A Random Roadmap: Lenses & Bodies [CR1]

[traveller]

24-70 f/2.8L II

Apparently 5 versions of this lens exist. A few even have IS. It is slated for a 2011 announcement. The patent we’ve seen says no IS.

Is it normal for manufacturers to produce 5 different, fully functional prototypes of different designs for a new lens? Surely in this age of computer design and modelling, lens manufacturers can just settle on one design and tweak it? Is anyone else a touch suspicious about this?

 

[dstppy]

24-70 f/2.8L II

Apparently 5 versions of this lens exist. A few even have IS. It is slated for a 2011 announcement. The patent we’ve seen says no IS.

Is it normal for manufacturers to produce 5 different, fully functional prototypes of different designs for a new lens? Surely in this age of computer design and modelling, lens manufacturers can just settle on one design and tweak it? Is anyone else a touch suspicious about this?

Apple does the same thing, though to obscure announcements. Guess what I spend my days doing besides waiting for Canon to release things?

 

[docsmith]

I'd much rather have a FF with crop options in-body. This should be standard, like choosing file quality.

Please forgive me if I am missing something obvious as I've never even considered this....but, other than file size on your CF card, how is this different from simply cropping a full frame image in post? ???

.... But, if an in-camera crop is applied, a smaller portion of the sensor could be used, allowing a faster frame rate. So a FF camera with a fast shutter could achieve higher frame rates that way, e.g. Nikon's D3 is 9 fps in FX mode (but that's only 12 MP, so a lot less data to manage than Canon's current 21 MP FF), but edges up to 11 fps in DX mode (1.5x crop).

Ok, that makes sense and I can see the benefit. But honestly, at least for me, that is pretty far down my list of desires. Better dynamic range, less noise at high ISO, etc.

 

[WarStreet]

Ok, that makes sense and I can see the benefit. But honestly, at least for me, that is pretty far down my list of desires. Better dynamic range, less noise at high ISO, etc.

But it is on top of the list for others, while still expect the better Dynamic range and noise at high ISO too

 

[Justin]

IMHO, the lenses would still significantly differ, e.g.

* The f/1.2 has metal body with weather sealing, where the f/1.4 has plastic body with no weather sealing.

* The f/1.2 has 8 circular aperture blades, while the f/1.4 has 7 straight blades.

* The f/1.2 has an aspherical element, the f/1.4 has none.

You know the sad thing is that even the new nikon 50 f1.8 has an aspherical element. And its 200$ not 1600$...
Canon screwed this when they released the new 50 f1.2, because its became the worst L prime in the lineup, its so weak that even the 1.4 can match it. So canon can't (and won't) upgrade the 1.4 until the 1.2 is not getting its refresh. And its not coming soon...

Yeah sad. Buy sigma and bring it back to the store five times before you get a good copy.

 

[Justin]

24-70 f/2.8L II

Apparently 5 versions of this lens exist. A few even have IS. It is slated for a 2011 announcement. The patent we’ve seen says no IS.

Is it normal for manufacturers to produce 5 different, fully functional prototypes of different designs for a new lens? Surely in this age of computer design and modelling, lens manufacturers can just settle on one design and tweak it? Is anyone else a touch suspicious about this?

Apple does the same thing, though to obscure announcements. Guess what I spend my days doing besides waiting for Canon to release things?

Me too brotha. Oh that and working and having fun and eating and stuff. But lots of mac and canon and photography rumors.

 

[pgabor]

You know the sad thing is that even the new nikon 50 f1.8 has an aspherical element. And its 200$ not 1600$...
Canon screwed this when they released the new 50 f1.2, because its became the worst L prime in the lineup, its so weak that even the 1.4 can match it. So canon can't (and won't) upgrade the 1.4 until the 1.2 is not getting its refresh. And its not coming soon...

Yeah sad. Buy sigma and bring it back to the store five times before you get a good copy.

Thats the problem. I don't want to buy sigma, or any other 3rd party lens, i want to buy a GOOD canon 50mm, but they don't offer one : ( I have the 1.4, but it has it's limitations (and it's AF started dying nowadays)
(btw i don't know with what i deserved this attitude... i don't know why you have to be offensive)

 

[Justin]

You know the sad thing is that even the new nikon 50 f1.8 has an aspherical element. And its 200$ not 1600$...
Canon screwed this when they released the new 50 f1.2, because its became the worst L prime in the lineup, its so weak that even the 1.4 can match it. So canon can't (and won't) upgrade the 1.4 until the 1.2 is not getting its refresh. And its not coming soon...

Yeah sad. Buy sigma and bring it back to the store five times before you get a good copy.

Thats the problem. I don't want to buy sigma, or any other 3rd party lens, i want to buy a GOOD canon 50mm, but they don't offer one : ( I have the 1.4, but it has it's limitations (and it's AF started dying nowadays)
(btw i don't know with what i deserved this attitude... i don't know why you have to be offensive)

I was agreeing with you. What attitude?

 

[Loswr]

Canon screwed this when they released the new 50 f1.2, because its became the worst L prime in the lineup, its so weak that even the 1.4 can match it. So canon can't (and won't) upgrade the 1.4 until the 1.2 is not getting its refresh. And its not coming soon...

What, specifically, is 'weak' about the 50mm f/1.2L?

 

[lol]

What, specifically, is 'weak' about the 50mm f/1.2L?

I'm guessing it is the relative lack of sharpness in border regions on full frame, from a quick glance at Photozone results. Seems to fall off significantly there. Not an issue on crop sensors. It doesn't sound like an issue in real world shooting, since if you're getting the f/1.2, you're wanting the current shallowest DoF at that length and unlikely to get everything sharp in frame anyway. There is the even older f/1.0 it replaced that's even softer.

 

[Loswr]

What, specifically, is 'weak' about the 50mm f/1.2L?

I'm guessing it is the relative lack of sharpness in border regions on full frame, from a quick glance at Photozone results. Seems to fall off significantly there. Not an issue on crop sensors. It doesn't sound like an issue in real world shooting, since if you're getting the f/1.2, you're wanting the current shallowest DoF at that length and unlikely to get everything sharp in frame anyway. There is the even older f/1.0 it replaced that's even softer.

I suspected that might be the perceived issue. Everything in photography involves tradeoffs and compromise. Want a faster shutter speed in low light? Fine, open the aperture and sacrifice DoF, or raise the ISO and add noise. Tradeoffs. In the case of the 50/1.2, Canon's designers chose to produce a lens with superior bokeh. The way to do that is to intentionally undercorrect speherical aberration, meaning sharpness is reduced a bit. But the lens is intended for portrait work ("Taking advantage of shallow depth of field allows the photographer to capture subjects with impact, such as for wedding portraits," from Lens Work III) - as such, especially with such a wide aperture, bokeh is critical - and in most portraits, it's ok for the resulting image to lack razor sharpness (all the portrait picture styles, from Canon and in 3rd party software, have the sharpness set lower than for general shooting for a reason).

 

[Lawliet]

What, specifically, is 'weak' about the 50mm f/1.2L?

The focus shift is an annoyance. Raw performance doesn't matter if systematic errors spoil the game. That puts both the 50/1,2L and the Sigma 50/1,4 rather low on my list.
For example the difference between a 50/1,2 and the Nikkor 50/1,4 is for the connoisseur, but flawed focussing ruins the picture for everybody.

 

[epsiloneri]

The focus shift is an annoyance.

Sounds like the solution is to not step down this lens. The argument can be made that if you're stepping it down anyway, you might just as well use the 50/1.8 or 50/1.4 (or a zoom at f/8). Of course it would be nice if you had the flexibility to use the 50/1.2L for smaller apertures as well. It shouldn't really be so difficult for Canon to calibrate the focus shift in software, I think. I wonder why it's not done.

 

[Lawliet]

F/1,2 only would be a bit to constrained for my taste.
At least its less severe with the more peripheral sensors of the 1D(s), but with the 5D(II) you just replace systematic error with randomness.

Well so much for why it is the weakest of the fast prime Ls - the others don't have such real life weaknesses.

 

[epsiloneri]

At least its less severe with the more peripheral sensors of the 1D(s), but with the 5D(II) you just replace systematic error with randomness.

I assumed the focus shift was due to the intentional spherical aberration (to get a nicer bokeh). That should give you a systematic shift as you step the lens down, because the focus distance for different radii from the central optical axis of the parallel beam will be different, and when closing the aperture you are selectively ignoring the outer parts of the beam.

But you seem to imply that the focus shift is random (for the 5D2 anyway), which doesn't make sense if the focus is still always retrieved at the same aperture. Can you explain what you mean by the sensors being more peripheral on the 1Ds?

 

[lol]

It shouldn't really be so difficult for Canon to calibrate the focus shift in software, I think. I wonder why it's not done.

I'd love that. For my style of fast prime shooting, focus shift is a significant hindrance.

As to why... I'd guess the technology wasn't in place when the lens came out. When did AF micro adjust come in? Assuming it wont be retro-fitted into lenses, a body could do it using a built in lookup table in a similar way to peripheral illumination correction.

 

[Lawliet]

But you seem to imply that the focus shift is random (for the 5D2 anyway), which doesn't make sense if the focus is still always retrieved at the same aperture. Can you explain what you mean by the sensors being more peripheral on the 1Ds?

I suspect that field curvature and focus shift cancel each other out.
With the 5D2 that effect applies at well, but the focus points are not accurate enough to nail the focus.

 

[skitron]

The focus shift is an annoyance. Raw performance doesn't matter if systematic errors spoil the game. That puts both the 50/1,2L and the Sigma 50/1,4 rather low on my list.

The newer "smooth finish" Sigma 50 is apparently much better in this regard. I have not tried the old one but took a chance on the new one when I saw good reports on it...tested it and found it has very little shift at all. My 50 tested to look virtually identical to the focus charts for the Sigma 85 at photozone.de.

 

[epsiloneri]

skitron, how consistently does your Sigma 50/1.4 focus over different distances? I have the problem that my ("old") copy focuses well on near distances (within 5 meters, say), but inconsistently further away, which is extremely annoying because I think the lens is splendid optically (with lovely bokeh). On the other hand, I have tended to use it indoors mostly, which normally limits it to close distances, so it hasn't been a huge problem for me yet, but if there is a better AF performing version I would trade in an instant.