Sigma vs Zeiss vs Canon

[sanj]

But but what about the people who want a good quality 50mm and are willing to pay appropriate money?

And don't you think lots of people will buy the Sigma?

Canon DOES have the 50/1.2 lens. You can't deny the quality of that lens, despite it's spherical aberration, which as it so happens to be, is a DESIRABLE trait in a portrait lens for many photographers. Not everyone screams for perfect corner to corner sharpness. Sometimes, having soft corners is beneficial to guiding your viewers eyes to the subject...which tends to be near the center of the frame.

I've always admired photos taken with the Canon 50/1.2 and 85/1.2 lenses. They have a specific aesthetic appeal that is just WONDERFUL for portraiture specifically, and for a variety of other types of photography as well (such as street.) I find it ironic how so many people write off the Canon 50/1.2 and 85/1.2 lenses as if they don't even qualify to be included in the lineup for comparison.

So, what about the people who want quality? Canon offers a VERY high quality 50/1.2 lens that offers STUNNING and very aesthetically appealing results. You should give it a try sometime. Oh, and you'll spend about half as much on that as you would on an Otus...you won't get razor sharp corners, but it's HALF as much as an Otus.

So. Am I to infer that if Canon comes out with 50/1.2 II that is sharper and has better corner to corner sharpness then you would not DESIRE to use it?

I'm sure a lot of people would. I'm also sure that a lot of the people who currently love the soft-focus traits of the current 50/1.2 would be bummed if Canon copied the Otus design with razor sharp focus corner to corner. It's better to have a DIVERSITY of lenses with different traits, than for all manufacturers to make exactly the same things that behave exactly the same way.

I think the center performance of the 50/1.2 and 85/1.2 needs to be improved...in the grand scheme of things, it's a bit soft, and doesn't need to be. I do, however, hope Canon keeps the soft focus traits in place if they release a 50/1.2 II and 85/1.2 II. If I want a lens with perfect sharpness, I can always get the Otus...if Canon copies the Otus, then I'm suddenly left WITHOUT the option of buying a lens that purposely leaves in a certain amount of spherical aberration for artistic flare.

I'm pretty sure that you can introduce spherical aberration through plugins or other software components if you really so desire. What you can't do is correct for poor image quality at capture time.

Anyway, in the main the comments above about justifying Canon's current design and product are more about trying to ensure that people who worship Canon find a way to present Canon's offering as good and justified so that they feel good about owning Canon products. That's it. I'm sure someone will argue here that this comment is wrong, but you don't see anyone saying that they wish the 70-200/2.8 II had soft focus like the 50/1.2L and so on.

Absolutely Dilbert. Perfectly said. Every word.

 

[Loswr]

I totally doubt that Canon makes lenses with soft edges on purpose. I think the soft edges are a result of technology limitations and cost saving.

In the case of the 50L (both of them), it was an intentional design decision by Canon.

I'm pretty sure that you can introduce spherical aberration through plugins or other software components if you really so desire. What you can't do is correct for poor image quality at capture time.

There are polarization effect filters for post-processing, but they cannot properly replicate the effects of having a CPL on your lens at capture. Similarly, adding spherical aberration in post will not correct for poor bokeh in the captured image.

Anyway, in the main the comments above about justifying Canon's current design and product are more about trying to ensure that people who worship Canon find a way to present Canon's offering as good and justified so that they feel good about owning Canon products. That's it. I'm sure someone will argue here that this comment is wrong...

Comments like the above are mainly about bashing Canon, made by people who have an inadequate grasp of the concepts behind lens design (and in one case, the inability to distinguish a lens from a camera).

...but you don't see anyone saying that they wish the 70-200/2.8 II had soft focus like the 50/1.2L and so on.

Not in those words, no. But plenty of people have said that they prefer the bokeh of the MkI versions of the 24-70/2.8 and 70-200/2.8 IS, or at least acknowledged the less-that-stellar bokeh of the MkII versions. When push comes to shove, they may not want to trade sharpness for better bokeh, but I suspect many people aren't even aware of the trade-off.

 

[Random Orbits]

Confirmation bias:
http://en.wikipedia.org/wiki/Confirmation_bias
".. is the tendency of people to favor information that confirms their beliefs or hypotheses."

Personally I'd love for the Canon 50/1.4 to be better than the Sigma 50/1.4 Art because then I wouldn't need to carry around the Sigma lens, but I simply can't justify that thinking given the results that have been presented. Same with the 50/1.2L.

I can't wait for the Sigma 50/1.4 Art to be tested by DxO and for it to wipe the floor with the 50/1.2L. I can already see the posts from those with Red Ring Fever putting down DxO, etc. What a laugh that will be to see.

Yes, but seeing schadenfreude in action can be just as ugly.

 

[Sporgon]

I'm pretty sure that you can introduce spherical aberration through plugins or other software components if you really so desire. What you can't do is correct for poor image quality at capture time.

There are polarization effect filters for post-processing, but they cannot properly replicate the effects of having a CPL on your lens at capture. Similarly, adding spherical aberration in post will not correct for poor bokeh in the captured image.

I'll just point out that I didn't mention polarization or bokeh, so I'll take your dalliance off topic as an indication that I was on the money but you can't admit it

Anyway, in the main the comments above about justifying Canon's current design and product are more about trying to ensure that people who worship Canon find a way to present Canon's offering as good and justified so that they feel good about owning Canon products. That's it. I'm sure someone will argue here that this comment is wrong...

Comments like the above are mainly about bashing Canon, made by people who have an inadequate grasp of the concepts behind lens design (and in one case, the inability to distinguish a lens from a camera).

No. The problem that we're seeing here is something called "confirmation bias", where people find any reason at all to support the idea that the Canon 50/1.2L is better. The "in one case" comment is simply someone's inability to move on, which is a rather sad reflection of said person.

Confirmation bias:
http://en.wikipedia.org/wiki/Confirmation_bias
".. is the tendency of people to favor information that confirms their beliefs or hypotheses."

Personally I'd love for the Canon 50/1.4 to be better than the Sigma 50/1.4 Art because then I wouldn't need to carry around the Sigma lens, but I simply can't justify that thinking given the results that have been presented. Same with the 50/1.2L.

I can't wait for the Sigma 50/1.4 Art to be tested by DxO and for it to wipe the floor with the 50/1.2L. I can already see the posts from those with Red Ring Fever putting down DxO, etc. What a laugh that will be to see.

Hmmm.......hardly fair. The 50 L is a niche lens designed for a very specific purpose. It was never intended to be a GP standard lens. The comments made against it are very much in the 'test chart specialists vs specialist practical users'. However because of the extremes in construction and pricing between the 50L and the 50 f1.4, the former is often misinterpreted as the 'high - end option', and to correct this Canon should introduce a better constructed, higher end 50mm than the 50 f1.4 to fill the gap. ( 50/1.8 IS perhaps).

The new Sigma and Otus should be seen as a different to the 50L.

 

[Loswr]

I'm pretty sure that you can introduce spherical aberration through plugins or other software components if you really so desire. What you can't do is correct for poor image quality at capture time.

There are polarization effect filters for post-processing, but they cannot properly replicate the effects of having a CPL on your lens at capture. Similarly, adding spherical aberration in post will not correct for poor bokeh in the captured image.

I'll just point out that I didn't mention polarization or bokeh, so I'll take your dalliance off topic as an indication that I was on the money but you can't admit it

Wrong again (or perhaps that should be, wrong as usual). Obviously, you don't comprehend the relationship between spherical aberration and bokeh, and the analogy of polarization failed to enlighten you. You're no more correct in this case then when you thought a lens was a camera, although I must say that was a particularly egregious example of your ability to totally miss the point. Frankly, you have made dozens of similar, if less colossal, factual mistakes in this forum, and your credibility is basically nil.

No. The problem that we're seeing here is something called "confirmation bias", where people find any reason at all to support the idea that the Canon 50/1.2L is better.

Again you miss the point. Is anyone saying the 50/1.2L is sharper? Not that I've seen. If you want to define "better" as sharpest, that's a judgement by you.

The problem that we're seeing here is something called "false-consensus effect," where people believe that everyone's definition of 'better' or 'best' is the same as their own, personal definition.

False-consensus effect:
http://en.wikipedia.org/wiki/False-consensus_effect
"...a cognitive bias whereby a person tends to overestimate the extent to which their beliefs or opinions are typical of those of others."

I can't wait for the Sigma 50/1.4 Art to be tested by DxO and for it to wipe the floor with the 50/1.2L. I can already see the posts from those with Red Ring Fever putting down DxO, etc. What a laugh that will be to see.

DxO already generates plenty of laughs with their Lens Scores. Is the EF 100mm f/2 really Canon's best lens? Is the 70-200mm f/2.8L IS II really worse than the MkI version it replaced?

 

[sanj]

"In the case of the 50L (both of them), it was an intentional design decision by Canon."

Neuro could you please guide me where I could read more about this? Find this so difficult to believe. Thx.

 

[sanj]

"In the case of the 50L (both of them), it was an intentional design decision by Canon."

Neuro could you please guide me where I could read more about this? Find this so difficult to believe. Thx.

In the press release Canon state that their target market is portraits, etc, which accounts for weak corner sharpness.

Hmmmm. Ok.
But they do not provide a sharp lens to people who want to shoot sharp portraits and sharp landscapes and sharp street and sharp journalism photos at wide apertures?

Are there not many uses to a sharp 50mm lens at wide f stop?

This is just not going down well with me. Am not being obstinate but find this logic incomprehensible.

 

[jrista]

"In the case of the 50L (both of them), it was an intentional design decision by Canon."

Neuro could you please guide me where I could read more about this? Find this so difficult to believe. Thx.

In the press release Canon state that their target market is portraits, etc, which accounts for weak corner sharpness.

Hmmmm. Ok.
But they do not provide a sharp lens to people who want to shoot sharp portraits and sharp landscapes and sharp street and sharp journalism photos at wide apertures?

Are there not many uses to a sharp 50mm lens at wide f stop?

This is just not going down well with me. Am not being obstinate but find this logic incomprehensible.

I'm curious why Canon HAS to make such a lens. Do you think Canon could do it cheaper than Zeiss, and that's why you want Canon to make one? There are SIGNIFICANT difficulties in making a lens sharp, corner to corner, at f/1.2. It would be EXTREMELY difficult to do so. It is even difficult to do it at f/1.4, which is clearly evident by the $4000 price tag the Otus has.

Why is it that you can't simply be satisfied with the fact that Zeiss has offered the exact kind of lens you want? Too expensive? If Canon made something similar, it wouldn't be any cheaper. If they made and f/1.2 version of the Otus, it would likely be significantly more expensive.

As for the rest of Canon's 50mm lenses, the 1.8 and 1.4 are VERY old lens designs, and the 50/1.2 is even getting a little dated. They were designed and built in an era where sensor resolution was lower than it is today. Canon surely has updates in the pipeline, and I'm sure when those new lenses hit the street, they will be competitive. Whether Canon chooses to compete with the Otus, or with the Sigmas of the world, is yet to be seen...but I would bet money that Canon ignores the Otus and sticks with what will sell in massive amounts: Something cost effective and affordable.

 

[jrista]

I'm pretty sure that you can introduce spherical aberration through plugins or other software components if you really so desire. What you can't do is correct for poor image quality at capture time.

Actually, spherical aberration is an effect in three dimensional space. You can simulate some aspects of soft focus in post, however those effects never fully replicate a TRUE soft focus. Neither can you change the nature of boke blur circles in post. Blur circles created by a lens with spherical aberration have a very specific aesthetic (brighter outer ring, with a clear spherical gradient to the center...it's a highly desirable trait for many photographers and cinematographers.)

The aesthetic effect caused by a lens with spherical aberration is not one that can be fully or easily replicated artificially in post. You can approximate some aspects of it, but for someone who likes the effect, those approximations NEVER measure up, and it is always obvious when it is a post-processed effect vs. a real optical effect.

Anyway, in the main the comments above about justifying Canon's current design and product are more about trying to ensure that people who worship Canon find a way to present Canon's offering as good and justified so that they feel good about owning Canon products. That's it. I'm sure someone will argue here that this comment is wrong, but you don't see anyone saying that they wish the 70-200/2.8 II had soft focus like the 50/1.2L and so on.

It has nothing to do with justifying or worshiping Canon. Your assuming something, then using your assumption to put words in peoples mouths as an attempt to win an argument. That's kind of you staple there, Dilbert. Why not try to put up a legitimate argument sometime, eh?

It simply has to do with exposing people to opinions other than their own. There is more than one way to design a lens, and there are reasons for designing lenses differently. I honestly do not think it would be good for every 50mm lens on the market to have exactly the same specs, offer the same exact IQ, produce the same aesthetic. It's better to have a diversity of options, because not everyone photographs the same things in the same ways that you do.

 

[jrista]

My 35 Art´s AF is drifting again (a third AFMA with Focal showed a further 4 step adjustment, on top of the 7 steps I got between the one I did when I got it and Christmas), so I must admit I am a bit skeptical to that part of sigma. But since so many are happy with it, I hope my AF problem is a one-off.

I have a technical query here:

As far as I understand, the purpose of AFMA is not to 'fix' defective lenses, but calibrate a specified lens to a given camera to account for manufacturing tolerances.
Once the AFMA is done, the camera knows how much to compensate for this lens, and everything is hunky-dory.

But in what condition can AFMA drift as is happening in Eldar's case? Is it because something is moving within the lens and a gap is getting bigger or a cog is becoming more loose?

I am particularly interested since I just acquired a 35A (so far it looks like it is focusing right on target as shown below- spot focused on "6" using a peripheral point and center point respectively), I haven't run it through FoCal yet.

AFMA is purely a camera body firmware thing. It only reconfigures the body, it does nothing with the lens. Drift is a pretty odd thing, but I'd like to know more. Spherical aberration can result in the focal plane shifting when you stop down or open up. Since lenses usually focus wide open, then stop down for the shot, spherical aberration can result in your focal plane ending up in an unexpected place.

The Canon 50mm f/1.2 and Canon DSLR bodies include firmware that compensates for this. There is a known component of spherical aberration in that lens (by explicit design), so the focus shift caused by it can be mathematically compensated for. When you have your aperture setting tighter than f/1.2, the firmware will focus the lens with a compensation shift to ensure that once stopped down, the focal plane is where you want it to be.

If the Art 35 has some spherical aberration, it is highly unlikely that such a focus shift is compensated for. That would require paired firmware between the lens and body. Assuming that is actually the problem. If the focal plane is shifting at the same aperture, then that is a different problem, and likely due to the lens, rather than the body.

 

[NancyP]

There. Is. No. Such. Thing. As. A. Perfect. Lens.

There. We have that settled.

Even the almighty Otus may not be perfect for some esthetics. I am starting to shoot LF (4 x 5 B&W film) and I am amazed at some of the interesting effects that people get by using the crude 19th century lens formulas ("Petzval lens" and the like). I am starting out with a modern era (1960s multicoated) lens, but there seems to be a large group of experimenters in LF.

 

[mackguyver]

There. Is. No. Such. Thing. As. A. Perfect. Lens.

There. We have that settled.

Well said

 

[jrista]

There. Is. No. Such. Thing. As. A. Perfect. Lens.

There. We have that settled.

Well said

Oh, if one could wish... These debates are never settled...closed minds cannot handle alternative points of view.

 

[mackguyver]

There. Is. No. Such. Thing. As. A. Perfect. Lens.

There. We have that settled.

Well said

Oh, if one could wish... These debates are never settled...closed minds cannot handle alternative points of view.

True. The closest lens to perfection I've found is the 300 2.8 IS II, but it still has some vignetting wide open and it's not exactly small or light. And it doesn't do macro. And it's expensive. Maybe the 50 1.8 is actually the perfect lens. Sharp, fast, not many flaws considering the price, and cheap enough to replace a whole lot.

 

[Loswr]

Would you like to address the comment I made about spherical aberration without referring to bokeh or polarisation?

Polarization was merely an analogy as an effect, like bokeh, that cannot be fully replicated in post-processing. As for addressing your comment about spherical aberration without referring to bokeh, the point is that the residual spherical aberration designed into the 50L is there because the lens designers chose to emphasize bokeh quality over sharpness for the designof the lens. If you're going to slam the 50L for not being as sharp as other 50mm lenses, the reasons behind that somewhat reduced sharpness are an integral part of that discussion.

Next up, let's discuss the interactions between planets…but we must avoid referring to gravity in that discussion. :

If spherical aberration is so important and necessary to photographers then why do lens manufacturers go to such great lengths to eliminate it?

As has been established and accepted by many people (other than you), Canon intentionally chose to not eliminate spherical aberration from the 50L design.

 

[Loswr]

btw, let me refer you to the wikipedia page here:
http://en.wikipedia.org/wiki/Spherical_aberration
what does it say about the top lens?
Is that the word perfect used to describe a lens without spherical aberration?

Gee, dilbert, yes it sure is. So, you're saying that you think a 'perfect' lens has only one optical element that focuses only red light? Do you have many lenses like that? : :

It's a diagram - a pictorial representation to illustrate a concept. The use of the word "perfect" is in contrast to the use of the word "real" to describe the bottom lens. A perfect lens is, "...an ideally corrected glass element that is free of aberration and focuses light onto a single point." Similar to the thin lens approximation, it's a way to simplify the relevant concepts so that most people can understand them…clearly, it's not a universally effective method.

"In the case of the 50L (both of them), it was an intentional design decision by Canon."
Neuro could you please guide me where I could read more about this? Find this so difficult to believe. Thx.

The relationship between spherical aberration and bokeh has been frequently discussed (for example, see this).

As for deliberately designing a lens with undercorrected spherical aberration, Nikon states, "When spherical aberration is left a bit undercorrected, flares surround the out-of-focus background, thereby resulting in a close-to-ideal out-of-focus background."

If you'd prefer to hear it from Canon themselves, in a white paper on Cinema EOS lenses, Canon states, "By a tightly controlled design, a small and precise amount of spherical aberration is introduced into these 4K lenses that has precisely the softening effect shown in Figure 7."

Yes, Canon designs and produces lenses that aren't as sharp as they could be, on purpose, by intentionally undercorrecting spherical aberration, and does so even with lenses costing $5,000 - $46,000.

 

[mackguyver]

"In the case of the 50L (both of them), it was an intentional design decision by Canon."
Neuro could you please guide me where I could read more about this? Find this so difficult to believe. Thx.

The relationship between spherical aberration and bokeh has been frequently discussed (for example, see this).

As for deliberately designing a lens with undercorrected spherical aberration, Nikon states, "When spherical aberration is left a bit undercorrected, flares surround the out-of-focus background, thereby resulting in a close-to-ideal out-of-focus background."

If you'd prefer to hear it from Canon themselves, in a white paper on Cinema EOS lenses, Canon states, "By a tightly controlled design, a small and precise amount of spherical aberration is introduced into these 4K lenses that has precisely the softening effect shown in Figure 7."

Yes, Canon designs and produces lenses that aren't as sharp as they could be, on purpose, by intentionally undercorrecting spherical aberration, and does so even with lenses costing $5,000 - $46,000.

You reference Nikon and say that Canon would intentionally make a lens less than perfect??? That's heresy, I say, heresy ;D ;D

 

[jrista]

@Dilbert: I refer you to Neuro's answers for all the spherical aberration stuff. Canon DOES purposely leave in spherical aberration by design, as it is a desirable effect in many circumstances.

As for your assuming, you assume that people are trying to justify what Canon does, when in actuality people are simply explaining what Canon does. You assume that people here "worship" Canon, when in fact some people are simply fans, others are simply customers and might otherwise not care about the brand. You assume a whole hell of a lot about people here man, and then you lash out at them with thinly veiled hostility and nasty words based on your INCORRECT assumptions.

All I'm saying is...might not want to assume, you would look like less of a donkey's rear end in the end.

 

[Eldar]

Up to a certain point it does have its entertainment value, but then ...

 

[Sporgon]

I guess it's the same reason why Canon have kept both the old 100 f2 and 85 f1.8 in the catalogue. You'll hear many people stating that the 100 is the better of the two because it's slightly sharper and has less purple fringing wide open. But the out of focus transition isn't quite as smooth and dreamy as the 85 - because the 100 is better corrected for chromatic aberration.