Canon 11-24/f4L

[Tannin]

Canon have a new ultra wide, the full frame 11-24mm f/4 L. I'm surprised there is not a thread about it yet.

The good news:

• 11mm at the wide end on full frame! That's even wider than the Sigma 12-24. In fact, the widest full-frame rectilinear lens ever made.
• Constant f/4.
• From specs and early previews, stunning sharpness, outstanding flare and and distortion control.

The bad news:

• Rear-mount gel filters only. The front element protrudes too far for orthodox filter use.
• Very big and heavy
• Jaw-dropping price. Apparently, it contains the largest series-production ground aspherical lens element ever made. At (from memory) 87mm, it's way larger than anything similar, and it's ground, not moulded. Grinding aspherical lenses is very difficult and expensive. Doing one this size ... well, that's no doubt why you are looking at over $4000.

More here: http://www.the-digital-picture.com/R...-USM-Lens.aspx

[arthurking83]

I'm sure that Lee(or others) will make a filter system to fit over the front of the lens, once some of these lenses get into the wild.

I have the Sigma 12-24, and it's so much fun to use.
I've adapted a Samyang filter holder(made for the 14mm f/2.8 lens) and use 150mm(wide) Hitech grads.
At the moment it's a bit fiddly to use, but this is due to my inability to cut and drill perfectly straight.

As long as ND gels are available for the lens, and those that need NDs are catered for then the gel holder fits a specific purpose (I guess).

At about 14-15mm (or so) and above a CPL could be a handy addition to have, but below this focal length range I can't see that a polariser will be of benefit .. other than to remove reflections off reflective surfaces(glass/chrome/water/wet surfaces/etc).
But for polarising a landscape, I think that the polariser at the really wide end is probably going to be superfluous.

I don't know if other Siggy 12-24 owners have commented on this possible occurrence, but at between about 12-15mm or so with the Siggy(definitely at 12mm tho!), it seems to polarise the scene in a slight way.
It's obviously not a proper polarised look, and as I don't own a polariser large enough to compare properly it's hard to explain.
But the sky takes on a polarised rendering in many situations when shooting at the wide end of the focal length range.
In fact so much so, that a few times I've done a panorama, the sky has been difficult to blend with a non polarised rendering .. ie. unevenly blue across it.

I'm wondering if other Siggy 12-24 have seen this look, and possibly successfully PP'ed it out.
I'm also wondering if this is actually the case, and if so, is it the lens itself(ie. the design Sigma has created) or the possibility that the wide focal length does this in some way.

Anyhow, back to the filters .. you can get 175mm, and I believe even larger, filters from Hitech, so I'm guessing that these filters may just allow the possibility to fit onto this new Canon lens and still give you an obstruction free image to the widest periphery.


I know Nikon made their 13mm f/5.6 which carried the title of widest recilinear lens ever made for years, and it's lens is larger than the new Canon(even tho it's Aperture rating is another stop smaller too!).
Lens diameter is over 100mm.
I'm not sure tho if it uses an aspheric lens element tho(can't find any info to confirm it)

EDIT: just checked .. the 13mm Nikkor isn't aspherical.

[wideangle]

A very interesting release by Canon, would be a great lens to check out for architecture photography as shooting at 16mm on FF is great but would be interesting to see the possibilities opened up by 11mm.

[Tannin]

I don't know if other Siggy 12-24 owners have commented on this possible occurrence, but at between about 12-15mm or so with the Siggy(definitely at 12mm tho!), it seems to polarise the scene in a slight way.
It's obviously not a proper polarised look, and as I don't own a polariser large enough to compare properly it's hard to explain.
But the sky takes on a polarised rendering in many situations when shooting at the wide end of the focal length range.
In fact so much so, that a few times I've done a panorama, the sky has been difficult to blend with a non polarised rendering .. ie. unevenly blue across it.

I'm wondering if other Siggy 12-24 have seen this look, and possibly successfully PP'ed it out.
I'm also wondering if this is actually the case, and if so, is it the lens itself(ie. the design Sigma has created) or the possibility that the wide focal length does this in some way.

I have frequently noticed the same thing, Arthur, but with a Tokina 11-17mm fish. Many is the time I've wished I could put a filter on the Tokina, then taken the shot anyway (having no other choice) only to find that it looks just the way I wanted it to when I was wishing for a CPL. I don't understand this. At a wild, wild guess .... no, that makes no sense. I was going to suggest some kind of effect going on at sensor level to do with the angle of light striking the elements, but that makes no sense - the lens has already bent the light such that it strikes at something close to perpendicular .... er .... or has it? I need to look at a diagram before my brain starts hurting again.

[arthurking83]

Funny thing is, I've never noticed this on the Siggy 10-20(on APSC of course) .. and even for a while when I was using it as my UWA on the D800E, I still didn't notice any strange CPL looking blue-ing of the sky effect.

Again, another strange thing I also just noticed when looking into more info about the (non aspherical) Nikon 13mm f/5.6 lens.
There's a site on the history of some of Nikon's lenses known as The Thousand and One Nights.
Some commentary about some of Nikon's lenses .. the 13/5.6 is one of them.
To add to the commentary, the author also uses some of the lenses(where he can) and posts sample images of various scenes with a good level of appropriateness.
The landscape scene sample posted with that lens used on a film camera seems to show this same CPL looking UWA partial blued sky effect too.

The effect can be nice looking, but can also be that annoying looking half dark blue sky effect.

So, if you are thinking of getting a super UWA lens such as this be aware that this effect can sometimes reveal itself.

ps. I just had a look at the Canon site for this lens, and they have an MTF chart posted for it's performance.
MTF looks good for example if you are into landscapes or any widescapes, but it doesn't look very good for something like widefield starry sky purposes(if you want critical point light source rendering).
If all you want is a really wide milky way image, it'll suffice, but the lens seems to be affected by astigmatism across most of the image area.
(at least that's how I'm reading the MTF graphs .. I'm not sure what the graphs lines represent as they aren't listed, but the traditional way they are set out indicates astigmatism from this lens)
But it does appear to be very sharp!!

Disregarding all this tho, it's still an impressive (lens)engineering feat!

[Morgo]

An interesting new lens for sure but I think for the price I'll be going with the new 16-35 f4 IS. Still, I might hold off until I see some reviews of it

[Steve Axford]

I don't think you mean astigmatism, Arthur. If it was, you could correct it by putting another lens in front of it. And I don't know how you decided it wouldn't be good for stars???? What lens would you recommend for Milky Way shots? It seems like an excellent MTF for an ultra-wide angle lens - almost unbelievably good. But we'll have to wait and see how it really performs.

[arthurking83]

Actually, you're right .. I should retract that statement about it not being an ideal lens for astro .. I just thought about it more and it may well be quite good.
And I did say it'll be great for general milky way shots if this is all you're after.
My comments were made about more serious wide field astro(where you're trying to capture every last ounce of detail in the stars.(but I am retracting that comment anyhow, as I thought about it more partly due to some of my experiences with the Sigma 12-24 I should have taken into account.


And yeah, definitely astigmatism.
(although I can't imagine adding another lens to such a lens!)

The astigmatism hypothesis will be common, due to the difference in the respective lines indicating resolution.

One line(usually dashed) shows sagital resolution(the contrast rendering of lines with an angle like this -> /
The solid line (usually)shows the same resolution figures but this time for the meridonal line line which is basically the contrast rendering of details at 90° to the sagital .. ie. detail in the \ direction.

If these figures are widely different, you may get sharply rendered detail if the subject is angled one way, it's a bit more smeared when the detail is rendered the other way(ie. at 90°).

It's not really unsharp, it's unsharp in one direction(that is, either / or \ direction) and so it's just rendered in a smeared kind of way.
Usually(from what I've seen) this circular point detail is usually rendered like a smeared arrow looking manner.

For widefeild star images, the Samyang 14mm f/2.8 has very good(theoretical) detail rendering ability.
of course a manufacturer's MTF charts will always be as a result of the most favourable conditions they can achieve
But if you look at the Samyang 14mm lens MTF chart, the two respective detail lines(ie. either the 10 ro 30 lp/mm lines) both meridonal and sagital graphs don't diverge too far apart.
Nikon's 14-24 lens has an almost similar graph to the Samyang ..

So if you happen to acquire any of these lenses iin ideal conditions(that is perfectly lined up lens elements), then in order of fast aperture wide image capture, the Samyang will render better and then the Nikon will render next best at the edges of the image.
At it's 11mm setting(according to Canon MTF) the detail won't be as sharply rendered wide open.
(I did eventually find some info on Canon's MTF charts tho)
Canon also show an f/8 MTF graph line, and this one(if I understand it correctly that is) shows OK detail rendering.
The problem with f/8 when doing astro/star images ... it's too slow. Therefore requiring either a tracking mount or really high ISO values(which kills detail rendering).

One thing I will predict about this lens tho, which gets me back to why I'll retract my comment about astro work.

I reckon Canon will do a better job of making this lens, than Sigma do of making their 12-24 ... so that at about 13-15mm mark this lens will make for very good images.

With my Siggy 12-24, while at 12mm it's not an ideal lens due to a lack of detail rendering at the edges of the frame, when zoomed into about 14-15mm, it's sharpens up nicely.
(I've yet to find the time to head out some time and do some astro images with it tho).
But from other more landscapey and static images I've got with the Siggy, it definitely gets better at 14or so mm .. still wide enough for most folks.
for astro work, f/4 is definitely the absolute minimum you'd want to use tho, unless you're ISO12800 or 25600 images come out clean as a whistle!
D800 is good to about ISO6400 tho beyond that noise clean up kills start detail .. so it's a losing battle. (I want a Df )

Note that the Siggy 12-24 has been much better than many sources on the net have made it out to be.
Yeah , it's a bit soft at the very edges(ie. corners) at 12mm wide open, but who shoots landscapes wide open anyhow?
Zoomed in a touch, it's even better again.

I should have taken this into account when I made that comment tho ..
It'd be nice if manufacturers gave us more detailed MTF charts that is at more focal lengths for zoom lenses and more aperture ranges too .. while they aren't the final word in how good a lens is for an individual, they do offer some indication as to what a lens can do.
While I think I have some differences between my Sigma 12-24, Sigma's MTF chart is quite close to how I see it's performance. (BTW, I think my 12-24 may have some slight defects which I'll get checked one day).

[Steve Axford]

I agree that MTF charts are very hard to interpret. People will often compare an f2.8 lens with an f4 lens based on an MTF chart. The charts show lens A at f2.8 and lens B at f4.0 and there could be a big difference. Often, it is concluded that the f4.0 is a better lens wide open, but if lens A is shot at f4.0 it may in fact be significantly better than lens B.
I was about to say that astigmatism is incorrect for camera lenses, but I see they do use the term. Generally in optics it means different magnification in 2 meridian axes, not a contrast difference in the sagittal vs the meridional. In the first case the effect can be corrected with glasses, in the second it cannot. I find it odd that they use the term astigmatism, because it is no the astigmatism that we normally associate with eyes.

[arthurking83]

..... because it is no the astigmatism that we normally associate with eyes.

Strangely I have no idea what astigmatism in eyes is .. other than it exists as a condition.
(I assumed that it was similar to lens astigmatism, but from your explanation it's different).

Zeiss have the best MTF chart information for their lenses .. using multiple apertures as well as many more resolution values.

But again, this is theoretical/hypothetical/predictive commentary.

MTF charts are only a guide on the potential performance of a lens ... so even then in your example of lens A at f/2.8 and lens B at f/4, because Lens A being an f/2.8 type, and hence by definition a more difficult engineering effort, there is usually a higher probability that Lens A is not perfectly made .. where Lens B may be better put together.
So in effect, Lens B can actually render better images even at f/4.

This is one of the reasons I like Lens Rental's analysis of lenses .. they tend to test multiple copies of lenses .. ignore the really bad copies completely.
Variation in lenses can be very significant .. even to the point as in your lens A-B example can show that the slower lens even at f/4(ie. wide open) can appear to be a sharper lens than the faster lens stopped down by one stop.

[Steve Axford]

Yep. To give a real world example. Most lens measurements say that the best aperture is f8-f11 (for ff) and diffraction cuts in after f11. Consequently, almost all internet instructions on focus stacking tell you to use f8-f11. I use f18, and I have worked this out by looking at photos at max resolution and deciding when diffraction cuts in (for my lens/camera). By using f18 I can get stacks that most people cannot see that there is any focus stacking, apart from the huge dof. I can also do it in the field using the focus ring. Most people don't realise that you can do this.
But back to the MTF charts. You have implied that an f2.8 lens is somehow compromised by being f2.8. I can't agree with that as my experience has been that f2.8 lenses are generally better at f4.0 than f4.0 lenses. There are exceptions of course, but generally the f2.8 lens is a better lens than the same brand f4.0 lens.

[arthurking83]

If it wasn't clear, I meant f/2.8 lenses are compromised more so than f/4 lenses only in that construction of these lenses is more difficult, than it is for an equivalent f/4 lens.
Large lenses required, tighter tolerances .. etc.

If you look at sample variation examples, I think you'll see more of the higher performance lenses will suffer from more variable build quality(to a degree).

Of course prices also comes into it, as the higher end f/2.8 lenses that are more difficult to produce are usually made in higher quality factories or assembly lines .. hence higher prices.

ps. apologies for taking this thread a bit OT.

Curious too tho .. are there many Canon folks out there interested in this lens .. at this price?

Sounds like a great lens for the Canon landscaper .. coupled with the 5Dr/s cameras .. great landscape combo for those times when printing large is required.

I dare say there will be a few AP members shooting with this combo pretty soon.

[Steve Axford]

I currently have a Canon 14mm F2.8 II and it is one of my favourite lenses. The new lens would be a possible, but it is expensive and where I make my money is macro, so it would be a luxury.
Incidentally, I do not use the 14mm wide open as it is difficult to use like that (the sharp areas are hard to predict with a very wide lens) and the bokeh is not particularly attractive (does any ultra-wide have a beautiful bokeh?). It is a great lens to use for all sorts of things and I'm not sure the 11-24 would be that much better.

[Ventureoverland]

Ive been looking for a good wide, full framer for my 5DMkIII for some time. The reviews that are now coming in for this lens suggest its spectacular and its only real draw back being the filter issue. Of course the other big issue for me is the cost, just cant justify it right now... which is a shame, because knowing that this lens is available will make it next to impossible for me to justify buying anything else as a stop-gap

[arthurking83]

@ Steve:. wide angle and bokeh seem to be mutually exclusive terms!
I've never seen any wide angle, let alone UWA, lens with good bokeh rendering. They all have a slightly nervous blur to them.

@ Vetureoverland: Don't completely discount the Sigma 12-24mm lens .. version II tho, which is what I have.
It depends muchly on your desired use for it too tho. As a get it all in with a single shot type lanscaping lens .. it's very good.
Central sharpness is on par with the best .. in fact blew me away at how well it renders at the centre of the frame.
Of course it is a slow lens, being f/4.5-5.6 .. but central sharpness is still high across the focal length range.
It's only at the periphery that the lens isn't so good when shot wide open, but from about f/11 it's probably good enough for most folks *unless you print in ultra high definition modes all the time.
Because I use it as a landscape lens only, I tend to be shooting at f/11 or f/16.

It can(or does) allow to do some astro/night imaging .. but not the most appropriate lens for doing so.
Slow aperture to start with .... and again, the not so sharp periphery rendering meaning that it needs to be stopped down a fair bit ... f/11 milky way shots therefore mean ISO104K if sharp points are wanted ... or trails only at 5min exposures.

Lens Rental have a review of this lens.
I'd love to see them test the Sigma lens too.

[Bennymiata]

I use my EF16-35 f2.8 a lot for real estate photography, but 11mm would be nice, especially for shoebox apartments.
I'll wait a year or 2 when the price will go down by 30%.

[arthurking83]

I use my EF16-35 f2.8 a lot for real estate photography..... .

Something to note with these types of lenses.

At some middled focal length this type of lens will produce zero distortion.

So for architectural type photography where straight lines look better than when they are bowed like a banana .. this will be a significant advantage in that less processing will be required(no distortion compensation needed).


The Siggy 12-24 produces zero distortion at about 14mm, so that when used at that focal length, it has an advantage over the Nikon 14-24 where the resultant files don't need correction.


For this Canon lens, an estimated guess that at approximately 13-15mm will be the sweet spot.

[wideangle]

Has anyone got this lens yet?

[Dylan & Marianne]

Has anyone got this lens yet?

noone replied to this for a while but Ian plant has his thoughts here if interested!
http://www.ianplant.com/blog/2015/03....6BBlTR4V.dpbs

[arthurking83]

LOL!
Thanks for the link Dylan.

Made my day.
Had a quick read of this Ian chap's site and got to this point and laughed ...

.... Soon my wide angle kit got too bulky for easy carrying and travel, so when Canon upgraded its 16-35mm f/2.8L lens, I found myself using my Nikon 14-24mm less and less. Sure, the Canon lens wasn’t as sharp as the Nikon, but it was small, lightweight, and filter use was easy....

I love the perspective he offers .. Canon 16-35/2.8 .. small and lightweight!

I like the way this guy thinks