Lens design comparison: Canon EF 50mm f/1.8 STM and the Canon RF 50mm f/1.8 STM

SteveC

R5
CR Pro
Sep 3, 2019
2,678
2,592
Do designs that have worked well in the past still meet the demands and expectations required to be considered as working well to day ?

For photographers, almost without exception, YES. For forum nerds, self appointed YouTube experts, ‘reviewers’, and people that just love to argue, no.

But what about lens design rot? I'm sure I've seen designs rot with age.

(And yes, I guess that makes me a "people that just love to argue" even if only in jest.) :ROFLMAO::p
 
  • Haha
Reactions: 1 user
Upvote 0

WoodyWindy

On the road again!
Jul 20, 2010
105
32
There are two reasons to change a design that's worked well in the past - one is to improve quality and the other is to reduce production cost. Or, very occasionally, both.
The classic 50 f1.8 optical formula is as old as the hills, and has been applied to/reused in multiple generations and iterations of this lowest-cost prime lens in the EF series, including the post-digital age. You could say, it is "well proven", and is probably VERY cost effective to produce. :) Therefore, the low-cost option for a native RF mount version would have stopped at shifting the optics out, and fixing up the mount.

Canon didn't do that. Instead, here there is clear evidence of substantial work on the optical formula. That bodes well for improvement in the output, including both the image quality (aberrations, vignetting) at the edges, probably overall color/contrast, and I would guess minimum focus distance as well. (how many folks have inverted a 50 f1.8 to create an "instant" macro lens?)

In the end, I'm excited to see what this baby can do.
 
  • Like
Reactions: 1 user
Upvote 0

Sporgon

5% of gear used 95% of the time
CR Pro
Nov 11, 2012
4,717
1,532
Yorkshire, England
The classic 50 f1.8 optical formula is as old as the hills, and has been applied to/reused in multiple generations and iterations of this lowest-cost prime lens in the EF series, including the post-digital age. You could say, it is "well proven", and is probably VERY cost effective to produce. :) Therefore, the low-cost option for a native RF mount version would have stopped at shifting the optics out, and fixing up the mount.

Canon didn't do that. Instead, here there is clear evidence of substantial work on the optical formula. That bodes well for improvement in the output, including both the image quality (aberrations, vignetting) at the edges, probably overall color/contrast, and I would guess minimum focus distance as well. (how many folks have inverted a 50 f1.8 to create an "instant" macro lens?)

In the end, I'm excited to see what this baby can do.

Well maybe. I think you might be giving Canon more praise for working a new optical variation of the classic double gauss than they deserve.

If the block diagram at the beginning of this thread is correct it looks like Canon have reused their lens design from the 1964 Canonet rangefinder camera ( " the poor man's Leica") that has a 45mm f/1.7 lens, and replaced one element with the aspherical one. If you go to the wiki page for Double Gauss lens that PBD linked to at the beginning of this thread you can see the Canonet block diagram at the bottom of the first diagram page.

Screenshot 2020-11-01 at 12.17.17.jpg

Canon have given the first doublet a curved matting which is much more expensive to do than the straight bond in the 1964 lens, and this should give a more pleasant rendering. However my guess would still be that the lens is going to be sharper at the expense of rendering / bokeh. In other words it will chart well.
 
Upvote 0

WoodyWindy

On the road again!
Jul 20, 2010
105
32
Well maybe. I think you might be giving Canon more praise for working a new optical variation of the classic double gauss than they deserve.

If the block diagram at the beginning of this thread is correct it looks like Canon have reused their lens design from the 1964 Canonet rangefinder camera ( " the poor man's Leica") that has a 45mm f/1.7 lens, and replaced one element with the aspherical one. If you go to the wiki page for Double Gauss lens that PBD linked to at the beginning of this thread you can see the Canonet block diagram at the bottom of the first diagram page.

View attachment 193747

Canon have given the first doublet a curved matting which is much more expensive to do than the straight bond in the 1964 lens, and this should give a more pleasant rendering. However my guess would still be that the lens is going to be sharper at the expense of rendering / bokeh. In other words it will chart well.
Reposting the original for easy reference. (I actually had one of those old Canonet cameras, by the way). The differences are still pretty significant. I'm sure it will chart, the question is, will this dog hunt? We'll see in a few weeks. :)
rf50lensdiagram.jpg
 
Upvote 0

gruhl28

Canon 70D
Jul 26, 2013
209
92
With IBIS now becoming a thing in Canon land that will dampen some of the demand for IS in the lenses.
True, but they put IS in the 35mm and 85mm. Granted, those were released before a body with IBIS, but this just leads to the question of why they decided to release low to mid-price 35mm and 85mm lenses before the 50mm.
 
Upvote 0
I'm far from being an expert, but shouldn't it be the aperture, rather than the center of the lens elements?



My impression was the additional glass would not be needed for the FL range between RF's 20mm and EF's 44mm.

FL is from the optical center of the lens to the optical center.

They cover it quite well here: https://photographylife.com/what-is-focal-length-in-photography#definition-of-focal-length

And just as they imply in there, things can get complicated real fast once you dive into the physics of it all. Simply based on this explanation you can see why lenses end up being what they are. The flange distance directly contributes to the distance of the optical center vs the sensor/film plane.

My point about additional glass had to do with WA lenses. A 14 mm lens is already by definition further in mm regarding its optical center compared to the optical plane. And thus additional glass is required to shift that projection. That is why WA lenses tended to be huge on DSLRs. With smaller flange distances less glass is needed to shift this focal point.

50s and 40s typically were relatively small on DLSRs due to the flange distance and lens FL. But for MILC with the flange being further back you end up with a longer lens. So when people talk as if the 50 mm lenses aren't taking advantage of the shorter flange distance they simply don't know what they are talking about.
 
  • Like
Reactions: 1 users
Upvote 0

jolyonralph

Game Boy Camera
CR Pro
Aug 25, 2015
1,423
944
London, UK
www.everyothershot.com
True, but they put IS in the 35mm and 85mm. Granted, those were released before a body with IBIS, but this just leads to the question of why they decided to release low to mid-price 35mm and 85mm lenses before the 50mm.

Probably because a significant percentage of upgraders already have an EF 50mm of one kind or another (I have four: 50mm f/1.8II, 50mm f/2.5 macro, Yongnuo 50mm 1.4 and the EF 50mm f/1.2L), and even with the EF-RF adaptor it's still a lightweight small combination.

Moving this to RF makes sense now if it's cheap enough most RF mount users will upgrade.
 
  • Like
Reactions: 1 user
Upvote 0

Del Paso

M3 Singlestroke
CR Pro
Aug 9, 2018
3,276
4,158
Do designs that have worked well in the past still meet the demands and expectations required to be considered as working well to day ?
I'm using on a regular basis a 1969 Leica M 50mm f2 Summicron.
Sharpness and contrast (edge to edge !!!) are still outstanding, even on EOS R cameras, and , for closeups, on high MP EOS DSLRs.
So: a BIG, AN ENORMOUS YES
 
Upvote 0

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
497
333
OK, this is an old thread but given that anyone commenting here was probably interested in how the 50/1.8 would turn out, here's my comparison of hand-holding vs. the 24-105/4.


 
Upvote 0

Sporgon

5% of gear used 95% of the time
CR Pro
Nov 11, 2012
4,717
1,532
Yorkshire, England
Looking at that I'd expect this RF lens to probably be very sharp at the expense of subject separation / bokeh.
I've quoted my own post here from this old thread.
I bought an RF 50/1.8 as a lightweight lens for my RP. Can produce very nervous, busy bokeh when the OOF background has many highlights. Sharp in the centre though, with huge vignetting. Basically I think I was right in my original post. Not really very keen; the EF 50/18 could hold up with expensive lenses when it can to shallow DOF bokeh; this one can't.
 
  • Like
Reactions: 1 user
Upvote 0

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
497
333
Can produce very nervous, busy bokeh when the OOF background has many highlights
I'd love to see some comparisons if it's convenient for you. I had the EF 50mm f/1.0, f/1.2, f/1.4, and the first-generation f/1.8 that was built as a pro lens. Honestly they were all pretty soft but images looked typical. I've had the RF50 a year or so and it seems to be much like the EF 1.4 and 1.8, somewhat sharper but I haven't noticed any problem with the highlights. (Except that the aperture blades are an improvement.)
 
  • Like
Reactions: 1 users
Upvote 0

Sporgon

5% of gear used 95% of the time
CR Pro
Nov 11, 2012
4,717
1,532
Yorkshire, England
I'd love to see some comparisons if it's convenient for you. I had the EF 50mm f/1.0, f/1.2, f/1.4, and the first-generation f/1.8 that was built as a pro lens. Honestly they were all pretty soft but images looked typical. I've had the RF50 a year or so and it seems to be much like the EF 1.4 and 1.8, somewhat sharper but I haven't noticed any problem with the highlights. (Except that the aperture blades are an improvement.)
I've never done a direct comparison, but your post got me thinking, so I quickly did one with lenses I have to hand. Maybe I've been a little harsh in my criticism of it; the 50mm lens I've judged bokeh by is the old Takumar 55mm f/1.8. Actually the RF50/1.8 looks to hold up well against that lens, at least in this test, maybe a little better actually. :censored:
I shot the same scene with same exposure settings on 55mm Takumar, 45mm Tamron SP VC, and RF 50mm 1.8, all at 1.8, 1/640, ISO 100. I forgot to change distance for the 55 mm :rolleyes: as I was in a hurry to get the three shots before the light changed. Looking at the exposure, so much for DXO measuring the 45 Tamron as a poor T stop. _MG_2707.jpg Above is Takumar 55mm f/1.8_MG_2708.jpgTamron 45 f/1.8_MG_2709.jpgRF 50mm f/1.8


_MG_2707-2.jpgTak 55/1.8_MG_2708-2.jpgTamron 45/1.8_MG_2709-2.jpgRF 50/1.8
The Tamron is a great lens, so good it's now been discontinued by Tamron due to lack of interest ;) But you can see how the lens is sharp at 1.8 whilst managing to keep a soft bokeh.
So I think I'm being unfair is saying the RF's bokeh is too busy, I've not been looking at a combination of this focal length and aperture for a while. You can't see it here, but vignetting is very severe, and the extreme corners are mushy even at smaller apertures. But you can't complain for the price.
 
  • Love
  • Like
Reactions: 1 users
Upvote 0

SwissFrank

1N 3 1V 1Ds I II III R R5
Dec 9, 2018
497
333
My main takeaway is that you have a really good eye and a beautiful yard :-D And are very good natured to go make such an excellent comparison.

The Tamron might be the most pleasant, with the other two giving very slightly the "glass bubble" effect that normally I don't like but you use to good effect here. Tamron has a softer edge even if perhaps with a color tint. But I also think the bubble edges go away at f/2 or so, so the lens designer is actually giving you your choice of looks. By the way the 100 Macro's spherical aberration control can give you your choice of glass bubbles in front, or behind, your subject. I haven't done much with it. RF seems to have the best CA (the wires of the lamp's cage are more nearly the same color in front of and behind focus, while the other two are distinctly green and purple).

I've written elsewhere (maybe in this thread actually!) that makers can make the 50s either in a plain old cheap unsharp compact double-Gauss, or go with a much bigger, sharper, more expensive modern design. Canon's done the former with the 1.8 and latter with the 1.2, which seems like a reasonable mix. I think they should also make a few $10,000 50/0.7's as sort of a halo model.
 
  • Like
Reactions: 1 users
Upvote 0