Another 'Why I Like the Canon 50/0.95' post

jlw

Rangefinder camera pedant
Local time
1:54 PM
Joined
Aug 27, 2004
Messages
3,262
Lately on RFF we've been having another outbreak of the recurring "Why do you guys only talk about gear?" lament. And I can see the lamenters' point to some extent... but I figure that if we talk about how we use gear, that's a bit different, because it might help someone else solve a photographic problem, right?

So, here's an account of my latest outing with my R-D 1 and M-adapted Canon 50mm f/0.95 lens...

As with a previous writeup I posted on a similar topic, the occasion was a modern-dance concert (Arcanum Contemporary Dance, in this case.)

One thing about photographing modern dance is that you almost never know what you're going to get lighting-wise -- but in this case I had two evenings' worth of chances to take pictures, and I had seen a technical rehearsal that seemed quite well-lit.

Since I expected the actual lighting to be similarly bright, and since I was going to have to stand mostly in the back of the theater, I decided the safest outfit to take on my first evening would be my Nikon D80 DSLR and a few lenses. I figured I'd be doing most of my shooting with my 70-200/2.8 VR Nikkor, with an 85/1.8 AF Nikkor as backup for dimly-lit scenes.

Oops. What I hadn't counted on was that between the tech rehearsal and the final, the lighting designer had decided to get all murky and atmospheric. The lighting on the stage was beautiful and expressive, what there was of it -- but that wasn't much. I had to crank the D80's sensitivity up to the "H1" setting -- equivalent of ISO 3200 -- to get even the 85/1.8 into three-digit shutter speeds, and the 70-200 was basically useless, vibration reduction or no; 1/50 sec. just doesn't get you very far photographing dance, even if your lens is electronically stabilized.

Examining my take late that night, I considered myself lucky to have gotten some reasonably exposed images -- but I realized a change of plan was in order.

So the second night, I left the 70-200mm grenade launcher at home. Instead, I took the D80 with just the 85/1.8 lens, and my R-D 1 with the Canon 50/0.95 and Voigtlander Nokton 35/1.2 lenses. I figured this combo would both improve my mobility, making it easier to get to different vantage points around the theater, and improve my chances of sucking in enough photons to get decent exposures.

Technical aside: Also in the interests of mobility, I had considered taking along my screwmount Canon 50/1.2 lens instead of the chunky 50/0.95. The 50/1.2 is a very well-liked lens on RFF's Canon forum, and I like mine. But an ad-hoc comparo during the afternoon confirmed what I had found before: contrary to conventional gear-head wisdom, the 50/0.95 at f/1.2 is sharper than the 50/1.2 at f/1.2!​

My plan was to use the 50 at f/1.2 whenever possible; this cleans up its performance noticeably compared to using it at f/0.95, and it also meant both my R-D 1 lenses would be exactly 1 stop faster than the lens I had on the D80. Since the D80 would be set to ISO 3200 and the R-D 1 to ISO 1600, I should get the same shutter speed on both -- allowing me to use the D80 as a "spot meter" for my R-D 1 shots.

Well, I could tell you how happy I was that the second night's setup worked better... but you want to see pictures, don't you? Okay, here's a typical D80 shot from the first night; this wasn't one of my best pictures from the concert, but I chose it to illustrate some of the problems typical of this type of shooting:

07-06-07_135.jpg


It doesn't look too bad at this size, and I'd consider it usable. Now here's a shot from the second night of the same piece, with the same lighting, only with the 50/0.95 at f/1.2 on the R-D 1:

07-07-01_123.jpg


Not so different from the first one at this size, but let's compare the 1:1 crops -- D80 on the left, R-D 1 on the right:

07-06-07_135-detail.jpg
***
07-07-01_123-detail.jpg


No, I didn't get the files mixed up! The 85/1.8 AF Nikkor is an excellent lens for sharpness, and I did get better results with it on the pieces that weren't quite so dimly lit.

But remember, to get usable shutter speeds with its f/1.8 maximum aperture, I had had to crank the D80 up to EI 3200 -- and its results, especially in the dark areas, are so "rocky" that they negate some of the lens' performance. Meanwhile, stopping down the Canon 50/0.95 to f/1.2 has tamed some of its legendary haziness; its full-stop-faster maximum aperture let me get the same shutter speeds while taking advantage of the R-D 1's much smoother EI 1600 imaging.

One interesting real-world consequence of this had to do with cropping. Since the R-D 1 had only a 50mm lens while the D80 had an 85mm lens (both have the same size sensor) I had to crop the R-D 1 shots more tightly to get similar framing. And of course the D80 gives me 10 megapixels to play with, while the R-D 1 provides "only" 6 megapixels.

But scroll up and look at those 1:1 crops again -- wouldn't you say that the R-D 1 shot is more "enlargeable"? I certainly would -- and in fact in the photos I've worked with so far, the R-D 1 images seem to stand up better at the same final size, despite the D80's advantages in pixel count and image magnification! Good to know, but not what I had expected at all...


For those who wonder, I used the R-D 1 with a 1.3x eyepiece magnifier, which makes focusing the 50/0.95 much more confident in dodgy conditions, and in most cases I wound up setting both cameras' shutter speeds manually based on a quick test shot whenever the light changed (this turned out to be both quicker and surer than my idea of using the D80 as a spotmeter for both cameras.)​

If you want some takeaway talking points, here are a few:



-- Ten megapixels aren't always better than six; when it comes to final results, six smooth megapixels can outperform ten rough ones.

-- In very dim light, using a faster lens will improve image quality more than using a slower one and cranking up the ISO.

-- Stopping down the Canon 50/0.95 even a tiny amount makes a big difference in how it images!


The floor is open for anyone who wants to debate, but be prepared to show examples!
 
Last edited:
Fantastic post, great results - thanks very much. I just recieved an r-d1 from melbourne and I'm using an old serenar 50 1.8 on it. So far I've been really impressed with it as far as handling and images go.

At some point, I'd LOVE to have a canon 50 .95 - it's just the final frontier in fast glass.

That nikon sensor really is quite bad in low light, I have to say.... My 30d is still very very useable and smooth @ 3200iso, however I think the r-d1 makes nicer noise at 1600 compared to the canon.

The ideal camera for this would be a 5d or 1d mk111 with 6400iso, and a canon 50 f1L or 50 1.2L + 85 1.2L


But my, the r-d1 performed admireably.

Thanks again,
Gavin
 
Wow, was that f0.95 at f1.2? That is amazing focus! Did you calibrate your RD1 to the lens?

Good work.

Vick
 
fdigital said:
The ideal camera for this would be a 5d or 1d mk111 with 6400iso, and a canon 50 f1L or 50 1.2L + 85 1.2L

Maybe in terms of high-ISO noise quality. But I get better photos of action subjects in general when I'm using an RF camera.

Still, that's a subject for another post...
 
Vickko said:
Wow, was that f0.95 at f1.2? That is amazing focus! Did you calibrate your RD1 to the lens?

The lens has been calibrated to the camera. The technician who did the M-mount conversion did an initial setup using a Leica M2 film body. Later I fine-tuned the lens to the R-D 1 by taking a few thousandths of an inch off its collimation shim. And the eyepiece magnifier helps a lot.
 
Wow! Nice shots of the dancer, and really impressive performance from the Canon/RD-1 combo. Amazing how good the Canon is stopped down a little
 
Wow, that's pretty excellent, I totally enjoy seeing your posts with this lens. Did you end up using your Nokton at all then? I wonder how it compares at 1.2.
 
SteveM(PA) said:
Wow, that's pretty excellent, I totally enjoy seeing your posts with this lens. Did you end up using your Nokton at all then? I wonder how it compares at 1.2.

No, I never wound up using it at all, although that had to do more with the theater than with photography. In this particular theater, the "house" (seating area) is unusually wide compared to its depth; getting close enough to get good framing with the 35 would have put me too far off to the side.

It would be interesting to see how the two lenses compare (not in a "which is better?" sense, but just how the images differ.) Sometime I'll have to try that...
 
jlw said:
The lens has been calibrated to the camera. The technician who did the M-mount conversion did an initial setup using a Leica M2 film body. Later I fine-tuned the lens to the R-D 1 by taking a few thousandths of an inch off its collimation shim. And the eyepiece magnifier helps a lot.

Who did the actual conversion?

I have one being converted by Ken Ruth.
 
colyn said:
Who did the actual conversion?

I have one being converted by Ken Ruth.

Eastcamtec did mine. You can tell it's a hand-made conversion, but nicely done and effective.

I'd be curious to see some pics of your finished conversion, so I can see how he did it. I've got some pictures of mine somewhere; will post them if I can find them.
 
jlw said:
Eastcamtec did mine. You can tell it's a hand-made conversion, but nicely done and effective.

I'd be curious to see some pics of your finished conversion, so I can see how he did it. I've got some pictures of mine somewhere; will post them if I can find them.

I'll post some pics when it comes home.
 
Nice pictures!

That said, the question I have, however, is what are they supposed to show? The main thing I see is that the one extra ISO stop degrades performance a lot and that hence there is a reason for using fast lenses. OK, that's a valid point. However, it's very much the result of your particular comparison; the main advantage of the 50/0.95 at 1.2 is that you're comparing it to a quite necessarily slower 85mm lens, and that you're comparing the R-D1 to a Nikon DSLR, which doesn't present you with a lot of lens options.

Comparing 85s to 85s, if you put any of the 85/2s or 85/1.8s on your R-D1, you have to shoot at ISO 3200 as well; in faster lenses you have the choice between the Nikon 85/1.5 and the ridiculously priced Leica 75/1.4, which you'll have a hard time focusing while gaining only half a stop.

Comparing 50s to 50s, on the Nikon you can always use a 55/1.2 or 50/1.2 Nikkor and leave the ISO at 1600. With some of the old lenses you have to focus manually (but that's what you're doing anyway), meter manually (once), and maybe get a better focusing screen (like you got an eyepiece magnifier to focus reliably).

And as others have pointed out, if you do that kind of photography often, a D80 might not be the best camera; even if you don't go into 5D+L territorry, a Pentax K10D or Canon 30D will leave you with a lot more lens options.
 
Last edited:
probably a Nikkor or Zeiss 85/1.4 with lower ISO-settings would have performed better in that situation than the 85/1.8 did (if you are able to keep the focus sharp!), and probably a Nikkor 50/1.2 or Canon 50/1.0 or /1.2 would have performed well. Is it the question? Don't think so.

You take the best what you have on such a shooting as a photog. If I had the opportunity I would take my Canon 50/0.95 as well (or maybe the FD 50/1.2AL as a backup). What the two pictures show is, when to choose between high ISO setting on a digital body in low light situation, with a normal high speed lens, and using a *very high speed* lens (even of vintage age) with normal ISO-setting, the second choice is best.

Or, even simpler: there is a right for the ultra high speed lenses to be there, even in 2007. Even if most people would be satisfied with the first picture.

The monster shows it's quality again. Nice skin tones! Anyone cares actually in digital age?
 
I really wish I hadn't read this thread :(

How much does the conversion cost? I use my Canon 50 f/1.2 a lot wide open and have always fancied trying its big brother...
 
Sonnar2 said:
Or, even simpler: there is a right for the ultra high speed lenses to be there, even in 2007. Even if most people would be satisfied with the first picture.

Or even better.. get a canon or fuji dslr with less noise than the nikon equivs :D :angel:

*ducks* :eek:
 
rxmd said:
Nice pictures!

That said, the question I have, however, is what are they supposed to show?

I thought I made this clear, but obviously I should have made it clearer: My goal wasn't to say, "X is better than Y." What I was trying to say was, "Here's an experience in which the gear choices I made gave results different from what I had expected."

Certainly there are other equipment combinations that might have produced the same or better results. But we're all often faced with the problem of choosing among the gear we've got to achieve the best results in a particular situation -- and I figure that the more decision-making information we've got, the better!

I felt this experience was worth posting because the choices that worked better for me were a bit non-obvious, counterintuitive, and surprising (and apparently you did too, or you wouldn't have bothered to post!)

In fact, it often seems nowadays that my experiences in the admittedly specialized arena of fast-action-in-crappy-light contradict the photographic Conventional Wisdom that you read on RFF, on other online forums, and in photo magazines. Here are examples of stuff that, while it might be true in many cases, I no longer automatically believe anymore:

"Zoom lenses are all you need. That's what the pros use."

"Nobody needs wide-aperture lenses anymore. With digital, if the light gets dim, you just turn up the ISO."

"A 10-megapixel camera will give you sharper, more detailed images than a six-megapixel camera."

"A slower, sharper lens will give you more detailed pictures than a faster, less-sharp lens."

"SLRs are better than RF cameras because they're more versatile."

"For action subjects, a camera with auto focus, auto exposure, and auto advance will always outshoot a manual focus, manual exposure, manually wound camera."

"Camera technique doesn't matter with digital because you can fix anything later in Photoshop."​

I'd admit that many of these statements have value IF accompanied by the caveat, "all other things being equal." But the photographic process has so many inputs that it's very seldom that all other things ARE equal!

And I guess that is what my post was supposed to show...
 
Or even better.. get a canon or fuji dslr with less noise than the nikon equivs

Yeah, but then I'd still be shooting with a (blecch!) DSLR!

Obviously this won't be true for everyone -- but for me, the fact is that in action situations especially, I just get better results using an RF than an SLR. I've discussed this belief and some possible reasons for it before in other threads, such as this one (click.)

I don't claim the same will be true for everyone. But it always surprises me that even when I say it here on Rangefinder Forum, someone always comes back and says, "You'd be better off getting a DSLR..."

Maybe those people are right -- but as Chico Marx said to Margaret Rutherford in the movies, "Who ya gonna believe: Me, or your own eyes?"
 
Back
Top Bottom