Lens hood length - avoiding vignetting

[Helen B]

Thanks George,

I'm finding with these wider lenses and long lens hoods that your hood is well within the field of view. For the last case, the 52 mm lens, I measure a side (A and B in your notation) of about 115 mm at 71 mm from the front vertex as being just in the field of view, definitely not 88 mm. The 55 mm and 80 mm examples are also much too small. I haven't tried the 400 mm lens.

Any thoughts?
Best,
Helen

 

[George]

How are you finding it is in the field of view? Mathematically, on a drawing or practically? If on a drawing someting must be wrong with the drawing, sorry to say. I calculated and constructed under a 100 lens shades with the program and didn't have any encrouching lens shade at all. The program is bullet proof. I have lens shades for all my film formats (24x36, medium formats, panoramic formats, LF and special formats of my special homemade cameras) all my friends who wanted them have them too and I had never 1 simple case of a vignetting shade. All of them were right on the first use. Try it practically on you lenses and I am sure you will not get vignetting. Otherwise we don't speak the same language.
The shade cuts all the image circle that is not on the film format.

 

[George]

Let's check it the other way - what is your proposed dimension for 600mm focal length, 58mm diameter front element, lens hood long 92mm, for 56 x 82mm film format?

 

[George]

I correct the length of the shade in the last example on 95mm...

 

[George]

Helen,
I know where the devil lies - in details, of course. I made a mistake in the results I gave you for you examples. I mistankenly took horizontal values for vertical ones and it gave smaller dimension than necessary. So here it goes corrected:
1) 55mm lens: length of the shade 54mm, sides AxB = 129.1 x 101.1 mm.
2) 80mm lens; length 80mm, sides AxB = 129.1 x101.1 mm.
3) 400 mm lens; length 98.9m, sides AxB = 77.7 x 60.9
That is of course bigger than what I wrote before. Hope we get closer now in our methods? George

 

[George]

Helen,
I just checked your 3rd example and I see that your method gives consistenly bigger dimension than mine. So there is discrepancy in both methods.

 

[Helen B]

George,

I'll re-try those later today - I am doing practical tests, not just drawing. I chose the 52 mm lens becuse I had sufficient information about the lens and the entrance and exit pupils to draw it out as well. With the practical tests I'm looking at the obscuration of the exit pupil as well as a simple check on the ground glass.

When I set up again I may as well try another lens - a 210 mm with a 57 mm front element, and 94 mm x 120 mm film. Could you give me the hood dimensions for that as well? Thanks.

The difference appears to be that the method I gave (as I've said before it is not 'my method') keeps clear of the rays that could enter the entrance pupil and then arrive on film if there were no further obstructions, while yours allows some degree of obscuration of those rays. It would help a great deal if you were able to give some indication of the theory behind your method - it would be a valuable contribution.

Thanks,
Helen

 

[George]

Here it goes, for the given 210mm lens:
Lens hood length L - 86.5mm, sides AxB - 65.7 x 83.8mm.

I know it would be a "valuable contribution" but the whole thing is too complicated for verbal description. Yesterday I foud some of the drawings used to do (15 years ago) the computer program (not the final edition) - it was so full of angles and trigonometry that I didn't try to get into it again, it would take me too long. Anyway, if someone is interested I will give him dimensions for his specific lens and film format. A less complicated way for me.

 

[George]

George,

It would help a great deal if you were able to give some indication of the theory behind your method - it would be a valuable contribution.

Thanks,
Helen

As I said, with the lens shade I simply cut off all the area of the image circle that doesn't show on the film area. That was the original idea. Then it's a lot of drawings, then comes the maths, then the algorithms for the program and then the program. After that it's just the happy clappy on the keyboard and only then comes the hard work with nice bellows formed lenshoods - 1 piece takes me about 2 full days to make it on the lens. The results are well worth of it all - staggering!

 

[George]

One more indication. Of course, it's not just enough - for the lens hood - to cut off the image circle around the film area. It must be cut off by the hood at such a distance from the lens that the flare is reduced to the desired level.

 

[Helen B]

George,

I've tried the 210 mm and re-tried the 80 mm. The hoods obstruct the exit pupil between about 30% and 50% along the edges and in the corners. These were practical tests, not theoretical.

I'll do a ray diagram for the 80 mm and the 52 mm I tried yesterday, and post them here when I have time, probably tomorrow evening. You could then give me some suggestions about what I'm doing wrong.

Thanks,
Helen

 

[Greg_E]

It's in the old Qbasic (remember?)

Yes, sort of. It's been almost 20 years since I last delt with basic. Care to share the text file of it? If not I truly do understand, a lot of people could make some money off your hard work.

 

[George]

Helen,
don't worry, there must be some other misunderstanding in it. If you answer my "test" dimensions we'll see it immediately - I have than lens shade on my lens and it works perfectly (as all the others too). With patience we can find it.
Greg,
Qbasic 5.1 I think, was the last version of it before they killed it (to sell better the Visual basic). It was still living 10 years ago, some nostalgic guys keep sites about it still alive but it gets more difficult to download from the net as the time goes. Share the text file? Really not keen for it, you got it. BTW the last time I was putting that on my XP home it took me 1hr to load there the program from my floppy disk - shame, I knew how to do that in a few seconds - 10 years ago...

 

[Helen B]

Let's check it the other way - what is your proposed dimension for 600mm focal length, 58mm diameter front element, lens hood long 92mm, for 56 x 82mm film format?

Length later corrected to 95 mm.

The classic method gives a hood 42 mm x 61 mm.

Do you have any comparisons for square format we could try?

Thanks,
Helen

 

[George]

Surprisingly, here we are on the same -62 x 41mm for me measured on my lens hood I use (the 1mm diference can be imprecision). I use it for my Fuji600C with Mamiya RFH. I don't have a square format hood but of course, I could calculate them. Strange, this is the only one we agree on.

 

[Helen B]

George,

The narrower the angle of view and the smaller the front element (relative to the focal length), the closer the two methods come.

Here is a square-format example to try:

Film 56 mm x 56 mm
60 mm lens, 48 mm front element

Let me know what lens hood length you would use, and I will attempt to give the other dimensions.

Best,
Helen

 

[George]

Helen,
for your square example above with 60mm lens: length of the hood L= 81.6mm, sides 87.7 x 87.7mm.
Cheers, George

 

[George]

Helen,
once again, let's do the "backward" test. What's you suggestion for 100mm lens, 43mm lens diameter, 75mm long hood on 24x36 mm film format? Thanks, George

 

[Helen B]

George,

The classic method gives sides of 42 mm and 63 mm. I think that your assumptions would lead to slightly smaller dimensions: 40 mm x 60 mm maybe. I think that I now understand your method for square format, but I'm not yet convinced that I fully understand what is going on with non-square formats.

Best,
Helen

 

[George]

Helen,
sorry, I had a little confusion in my papers. Can you let me know, once again, your dimension for: 100mm focal length, 45mm lens diameter, 78.4mm long hood and 24x36 film format? Sorry for the hassle. Thanks, George.

 

[George]

My dimension for your calculated example that gave you 42x63mm, are something like 58x39 so its all the time the same melody...

 

[Helen B]

Helen,
sorry, I had a little confusion in my papers. Can you let me know, once again, your dimension for: 100mm focal length, 45mm lens diameter, 78.4mm long hood and 24x36 film format? Sorry for the hassle. Thanks, George.

George,

The method I use gives a hood 44 x 66 in that case. The smaller the angle of view, the less discrepancy there is between the methods. The larger differences seem to occur with wide angles of view.

I wonder whether anyone else is following this - I'm just curious, and this isn't a subject that is covered much in textbooks or on the web. Drawing the results out is enlightening, and I think that I have a reasonable understanding of your method, especially for square format. If you wish to continue, should we do it via private messages?

Best,
Helen

 

[resummerfield]

I am very interested in this thread, and am following it closely. George designed a shade for my Voigtlander 420mm f4.5 lens, and I also used Helen’s formula to make another shade so I would have 2 shades for a practical test. The best test will be to expose some film and check densities of a uniform area, but I’m still waiting for a uniform sky. Until then I’m trying to evaluate these shades by looking at the ground glass, and I would like to expand the discussion to this question:

How is the best way to test for vignetting without exposing film? I was taught to observe the corner openings on the gg, and check for the round appearance of the aperture. Any distortion or obstruction indicates vignetting.

Now, when I examine both shades at the gg corners (not on the gg, but at the cutout), at infinity, there is a pronounced image of the shade. However, on the actual gg there is no indication of an obstructed image from the shade (dark line) against a uniform sky until I move the lens off center by about 70mm!!! If I understand George correctly, I should be seeing a slight darkening of the gg when the lens is moved about 10mm off axis, and looking very carefully I do detect a slight hint of darkening of the gg when moving the lens about 10-20mm off axis.

Jim Jones’ suggested a method “The quick and crude way to check lens hood vignetting on a camera with ground glass or a focusing screen is to illuminate the screen and check its corners through the lens and hood. This might not be perfectly accurate with SLRs with less than 100% viewing, but works good enough for practical use”. When I used this method, both shades obstructed the image of the illuminated gg.

One more observation. By examining the corner openings on the gg, without any lens shade, I noted a portion of the lens barrel distorting the round appearance of the aperture (at wide open aperture). I had never noticed this before. When closing the aperture, this image of the lens barrel is fully obscured by the aperture at f11.5. I mounted a Nikon M 450mm f9.0 and observed similar, but not as significant, results, which also disappeared by f11. In practice, I’ve never noticed any darkening of the sky edges with the Nikon 450mm (but then I don’t think I ever used this lens wider than probably f16).

So I’m wondering, how is the best way to test a shade for vignetting, without exposing any film?

 

[George]

Helen,
my experience is that most of photographers indeed don't care about high level lens shades. I think the reason is that nobody sells them so they never had any contact with it. In one word - ignorance. Surprisingly, as you say, there is not too much about it in text books either. Anyway, to answer you question - let's move it to the PM area. Those who are interested can always ask me for a dimension for their lenses. Just last public word about the "methods". The square format uses the same algorithms as all the other formats. Cheers, George.

 

[George]

Eric,
I myself don't need any ways whatever to test the shades if they work correctly or not - they do. All of them, no vignetting ever in 15 years of their production. Of course, you can check it (by means of the taken picture) on your specimen. The proof is in the pudding. Just let us know when you took the first pictures with the shade so that it dispels doubts. Cheers, George.