mh2000 wrote:
The purpose of this lens seems more to push the speed envelope than anything else...
Why do you say that? It seems to me that it's actually their attempt to give an extremely refined 50 (ostensibly the most popular focal length) for the next generation of fullframe shooters. It's got (presumably) great glass, L-build quality, a wide aperture, state-of-the-art focus processesing/motor speed and is fully weather sealed. The aperture of the lens is actually the least notable aspect of the lens, imo. They could have do a 50/1.4L with the same set of features and I would still be buying it.
"With Canon volume, plastic design and manufacturing I bet they could get the 50/1.8 up to near Summicron IQ level for under $200 and the 50/1.4 up to near Summilux IQ for like $500."
According to Photodo the EF 50/1.4 is already better than the Summilux.
Canon's score is 4.5 compared to Leica's 4.4.
cprofit wrote:
I'll hold off on judgment until I see pictures from the lens. Without seeing pics, if the price approaches 1200 to be more in-line with the 35L, which isn't likely to happen since it is faster, I will strongly consider it.
I don't know. For the fact that the elements are molded, we may see it dip closer to that range once its released.
In the sequence of four photos in that URL (17, 35, 50, 100) taken with a variable distance from subject, there _is_ something that does not change. You mean that's not perspective not changing. What's _not_ changing in that sequence? We have both the lamp-post and the church bell in all pictures. So what doesn't change? Not the FoV, not the perspecitve, then what? That's what I called "perspective".
Besides, what I was really try to point out is that even one tries to do that comparison with a 50 and an 85 one can't deduce anything about one lens being sharper. Note how the church bell moves from the center of image to upper left corner of image. It's known most lenses are sharpest at center (following the MTF curves). One can't compare the sharpness of the 50 at edge with the 85 at center, nor vice-versa.
Then, if one needs to crop the image to bring the church bell at about the same size (between 50 and 85 shots) then again one can't compare sharpness, because resolution changes.
I think I'll wait for more reports on this, but mostly looking for a 50-to-50 comparison, not 50-to-85, nor Leica-to-Canon.
Ok, my direct experience is only comparing a Summicron with all 3 current Canon 50's... but overall (not just MTF), the Summicron is just "better" IMO. Among Leica shooters, it seems to be too close to call between the Summilux and Summircon... so I just take it on faith that I would find a Summilux superior to my EF 50/1.4. Leica gets by with 6 blades in the Summicron so I image Canon could to. Zeiss fixed its 50 to get better bokeh (adding "Ninja stars") and I'm sure Canon could do something to improve the bokeh in its 50s if it had the initiative to do so (why don't they?). All I'm saying is that the 50L might end up being really nice, and if the price comes down I'll probably buy one, but for me the f1.2 isn't the driver, it's overall IQ and it would be nicer if their more "standard" 50s were built to a higher imaging level. Also, a small lightweight 50/1.8 is sometimes nicer to carry than something big and heavy. I generally shoot my EF 50/1.4 over my Summicron for the AF and not for that extra aperture...
gdeliz2 wrote:
...
According to Photodo the EF 50/1.4 is already better than the Summilux.
Canon's score is 4.5 compared to Leica's 4.4.
...
George Deliz
Sam Bennett wrote:
So, reading the press release Canon says "an almost circular aperture diaphragm for attractive bokeh..." - are we assuming at this point that the aperture blades are curved a la the 85/1.2 II, or is this statement just based the number of blades and width of the aperture?
"Circular aperture" means that when the aperture gets more and more closed (f/11 and f/16) it does not form an octogon, but a circle. With usual blades, when the aperture is wide open (f/1.2, f/5.6) the hole looks as a circle but when closing down it looks more and more as precisely edged polygon. That aspect is considered as not good for some pictures. Having lens blades that close in a way that forms more of a circle than a edged-polygon involves better computing the blade geometry, need computers for that.
"Bokeh" is usually relating to thin DoF and creamy OOF subjects. But circular aperture diaphragm is not for that. It's for f/11 and above and some pictures are better with it.
The fact that a blade is more polished on its edge ("rounded") does play some role at f/11 and above (f/16) to deal with diffraction. But I'm not sure Canon means that in its announcement.
And I speculate it's the same blades and blade technology as in the 85/1.2 II blades.
petrescu wrote:
What's _not_ changing in that sequence? We have both the lamp-post and the church bell in all pictures. So what doesn't change? Not the FoV, not the perspecitve, then what? That's what I called "perspective".
What you're calling "perspective" is typically called "framing". It's the lamp post that is staying the same size. If you were shoooting a headshot of someone, it would be the size of their head not changing.
petrescu wrote:
"Circular aperture" means that when the aperture gets more and more closed (f/11 and f/16) it does not form an octogon, but a circle. With usual blades, when the aperture is wide open (f/1.2, f/5.6) the hole looks as a circle but when closing down it looks more and more as precisely edged polygon. That aspect is considered as not good for some pictures. Having lens blades that close in a way that forms more of a circle than a edged-polygon involves better computing the blade geometry, need computers for that.
Actually, I think it's just achieved by adding a curve to the aperture-facing blades. This is a well known design difference between lenses, and is precisely what I was asking about. After a certain point of closing the aperture down, the aperture gets less and less "circular". There's unfortunately no way to avoid this - the "circularness" of the aperture will only apply to a limited range of apertures - usually close to wide open.
So... my question was, is Canon doing curved blades like they did on the 85/1.2 II, or are they just attributing that characteristic to the number of blades - since the more blades, the more "circular" the aperture will be wide open.
petrescu wrote:
"Bokeh" is usually relating to thin DoF and creamy OOF subjects. But circular aperture diaphragm is not for that. It's for f/11 and above and some pictures are better with it.
No, you're completely wrong. Circular apertures are precisely what give you "creamy OOF subjects" - to over simplify matters, a photo is composed of billions of points of light - circular apertures make these points circular, hence more creamy, hence more pleasing OOF areas. As apertures close down, the points become "rougher" giving a less pleasing image.
petrescu wrote:
The fact that a blade is more polished on its edge ("rounded") does play some role at f/11 and above (f/16) to deal with diffraction. But I'm not sure Canon means that in its announcement.
As I understand it, curved blades play very little role in small apertures since at that point maintaining a circular geometry is essentially impossible.
Sam Bennett wrote:
Actually, I think it's just achieved by adding a curve to the aperture-facing blades. This is a well known design difference between lenses, and is precisely what I was asking about. After a certain point of closing the aperture down, the aperture gets less and less "circular". There's unfortunately no way to avoid this - the "circularness" of the aperture will only apply to a limited range of apertures - usually close to wide open.
So... my question was, is Canon doing curved blades like they did on the 85/1.2 II, or are they just attributing that characteristic to the number of blades - since the more blades, the more "circular" the aperture will be wide open.
No, you're completely wrong. Circular apertures are precisely what give you "creamy OOF subjects" - to over simplify matters, a photo is composed of billions of points of light - circular apertures make these points circular, hence more creamy, hence more pleasing OOF areas. As apertures close down, the points become "rougher" giving a less pleasing image.
As I understand it, curved blades play very little role in small apertures since at that point maintaining a circular geometry is essentially impossible.
We both agree that stopping down the aperture f/11 has a risk of edgy polygons (instead of circular). Adding more blades and and of better shape can avoid that.
We may disagree because of the following: at f/1.4, f/4 the diaphragm is easily a circle with many blades; at f/9, f/11 it's easily an octogon; at f/22 it's again easily a circle.
Bokeh OOF creamy is for f/4 and f/2, not for f/11 nor f/22.
Being right or wrong... I don't know, but thanks about the "framing" term... so the 4 images with 17, 35, 50 and 100 have all same framing.
Edited by petrescu on Aug 25, 2006 at 06:06 PM GMT
While not my personal favorite, the Super-Tak is a very well respected lens... the sample image shows how narrow the DOF can be and also shows some pleasing bokeh IMO... granted, it looks like a really forgiving background etc... and we'll have to see how it handles highlights and chunky textures etc... but I think it provides at least a hint that it will be a pleasing lens. It also shows why f1.2 isn't that important to me... DOF that narrow gets to look like tacky effect real fast.
Clayphish wrote:
Looking at Canon's online sample image:
...
it looks a lot like what I see coming from my Takumar SMC 50mm f/1.4. Lets see if I end up wrong once the full size images hit the net.
Have you ever actually looked at the aperture of a lens while stopping down? Try it and say what you just said again.
Checking 50/1.4... circular at f/1.4 (no blades edges visible), somehow circular at f/1.6, less circular at f/2.2, really octogon at f/6.3, approx circular at f/22...
Whatever circular means... I give up, your desk is valuable
petrescu wrote:
Checking 50/1.4... circular at f/1.4 (no blades edges visible), somehow circular at f/1.6, less circular at f/2.2, really octogon at f/6.3, approx circular at f/22...
Whatever circular means... I give up, your desk is valuable
Circular means... uh, circular. Care to explain how the aperture magically goes from octogon to circular between f/6.3 and f/22? And are you using some sort of maginifying glass to view the aperture at f/22?
From the Canon lens museum literature on the 85/1.2 II:
Ideal for professional photographers who shoot portraits and weddings, Canon’s EF85mm f/1.2L II USM medium-telephoto lens features a circular EMD (electromagnetic diaphragm) that achieves excellent background blur.
The circular aperture (which only really applies to apertures wider than f/2.8) is designed to give more pleasing background blur.
Sam Bennett wrote:
Circular means... uh, circular. Care to explain how the aperture magically goes from octogon to circular between f/6.3 and f/22? And are you using some sort of maginifying glass to view the aperture at f/22?
Not magnifying glass, just naked eye, looking from the front element direction. At f/22 it's really small, only a very small fraction of the blade edge contributes to the hole... a small circle made of small points is that... maybe not good eye.
To me the most perfect octogon is at f/6.3 . The more I open towards f/1.4 (from f/6.3) the more it becomes a circle.
But I'm probably wrong. I don't know how I got into this detailed comparison, aperture and so on. I'm very happy with the pictures I take, and I need more time to transfer those on dvd.
Sorry, I can't stop myself... EMD is old technology, it was there long before 85/1.2 II... maybe that's why my 50/1.4 has it and I see it circular wide open...
EMD isn't the circular technology dude, and I know circular apps have been around for a while - which is why so many people like German glass. But your 50/1.4 doesn't have them, and it wouldn't change the way things look at f/22 anyway:
EMD is a technology that has nothing to do with the aperture blades forming a circle. I checked EF lens work iii, and all it says is about how EMD controls the blades, DoF preview button, electronic control, reliability, stepping motor, digital control, in-lens only, in XL1 videocam too, etc. Nothing about the hole being circular...
Sam Bennett wrote:
EMD isn't the circular technology dude, and I know circular apps have been around for a while - which is why so many people like German glass. But your 50/1.4 doesn't have them, and it wouldn't change the way things look at f/22 anyway:
Edited by Sam Bennett on Aug 25, 2006 at 01:38 PM GMT
So if the EMD isn't the circular technology how can the Canon announcement claim it gives beatiful background blur?
? 
