Wide angle lenses on R-D1s

[yoshimura]

Hello, I am new to the forum and glad to join. I received a few days ago my new R-D1s purchased from the web in Japan @ 1500 euros or about US$1900, quite a bargain.
In an old thread on this forum, a member reported that the actual FOV of a Kobalux 2.8/21mm was actually closer to 24 mm instead of the 32mm induced by the 1.53 crop factor. Being the happy owner of a Kobalux 21mm 3rd generation , I wanted to test that result and I came to the same conclusion: the Kobalux's actual FOV is about 24mm, ie a crop factor of 1.14!. Of course this is not based on a scientific experiment and is purely empirical. It appears to be due to the deep receding of the lens into the body of the camera (17mm).
Encouraged by this result, I wanted to check with other WA lenses, and I came with the following very surprising results:
Zeiss ZM 2.8/25: actual FOV 31mm, crop factor 1.24
Konica M Hexanon 2.8/28: actual FOV 36mm, crop factor 1.28
Konica M Hexanon 2/35: actual FOV 50mm crop factor 1.42.

Of course this is very approximate and purely empirical. I also tested my 'Cron 2/50: actual FOV 72mm, crop factor 1.44. All this is perceptible when comparing the images between the frames in the viewfinder and on the screen. Am I getting too enthusiastic or is there some ground to the idea that the crop factor in the R-D1s is theoretical (ratio between the film image and the CCD) and must be differentiated with each lens?

 

[rvaubel]

Yoshimura
Don't trust the viewfinder framelines to match the theoritical field of view as marked. Because of parellex issues and other reasons, camera manufacterers factor in about a 15% "safety margin" so you don't cut off part of the shot by mistake. Then, in addition, the RD1 has a issue with the crop factor which further complicates matters. Personally, I wish the framelines would be designed to cover 100% of the FOV of the nominal focal length.
Practically speaking, the bestg thing to do is to experiment with your different lenses to determine for yourself what framelines or finders to use with the lenses you have. In my case (with the RD1) I use

- 21mm Zeiss finder with 15mm Heliar,,,this is a pretty good match
- Outter frameline of 28-35 CV mini dual finder with 21mm Kobalux,,,, almost perfect
- Inner frameline of 28-35 CV mini dual finder with 28mm CV f1.9 ... a little tight but workable
- 35mm RD1 viewfinder frameline with 40 mm CV Nokton,,, almost perfect
- middle part of 35mm RD1 viewfinder with 50mm Canon F1.2 .... its workable
- 50mm RD1 viewfinder with 75mm f2.5 CV ..... almost perfect

For 80mm or 90mm lenses, I use the center part of the 50mm RD1 viewfinder,. , it works OK'

Basically I keep the 28-35 dual minifinder on the camera and never change finders except for extreme wide angle. I like to work fast, so this works for me.


Rex

 

[Mark Norton]

The angle of view of a lens is 2 arctan (a/2f) where a is the frame dimension and f is the focal length.

If you take a 21mm lens, and a 35mm frame, the diagonal is 43.27mm and the diagonal angle of view is 91.7 degrees. If you have a sensor crop factor of 1.53, the diagonal reduces to 28.28mm and the diagonal angle of view is 67.9 degrees.

The TAN function is not linear, so that 91.7/67.9 is not 1.53, it's 1.35. That means that a cropped sensor does not reduce the angle of view by as much as you might first think with wide angle lenses.

Not the same with long lenses. At 90mm, the angle of view full frame is 27.03 degrees. Apply the crop factor and the angle of view is 17.86 degrees. The ratio of angles of view is now 1.51, pretty much the same as the crop factor.

Going back to the 21mm lens with its 91.7 degree angle of view and instead ask the question: "With a crop factor of 1.53, what focal length do I need to give an equivalent angle of view?"

Answer: 13.73mm. Ratio of focal lengths: 1.53, same as the crop factor.

You don't say how you arrive at your results, but to measure FOV in mm is bizarre and I agree trying to use the framelines is like trying to make measurements with a hand-drawn ruler.

 

[yoshimura]

Ok, I understand better from both your answers. Not being of a geometrical mind at all, my method was quite clumsy (from a 2m distance, shooting a tiled wall and comparing the field covered from the finder and from the pic). Actually, I forgot to say that I also used external finders, not only the R-D1's, to compare field covered from "genuine" 21, 28 or 35 mm lenses with the internal R-D1 frames. I admit this is not very scientific, but it sprang out of my enthusiasm for that long sought after camera!

 

[sebastel]

hi,
this issue is quite interesting ....
in order to find the actual FOV, what about measuring the actual angle?
since you know the distance d of your camera from the object, you only need to find out how wide w the field of view is (diagonally, horizontally or vertically), and then determine the angle as FOV(angle) = 2*arctan(w/(2*d))

i assume, the actual values will be a lot closer to the "official" crop factor of 1.53 ...

cheers,
sebastian

 

[yoshimura]

Good idea, I'll try this out and keep you posted.

 

[Mark Norton]

You can't use the coverage from the viewfinder as a reference because they are pretty inaccurate. Take some pictures, move further back and calculate the angle of view based on the formula (which works both when thinking about in front of the lens and behind it). The angle of view you get should then be closer to that predicted by the formula.

 

[jaapv]

. Personally, I wish the framelines would be designed to cover 100% of the FOV of the nominal focal length.
Rex

Oh, but they do. However, they are corrected for the minimum focussing distance. Your lens has a longer focal length then, as it has been extended to focus. And they figure in the frame of the slideframe (as SLR viewfinders do as well) adding up to a frameline that is 10 to 20 %, depending on focal length and closest focussing distance, "too small" at infinity focus. But they are exact.

 

[rvaubel]

Oh, but they do. However, they are corrected for the minimum focussing distance. Your lens has a longer focal length then, as it has been extended to focus. And they figure in the frame of the slideframe (as SLR viewfinders do as well) adding up to a frameline that is 10 to 20 %, depending on focal length and closest focussing distance, "too small" at infinity focus. But they are exact.[/quote

jaapv

I forgot that the framelines were corrected for minumum focus distance. But why? It seems to me that they should be corrected at normal focus distances, not extreme closeup. In any case most framelines include about 15% or so less than you are going to get at 5-7ft to infinity. In addition many lenses of nominally the same focal length actually cover slightly different fields of view.

There are so many conflicting philosophies of "correct" rangefinder coverage between manufactures that, especially with a crop factor camera like the RD1, that I simply prefer to find the frameline that agrees most closely with 100% coverage @ infinity. When shooting close ups, I just give myself plenty of room as parrelex errors are always a problem especially with the RD1 as the parralex correction doesn't know anything about it being a "crop" camera.

When using accessory finders, especially close up with wide angle lenses, parrelax error is severe. However, I learn little tricks that have been developed imperically.

I don't see any reason to try and match rangefinder framelines with lenses in a theoritical way. I tried, at one point, and finally gave up. Thank God for digitals instant feedback or my trials and tribulations would take forever.

I would be suprised if other RD1 owners haven't had to do roughly the same thing as I have. It's interesting that coming from an SLR background, my style of shooting has become much more spontanious with the RD1. I find that I must do my final cropping in Photoshop as I have come to regard the framelines as simply a rough approximation. I used to think that would be a bad thing but I'm not so sure now.

Rex

 

[andyturk]

R-D1 framelines are only a rough guide to what you're framing. When I shoot with a 21mm, I use the corners of the 35mm frame as "rule of thirds" points. That works pretty well, but I never know what actually made it into the shot until afterwards.

Shooting up close is really challenging though. The parallax error can be huge. Getting this image of a seed pod less than a meter from a 75mm lens took a dozen attempts. There's a reason people don't use rangefinders for macro shots.

 

[Mark Norton]

Shooting up close is really challenging though. The parallax error can be huge. Getting this image of a seed pod less than a meter from a 75mm lens took a dozen attempts. There's a reason people don't use rangefinders for macro shots.

I agree. Fan as I am of rangefinders, they're not well suited to macro where the precise focussing plane and depth of field can dramatically affect the shot. Never seen the point of the Leica macro adapter, for example.

 

[yoshimura]

RVaubel,
Your suggestions were quite good, but the question remains a mystery for me. I received today my second hand M Hexanon 21-35/3.4-4 from Tokyo: using the CV mini finder, I found that the 21 focal length matches the 28mm (normal) frame of the mini viewfinder, and the 35 focal length matches exactly the R-D1's internal viewfinder 28mm frame. This does not fit with the theoretical values due to the 1.5 crop factor (the 21 should be closer to the standard 35 frame, the 35 should be beyond the standard 50 frame).
So I still wonder if that 1.5 factor is not a theoretical calculation derived from the respective areas of the CCD and of a 24x36 mm film frame, but which should be compensated by the specifities of each lens. Anyway, my Kobalux 21 and the Konica 21 do not seem to be limited to the cropped 32 mm they should nominally show. I don't know if this clear, but I think the question remains open.

 

[rvaubel]

Yoshimaru

This is one of the most puzzling aspects of matching viewfinders to lenses i.e. the theoretical and the emperical don't match....not even close. I'm not the type that can take this discrepancy lying down forever, I'm sure there's an answer. But until I can rethink all the logical steps to find the reason for this logical conundrum, the fact remains that one cannot deny the evidence of their senses. I have emperically determined which frameline/finders work best with which lenses and thats it, logic or no logic. In my current RD1 setup, I keep the VC 28/35 minifinder on the hotshoe at all times. In conjunction with the RD1 framelines, these suffices for 6 lenses without ever swapping finders. The list once more

21mm f2.8 Kobulux 28mm mini
28mm f1.9 Nokton 35mm mini
40mm f1.4 Nokton 35mm framelines
50mm f1.2 Canon 50mm framelines
85mm f2.0 Nikon 50mm framelines (center part)

I might add that I find telephoto finders to be almost useless anyway because parallex errors are always worse than estimating the center of an actual corrected frameline. Not to mention the fact that you basically have to keep your eye plastered to the focusing patch to maintain critical focus if your subject is at all mobile.

Using this strategy, I never have to change finders, except when using my 15mm heliar. That is a very good thing since I hate fumbling with finders.

I'm thinking about getting the VC 75mm f2.5 for the telephoto end as I am finding the 85mm f2.0 Canon to be a little too much, both power wise and weight wise. The 75mm would be a better fit with the 50mm viewfinder also.

By the way, I carried the entire kit of the above mentioned 6 lenses in a small camera bag for a 2 week vacaction in Canada with no problem. The last time I tried to carry a range of lenses with my DSLR it almost broke my back..Never again.

Rex

 

[rvaubel]

Yoshimaru

This is one of the most puzzling aspects of matching viewfinders to lenses i.e. the theoretical and the emperical don't match....not even close. I'm not the type that can take this discrepancy lying down forever, I'm sure there's an answer. But until I can rethink all the logical steps to find the reason for this logical conundrum, the fact remains that one cannot deny the evidence of their senses. I have emperically determined which frameline/finders work best with which lenses and thats it, logic or no logic. In my current RD1 setup, I keep the VC 28/35 minifinder on the hotshoe at all times. In conjunction with the RD1 framelines, these suffices for 6 lenses without ever swapping finders. The list once more

21mm f2.8 Kobulux 28mm mini
28mm f1.9 Nokton 35mm mini
40mm f1.4 Nokton 35mm framelines
50mm f1.2 Canon 50mm framelines
85mm f2.0 Nikon 50mm framelines (center part)

I might add that I find telephoto finders to be almost useless anyway because parallex errors are always worse than estimating the center of an actual corrected frameline. Not to mention the fact that you basically have to keep your eye plastered to the focusing patch to maintain critical focus if your subject is at all mobile.

Using this strategy, I never have to change finders, except when using my 15mm heliar. That is a very good thing since I hate fumbling with finders.

I'm thinking about getting the VC 75mm f2.5 for the telephoto end as I am finding the 85mm f2.0 Canon to be a little too much, both power wise and weight wise. The 75mm would be a better fit with the 50mm viewfinder also.

By the way, I carried the entire kit of the above mentioned 6 lenses in a small camera bag for a 2 week vacaction in Canada with no problem. The last time I tried to carry a range of lenses with my DSLR it almost broke my back..Never again.

Rex

 

[yoshimura]

VRaubel, I think your method is the best and I follow it since you first mentioned it : I also permanently attach the CV 28-35 mini finder so I cover every focal distance from 21 to 90 mm without further manipulation. I also have the Leica 21-24-28 finder, which is a marvel but which is so huge that I don't use it, I might sell it some day. Nevertheless I think that there is a point on the R-D1's actual crop factor, another Erwin Puts should address the problem! When I have time, I'll shoot identical pix (I mean same lenses, same subject, same distance etc) with the R-D1 and with my good old M3and compare them, that may be a good way.

 

[jlw]

I forgot that the framelines were corrected for minumum focus distance. But why? It seems to me that they should be corrected at normal focus distances, not extreme closeup.

If the framelines were corrected for a "normal" distance (or any distance other than the closest) then whenever you focused closer than that distance, the framelines would show MORE than would be in the final picture. If you carefully composed your shot so a particular detail was right at the edge of the frame, you'd find it was cropped out of the actual image, and you'd be justifiably annoyed.

This unpleasant situation doesn't have anything to do with frameline design, but with the fact that the focal length of a conventionally-constructed lens gets longer as you focus closer -- in effect, it "zooms in" slightly. There's no way to avoid it.

Since it's easy to crop a digital image that contains too much "stuff," but impossible to put back "stuff" that wasn't in the picture in the first place, it makes sense for the designers to err on the side of caution. That means designing the framelines to show slightly less than the actual picture area even at the closest focusing distance.

An unavoidable side effect is that at longer and longer distances, the lens' actual focal length will get shorter and shorter, and the frameline view will be progressively tighter and tighter relative to what's in the actual picture.

Stinks, yes, but there's no way around it short of designing a field-size-correcting frameline system, like those found on some Konica and Polaroid cameras. In these, the upper framelines can move relative to the lower ones as the rangefinder is focused, allowing the framelines to have a closer match to the actual coverage of the lens. This adds complexity even for a fixed-lens camera with a single frameline; I don't believe anyone has ever produced a camera with both field size correction and multiple framelines for interchangeable lenses. (If you know of one, post a picture!)

 

[LCT]

...I don't believe anyone has ever produced a camera with both field size correction and multiple framelines for interchangeable lenses...

Wonder if Leica could do it with coded lenses and electronic frame lines in the M8 viewfinder. Just a dream i guess. 🙄

 

[Mark Norton]

Wonder if Leica could do it with coded lenses and electronic frame lines in the M8 viewfinder. Just a dream i guess

The position of the cam follower is related to the focussing distance in a fixed way, irrespective of the lens mounted. The profile of the lens cam on each lens sets the cam follower to the correct position for a given rotational position of the focussing ring which in turn is related to the lens to sensor distance and hence subject to lens distance.

If the camera were able to measure the cam follower position accurately - there's only about 20 degress of travel - and with a knowledge of the actual lens mounted, a single accurate frame line, precisely corrected for parallax could be displayed to the user through an LCD panel, like an EVF, used to inject the framelines into the viewfinder.

Without the lens coding, the camera could only display a pair of frame lines as now using the bayonet lug to select them and the parallax correction would inevitably be "one size fits all".

 

[rvaubel]

RD1 frameline only a rough quide

RD1 frameline only a rough quide

Everyone

I believe the RD1's framelines are less accurate than the Leica's for two reasons.

1) The design was simply copied from the R3a, a full frame camera. It would have been better for Epson/CV to be a little less expediant and develop framelines that were more appropriate

2) For the same reason, I think the parallex correction is keyed for a full frame camera and isn't quite right for the crop sized sensor

I could be wrong about both these things but it really doesn't matter much because framelines on rangefinders are a bit of wishful thinking anyway. If you want precise framing in a camera, get an SLR.

Yoshimura: How do you like your Konica 21-35mm (Bi-Hexagon?😱 ). I understand, they were only released in Japan. If you happen to hear of anotherone available, I would be most interested.

Rex

 

[yoshimura]

RVaubel,
I tested the Dual M Hexanon all week end: I was fearing to be disappointed because of the reports I read on the web, particularly Erwin Puts' one. I must say that I am happily surprised, it seems very good. The 21 looks very sharp, no vignetting and the 35 behaves quite well also, much better than I expected, although I have to try in low light and compare it with the fixed 2/35 M Hexanon. IMHO, it makes an ideal outfit for the R-D1, more adapted than the Tri Elmar as far as focal lengths are concerned. I will post pix, this is better than words. The one I purchased is a second hand one, I was lucky to find it, but there are at the moment 2 new ones for sale on the major selling portal in Japan , Rakuten. Follow this link: http://esearch.rakuten.co.jp/rms/sd/esearch/vc?sv=2&item=m+hexanon.
If the link does not work, log on to Rakuten's site (www.rakuten.co.jp) and type m hexanon in the search window and you will find them. They sell overseas and accept credit cards, the only problem is you have to read Japanese...In spite of my pseudo, I am not Japanese but I used to live in Japan, a few years back.
Coming back to the Dual Hexanon, there were only 800 produced. The two for sale at the moment go for 187000 and 198000 yen, that's approx US$ 1600~1700. I hope this helps...