Lens hood length - avoiding vignetting

[DBP]

I'm looking for a lens hood for my 180mm Sonnar (P6 mount). It takes an 86mm thread, so options are a bit limited. The one I am currently considering is 73mm deep and I wonder whether it will vignet. Does anyone know, or have a method for figuring it out?

 

[Dave Parker]

Cut a simular size piece of stiff cardboard and make a hood of approx the same dimentions and then you can hold on the end of the lens and should be able to see if it would vignette, you could also lightly tape to the end of the lens and take a few test shots and have them developed..

Dave

 

[DBP]

I'm feeling stupid now, should have thought of that myself. Thanks.

 

[George]

If you want, let me know the diameter of the front glass, the film format in use and I will give you the lens shade's size for different shading efficiency.

 

[Dave Parker]

Also, forgot to point this website out, they have downloadable templates in .pdf format to make your own lenshoods for many lenses as well as a custom design template, it could be helpful if you have to design your own.

http://www.lenshoods.co.uk/

Dave

 

[DBP]

If you want, let me know the diameter of the front glass, the film format in use and I will give you the lens shade's size for different shading efficiency.

Approximately 82mm diameter and 6x6.

 

[George]

For your 180mm focal length lens, 82mm diameter of the front lens and 6x6 film format, you can use a lens hood:
1) 74mm long, 76x76mm opening - medium efficiency, but better what you find on the market;
2) 98mm long, 81x 81mm opening - very good efficiency;
3) 118mm long, 86x86mm opening - better than very good efficiency;
4) 138mm long, 91x91mm opening - excellent efficiency.

In case the 82mm front lens diameter was underestimated you would need to contact me again. The length of the lens shade is measured strictly from the front lens appex. If you screw the lens shade on the filter thread you have to lessen the length with the proper number of mm.

 

[DBP]

Thanks.

 

[Helen B]

George,

I'm a bit puzzled by your lens hood calculations. Could you explain how you did them? Your opening sizes seem to be rather small, unless I've misunderstood what you meant by 'apex' - I assumed the front vertex. If you meant the entrance pupil, wouldn't you need to know the lens' aperture?

Puzzled,
Helen

 

[George]

Opening sizes rather small? They are the smallest usable for the given lens shade length, the film format and the front diameter. Anything smaller and you would have the beginning of vignetting. Anything biggger and you loose the efficiency. The longer the lens shade, the better the efficiency, but it's not a linear function. After a certain moment you would gain very little with a huge increase of the shade's length. The apex means indeed the front vertex. The calculation is for the maximum aperture - so that you are not limited with your choice of working apertures. To explain the calculation would be simply too long. I build my own lens shades for all my lenses - I did so for the last 15 years.

 

[Jim Jones]

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.

 

[Helen B]

Here's a simple way of calculating the diagonal of a rectangular or square hood, or the diameter of a round hood. It assumes that the front element is only just big enough to avoid vignetting when the lens is wide open (this may not be true, and the assumption may lead to hoods that are slightly larger than they need to be).

f = lens focal length
S = film diagonal
L = diameter of front lens element
H = diameter or diagonal of lens hood
d = distance lens hood protrudes in front of front element, measured from the periphery of a convex front surface, not the vertex.

All measurements should be consistent, eg all in millimetres. It gives hoods that are slightly larger than George's. The formula is based on similar triangles, with the fundamental one being the one formed by the film diagonal and the focal length.

Strictly speaking it only applies for a round hood, but the correct calculation for a rectangular or square hood would require the aperture, and it would be more complicated. The error should not be great, though the error does reduce the size of the hood to slightly below what it would be with a rigorous calculation.

H = S.d/f + L

Best,
Helen

 

[George]

Sure -your calculation gives a lens hood with bigger than necessary opening so it increases the size of the hood, not reduces as you say. That leads to reduced efficiency - the common defect of all the shades on market. Also, calculating a lens hood according to the aperture in use has no practical meaning - you don't want to change your lens hood when you change your aperture. In itself calculating a square or rectangular lens shade (and only such a shade can be really effective in all situations) doesn't require aperture input more than a calculation of a round lens hood.

 

[vanspaendonck]

This site may be helpful:

Dead Link Removed

 

[George]

The article has some imprecise statements. There is nothing like an "optimum" lens hood length. That simply because there is nothing like "optimum" lens shade efficiency. You can allways make a lens hood more efficient, when you make it longer (with the corresponding non linear increase of its opening). What value is optimum is just in your idea of the sufficient efficiency of your lens shade. The most efficient lens shade would need to be protruded beyond the light source causing flare. Of course, that is not possible with Sun light nor is it completely possible with other light sources. Therefore you have to decide what kind of efficiency is acceptable or wanted for you and you stop there. It is unfortunate that the inefficient lens shades often associated with sold lenses give photographers a sense of accomplished shading - most of these shade are far, far from being even good, not to say excellent for shading a lens from the flare causing light. If, f.ex. the Fuji 690GW has the correct lens shade length (for its given diameter) it looses a lot by not being rectangular. Few photographers know what they are loosing in terms of contrast and color saturation using insufficient lens shades - simply because they never had a really efficient lens hood on their lenses. Yet, the cinematography knows better - just think of their barn doors lens shades.

 

[George]

One more detail. In my calculations I use a way how to express the efficiency of the designed lens shade. I express it in the percentage of the eliminated flare causing light, given certain general way of illumination. In such a way I know where to stop prolonging the lens shade, as after a certain point you get (in a non linear way) too little (efficiency increase) for too much (shade's dimension). Even if in the real situation the efficiency could be less than calculated in the very majority of cases it is much, much more (given the way I pressume the lighting).

 

[George]

Helen,
sorry to say, but people must be warned ;-)
As you say, your calculation is just for a round hood. This means that in some cases they are of very small efficiency. In many cases, like for longer focal lengths, you can make a lens hood with a rectangular opening that has one side (the vertical if it is used on a panoramic format) even smaller than the front lens diameter! In the case of your calculation this dimension would be increased with several cm - making the shade insufficient almost like those rubber shades sold in shops (I call them "rubbish" shades, for that's what they are). Sorry to say, but someone who doesn't know could think that calculating with you equation he gets a good lens shade while obteining a nothing close to it. Only shades that copy the film format shape can be efficient - provided they are well calculated or tried. Com'on, you know better than that - give them the true stuff ;-)

 

[Helen B]

George,

You didn't read what I said: "Here's a simple way of calculating the diagonal of a rectangular or square hood..." The results are similar to yours, and, using that simple formula (I can't call it 'my formula') one dimension could be smaller than the diameter of the front element. I agree that lens shades should copy the film format for maximum effect.

Where did I suggest that the lens hood should be sized for the aperture in use? The relevance of the maximum aperture for rectangular hoods is to try to get as close to the extremities of the ray bundles as possible without any obstruction of them. You run into difficulties for a rectangular hood because the envelope of the ray bundles is not a simple projection of the film format - it's generally barrel-shaped to some degree. This can be seen in the design of the entry baffles for some lenses. Of course you need to know a lot more than just the aperture to design the ultimate baffle or matte.

I only offered the simple formula in the absence of anything better - you had turned down my request to explain your method - and I outlined the reasoning and assumptions behind the formula I gave. In reality I wouldn't use a purely theoretical method to determine the geometry of a lens hood - I'd use it as a starting point, then use a method similar to the one Jim describes to fine tune it.

I, for one, would be very grateful to you if you were to explain your method.

Best,
Helen

 

[George]

Don't worry, I read it all. If the poor guy gets just the value H of the diagonal you still have to give him a way how to calculate the correct rectangular dimensions as there are more different rectangulars with the same diagonal - only one of them corresponds relatively to the film format.
You yourself say that the correct calculation for a rect. or sqr hood would require the aperture. Sorry, the barrel shaped projection is just of a purely academic importance when you think of the practical issues involved in the lens shade use. You have to have some space for not absolutely precisely placed hood, not absolutely precise distance from the lens, etc. You don't make the lens shade with a micrometer either. And because the given lens shade has anyway just a limited efficiency you cannot strive for any "ultimate" efficiency or a hood shape.
I use the calculations exactly to avoid a plenty of Jim's "fine tuning" - not at all easy when you have to move all for sides and keep the centre of them on the axis. Actually, when I put the lens shade on the lens and try then to shorten one or the other side it shows vignetting immediately - such is the precision of the calculation.
The reason I don't want to go to lengthy explanations of the calculation is mainly practical. I wrote it as a computer program 15 years ago and I didn't care about it from then. It's in the old Qbasic (remember?) and I never cared to rewrite it again - on my XP I use the Qbasic taken from the history. Nobody will pay me the time for explaning the whole theory. Whoever wants, can ask me the dimensions. Cheers.

 

[George]

As an added bonus here's a real story. I once calculated a lens shade for a guy with his Alpa camera with a wide angle, 47mm if I recall it well. He made it(very nicely, machined from a block of some hard plastic, I was impressed) only to write to me later that when he started to use this lens hood he noticed that he didn't need a centre filter for the lens anymore. And he sent me pictures taken with and without the lens shade and the filter. I was surprised, and sincerely, I don't know the reason for that effect on the film. Helen, any theory?

 

[Helen B]

... If the poor guy gets just the value H of the diagonal you still have to give him a way how to calculate the correct rectangular dimensions as there are more different rectangulars with the same diagonal - only one of them corresponds relatively to the film format.
...

Well, I thought that it was too simple to have to explain. All you do is multiply the relevant film dimension (the height or the width) by the ratio of the hood diagonal over the film diagonal. For example:

Hood height = Film height x H/S

I’d be interested to see the precision of your method in comparison to the crudity of the one I gave. Could you give me a range of hood sizes for the following lenses:

55 mm lens with 46 mm diameter front element;
80 mm lens with 44 mm diameter front element;
400 mm lens with 69 mm diameter front element.

The film dimensions are 94 mm x 120 mm.

Many thanks,
Helen
PS If a good lens hood reduced the difference in illumination between the centre and the edges of the frame, then it is likely that a significant portion of the difference was caused by flare being greater in the centre than at the edges. That's not unusual. I've never felt the need for a centre filter for a 50 mm lens on 6x9.

 

[George]

Here it goes (doubble checked) -
1) 55mm lens - for L (length of the shade) 60.8mm, A and B sides = 92.4 x 118mm
2) 80mm lens - L = 86.2mm, AxB = 89.6 x 114.4 mm
3) 400mm lens - L= 99.3mm, AxB = 56.9x72.7mm
Don't forget that I measure the length L from the lens vertex. Let me know how far it is from your model.

You explanation seems to be logical. It somehow goes against the common sense that most of the flare could be in the center. Simply because the smaller the lens diameter the better chances against the flare the lens has... Anyway, I believe you. George

 

[George]

Curious, I just checked it for at least the first of your examples. Hm, something is not good in your model - it gives very unnecessary big dimensions for the hood. If I didn't make a mistake, your model in the first case gave me 214mm hood diagonal against my 148.9 mm. That is too a big difference, you would be loosing too much of the efficiency. In fact if I were not so lazy I would find my original algorithms for my program but I have too much to do in this period of time. I even sent, some 12 years ago, the computer program with all the geometry to the Lee hoods firm and they answered me with a letter, telling me that my hoods are more efficient than theirs (not in production yet), for several reasons. (As a matter of fact my prototype of the first self supporting lens shade preceded theirs with 1 year, yet another story).

 

[Helen B]

Your maths is correct: there is a huge difference in the two methods. Could you do one more to help me understand what is going wrong with the classic simple method?

Lens focal length 52 mm
Film sides 55 mm x 55 mm
Diameter of front element 50 mm

Thanks very much,
Helen

 

[George]

Here it is: L=70.7mm; AxB = 88.2 x 88.2mm. Good luck!