[UPDATED] Can we get much wider? Canon EF 11-24mm F4 L USM samples gallery posted

For full disclosure, though, a grad ND filter might still be a better way to make an image like the one above, b/c by holding back highlights, you can increase the exposure on the shadows. This'll always yield cleaner shadows than underexposed shadows that have had to be pushed, even with high dynamic range cameras. This is b/c the shorter your exposure, the more shot noise, so you always benefit from collecting more light.

But a shot like the one above also doesn't have offensive noise in those rocks, so is perfectly acceptable. Grad ND filters of course have the disadvantage of potentially degrading your image (b/c of the resin), and darkening things you don't want darkened, b/c you don't have a perfect line of a horizon (cityscapes, for example).

So there are pros/cons to using/not using filters. Reason I bring all this up is b/c: it's no longer absolutely true that you absolutely need filters for 'spectacular' twilight light. Sensors have come a long way.

In film days I was a dyed-in-the-wool grad filter user, but come digital, I actually preferred the look of post-processing. My photoshopped images looked more natural than my grad-filtered images. So I no longer use the grads, and looking at my old film images, I get a bit of a cringe feeling now when I see those filtered skies, seas or whatever. Technology moves on, and our art benefits.

For horizon shots I use Reverse Grads or Strip Filters. But we have liveview, just use it and choose which filter will look natural..Thanks for the reply Rishi! Anyway we look at it, it is a very difficult lens to master! I'd love to shoot with it, maybe I will rent it first. I was shooting at Phuket Ban Naithon Beach against the sun I really did get the shot with my 14mmMk2 L with Lucroit Filters from sunset I used .9 reverse grad until the after glow, the light, the sky and cloud was spectacular! I have never seen light like that before and I captured each passing light. What made it spectacular was on the left side of me, the cloud and light closest to me was purple pink magenta with blue light! It was amazing! I was working frantically back and forth with my .6 reverse and .6 and .9 soft grad, I got natural shot because I shot them with Liveview, and no way that I was going to miss the shot! I have been to the same place dozens of time but last August 2014 was THE best.

Ecopix, I would like to know which sensor is able to shoot against the setting sun without using at least a 3 stop reverse grad. I have to use Benson 4 stops reverse grad. When printed, even my 7d looked spectacular on Hahnemuhle Fine Art Baryta Paper and nobody mentioned anything about me usingfilters. By the way, I use a 6d since June 2014. Just use the liveview, meter on the foreground and use filters for the photo to look natural. Sometimes I just use my X pro adapter ring and use blu tac to stick my grads so I can work quickly.

You're correct of course - we still often need different exposures for different parts of a scene to recreate what the human eye/brain can accomplish.

We're just talking about different ways of doing it. You prefer to do it in-camera; I found I got more natural results making two or more exposures and combining in post.

Two ways of doing the same thing, to overcome technical limitations so we can portray the world the way human vision would perceive it.

You've obviously perfected your technique to make inspiring images, so good on you. That's what it's all about.

EcoPix - I hear you. The grad filters make you make certain irreversible decisions ahead of time - such as where that hard edge will be. Which doesn't work very well for cityscapes, where you need a soft gradation so building don't immediately look harsh. But then the interesting light is usually when the sun is low in the sky, and that's where the Daryl Benson reverse 4-EV grad comes in handy - but then again that just eats into the city buildings... it's very difficult to do right.

Which is why you'd be much better off with just a 21 EV dynamic range sensor in Raw. We don't have that yet, but a D810 is pretty close. Keep it at ISO 64, and expose for the highlights (this is where I wish manufacturers had a Raw-based histogram). Bracket a bit around it, then pull shadows/midtones to taste in ACR. Do all the gradients/masking/luminosity masks you want there. Image average identical shots to decrease noise levels in the pushed shadows.

Are you referring to stacking in your last line, Rishi? That is a third way, and there is a quite technical discussion of that over in the Nikon 1 system forum at present, exploring full-fame IQ from those little cameras.
Turns out they trade DR for sheer shooting speed, which can facilitate median stacking in post. Many routes to the same destination.
Re skyscrapers, I work in forests and woodlands - same problem. It only takes one tree to upset a grad filter!
Thanks for the insights, and from abi, too.

Yes, EcoPix, exactly what I'm referring to. I often think it's easier to average a few of the same exact exposure than to average different exposures, b/c you don't typically have to worry about harsh transitions between two drastically different exposures, where masks become complicated (and I don't like the results of automated HDR software).

With multi-image averaging, as long as none of your tones are at/near the low read noise floor of the D810, you shouldn't have to image average that many shots to get clean, ISO 100-like pushed shadow performance (that looks like unpushed ISO 100, that is).

While it's rare that you'll need multi-imaging to reduce shadow noise w/ the D810, sometimes for extremely high DR scenes that you want to make large prints of, you may wish to clean up noise a bit. In which case, it helps that with the D810 few of your tones, if any, will be at the noise floor where SNR=1.

...

...

B/c to get SNR=1 tones up to, say, SNR=10, you'd need to average 100 shots (SNR goes up as sqrt of # images averaged). If those same tones are SNR=5 (much more likely for a D810, given its incredibly low noise floor), you only have to average 4 shots to get to SNR 10.

With my 5D Mark III, tones would easily run down to SNR=2 or 1, in which case multi-image averaging becomes of limited value, and it's easier to just to HDR different exposures. Otherwise, you'll find yourself averaging hundreds of shots... You're better off just increasing your exposure considerably to get all tones above the high read noise floor.

Probably a similar case for the Nikon 1 cameras, since they also have high downstream read noise, certainly not helped by the fact that shot noise is already pretty high because of the small sensor.

Thanks Rishi, that's a very good tip - I hadn't thought of combining the two like that. You should be paid more!

Classy comment!

I guess he is sad with the nikon-lineup ;)

Like all of Canon's products, it's still very overpriced.

Tyler C, why shouldn't Canon price it's product as it wishes? There are people whom will pay for a brand and Canon is a reputable brand. I see people throwing their money on Apple brand and why is that? For instance, iPod... they sound quality is mediocre compared to some competition portable players like Sony but yet people (the standard Joe) are willing to throw money on it because of the brand.

Clearly you don't know what f/2.8 is or why you would need it.

The 5Ds is already announced. Don't troll.

Better idea: Use Zeiss OTUS lenses on EVERY camera tested to get a REAL idea of how those cameras really perform.

Uhm, the 50MP Canon 5Ds and 5Ds R are going to available in June. Chill dude.

So there is no correlation between ISO and visible noise. Interesting.

This shot was exposed for the highlights (ETTR), and shadows were raised a lot. You should see the original shot actually... perhaps I'll post it in my gallery for fun tomorrow :)

Those rocks were probably pushed something like 4 stops or more. This scene was a very high dynamic range one, so you'd certainly have seen read noise in those rocks.

Considering how m43 is stuck at 16 MP while APS-C has now reached 28mm and will probably take off from there I don't see m43 making as much inroads as you think.

A lot of buyers are hobbyists and first time buyers who will see mega-pixels above anything else.

In the interest of correctness ;
APS-C has now reached 24mm (on some).
http://www.bhphotovideo.com/explora/news/new-rokinon-16mmじゅうろくみり-f20-wide-angle-lens-aps-c-and-micro-four-thirds

^^ What the heck are you talking about ? :P

I didn't think you would reveal yourself so quickly. You must be a sad person that the only authority you are able to practice is in a virtual world.

A nice and also informative reply, Samuel, thank you

I've been off the internet moving (actually an enlightening pause...) and wanted to look at the images you mention before replying.

Actually, I think both of those images could be interesting in the right context, although both are 'pushing it' as far as normative viewer response. At the least, the Porsche one gives opportunities for crop, another important point for say a journalist or other in need of a quick take as I think you were.

Indeed, I have been looking and want to look more at the dimensions of what happens with wide angle geometric and perspective distortion. Even with good spatial perception and all the academic backgrounds, the whole picture has not yet clicked into place, and I think this would be useful.

One further point I could mention on your interesting 'new realm' would be the work of a fine journalistic photographer who used to post at least here on DPReview.

(cont'd)

...this limit is really dumb, #2 here of 3)

He'd moved to Alaska and started a paper there (think he moved back out). His choice for wide angle was a mild fisheye Canon, on a 10D if I remember.

This actually gave wonderful shots for many of his subjects, indoors or out. I remember an extended family dinner, a basketball team, and group of fishermen showing off their very large salmon. Also a boat along a shore, other things. The look looked right, perhaps because it was something like what our eyes actually see and correct for; don't know, but it worked. It probably helped not to have to spend time on postprocessing also, if you think about it, and caught a lot of interesting details.

I was interested enough to try some work with a Sigma and one of their fisheyes on loan from a friend as we walked around interesting places in Basel one afternoon.

(also cont'd)

not really.

there's a few UWA's that cost similar or more.

and btw, Samyang did - it's the 14mm 2.8.

I'm aware of the 14mm Samyang. What they don't have is a rectilinear 11mm or 12mm prime, but I do see that coming as they already make them for mirrorless formats (a 10mm and 12mm)

Yes, it costs $1.30 Can. to buy one U.S. dollar right now. I think a lot of cross-border shopping has been curtailed!

That's what I implied. $3000 USD ~= $3900 CADきゃど. Even with the slightly higher tax rate in most provinces, you're paying much more with cross-border shopping.

Well, if you compare to Canon itself, this is the new high.

Is that the same "fluoride" in our toothpaste, thanks?

There are two areas where fluoRITE is used in lenses. As an anti-reflection coating mixed with calcium, OR as an crystaline material used to make THE ENTIRE LENS ELEMENT. The latter is used because of its ultra-low dispersion characteristic, to correct colour error or chromatic aberration. Fluorite is still the best material to accomplish this, despite the growth of crown-fluorite glass (ED) elements in recent years.

Fluorine is an element (F). If it is contained in a coating, like a fluorite coating, it is technically correct to call it fluorine coating (well "coating containg fluorine" to be precise). Less accurate than fluorite (CaF2) coating, but not wrong.

Fluorine is a yellow toxic gas. Fluoride and fluorite are compounds incorporating fluorine. They are what is used as coatings or lens element compositions.

RichRMA, yes, correct. But as I said, fluorine simply is the more generic term for anything containing the element F. Fluoride can be more accurate in some instances, fluorite in others (e.g., if used for the element's glass composition itself).

It is like talking about a magnesium alloy shell. Nobody thinks the shell would be made of pure magnesium. Same for a fluorine coating, nobody thinks it is made of pure fluorine anyway ...

It's not a generic term, it never was. i never even saw it used in conjunction with lenses until very recently. Magnesium-fluorite is an anti-reflection coating for lenses. Calcium-fluorite is a lens making component to eliminate chromatic aberration.

fluorine wasn't used on lenses until very recently.

Fluorite elements are entirely separate and different. canon started using fluorine coatings with I believe the fisheye zoom.

fluorite has never been used as coatings.

Fluorine gas an element, in its native state is never used. Entire lens elements are made by Canon and others from calcium-FLUORITE. Lens COATINGS for anti-reflection are made of magnesium-FLUORIDE and have been since the late 30's.

not the same, this is probably something similar to polymer fluorine combinations that allow for substances to slide off of lens elements. THIS is very recent - only pentax, canon and Nikon have fluorine coated lenses.

has nothing to do with the super spectra coatings,etc.

putting it in caps doesn't make your point any more informed, this is not a reflection minimization coating and most certainly this coating has NOT been around since the 30's.

Pentax I believe came out with it first, then canon and then Nikon. canon has it on their latest super tele's, and most of the newer L's.

maybe some RTFM first may be a good idea.

They need new applications for teflon since it's being discontinued in cookware. Mores the pity; the substitutes are crap.

i do belive you, but when you get some nice sample (and this is not 50-50 scenario, i would say something like 25% chance or even less for real good one), sigma is not shabby at all. I do own for ex. now dinosaur 15-30, its heavy, it's loud, but sharp and all that, for fraction of the price of some other lenses in this area.
anyway, nice to hear that new canon is great (it should be for that kind of money), how are corners doing, barrel distortions, flare resistance etc?

My first ultrawide zoom was the Sigma 10-20mm on my A100. My benchmark now is the Sigma 8-16mmじゅうろくみり on my NEX-7, and this isn't looking any better (or even as good?) on FF compared to the 8-16 on APS-C, except wide open in the corners. I'd consider the Canon 11-24mm for my A7II, but it's a lot more pricey than the Sigma 12-24mm (which itself is 1.5X pricier than the slightly better 8-16mmじゅうろくみり).

What I'd really like is a smaller, cheaper FF lens with the field of view and IQ of my APS-C 8-16mmじゅうろくみり -- it can be slower (even f/5.6 or f/8 wide open) to meet those goals. However, I will admit that this Canon is the most expensive lens I've ever seriously considered buying....

You're right :) exactly the same as you do.

@wedding photog. I wasn't talking to you.

@Cane. You will still be right, because it concerns me too. Anyway, read the news interesting, at least for self-improvement and for spending the time too :) I hope that one day I will be able to upgrade my cheap Samyang to something better :)

Cane ure talking to the world dear boy. Carry on Wedding Photographer

Probably not as wide as my Sigma 8-16mmじゅうろくみり

I thought it was your pet name .. I've seen worse in dpreview :p

Well, it is possible with some editing. Another possibility is to use several cameras at once which would probably still be cheaper.

"Cheaper"? As expensive as 11-24L is, it still costs less than two 6Ds with 24-105L kit lenses, which is probably the minimum you'd need to replicate a 11mm FOV using multiple cameras. And don't forget the extra tough tripod you'd need to support all three cameras at once.

Why do you need 6D for that? APS-C will do just nicely. And pretty much every professional already has some backup camera(s).
Sure it is inconvenient way of taking pictures, but if you do not take very wide FOV pictures very often it is better than paying $3000 in my opinion.

The moment you are willing to compromise on quality by mentioning APS-C that means you aren't the intended buyer for this.

Stitching is a big pain in the ass and I have given up on it. Prefer to get it as right as possible in the field itself than spend hours at home stitching.

Of course I am not gonna buy this lens either but if I had a ton of money I would as I appreciate UWA lenses for their creative distortions.

This is NOT a lens for penny pincher's like us. This is for people who want the best and are willing to pay for it.

You are right about me not being intended buyer for this - I want UWA only rarely. But you are not going to compromise quality by using APS-C this way. You can probably get even better results because you will use several cameras, which means not only more resolution, but also more sensor area used.
Personally, I find stitching acceptable on some occasions. It doesn't take much time on a good computer.
But I agree, if UWA is someones favorite style then this lens is a nice addition.

If you find a bird who doesn't mind it & a 5D strapped on its back...

Takes the cake!

Kicked out?!!! Remind Dan that he's not in New York anymore and that people are much more reasonable in the rest of the country—if he wants to shoot in a location he should call ahead for approval, or just ask politely when he gets there. You'll be surprised how accommodating people can be in normal cities like Seattle if you use an old-fashioned approach.