Comparing RF & SLR wide open

[Timmyjoe]

This is one of those times when I really miss Tom A.

Did a test today, shooting the same subjects, with a Nikon S2 w/3.5cm W-Nikkor-C f2.5 lens and a Nikon F w/35mm Nikkor-OC f2.0 lens (and Kodak 5222, Double XX). Both lenses were in very good condition.

The Nikkor OC is about a twenty year newer design than the W-Nikkor-C, but I was surprised to see the W-Nikkor-C was sharper wide open, sharper stopped down to f2.8 and even sharper on some shots stopped down to f4. I was wondering what the theory is as to why rangefinder lenses are sharper wide open than their newer SLR counterparts.

Does it have to do with the lens being closer to the film? Are rangefinder lenses easier to design to be sharper wide open? Wasn't expecting this result.

Best,
-Tim

 

[Moto-Uno]

Would you have any examples we might be able to compare ? Or are the differences kinda lost on web postings ? Peter

 

[Robert Lai]

SLRs have the weakness of poor focus accuracy for wide angle lenses. Even if you use the split image RF in the screen of the SLR, the effective baselength is far less than that offered by your rangefinder camera.

At the other end of the spectrum, the RF camera is weak with focusing long focal lengths. Here, the SLR shines. Hence, even Leica makes you use a Visoflex to focus a 180mm+ lens on a rangefinder camera.

Erwin Puts has a detailed mathematical development to illustrate this. I can't find the actual text at the moment, but my recollection is that somewhere around 75mm or so, the SLR and RF are equivalent in focusing accuracy.

So, I suspect that your lenses are equally sharp, but the SLR may be very slightly misfocused.

Here's Erwin's public domain article on RF accuracy:
http://leica-users.org/v04/msg07362.html

 

[Godfrey]

The fundamental difference between RF and SLR lenses that are less than 50mm focal length is that SLR lenses need to be retro-focus designs to clear the swinging mirror where RF lenses can be simpler designs. The retro-focus lens requires more lens elements, better lens coatings to eliminate flare, and more precision in centering to work well. Faster lenses require even more lens elements, more precision, and better coatings to perform well.

For example, your W-Nikkor-C 3.5cm is a basic symmetrical design with four elements in three groups. Excellent examples of this lens type have been made since the 1930s. The Nikkor-O 35mm f/2 for the SLR is not only a stop and a half faster, it has eight elements in six groups to work on the Nikon F SLR body. This is quite an early 35mm SLR lens: it was made long before multicoating existed and so suffers in comparison to the simpler, slower RF lens.

Lenses for RF cameras were getting excellent quality results in the 1930s and 1940s before SLRs were even available... SLR lenses with their more complex designs hit their stride in the 1970s and later, after the invention and popularization of multicoating by Zeiss and Pentax became generally available on all high quality lenses.

G

 

[michaelwj]

This is one of those times when I really miss Tom A.

Did a test today, shooting the same subjects, with a Nikon S2 w/3.5cm W-Nikkor-C f2.5 lens and a Nikon F w/35mm Nikkor-OC f2.0 lens (and Kodak 5222, Double XX). Both lenses were in very good condition.

The Nikkor OC is about a twenty year newer design than the W-Nikkor-C, but I was surprised to see the W-Nikkor-C was sharper wide open, sharper stopped down to f2.8 and even sharper on some shots stopped down to f4. I was wondering what the theory is as to why rangefinder lenses are sharper wide open than their newer SLR counterparts.

Does it have to do with the lens being closer to the film? Are rangefinder lenses easier to design to be sharper wide open? Wasn't expecting this result.

Best,
-Tim

Adding to what Godfrey said, you say the RF 3.5/35mm lens was sharper than the SLR 2.0/35mm lens when the SLR lens was at f/2.0, f/2.8, and "sometimes" when they were both at f/4.0? Can we say that at f/4.0 they perform similarly, while at f/2.8 and f/2.0 the SLR lens is obviously far better as the RF lens doesn't open that wide.

In other words, at f/2.8 there is only one choice (the SLR lens) and at f/4.0 they are similar.

So while your premise that RF lenses are sharper wide open than SLR lenses (for wide angle lenses anyway) may be correct, your test does not actually test that hypothesis. Your will need to test f/2.0 vs f/2.0, rather than f/2.0 vs f/3.5.

 

[Corran]

In other words, at f/2.8 there is only one choice (the SLR lens) and at f/4.0 they are similar.

Re-read, he is using the f/2.5 variant.

However a comparison to the 3.5cm f/1.8 would be interesting.

Either way, comparing on a fast, grainy film to me would be the weak point here.

 

[Timmyjoe]

Thanks for everyone's input.

Yeah, it's the f2.5 W-Nikkor-C. The test was both shot wide open, so f2.5 compared to f2.0 (not an exact comparison I know), and then both shot at f2.8, and then both shot at f4.0. Then I shot a number of pictures with both lenses set at f2.8. The rangefinder lens was sharper in all those shots.

The 35mm SLR lens is a Nikkor-OC, so I assumed that meant it was coated. Maybe it doesn't have the multi-coating like the current SLR lenses do, but I still thought it would outshine a lens designed in the late 1940's, as it was new in the early 1970's.

And I don't find the Double XX to be that grainy when processed in HC-110 at a dilution of 64:1. Comparable to Tri-X.

Again, thanks for the input.

Best,
-Tim

 

[Corran]

I would love to see your results.

As for film, funny you mention XX in HC-110 as I just developed my first roll of that, and for the first time used HC-110 with it, so I'll reserve judgement till I scan the negatives. However, I've shot plenty of Tri-X and at least 1000' of T-Max 100 though, and there's no comparison obviously. Any serious "test" IMO should be on a 100 speed T-grain film if you want to actually see the real resolution differences.

 

[Timmyjoe]

OR . . .

If the difference shows up with Double XX, then it would show up much more with 100 speed T-Grain.

😉

I'm looking at a project that will be shot on Double XX, and trying to determine which camera/lens combo to use. That's why I'm shooting the test on Double XX as well.

Best,
-Tim

 

[Timmyjoe]

Okay, a couple of samples. Again, Double XX, processed in HC-110. 100% crops of the part of the image I used for focus (which also happened to be the center of the image). Yes I know it's grainy. Both lenses set at f2.8 for all shots.

Close up shot (approximately 4 feet from camera)
Nikon F w/35mm Nikkor-OC f2.0 lens @f2.8

Nikon S2 w/3.5cm W-Nikkor-C f2.5 lens @f2.8

Distance shot (approximately 100 feet from camera)
Nikon F w/35mm Nikkor-OC f2.0 lens @f2.8

Nikon S2 w/3.5cm W-Nikkor-C f2.5 lens @f2.8

Best,
-Tim

 

[Corran]

I don't necessarily disagree with your statement, having shot both the lenses in question, but honestly the images shown from the SLR lens look out of focus to me. Or a really bad sample?

 

[michaelwj]

Thanks for the images, the SLR lens looks well out of whack, mis focus maybe? Something else?
Apologies on the 3.5-2.5 misread, but you still need to compare apples to apples...

 

[ssmc]

I agree, I'm surprised the SLR lens looks this bad. In the first sample I didn't even notice the writing under the rock until I scrolled down to the RF version.

Is anything in the SLR frame as sharp as you expected? Or is it like that everywhere?

I wonder if your F body needs to have the focusing screen shimmed... I don't know if it's possible on an F, but I had to do it on a D700. That problem drove me absolutely nuts until I worked out what was going on.

Regards,
Scott

 

[radi(c)al_cam]

I wonder if your F body needs to have the focusing screen shimmed...

Yes, or the mirror, etc.etc.

 

[Spanik]

Then there is also to consider differences in development (won't affect sharpness but could influence grain and perception of sharpness), flatness during exposure, flatness during scanning, possible auto settings that cannot be changed (of scan and/or lab if used) and very likely a dozen other thing I'm forgetting. And a tripod was always used, just like a remote release I assume.

What were we comparing?

 

[Timmyjoe]

There seems to be a reluctance here amongst some to see that maybe the old rangefinder lens is sharper than the newer SLR lens when shot at the wide open end. I too was surprised, but after looking at the results, shot at the same time, using two cameras and lenses that had been recently serviced by folks who are experts in the Nikon SLR & Rangefinder fields (APS & DAG), using the same batch of Double XX film, processed in the same chemicals, in the same tank, at the same time, and scanned with the same film scanner, actually on the same film tray, at the same time. And the results above are what came out. They are crops of the VueScan output with only slight curve adjustment.

I think what Robert and Godfrey are talking about is coming in to play here.

Best,
-Tim

 

[radi(c)al_cam]

There seems to be a reluctance here amongst some to see that maybe the old rangefinder lens is sharper than the newer SLR lens when shot at the wide open end. I too was surprised, but after looking at the results, shot at the same time, using two cameras and lenses that had been recently serviced by folks who are experts in the Nikon SLR & Rangefinder fields (APS & DAG), using the same batch of Double XX film, processed in the same chemicals, in the same tank, at the same time, and scanned with the same film scanner, actually on the same film tray, at the same time. And the results above are what came out. They are crops of the ViewScan output with only slight curve adjustment.

😡
How do we deserve that a «cute» comment?

Additionally: You hadn't mentioned before that the F would have «been recently serviced by folks who are experts in the Nikon SLR & Rangefinder fields (APS & DAG) […]».

 

[michaelwj]

There seems to be a reluctance here amongst some to see that maybe the old rangefinder lens is sharper than the newer SLR lens when shot at the wide open end. I too was surprised, but after looking at the results, shot at the same time, using two cameras and lenses that had been recently serviced by folks who are experts in the Nikon SLR & Rangefinder fields (APS & DAG), using the same batch of Double XX film, processed in the same chemicals, in the same tank, at the same time, and scanned with the same film scanner, actually on the same film tray, at the same time. And the results above are what came out. They are crops of the ViewScan output with only slight curve adjustment.

I think what Robert and Godfrey are talking about is coming in to play here.

Best,
-Tim

We all agree that the RF lens can be sharper (from a pure design standpoint at least), it's just that those two shots are like night and day. It's almost like you smeared the front of the SLR lens with vasoline 😀

If it's not too much effort, it would be interesting to see the f/4 shots you said were not so clear cut.

 

[Timmyjoe]

Here's both camera/lens set ups shot at f4.

Close up shot (approximately 4 feet from camera)
Nikon F w/35mm Nikkor-OC f2.0 lens @f4.0

Nikon S2 w/3.5cm W-Nikkor-C f2.5 lens @f4.0

The SLR lens sharpens noticeably when stopped down to f4.0.

Best,
-Tim

 

[Moto-Uno]

^ I think we can agree that the older RF lens is certainly a competent performer . Peter