Dear Canon Santa: I'd like RF 28-70mm f/2 MK2 with tripod collar, optical image stab, no dust sucking, no pealing coatings, and in body focus breathing correction.
Upvote
0
The hybrid primes won’t be all we get for fast wide-angle L prime lenses
- Thread starter Canon Rumors Guy
- Start date
Huh? The Bovine Scat meter just pegged on that one.
First, with a body that has a viewfinder with 100% coverage, the jpg output or the RAW output from DPP is exactly what you see in the viewfinder with a ‘forced correction’ lens.
Second, if you shoot RAW and convert with LR or DxO, you get a slightly wider framing (usually desirable with WA lenses, easy to crop if not).
Third, in DSLR and film days when all lenses were ‘optically corrected’ (yet still had 3 to >4 stops barrel distortion) only the very high end bodies had 100% viewfinders, most were in the 90-95% range so getting the exact framing you saw when you took the photo has really never been a thing.
You responded to a request for an example image with one line of BS text. Well done.
Upvote
0
Well the software chromatic aberration correction may eliminate the color fringing it doesn’t eliminate the fringing so you still get a halo effect just white instead of purple or green. I prefer an optically corrected lens, much better.
Upvote
0
I know about this lens. One of the few Nikon lenses that are really attractive.
Upvote
0
I brought RF35 1.4 to Mt.fuji last week and it deliver very good results. Size and weight also nice for hiking trip.
Upvote
0
Me too! with the exception of IS or a collar. Both not a deal breaker. I do really like the original and use it more than any other lens in my bag.
Upvote
0
OK, but again, why do you believe that?
OK, but what makes you think that the soft corners in the RF 16mm are due to software distortion correction, and not due it being an extremely compact sub-US$250 lens? Can you show me another $250 wide-angle made by anyone on any platform that doesn't need software distortion correction and has notably sharper corners? And can you post some side-by-side photos showing those sharper corners? Or just share a link to such, I don't want to ask you to actually do the experiment but just show me where you learned this.
Upvote
0
For the heavily distorted lenses made by Canon you cannot disable the distortion correction in the body, so the EVF and JPEGs will always have corrections applied. I’m not sure if they are forced on for things like RAW movies.
And to state the obvious:
For RAW stills (CR3 files), you can choose if you want correction or not in almost all RAW converters. DPP4 will likely force corrections on.
Upvote
0
but probably none of us would have liked said optically perfect 35mm's price.
Upvote
0
This is exactly what I meant: It's a matter of cost! You just cannot achieve everything with software.
Well designed lenses need less extreme corner stretching, which results in a loss of sharpness. An inexpensive lens needing extreme correction won't have corners as sharp as a better optical lens design.
Lenses without software correction and sharp corners? Summilux 21, 24, 24, 35, Summicron, 28, 35 etc...
EF 35 F/1,4.
And yes, the little 16mm is certainly worth its price, maybe even more. But it's a lens I would never buy, especially not for landscapes.
Upvote
0
I was never a big fan of the EF 24mm f1.4 II L. I've used it a fair amount when i shot a lot of UK weddings. I had the complete compliment of fast EF primes, 24, 35, 50, 85. 135. As well as f2.8 zooms and a few macro and fish eye options.
Optically, it' s a wee bit sharper than the EF 35mm f1.4 L (mkI) and similar contrast and colours. I'm not convinced it'll be particaulrly sharp on a R6mkII, let alone a R5.
For me and my shooting I like the EF 35mm IIL a lot more. It's more versatile for portrait work. I find that anything wider that 35mm doesn't seem to need the fast aperture that f1.4 glass brings. Creatively, the f1.4 doesn't do much compared to say a 50mm or 35mm and I don't need the fsater shutter speed either.
My old wedding combo was a 16-35/2.8, 35/1.4II and a 85/1.2II, each on a seperate caemra body. This worked well for me.
For landscapes, my 11-24 f4L, 24-70 f2.8 IIL and 100-400 f5.6 II L work really well and for wildlife / Birding my 100-400 f5.6 IIL with or without a 1.4x TC or my EF400mm f2.8 LIS II (with 1.4x or 2x TC) work amazingly well. I tend to find that for me, these three genres of photography have specific kit. Which for me don't include a 24mm f1.4.
I think it's really interesting news that Canon are bringing so many great options to the market place. The hybrid 35L, 50L and 85L will be very populr optics and might be of interest to me if I ever get more requests to shoot weddings again. However, I can also see the attraction of optically corrected f1.2 glass across the prime options.
Upvote
0
Again, how do you know that "extreme corner stretching" results in "a loss of sharpness?" I'm not arguing with you, I'd just love a link to an article that explains why, or an example photo or something. And what constitutes "extreme?"
Also not sure I follow you on "well-designed lenses..." The 16/2.8 is PHENOMENALLY well designed, to be sharper than my old EF14/2.8 from 1995, and far smaller as well, and now literally 10% of the price even ignoring inflation. That was $2499 in 1995 dollars. Now after 30 years of inflation $249 will get you the 16/2.8. If that's not good design I don't know what is.
Upvote
0
The thing is that lens design is a huge tradeoff design. You can make any one or even sometimes two or more aspects absolutely perfect, by simply letting the other things get worse:
- center sharpness
- corner sharpness
- astigmatism
- aperture
- OOF highlight shape
- vignetting
- distortion
- lateral chromatic aberration
- longitudinal chromatic aberration
- spherical aberration
- coma
- field curvature
- close focus ability
- autofocus speed
- dust/moisture ingress control
- size
- price
- weight
So the beauty of distortion is that it seems to be nearly PERFECTLY correctable in software. And if so, then the lens designer can improve any and all other aspects of the lens, at the cost of distortion, then fix distortion in software. This is like the one free lunch there is in photography. (A couple other aberrations seem to be fixable in software but frankly I don't know how they are done and whether the fix really is perfect***, as distortion is.)
People keep saying that a lens without distortion would be sharper but I am pretty sure (though still willing to hear people out) that the opposite is true. If you bend over backwards to get distortion perfect, you're going to be screwing up OTHER aspects of the lens.
*** as far as shapes, there's no reason software cannot get distortion to less than a pixel's width, so I'd call that perfect. I say that as a senior software guy.
As far as the cost it imposes, it has a theoretical cost in sharpness, by blurring the image at MOST a half pixel. Let's take the far right edge of an R5 picture as an example. The pixel at (8100, 2732) has to be moved to (8190.5, 2732), let's say. So the pixel at (8191, 2732) will be composed of half of (8100, 2732) and half of (8101, 2732). So it's blurred a half-pixel linearly. And yet is that visible? That's 227.5 pixels per mm. Or 113 lp/mm. So we're saying your contrast at 110lp/mm can fall to zero. Well what do you think your starting contrast was at 110lp/mm before correction? It was close to zero anyway. MTF charts only are 10, 20, 30, 40lp/mm from manufacturers and even on geeky sites like lensrentals I've never seen more than 60lp/mm discussed.
And that's the worst case scenario. You might find The pixel at (8000, 2732) has to be moved to (8090, 2732). So the destination contrast at 113lp/mm is literally the same as it was before distortion correction. So, half the "corrected" area isn't even going to have this theoretical loss of contrast.
Again if I'm wrong, just show me the example photo where the corrected version has somehow suffered for the correction, or is otherwise inferior to a similar lens that doesn't need correction.
Upvote
0
Similar threads
