The first review of the Canon RF 10-20mm f/4L IS STM is here

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
526
361
Which means that Canon is using software to correct the dodgy optics and we lose data due to the stretching and bending and adding of exposure to the far corners
Fair enough! Show me an image where this is the case! As I've written image processing software I understand the technical possibility, but I'm curious if there's even ONE actual real-world case you can show where a picture is detectably inferior for this reason.

E.g., you might complain that a highlight in the far, deep corner that still held a bit of detail was blown out thanks to this software correction. But if you hypothetically had the exact same lens spec with zero optical vignetting... it'd be blown out anyway!

E.g., you might have a picture with Zone III barely-detailed shadow across the frame, and a hypothetical "zero optical vignetting" would capture that in the far, deep corner as well, while the software-corrected image may be ZoneII (not quite black but still devoid of detail) or even Zone I (blackest black). And further, it needs to be a shot where simply increasing exposure about one stop would also notably degrade the image in other ways, such as by increasing camera shake or subject movement or noise or blowing out highlights in the center region by an amount people could detect. OK, fair enough, such a technical possibility exists. But that's basically your only possibility. Can show me a photo of such a subject? Doesn't have to be yours, doesn't even have to be Canon or a known lens. Doesn't really even need to be wide-angle! It just needs barely-detailed shadow across the frame all the way into the very, very, very corner, where adding a stop of exposure also would visibly degrade, so I can see what kind of image may possibly be less shootable with this lens.
 
Upvote 0

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
526
361
I'm not happy about the rear gel filter holder. After 20 years of this type of optic (the Sigma 12-24mm being the first of this type), some one really needs to develop a better filter system than rear gels.
I had it on my EF 14/2.8L in 1995, and I'm not sure it was the first. I think the FD system had the same filters, no?

What are these filters you're using, anyway? How is Photoshop not sufficient? (If it's ND, I'm predicting the R1 or R5.2 will have electronic ND.)
 
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,657
4,237
The Netherlands
Doesn't LR by default have profile correction applied to all new images? I've tried turning it off, and then an update puts it back on by default. This is not how LR behaved several years ago.

It will be interesting to see how long it takes for the 10-20mm to get added.
You can assign presets to get applied on import based on criteria like camera model. I’ve set it to always apply the profiles, since LR didn’t do that by default.

So you can set it to not apply it during import if you want.
 
  • Like
Reactions: 1 users
Upvote 0

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
526
361
Most folks will be better served with the RF 14-35/4... Is 14 mm wide enough?
Hear hear. I've shot the 14mm a LOT and I've never lost a shot or had a photo simply be dissatisfying for not being wide enough. (I also had the 20-35, 17-35 then 16-35, and I felt 14 was getting me something that 20mm or 17mm wasn't. At 16mm I start to wonder though.) I'm sure there are some people who have actual specific examples where 14mm wasn't wide enough and I'd like to see the shots.
 
Upvote 0

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
526
361
One of the things I've loved about the L EF lenses were their durability vs the reast of the competition and in recent years, Canon have been under-enginering the durability of thier lenses. Has anyone else noticed that the newer "easy to clean" flourine coated front elements are a lot more susceptable to scratching?
What I've noticed is that the plastic barrels show scuff marks more easily. As I was selling my decade- or two-decade-old EF glass and taking hi-res photos of the gear, I was surprised to see how there were no detectable signs of use, or maybe a few hairline scrapes, once I gave them a damp cloth.

On the other hand, Lens Rentals' lens teardown results--if I understand them correctly--say that the RF lenses have the best construction they've ever seen, that the quality is all inside where it really matter. I haven't seen any fluorine scratches on my RF gear yet.
 
  • Like
Reactions: 3 users
Upvote 0

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
526
361
I'm not finding a good chart that helps visualize the difference between 10mm and 14mm (or 15mm) on a FF body. Finding some charts, but most stop at 14mm and then jump to a fisheye. Also found a table of numbers on the Nikonians forum, but that's about it.
Say you're at the 20th floor of a skyscraper, shooting the skyscraper on the other side of the street.

With the 10mm lens, you will see exactly 14/10, in other words 1.4x, more floors of height, and windows of width, than the 14mm.

Another way to say that is that a flat thing that runs parallel to the sensor at any given distance will let you see 1.4x wider with the 10mm.

People try to complicate it by using trigonometry to calculate the width, but it doesn't really help when you're getting this wide. I'll show you how to derive it though. the sensor is 36mm wide, 24mm height, and has a 43.3mm diagonal. Pretend you have a fixed viewpoint of a point, and hold that rectangle 10mm in front of the fixed viewpoint. What angle you see? Get some graph paper and a pencil. Draw a point, draw the sensor as a line in front of it 36 squares wide (1 square = 1mm) and 10 squares above the point. Draw triangles from the viewpoint to the line center and from the viewpoint to one of the line ends. The angle subtended by the two lines attached to the viewpoint are half the view angle. So, we half the width of the line, use inverse tan to calculate the angle give the ratio of opposite to adjacent lines, and double the result to account for us measuring only half the angle.

In this case, it will be 2* inverse tan( (36/2) / 10 ) = 122 degrees wide. Likewise, 100 degrees tall, 130 degrees diagonal. In contrast the 14mm is 104 degrees wide, 81 degrees tall, and 114 degrees diagonal. So if you're in a circular room, you'll capture maybe about 18 degrees more, but I don't think this is really representative of how people shoot with non-fisheye lenses.
 
  • Like
Reactions: 1 user
Upvote 0

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
526
361
Very uninspiring images in the review, clearly not the way to use this lens creatively.
The interior with the steel and glass construction is Tokyo's International Forum, basically across the street from Bic Camera Yurakucho branch. I'm guessing the reviewer bought the lens there, and further, that he bought it at that specific shop because you can get down to Yokohama without changing trains, and given the other shots are in Yokohama I suspect he lives there so had them send his 10-20 to that particular branch of Bic. International Forum is used a lot in such lens sample shots. The dusk exteriors are the Yokohama Landmark Tower and the neighboring ferris wheel. The wooden house interiors are the Yamate Seiyo-do (Western Hall, an preserved old house) again in Yokohama.
 
Upvote 0
You can assign presets to get applied on import based on criteria like camera model. I’ve set it to always apply the profiles, since LR didn’t do that by default.

So you can set it to not apply it during import if you want.
I have never really got the use of presets. Maybe because I have developed my own look for different genres and happy to adjust each one. Similarly, the film simulations never appealed to me as (I guess) that I never really played with different film types for a "look"
I will try LR with a preset for that lens and see what happens.
Generally, I use the attribute for the 8-15mm lens and select all and sync the setting with profile turned off.
 
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,657
4,237
The Netherlands
I have never really got the use of presets. Maybe because I have developed my own look for different genres and happy to adjust each one. Similarly, the film simulations never appealed to me as (I guess) that I never really played with different film types for a "look"
I will try LR with a preset for that lens and see what happens.
Generally, I use the attribute for the 8-15mm lens and select all and sync the setting with profile turned off.
The presets I use are all technical: enable optical corrections and select the Camera Faithful profile for old bodies and colorfidelity profiles for recent bodies.

Nothing artistic like prosets of fropacks :)
 
  • Like
  • Haha
Reactions: 1 users
Upvote 0

YuengLinger

Print the ones you love.
CR Pro
Dec 20, 2012
3,784
2,325
USA
Jared has his review up
I never thought I'd say it, but thank goodness I watched Fro's first-look. Helped me understand why the images look great in DPP, but still wretched in LR Classic. Whoa! In LR Classic the petals of the lens hood are clearly visible.

And, without the lens profile in LRC, quite a bit of work.

Closest profile that does a fairly decent job is for the RF 14-35mm. EF 11-24mm not good at all.

Very, very sharp. Good AF in low light provided I up the exposure a bit.

Fortunately, DPP can transfer images to PS CC, as the lens-corrections issues are nearly non-existent that way.

Learning curve! The youtube videos I saw for creating a LRC lens profile involve a substantial time investment. Hopefully the LRC updates come soon.
 
Upvote 0

mpeeps

Lovin' life on the Central Coast
CR Pro
Dec 5, 2013
105
80
California
www.mpeeples.com
A 21mm and a 24mm gives a different look to an image. It's easy to see this focal range as just a number, and yet thet create different looking images. Another example is the difference between a 28mm and a 24mm view. They seem to create a slightly different look and feel to an image.

I agree, it's the general build that I was commenting on. My Old EF 135L has been dropped several times, it's a very front heavy lens that likes to topple out of my camera bag (these days i use a camera bag where I can aly this lens down on it's side). My copy has dropped out of my hands and bounced onto a cobbled pavement more times that I would care to mention. It has a polycarb body shell and the part the needed replaceing the most was the metal lens mount ring, three times I've had that part replaced on my copy. I dropped my 70-200/2.8 II LIS once on a hard floor and that was far more catestrophic. It's got a lot of metal in it's construction, but as another has commented, it's also a lot bigger and heavier. So it hit the floor with a lot more mass. I'm sure dropping 700g lens is less of an engineering problem than at 1.5Kg lens. I also think that the extra engineering that is required to make silent AF systems for Videographers and the inclusion of IS units and elements makes a lens slightly more fragile than the older and simpler designs. I wonder if the new (and rather spectacular) RF 135 L IS is as resistant to cobbled pavements as the old EF model, not that I would like to intentionally bounce a RF 135 L IS! Subjective, I know.


From my experiances of using EF glass professionally over the last 20 years it is apparent to me that the newer coatings are slightly more suspeptable to scratches than the older coatings, but way easier to clean. I far prefer the newer coatings, especially with the newer BR and SWC. The newer lenses flare way less.
One of the things I didn't mention in my observation of the test's photographs that were presented was the amazing lack of ghosting and flare when shot directly into the sun, really clean and spectacular Sunstars....possibly the best I've seen so far.

I also think that the newer EF and now RF lenses are generally a lot shaper wide open than the older lenses, so they can take the very slight Sharpness hit from a clear filter. The older lenses...less so. Squeezing every last drop of sharpness from the older EF lenses was a priority. I used to advise my collegues to just use the hood for protection....these days a clear filter is the better and safer option.


Yes I sold that rig about 6 months after investing in it (just after launch of the range, I was one of the early adopters of the system).
I had a 3 & 6 stop ND filters, plus a CPL. Some of the worse colour casts I've seen on a ND filter. The CPL was crazy sized.
The problem with the system was
a) dealing with the sheer size of the filters (it was like carrying a set of small dinner plates). It would take me so long to assemble the rig that often my moment had passed.
b) dealing with bizarre reflections and chronic additional flare that seemed to appear.
For me, both be 10-20 and the 11-24 are niche lenses offering an ultra ultra wide angle. The use case is different from, for example, the 14-35/4 which can serve as both a UWA and walk around lens.



A fair comment.
Let me explan my point of view in more detail and please feel free to dissargree and comment yours. It's an awesome forum and we all have different perspectives. Debating and sharing our views helps with our education and I'm more than happy to be proved wrong...it's a way of learning. I may be a bit of a luddite, but I'm not intransient. I'm always up for a re-learn.

If a lens needs heavy correction and several stops of vignette exposure correction, there where does that data come from? Obviously the image file's DR. If a lens needs 2 stops of edge brightening, then that's 3 stops of extra noise and three stops of less DR I have as a photographer to use in post processing. If the image is being stretched and pulled significantly, then again that can affect the post production. Sure, I grant you that Ai is getting better all the time, but at the moment...it's a simple negative vignette mask and a geometric distortion propfile that is being applied.
Sure this available technology isn't there to literally cut corners on their lens design but to help photographers get the best results possible in their photographs.

It's just occurred to me how crazy it is that we are dicussing the particular merits of 10-20mm and 14-35mm lenses....it's amazing how these focal lengths are so much easier to access these days. This new lens is a lot smaller and lighter than the older EF 11-24mm As you stated earlier, a RF 14-35mm offers the average photographer a massive range of ultra to wide angle options, far greater than the older EF 16-35mm f4 ever did. With my EF 11-24mm lens, I'm regularly seeing compositional requirements that are specifically 12mm, 14mm, 16mm, 19mm, 21mm, 24mm.
Such a civil conversation! I don't usually comment on these threads, but I'm obsessed and read them all. It's nice to see conversations that are not happening out there in our political/social worlds.
 
Upvote 0
I had it on my EF 14/2.8L in 1995, and I'm not sure it was the first. I think the FD system had the same filters, no?

What are these filters you're using, anyway? How is Photoshop not sufficient? (If it's ND, I'm predicting the R1 or R5.2 will have electronic ND.)
I'm using Meiki drop in filters because I'm using EF glass. Gels never worked well for me. I get to use the same filters across my entire lens collection. Having the filter behind the main elements vastly reduces flare and ghosting. It's a system that i really like.
I have a VND, CPL, 5 stop and 10 stop ND's. Their ND's are the most neutral that I've ever used. They blow away my expensive screw on Heliopan filters.
 
  • Like
Reactions: 1 users
Upvote 0
Jul 21, 2010
31,228
13,089
Just FYI I recently installed DXO PhotoLab 7 and they now have profiles for the RF10-20 mm f4 lens.
Good to know, thanks. FYI, lens profiles are typically version-independent, it's only camera bodies that require a version of DxO PL more recent than the camera. So now DxO has the 10-20/4 profile, but I don't yet have the lens. I do have the 24-105/2.8, but DxO does not yet have a profile. Sigh.
 
  • Like
Reactions: 1 user
Upvote 0

john1970

EOS R3
CR Pro
Dec 27, 2015
994
1,235
Northeastern US
Good to know, thanks. FYI, lens profiles are typically version-independent, it's only camera bodies that require a version of DxO PL more recent than the camera. So now DxO has the 10-20/4 profile, but I don't yet have the lens. I do have the 24-105/2.8, but DxO does not yet have a profile. Sigh.
I am still waiting on the RF 24-105 and RF 200-800 profiles as well, but I hope we have them in the next few weeks.
 
  • Like
Reactions: 1 user
Upvote 0
"I do keep a B+W clear filter on lenses (except the really cheap ones)"
You pay more than $2500.- for a premium lense to get the best possible sharpness and than you mount a clear glas filter to protect the front lense? The front lense is proteced enought if you mount the lense hood. I don't understand this kind of care understanding. Just keep care - that's is protection enought. If not, you should change your handling.
 
  • Haha
Reactions: 1 user
Upvote 0