film to make a print that looks like it could have been shot with digital, but you can't use a digital camera to take a shot that looks like Tri-X pushed to 1600. You might be able to add something in post-processing to simulate grain, or overlay grain from a film sample, but that proves the point right there when you have to take extra steps specifically to mimic something else to make it even half-believable. At the same time, I don't know of any film that looks as bad as high ISO digital :)
 
40oz said:
but you can't use a digital camera to take a shot that looks like Tri-X pushed to 1600.
Click to expand...

That's why I start with XP2 :), digital print from film on whatever they put into the Frontier and wet print from digital on heavy FB paper, take both in your hand and you're half fooled because you expect cheap paper from digital and heavy FB from film.

You might be able to add something in post-processing to simulate grain, or overlay grain from a film sample, but that proves the point right there when you have to take extra steps specifically to mimic something else to make it even half-believable. At the same time, I don't know of any film that looks as bad as high ISO digital :)
Click to expand...


Oh I know a lot of films which look very bad compared to digital, like E400 pushed one stop to ISO800 from a Canon 1D MkII, even at ISO1600 the Canon files look much better than the ISO800 C41 films I know, i.e. Fujipress 800 and Kodak Royal 800 (got two sample rolls at Photokina '02, never used it again).

In those cases I see a difference, but the people I know who shoot ISO800 and higher do that for faster shutter speeds and want a picture which doesn't show grain. PJs mostly and they consider grain (and noise) a bad thing.
 
It's the tonal range and special look of film that keeps me shooting it, even though I enjoy digital photography. Can't get the same kind of B&W from digital though I often prefer digi shots for colour unless we're talking medium format.

Gene
 
The high end digital is very comparable with high end film as long as the enlargment does not stretch pixels nor grain. What people 'spot' easily is poor digital and/or poor film images (ok, some 'poor' film images are made for specific effect -- grain).

If I could afford to capture digital images with my manual cameras I would. ;)
 
You must log in or register to reply here.
Back
Top