But for an 85mm lens? It doesn't need to be fast for a shallow DOF -- if you're wielding an 85mm lens on full frame and focusing on a subject 10 feet away, even at f/8 your DOF is less than 2 feet.
True, but the big differentiator is how out of focus things are. Although the background (let's say a ugly city street) will be out of focus from the subject (cute model of course), the difference is *how* OOF the background will be. With f8 it will be out of focus, but recognizable as a city street. At 1.2 it will be a beautiful mess of color (bokeh).
DrPablo wrote:
I plugged the values for both APS-C and full frame into the calculator, and the biggest difference it made was +/- 0.01 feet (1/8 inch).
The pixel density doesn't figure into the equations that go into DOF. It certainly matters if you're taking a full frame image and cropping to a smaller image, though.
well, crop makes a difference in the following sense. if i'm doing a portrait at 1 meter with the 85 on a 20d, then to take the "equivalent" portrait with an 85 on a FF, i have to be at, what, like 0.6 meter. pixel size makes a difference if you agree that's its gonna be correlated to print size, which says something about how big a "circle of confusion" is.
jcbenner wrote:
In summary, from the MTF, the IQ ought to be better than the 50mm f/1.4 at f/8 but will be roughly comparable wide-open (although wide open means f/1.2 as opposed to f/1.4). Of course again, real world performance remains to be seen.
If 50 f/1.2 wide open resolving power and image clarity is comparable to that of 50 f/1.4 wide open, then the new lens value will be mostly in that pretty L-type barrel packaging and weatherproofing which in my books does not make a lens.... but a pretty decorative accessory.
coffeeshakes wrote:
True, but the big differentiator is how out of focus things are. Although the background (let's say a ugly city street) will be out of focus from the subject (cute model of course), the difference is *how* OOF the background will be. With f8 it will be out of focus, but recognizable as a city street. At 1.2 it will be a beautiful mess of color (bokeh).
It mathematically depends much more heavily on the proximity of your focal point and the distance of the background. In fact if you look at the equations both focal length and focal distance are exponentially greater contributors to DOF than is aperture. Furthermore, with respect to the DOF, the difference between f/1.2 and f/1.4 or f/1.8 is insignificant. Other factors such as diaphragm leaves may contribute to the 'quality' of the bokeh, but the difference between 1.2 and 1.4 isn't the major contributor (unless you're scrutinizing the shape of out of focus highlights).
bad_doggie wrote:
well, crop makes a difference in the following sense. if i'm doing a portrait at 1 meter with the 85 on a 20d, then to take the "equivalent" portrait with an 85 on a FF, i have to be at, what, like 0.6 meter. pixel size makes a difference if you agree that's its gonna be correlated to print size, which says something about how big a "circle of confusion" is.
I hear what you're saying. Actually I don't -- 'I read what you're writing' would be more correct. But there is a strict optics definition of circle of confusion, and it doesn't take into account the resolving power of the lens, film/sensor, or printer that produces the image. If on my 4x5 I take the same shot with ISO 25 and ISO 3200 film, obviously the ISO 25 has far greater resolution in that it has far more grains per unit area -- but that doesn't change the CoC if all else is equal.
As I said already, it either performs really well wide open, or it becomes a sturdily built and and oh-so-purty weatherproofed decorative accessory.
That's a practical photographer's view.
now if the 50/L looked like THAT, i'd be "poised to buy" ... but otherwise, its a big yawn.
Edited by bad_doggie on Aug 24, 2006 at 10:04 PM GMT
Pitfalls: Comparing apples to oranges. MTF of a 55mm compared to a 50mm, and 1.6x image circle compared to ff, f/2.8 compared to a f/1.2. Expecting an excellent 50/1.2 to stick on the ceiling (it'll never happen, if you want that check the 300 at f/8).
Avoid the following pitfalls: (1) don't compare apples to oranges and (2) don't compare oranges to apples.
The only _approximatively_ meaningful MTF comparisons that could be done are: the 50/1.4 at f/8, FD 50/1.2 at 1.2 and Leica 50/2.8 at f/8 and only 10lpmm.
That's the only thing one could do with the published MTFs of Canon, Leica and Zeiss.
handheld wrote:
Realistically, you will be waiting one heck of a long time to get that lens, if $900 or less is your purchase point.
What on earth can make you believe that the lens will retail for $200 less than the 24L and 35L, to use those as examples ?
Well, but to mu knowledge, that lenses don't use moulded elements...which the 50/1.2 apparantly does
MTFs FD 50/1.2 vs EF 50/1.2. The only things comparable between the two MTFs is the f/1.2 curves, at 10 and 30lpmm. The other curves are f/5.6 for the FD and f/8 for the EF.
The EF f/8 curves look much better than the FD f/5.6 curves, and _usually_ f/5.6 performance is better than f/8 performance, so I would _speculate_ the EF 5.6 (not plotted) is much better than the FD 5.6 (plotted). But that's speculation.
At f/1.2, checking the red lines in the left figure and the black lines in the right figure (f/1.2, 10 and 30lpmm) I find the EF to be noticeably better than the FD.
