Lotusm50 wrote:
But her blog said she got you a new camera and lens! And I quote, "And what did my husband get for our Anniversary? A new camera complete with a new lens...both of which he used to take the picture of my ring with " Do you mean she lied on her blog? People don't lie on blogs, do they? ;-) Nice picture of the ring, BTW (a little bit of CA on the edge of the center stone, though).
Oh dear Lord!
I traded my already-owned Nikon system for a Canon system - I didn't get anything new really, it's just that I did the trade near our anniversary! She can't count that as any type of new gift! HA! I'm going to have to catch her on that; it's misrepresentation!
And yeah, I couldn't help the CA unless I manually edit it - Lightroom wouldn't get rid of it no matter what I did, but I was way too lazy to do anything about it
I could have probably quickly just done a negative saturation brush on the edge but most of the viewers of the ring photo won't know the difference
You can tell her that I will try her recipe the next time I make a big pot of sauce. It sounds decidedly artery-clogging, so it will be in measured quantities. You can also tell her that she needs to proofread the text ("adiment", "purposal", and "hamberger", for example).
Congratulations on the year 10 milestone.
Thanks and I'll let her know about the spelling! I think she normally relies on the built-in spell check so I'm not sure how she managed so many fumbles. I think she was in a rush today, packing up readying the kids and herself for a flight to visit her sister. I have two weeks of peace and quiet all to myself, whatever will I do!?!
carstenw wrote:
I think I see what you mean, but looking at your medium distance shots, I also see that they are not particularly fair examples, in that the foreground and background colours also blend much more than in the other shots, especially the first shot. I did exactly the same thing myself when I first got the 50MP, which is one reason I was disappointed with it.
They were not meant to be fair, but to show a couple of situations where the 50 MP has problems.
It could be that the car, bicycle and tricycle would have been rendered better with the 50P, I don't know, because I stopped shooting everything twice, and just work with the 50MP now, to see what I can get. I disagree that the car shot has 3D only because of the implied perspective, though the effect is probably milder than what the 50P can get. Olaf's last shot in the post before mine also has some mild 3D, IMO, around the decorative trim just under the edge of the upper floor. I can almost "feel" the bumpiness of the metal there.
Btw, another data point in our differing 3D perception is the two tube shots you posted. I see little 3D in the first one, and zero in the second one, which looks as flat as can be....Show more →
Add one more as I don't see any 3D in Olaf's last shot on the previous page. What is perhaps more interesting however that the 'implied perspective' kind of 3D that I see in your car shot is identical, albeit slightly weaker than the two shots of the pipes. If we don't count that kind of 3D (which you can get with any lens or medium, including a simple drawing) then I see no 3D at all in any of the shots mentioned.
The bike shot is IMO a variation on the theme - implied perspective in combination with shallow DOF. Something similar to this:
It's not the real thing though as the effect would disappear instantly had it been shot at a smaller aperture and you could get the effect with most lenses. The sharpness-to-blur transition type may be a contributing factor though.
I think one of the things that makes the whole 3D thing difficult is that if you stare long enough at a photo your brain will start interpreting stuff as 3D. That's why people have a stronger tendency to see 3D in their own photos than in other peoples'.
It would be interestig to do a bit of experimental reseach on this. Right now the data that we have is very disorganized and the only thing we can say is that there is little agreement on what constitutes 3D. Perhaps if one was to set up a web site where people could submit their 3D candidates and be shown randomized sets of images that they could evaluate for 3D.. The problem remains though that we would need some type of definitions though to avoid situations like here where you don't think implied perspective in a shot counts as 3D while I do count it as one type among several.
One of the things I learned from reading forums is that discussing 3D in a photo is quite useless.
Some see it, some don't, some see it differently, some say any photo with shallow DOF looks 3D, some say any really sharp photo looks 3D,...
My continued interest in this is partly based on the hope that we one day find a set of appropriate definitions, and can back them up with a series of examples so that we may all, or at least most, agree and be able to discuss the finer details with a shared vocabulary.
My hope is also partly based on the observation that nowhere else in the Internet have we come closer to being able to succeed in this than here in the Alt forum.
denoir wrote:
They were not meant to be fair, but to show a couple of situations where the 50 MP has problems.
The not-fair part was meant to imply that the 50P might also have difficulties with these shots, i.e. that it is not just the 50MP which has trouble there. The similar foreground/background colour removes one important visual clue of the transition.
What is perhaps more interesting however that the 'implied perspective' kind of 3D that I see in your car shot is identical, albeit slightly weaker than the two shots of the pipes. If we don't count that kind of 3D (which you can get with any lens or medium, including a simple drawing) then I see no 3D at all in any of the shots mentioned.
Not counting implied perspective is not necessarily a good strategy though. I simply meant that I don't think I see the 3D in the Citroën shot *because* of the implied perspective *only*.
I believe that convincing 3D in a photo is simply a sum of parts, where a sufficient number of strong enough parts will eventually add up to more perceived 3D. Conversely, removing enough parts would eventually remove the feeling of 3D. Some parts are stronger than others, of course. I am not convinced that any part is necessary, but some come pretty close.
Some example parts would be high micro-contrast, punchy colours, implied perspective, boke falloff characteristics, and so on. Some of these can be missing and 3D still perceived. On the other hand, if you have something which is the opposite of one of these characteristics, then it could counteract the feel so strongly that it disappears.
It's not the real thing though as the effect would disappear instantly had it been shot at a smaller aperture and you could get the effect with most lenses.
I disagree with this sentence twice First in your use of "real", as if implied perspective is a fake 3D effect, but something else (what?) more real. Secondly, in that most lenses could give a 3D effect simply by having implied perspective. I think the rendering of some lenses, for example most Canons, counter-acts the 3D look so strongly that no amount of implied perspective could yield a 3D look, even weakly.
I think one of the things that makes the whole 3D thing difficult is that if you stare long enough at a photo your brain will start interpreting stuff as 3D.
Interesting comment! To me the point at which the brain starts to interpret stuff as 3D is actually the crux of the matter. In other words, what do we have to present to our brain to reach the tipping point, where it starts to believe the 3D (which in a 2D image must by necessity be fake).
It would be interestig to do a bit of experimental reseach on this. Right now the data that we have is very disorganized and the only thing we can say is that there is little agreement on what constitutes 3D. Perhaps if one was to set up a web site where people could submit their 3D candidates and be shown randomized sets of images that they could evaluate for 3D.
Agreed! Although I think we have to do more footwork before such a survey would be feasible. We need more organized thoughts about this in our own heads before starting to ask others what they think and see.
carstenw wrote:
The not-fair part was meant to imply that the 50P might also have difficulties with these shots, i.e. that it is not just the 50MP which has trouble there. The similar foreground/background colour removes one important visual clue of the transition.
I could be wrong but I'm pretty sure the Planar would do a better job there as would some other 50mm lenses like the 50 Lux ASPH.
I believe that convincing 3D in a photo is simply a sum of parts, where a sufficient number of strong enough parts will eventually add up to more perceived 3D. Conversely, removing enough parts would eventually remove the feeling of 3D. Some parts are stronger than others, of course. I am not convinced that any part is necessary, but some come pretty close.
I agree with that. I would roughly divide the effect in three independent parts and you can get a spatial impression with any or all of them:
1) Implied perspective 3D
2) DOF 3D
3) Volumetric 3D
Number 3 is the most complex and typically what I associate with Zeiss rendering. It's a combination of factors but the end result is that the object(s) in the image in question appear to have an actual volume. For instance in your car shot I see 1) and 2) but no 3) and the same goes for my plastic pipe shots.
I disagree with this sentence twice First in your use of "real", as if implied perspective is a fake 3D effect, but something else (what?) more real. Secondly, in that most lenses could give a 3D effect simply by having implied perspective. I think the rendering of some lenses, for example most Canons, counter-acts the 3D look so strongly that no amount of implied perspective could yield a 3D look, even weakly.
My use of "real" in this case is to refer to the type of 3D we associate with the lens - i.e not just a geometrical optical illusion. As for Canon rendering, no offense, but there is no way you could differentiate Canon from Zeiss in the shots of aluminium/plastic pipes due to the absence of any color nuances and distinctive textures or any chance to evaluate the sharpness to blur transitions.
I may be wrong of course. For instance, can you tell me which one of these was shot with the 100 MP and which one with the Canon 135, or alternatively an explanation of what makes you say that the rendering in one implies 3D while the other one does not:
Although I think we have to do more footwork before such a survey would be feasible. We need more organized thoughts about this in our own heads before starting to ask others what they think and see.
Yes, I definitely agree. There would be no point in starting doing polls before we had a basic common nomenclature and some common foundation that we can build upon.
Time to jump over to one of the 3D threads again?
Yes, perhaps it is.
Apr 29, 2011 at 10:07 AM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
denoir wrote:
I could be wrong but I'm pretty sure the Planar would do a better job there as would some other 50mm lenses like the 50 Lux ASPH.
For some definition of "better", you could be right. I don't think it would show much 3D in the end though. Very similarly coloured foregrounds and backgrounds kill much of the spatial feel.
1) Implied perspective 3D
2) DOF 3D
3) Volumetric 3D
I could argue a bit with the choice of these three. Where is micro-contrast, for example? It would have to be 3), I would think, which would mean that 3) is itself a set of smaller parts. Anyway...
I may be wrong of course. For instance, can you tell me which one of these was shot with the 100 MP and which one with the Canon 135, or alternatively an explanation of what makes you say that the rendering in one implies 3D while the other one does not:
The second shot comes closer, having crisper detail, but note that I said "most", and this was done to deliberately not include some of Canon's stronger lenses, like the 135L, the 200/1.8, etc. I think that using "most" in this sense is not controversial.
carstenw wrote:
I could argue a bit with the choice of these three. Where is micro-contrast, for example? It would have to be 3), I would think, which would mean that 3) is itself a set of smaller parts. Anyway...
Micro contrast, color separation, contour contrast, sharpness etc are all contributing factors to 3). The three item list could of course be longer as there are a bunch of optical illusions that imply 3D in a 2D image.
The second shot comes closer, having crisper detail, but note that I said "most", and this was done to deliberately not include some of Canon's stronger lenses, like the 135L, the 200/1.8, etc. I think that using "most" in this sense is not controversial.
The second shot is the 135L I'm pretty sure that I could recreate the experiment with the Canon 50/1.4 with that type of subject. In the case of a simple perspective 3D illusion, you can get very far even with a hand drawn picture..
JulianDeLaRosa wrote:
Hi everyone. I was skimming through this thread and all the images are amazing! Going through the archives recently and found this.
denoir wrote:
The second shot is the 135L I'm pretty sure that I could recreate the experiment with the Canon 50/1.4 with that type of subject. In the case of a simple perspective 3D illusion, you can get very far even with a hand drawn picture..
Well, you did your best to take out all differences in order to present them as equal, and you succeeded. Congratulations. I still prefer the 100MP in most situations (all I have seen so far, except for this shot)
carstenw wrote:
Well, you did your best to take out all differences in order to present them as equal, and you succeeded. Congratulations. I still prefer the 100MP in most situations (all I have seen so far, except for this shot)
I didn't take out the differences at all - they're just resized using the same script. When the shots were taken they were not intended as comparison shots. There simply isn't enough information in the shots to tell them apart (no textures to speak of, no subtle color variations etc).
You are however missing my point. It was not to compare the 100 MP and 135L. They are two quite different lenses that are in many situations trivial to tell apart. The point was that for certain 3D elements such as implied perspective it makes no difference which lens you use. In the case of the two images above, I could have used just about any lens and I would have captured the geometric properties of the scene.
akul, like the soft transition of focus.
melbmanu, nice details.
rji2goleez, excellent series. Love the 4th one.
-OlAf-, great details on the 3rd image and 2nd upload.
Jochenb, Love the symmetrical composition.
Lars, another nice image with 100.
antifire, nice focus. Is it a bit overexposed? I'm not sure.
JulianDeLaRosa, Love those colors. Fantastic.
denoir, carstenw - thanks for nice discussion, learning a lot from it.
" Do you mean she lied on her blog? People don't lie on blogs, do they? ;-) Nice picture of the ring, BTW (a little bit of CA on the edge of the center stone, though).
