AhamB wrote:
That's plasticity and DOF separation and they don't have to occur at the same time to make a picture "3D".
A "gaussian blur bokeh" is the most importand part for the 3d effect. You can check that pretty simple: load a picture in photoshop do a selection of the background and apply a 50 px gaussian blur filter.
cavewalker wrote:
A "gaussian blur bokeh" is the most importand part for the 3d effect. You can check that pretty simple: load a picture in photoshop do a selection of the background and apply a 50 px gaussian blur filter.
Maybe the blur is the most important part of the "pop", but without the microcontrast/plasticity the 3D effect is weakened because the contrast between sharp area and blur is less. Look at Makten's first shot (without any blur) to see an example of plasticity. It's the crystal clear sharpness of the textures that cause the effect.
The blur may enhance the effect, but it is NOT the cause of it. Too much blur can help render that fake cutout look when the transition rate becomes too fast and our eye/brain detects this. Too slow a transition rate and we see it as flat / 2D.
Getting the right rate of transition associated with the right portion of the image is part of the key to creating the effect. Too many times, that DOF separation / blur can be evident between the subject and the BG ... yet the subject in and of itself still looks cutout/flat/2D. Thrice had a very nice image way back when that illustrated the criticality of focal point placement's impact on creating the effect.
As noted elsewhere, the 3D effect can be achieved without any blur in the BG (Makten and the 35/2 comes to mind).
BG blur and DOF separation are NOT the same as 3D-ishness (although they can often be associated with it as an enhancement to the effect in many images).
Different lens designs have different rates of transition. Within a given lens, knowing where those transition rates begin to change is key to understanding how to render a subject with the most 3D-ishness that lens has to offer.
It kind of reminds me of the need to understand what range a Beauty Dish functions best. The only real difference is with a BD, it is light going out, with a lens, it is light coming in. The output of light from a BD is not evenly spread and as evidenced by those donuts of light you can get from them and the falloff characteristics / variance between a BD and a softbox.
With lens design (which Zeiss frequently does) that uses un-evenly spread (non-linear) transition rates, the 3D effect is easier to achieve than with glass that has very even (linear) transition rates (i.e. Oly, Leica). Kinda like trying to surf on 0-2 foot swells vs. 5, 10 or 20 foot swells ... how much 'wave action' is present impacts how "WOW" the ride is when you catch it right.
Every mfr has some glass that is +/- capable of creating the effect, but it is a function of lens design (transitions), focal point, subject distance, more than aperture to create a skinny DOF / bokeh / blur. In fact, too small of an aperture is often times counterproductive to the effect as it renders those transitions to occur within the sujbect rather than near the edge of the subject (while DOF still encompasses the subject). In the case of the C/Y 35-70/3.4, most people suggest that 5.6-8.0 is the aperture range that renders the effect best, not wide open (as the 'blur theory' might otherwise suggest).
Study the MTF's of those lenses (various mfrs) that notably yield the effect and you'll start to notice a trend in design as compared with those that render more even imagery. It isn't brand specific ... although brands typically carry consistent design philosophies into the majority of lenses they produce.
On a related note, different thread ... the measured sharpness resolution of glass that has more non-linear transition rates can be a bit misleading compared with glass that has more linear transition rates ... and vice versa.
AhamB: Thank you I originally left them here to show a depth cue between the sheep in focus in front of it and the ones behind them. But now that you mention it, it is starting to bug me - will probably crop them out !
Hey Paul ... I wondered how long it would take for you to chime in on this one.
I think this might be one to illustrate not a lot of BG blur. C/Y 28/2.8 (guessing f11)
(RAW + PP as usual ... looks a touch oversharpened at this size, i.e sharpened for print)
and one that does have some blur ... noting the DUAL focal points.
+1 Makten's pic illustrate that the Zeiss pop has very little to do with narrow DOF. It's about the texture of the subject that makes it pop, rather than a blurry background. Actually very blurry background makes it less "3D" as it becomes less realistic to the human eye.
curses! i thought this debate had finally died with the archiving of the "what is 3d" thread. after raging endlessly the only thing to become clear was that many people have a different idea of what 3d means and that there is rampant disagreement about whether any given image has "3dness".
... i.e. made a detour while traveling through Tennessee. I live in a place that's closer to "Cornfield County" 
