comotionfilms wrote:
I was one of the few who were hoping for more of a statement 35mm, not a hybrid lens that seems like a nice lens, but one full of compromises. I own a lot of 35mm lenses, as it’s what I most often shoot, and I’m just not that excited about this one, honestly. Everytime I think I’ll warm up to it, I see an image like this that gives me pause.
The photographer (96whiteknight) did a nice job, but I don’t like the field curvature, the way the outside edges that approach infinity, seem more sharp than the center (look at the brick wall and the palm trees). I also typically turn off corrections with my lenses, and this one is a bit of a mess without them. To each their own, I’m just not a fan, yet. The closest lens I have to a perfect 35mm is actually the sigma 40mm, but it’s not a 35, so… the next closest is the Tamron 1.4, though I also don’t hate my sigma art 35 (it’s obviously softer, but in a nice way). I occasionally still shoot with the v1 of the L 35 1.4 when I’m interested in a bit of a look. But when it comes to chasing kids, it’s the 1.8 for me, shot at a t2. It’s really nice from about 10 feet in, and it’s very easy to carry without being in the way of all of the dad stuff.
Shooting Canon RF with corrections off is what created the framing problem with this image. Is it that with the corrections on, the image was simply just not wide enough? I am just trying to understand the logic of shooting with corrections off, is it due to correction warping and distortion close in?
As we know, this lens is at about 32mm uncorrected, framing for 32mm and then blaming an uncorrected lens that is designed to be used with electronic corrections might be the problem here. I suggest shooting RF corrected, and then you can always turn off corrections in Lightroom when needed.
The warp tool in photoshop will allow an acceptable correction of this image in photoshop BTW, but you probably already know this.
Planetwide wrote:
Shooting Canon RF with corrections off is what created the framing problem with this image. Is that with the corrections on, the image was simply just not wide enough? I am just trying to understand the logic of shooting with corrections off, is it due to correction warping and distortion close in?
As we know, this lens is at about 32mm uncorrected, framing for 32mm and then blaming an uncorrected lens that is designed to be used with electronic corrections might be the problem here. I suggest shooting RF corrected, and then you can always turn off corrections in Lightroom when needed.
The warp tool in photoshop will allow an acceptable correction of this image in photoshop BTW, but you probably already know this.
Its actually a good question...If I wasn't on travel I would test it. Does the EVF show the pre-corrected image? If so, this could make framing challenging since 3mm on the wide end with a 35mm is somewhat significant.
In this case, did the photographer miss the frame and then distort to correct the framing issue?
RustyRus wrote:
Its actually a good question...If I wasn't on travel I would test it. Does the EVF show the pre-corrected image? If so, this could make framing challenging since 3mm on the wide end with a 35mm is somewhat significant.
In this case, did the photographer miss the frame and then distort to correct the framing issue?
I have the lens, and the EVF shows the corrected image when corrections are on.
I actually like what Canon is doing here with corrections. By designing a "wider" lens, and then correcting, they are essential cropping the weakest areas of the image, and using a higher resolution inner part of the image for the corrected area.
Planetwide wrote:
Shooting Canon RF with corrections off is what created the framing problem with this image. Is that with the corrections on, the image was simply just not wide enough? I am just trying to understand the logic of shooting with corrections off, is it due to correction warping and distortion close in?
As we know, this lens is at about 32mm uncorrected, framing for 32mm and then blaming an uncorrected lens that is designed to be used with electronic corrections might be the problem here. I suggest shooting RF corrected, and then you can always turn off corrections in Lightroom when needed.
The warp tool in photoshop will allow an acceptable correction of this image in photoshop BTW, but you probably already know this.
Framing problem? Sorry, I’m not following… did you mean to quote me, because I was talking about the focus plane and not any framing issues, so I’m a bit confused.
I often shoot without corrections because I typically like the way lenses naturally falloff, vignette and even bend at the edges, but I’m more of a character lens guy.
comotionfilms wrote:
Framing problem? Sorry, I’m not following… did you mean to quote me, because I was talking about the focus plane and not any framing issues, so I’m a bit confused.
I often shoot without corrections because I typically like the way lenses naturally falloff, vignette and even bend at the edges, but I’m more of a character lens guy.
If you turn corrections on, and the slide the distortion slider, you can adjust the amount of correction applied. Perhaps that will get you, your desired look.
Planetwide wrote:
If you turn corrections on, and the slide the distortion slider, you can adjust the amount of correction applied. Perhaps that will get you, your desired look.
Honestly, I’m happy just leaving it off. If distortion will be an issue for me for me with a specific setup, I just chose a lens that suits what I’m after.
comotionfilms wrote:
Framing problem? Sorry, I’m not following… did you mean to quote me, because I was talking about the focus plane and not any framing issues, so I’m a bit confused.
I often shoot without corrections because I typically like the way lenses naturally falloff, vignette and even bend at the edges, but I’m more of a character lens guy.
If you are a character lens guy, I would think you would love the 35 vcm. I wouldn't judge the lens by a weird rendering of a brick wall at the top of one photo someone posted here.
rscheffler wrote:
@RustyRus@ as a fellow Leica shooter, I think you're familiar with the decreased blur towards the image corners that @comotionfilms@ does not like, given how common it is with wider rangefinder system lenses. I tried to explain it to the best of my layman's abilities in the 35 VCM thread, which you may have seen. Basically, high optical vignetting means that light rays forming the peripheral parts of the image effectively pass through a smaller maximum aperture opening than on-axis light rays. Smaller effective aperture means more depth of field in that portion of the image.
Correct.
In the example image in the post above, the cable and palm branch at the top left are sharp, which are probably close to the plane of the subjects, so I don't think it's traditional field curvature causing this (or only traditional FC). That said, the 35's MTFs suggest that the notable divergence of sagittal and tangential planes towards the image periphery might mean one of the two planes does have field curvature and will increase apparent sharpness away from the intended flat plane of focus.
Not quite correct. In this case the divergence is due to lateral chromatic aberration.
Planetwide wrote:
I have the lens, and the EVF shows the corrected image when corrections are on.
I actually like what Canon is doing here with corrections. By designing a "wider" lens, and then correcting, they are essential cropping the weakest areas of the image, and using a higher resolution inner part of the image for the corrected area.
From an engineering POV, makes sense.
However, I think Canon is cruising for a wake up call with this design approach. At some point, I expect the company to be hit with a consumer class action basically alleging that the company has not disclosed the resolution loss arising from the digital correction. Specifically, the issue will arise because the lens profiles not only cut off the edges but also rescales the image to the native resolution, but nothing I've seen from Canon's literature discloses what resolution the lens actually delivers before the profile rescale.
E.g., if the lens profile crop results in 95% of a R5's native resolution actually being shown, the fact that the lens actually delivers a rescaled 42.75 MP image could be deemed materially misleading.
To be clear, it's true other lens makers also handle profile corrections similarly, but Canon is clearly engineering the profile correction as part of the lens design in some cases and so IMHO sometimes seems to be pushing it pretty far.
fraibert wrote:
To be clear, it's true other lens makers also handle profile corrections similarly, but Canon is clearly engineering the profile correction as part of the lens design in some cases and so IMHO sometimes seems to be pushing it pretty far.
This has been a 'feature' of mirrorless lenses from nearly the get-go. Whether just stretching due to dark corners or actual mechanical vignetting (cutting off corners), lenses are being designed to requirements that aren't exactly 'covering the full sensor'.
Thing is, this is know through reviews - and output isn't so affected that the lenses aren't fit for purpose or even uncompetitive really.
fraibert wrote:
From an engineering POV, makes sense.
However, I think Canon is cruising for a wake up call with this design approach. At some point, I expect the company to be hit with a consumer class action basically alleging that the company has not disclosed the resolution loss arising from the digital correction. Specifically, the issue will arise because the lens profiles not only cut off the edges but also rescales the image to the native resolution, but nothing I've seen from Canon's literature discloses what resolution the lens actually delivers before the profile rescale.
I think the MTF charts do just that. Although in the case of Canon (and many other manufacturers) it is computed MTF, not measured MTF, so actual samples will be worse.
E.g., if the lens profile crop results in 95% of a R5's native resolution actually being shown, the fact that the lens actually delivers a rescaled 42.75 MP image could be deemed materially misleading.
The only resolution that counts is the ability to resolve fine detail in the final image. The rescaled 42.75 MP sensor image is a true 45 Mpx output image that (in the case of the RF 35/1.4) has a higher resolution across the frame than unscaled images from 35-mm lenses of the past.
Use these lenses as intended, i.e. with the profile correction, and only then you can complain if you don't like the results.
To be clear, it's true other lens makers also handle profile corrections similarly, but Canon is clearly engineering the profile correction as part of the lens design in some cases and so IMHO sometimes seems to be pushing it pretty far.
Other manufacturers are pushing it pretty far too. Sometimes at least, it depends on the lens.
fraibert wrote:
However, I think Canon is cruising for a wake up call with this design approach. At some point, I expect the company to be hit with a consumer class action basically alleging that the company has not disclosed the resolution loss arising from the digital correction.
Any such suit would be groundless, as using these lenses without corrections is neither documented nor supported by Canon. So, there is nothing to "disclose".
The published focal lengths and MTF charts reflect the expected performance *with* digital corrections applied. (This only makes sense, as the digital corrections are a part of the lens design.) Buyers have ample opportunity to judge whether the lens is suitable for their use cases prior to purchase.
I personally think that vignetting and lateral CA are bigger issues than (the corrected) distortion in some of these lenses I have used.
johnctharp wrote:
This has been a 'feature' of mirrorless lenses from nearly the get-go. Whether just stretching due to dark corners or actual mechanical vignetting (cutting off corners), lenses are being designed to requirements that aren't exactly 'covering the full sensor'.
Thing is, this is know through reviews - and output isn't so affected that the lenses aren't fit for purpose or even uncompetitive really.
Known through reviews probably isn't going to cut it in a legal case--courts aren't going to require buyers to read through third party resources to have a good idea of what they're purchasing. And for these kinds of things, it isn't necessarily a fit for purpose or competition thing. (Sometimes, it may be considered "common sense" or "known to the industry" or whatever but still misleading in connection with the average consumer.)
It may or may not ever happen. Just something that has stuck me lately, especially with the 35 VCM's lens profile being a topic of some debate.
Planetwide wrote:
Doesn't Canon quote "effective pixels" in their literature...
Yes-
This thread has taken such a silly turn-
Its a 50 Megapixel sensor in the R5mk2 to compensate for IBIS etc- Its very possible that the engineering team took into account the extended edges so you aren't loosing anything in terms of resolution.I am not engineer at Canon so I don't know how it works.
So the extra 5mp around the edges- Is that what is cut? Maybe-
fraibert wrote:
Known through reviews probably isn't going to cut it in a legal case--courts aren't going to require buyers to read through third party resources to have a good idea of what they're purchasing. And for these kinds of things, it isn't necessarily a fit for purpose or competition thing. (Sometimes, it may be considered "common sense" or "known to the industry" or whatever but still misleading in connection with the average consumer.)
It may or may not ever happen. Just something that has stuck me lately, especially with the 35 VCM's lens profile being a topic of some debate.
I doubt it will happen. If you buy an RF 24-70 F/2.8 you have more reason to complain, because you are getting an RF 25-68 F/2.9. There is no way to shoot at 24 mm, 70 mm, or at f/2.8. Lens manufacturers have been doing this for ages, and apparently very few people care. With the 35 VCM you can at least shoot in raw and retrieve your 45 native megapixels in a third party editor.
I preordered the RF 50mm F1.4 L VCM when it was announced, and it just arrived today, believe it or not! Haven’t had time to do anything other than open it up and throw it on my R6II to see if it works, but initially it seems like a great lens. I wish is was a little shorter, but oh well.
I’ve owned every non-L 50mm AF (1.8 and 1.4) lenses that Canon has made, but have been really wanting a new 1.4 for years! Glad it’s here. I learned photography shooting with just a 50 and I think they make for a great “all around” focal length. I am very hopeful this lens will perform wonderfully, especially hoping for faster AF with it, so will see….
filmjuicer wrote:
I wish is was a little shorter, but oh well.
These lenses are internal focusing making them good candidates for metal hoods (unlike the EF 50mm f/1.4 or the current set of RF STM primes). You could try a short, 67mm metal lens hood if you want to shave off a bit of bulk and length.
