You qualify Clarkvision site as utter nonsense, although you don't provide anythink than backups your claim. You could use a drum scan to get 5% (?) more resolution, or you could even stich microscopic captures to get the last bit of resolution of those exquisite amorphous grains, just to realize that the results wouldn't be significantly better that those from figure 5 at Clarkvision page. You could even try a different numerical approach to compare DR between film and digital, even one that show film has better DR, as claimed by someone informed on the matter, and then you would have a hard time trying to explain why that Velvia50 sample at figure 5 looks so bad.
Pablo, let me put aside that stupid OT discussion of ours. I still consider your insights into the main topic as the best contribution.
I also think that 3D effect is related to human perception. Although we see everything in focus, our vision is sensible to tiny diferences in contrast (or sharpness, microcontrast...) so we can perceive changes in texture as the subjects receed or aproach into the space, and this gives our brain depth clues. In fact, shallow focus in photos acts the same way, although exagerating this effect. Of course a good lens can contribute this effect, although I don't think it is the main factor, provided those differences are present in an image even if captured with a mediocre lens. Enlarging or reducing size, or intended viewing distance, can contribute by enhacing differences, or can detract, by destroying such differences. The same way, post-processing (sharpening). You can enhance appearence by selectively sharpening (or blurring) to make things to look more natural, as we see them, or you can oversharpen overall (or blur) and then you destroy the tiny contrast nuances.
I won't belabor the point, except that CCD scanners other than extremely high end Imacon ones have but a fraction of the Dmax and Drange of drum scanners and true densitometers. True densitometers are the only way to actually measure an exposure-response relationship on film, because they can encompass the entire film density range, including the baseline film base / fog density. If you PM me I can point you to some real expertise on the matter. As to the MTF, I think in the end we're saying largely the same thing, because at the level of fine detail resolution contrast and resolution are equivalent concepts.
I agree with everything you've said in the second post, except that in a flat photograph it's not human perception that creates the 3D effect. It's that the photo displays things that in the real world we use as textural and depth cues. The 3D-looking photographs remind us of 3D things, in other words.
StevenPA wrote:
Oh yes, and I think the kit lens shot (flower) I posted earlier speaks to this point.
Yes, and it's frequently done in photoshop. It seems to need the combination of clarity (looks real) and clear spatial information: The penny in focus is believed to be closer than the OOF pennies. We do not consider that paul may have made a giant penny. Or that the shadow is cast by an object out of view. We are extremely adept at processing visual information without our conscious mind.
I've always considered that a major factor in producing a 3D look is having the background as OOF as possible while still being identifiable.
A related subject to consider is why shadows are considered (visually) a break or division in a photograph, but not so in our 3D world.
Jonas B wrote:
Great discussion here, a joy to read a new chapter every evening!
Can it be correct to say that "per-pixel clarity" = microcontrast? It would fit fine for my brain and the way I think of the word microcontrast. Is there something I am missing?
I posted the above back at page 26. It would be interesting to hear you educated and well-tempered brains comment on this. Why as again? Because I would prefer not to have to add another definition to the long list of "must-knows" we allready have here!
It's till an amazing thread this: I have been away for a couple of days and got not only one, but two new chapters that I first (now) have skimmed and will reread...
A friend of mine, at another forum, posted this some time ago (roentarre at Pentaxworld.com and all credits go to him):
It's a simple effect, it can be reproduced by anyone walking by at that place at the right time. I looked at it for a while as the popping effect is immediate. I wonder if the 3D effect here is so pronounced thanks to good micro contrast, ordinary contrast and resolution but a good relation between the distance to the foreground and ditto background, or if there is another main single reason for it to be so pronounced here. Any thoughts are welcome!
Selective focus definately contributes to the so called "3D" effect in both Andi's shot and the images linked and many of the linked images also just happened to show warm colored flowers(which comes forward) with selective focus and minimal DOF.
It's in the lighting more than anything, IMO. Lenses can have better contrast and slightly higher resolution which will help, but in the end if the light is flat it's flat.
Some of you may have seen this image before so for you I apologize for the redundancy!
This picture of the Horseshoe Bend Island which Brett displayed so well in full view earlier in the thread was taken with the Leica M8 and the 28mm Summicron. I believe the 3D look here is due almost exclusively to the characteristics of the lens. It was tripod mounted and at an aperture of f11 set to get the maximum DOF to get both the bush and the far away walls of the Colorado River Canyon in great focus detail. I love it but YMMV
Funny, I see it almost entirely due to near/far relationships of objects and lighting with the darker less detailed background and cooler(which recedes) lower left 1/3 river juxtoposed to the brighter, reder(which comes forward) closer and more detailed right cliff. Nothing at all to do with an M8 or any other camera. Any decent lens of the same comparitive focal length would have given the same effect in identical conditions imho. You also have some leading lines and shapes at play in this photo.
Tariq Gibran wrote:
Funny, I see it almost entirely due to near/far relationships of objects and lighting with the darker less detailed background and cooler(which recedes) lower left 1/3 river juxtoposed to the brighter, reder(which comes forward) closer and more detailed right cliff. Nothing at all to do with an M8 or any other camera. Any decent lens of the same comparitive focal length would have given the same effect in identical conditions imho. You also have some leading lines and shapes at play in this photo.
Tariq
All due respect but I disagree. This shot came from my first day of shooting with the M8 back in November when I was attending the digital printing workshop. Not knowing what the results would be I also shot the same scenes with my D2X and Nikon 17-35 2.8. The files looked very different, particularly in terms of color balance and microcontrast. The 3D look seen in the posted image was not present in the Nikon files. Sadly I did not keep the Nikon files as the M8 did such a better job.
Woody / Tariq, I actually don't see it as a very 3D-looking picture at all. That's not a negative criticism, because it's a beautiful picture. But it does not look 3D to me in the slightest, in fact it looks more like a flat painting.
There are a few reasons. First, despite the wide angle lens, the huge size of the landscape features (and their proximity) makes it look extremely compressed, almost like a telephoto look. Its 3-dimensionality is very much lost by the almost superimposition of the foreground and background.
Secondly, there are no recognizable points of reference. Even the bush on the ridge doesn't help, because it's hard to tell how big it is. Recognizable things like a human on that foreground ledge would make the scale more obvious.
Finally, the contrast and colors both have a pretty narrow range. It has a pastel look such that the colors and contrast do not suggest depth, and the shadows are insufficient to suggest texture. I suspect that's more than anything the lighting you had to deal with.
Incidentally, jaapv's picture does not look particularly 3D to me either. I think this is because the background is very noisy, though, and the noise is made more prominent by the strangely high degree of saturation (more than I'd expect for the lighting) -- so the noisy background ends up being too distracting to really serve as a backdrop for a textured, 3D-looking picture, and the foreground colors are distracting in their own way (by virtue of the saturation).
I think those are some good observations DrPablo. The major thing which I beleive gives it some depth is the lighting and texture on the water compared to the near cliff but it does not stand out as a stunning example of a 3D look to me, though a nice image. I'm envious of your M8 nonetheless Woody.
It may be a little joke, but it is based in truth:
Ok...
I think the shot you took is all about the light and shooting fairly wide open, jaapv. It looks 3d because of the bright light in front and behind the subject and the great composition. I only have to disagree because I've shot 3d looking photos with a mamiya 645, a fuji gx680 with a imacon back, canon 20d, kodak retina etc... Its about having a decently sharp lens and then knowing what to do with it. Good photography isn't just equipment but mostly knowlege on how to use that equipment.
caleb condit wrote:
Good photography isn't just equipment but mostly knowlege on how to use that equipment.
I'd go even farther to say that all you really need is to know how to recognize good lighting. Obviously skill and experience make that easier. But someone who knows nothing more than pointing and shooting can take a vivid and perhaps 3D-looking picture if the lighting is right.
Here's my crowning example. I took this picture with a Rebel G and its 35-80mm kit lens about 7 1/2 years ago -- and at the time I only used it in fully automatic mode. I didn't know anything about photography at the time. In fact I was using some plain old Kodak Gold film, nothing fancy. Worse yet, I didn't even think about the lighting. I knew the little girl and the woman, though, I'd been taking care of them for about a month, and this was just a snapshot as they left. But everything from the lighting to the exposure happened to fall together for that picture, despite my complete technical ignorance and my super cheap lens. By the way, if it looks soft it's because this is a direct scan of the print.
3-D Components:
1. Lens resolving power (actuance/micro-contrast) - separates finest details from each other and not smooshing them together.
2. Aperature - depth of field
3. Lens Focal Length - the longer the focal length the more the subject is pull away from the background.
4. Format - the larger the format the less spatial compression and due to wide angle lens are at longer focal length in comparison to the 135 mm format. 8X10 is where you will see the most natural element of your image begin.
5. Dynamic Range - The higher the dynamic range the better the tonality changes and make your image appear real instead of flat.
6. Sharpening - total sharpening will compress the spatial information and create artifact; while selective sharpening will help to create the longer lens and the depth of field effects.
7. Light Effect
Thus, 5 out of 7 components is direct or indirect effect of lens and only dynamic range of the sensor and light effect as the other components to attenuate the flatness of 2-D image to appear 3-D.
