I agree Michael. I know I will eventually get rid of my D700 but just can't at the moment. TBH - I don't see many pics (using smaller format) with rendering similar to the one you posted.
H.Lux wrote:
But why is it so? Abberations? The blur of the light which is out of focus should be identical because the effect of f-stop and sensor size should cancel each other, according to the arrangement of the experiment.
Stefan
Stefan here is my understanding of this. Given that the light source remains constant between the images the smaller opening of the smaller fstop will reduce the size of the spectral highlight because the light is coming into the lens at a more direct angle. Google circle of confusion. It's been a long time since I've had to think about the science behind it. The smaller the diameter fstop the smaller the spectral highlight will appear. It's just something I know from experience.
Shooting latent finger prints (not dusted) on objects in the studio I often worked within the spectral highlight to bring out details of the print. With a view camera I used swings and tilts for depth of field and focus and the fstop controlling depth of field balanced with the size of the spectral highlight to cover as much of the fingerprint as possible.
Spyro P. wrote:
I read most responses and it seems that anyone who guessed right, did so based on lens characteristics or slight differences in distance or fov. In other words, testing flaws.
"Rendering", "look" and other mythical attributes of bigger sensors had no bearing whatsoever
One very large part of the "lens characteristics" that you talk about is just a natural effect of having to use lower aperture numbers (larger relative apertures) on smaller formats to keep DoF and light throughput constant. This is very much a format-related thing, and not a lens related thing [in itself].
Typically aberrations like SA, coma, LoCA grow at least twofold per aperture stop. That is, the SA aberration is twice as strong at F2.0 as it is at F2.8 and so on. Having to use two full stops larger apertures means at least four times larger aberration powers - if the lens quality is kept constant.
Regarding the specular highlights - looking back at the images, I'd say it's part an aberration, partly just small variations in the angle she holds her head. Ideally there should be NO differences in OOF specular highlight diameters in equivalent shots on different formats, since the base requirements for equivalence are "same absolute front pupil diameter" and "same image field of view".
Since the highlights are blown though, the diameters will vary with aberration strengths.
Jordan: Do you think you could go back and check the shutter speeds for the three comparison shots in the original post please? See the later part of this post for reasoning.
Other: And regarding diffraction, most of what you hear 'on da net' is pure BS.
Smaller formats have a small advantage for very deep DoF's in reasonable, but not extremely low (or "unlimited") light. But not for any reason even remotely close to how most people (even otherwise seemingly knowledgeable people) seem to want to "explain" it as being responsible for the effect.
The simple fact is that most small, new sensors are most of the time better per square mm² of sensor real estate. They are better, more exact, at handling medium small [absolute] amounts of light, and turning it into an image.
ISO1600 on an µFT sensor is (should be, if not Oly cheated the ISO spec with about a full stop) almost exactly the same absolute amount of light energy going into the image as ISO6400 on a FF sensor. But since the pixels on the µFT sensor will most of the time be smaller, they are inherently better at handling small charges of converted energy since their FWC values and read noise values are also lower.
So in the medium-high (not extreme-high or base ISO due to other losses) ISO range, smaller sensors will often be better when shot at equivalence settings. This entire line of reasoning is of course only valid for lenses well stopped down.
After another look at the images I still think size and shape of the highlights play a role. Even though the subject may have moved the highlight is still on approximately the same part of the orb. I'm not lucky enough to get it right by chance.
