Calculating a 'fixed' Compendium hood

[Tim Povlick]

How does one design a 'fixed' compendium hood. The meaning of fixed is the hood is CNC'ed from a solid block of material. This would be for use on a camera using only a single lens (not interchangeable), 150mm f/l and the film is 6x12.

It seems the opening should be a ratio of 12 : 6 or 2 : 1. That's straightforward. Let's say the hood will be 3" x 6" opening. How far from the lens should this opening be to have maximum shielding effect without causing vignetting?

One could calculate the angle of the opening (on the film side) knowing the 6x12 image is formed 150mm from the lens. But the hood goes on the opposite side, so does one use the thin lens equation?


I did a quick search of the web and this forum but did not see anything close to the topic.


_..--
TiM

[Jim Jones]

It is probably easier to measure the minimum front opening of the lens hood rather than calculate it. The opening will be close to, but not exactly equal to, a 1:2 ratio because the radius of the entrance pupil must be added all around a theoretically calculated opening.

[jb7]

The most efficient hood will be of a 'petal' type, you can find examples by searching on ebay...
If it's going to be made using cnc, then you will have to take into account the magnification and aperture...

[Bob Salomon]

The most effective compendium will have an opening the same size as the film format and a length equal to the distance the lens is extended from the film plane. That will only let in light that actually forms the image. Since that is impractical in use better compendiums have adjustable masks to cut off the stray light rays.

[ic-racer]

The rays enter through the front nodal point. So you need to know exactly where that is if you are going to design a lenshood using math.

[vinny]

In the motion picture world we have fixed mat boxes and "hard mats" which clip on the front which are specific to each focal length. Putting a 50mm hard mat on when using a 50mm lens only allows image forming light in like bob mentioned.
You could make the hood then add a hard mat which would have a smaller hole. I'd make a long prototype from foam core, mount it, and cut away the excess material until it doesn't vignette.

[Tim Povlick]

Thanks very much for the helpful replies.
As ic-racer points out the nodal point is needed and since this is not known toany accuracy I will skip the math exercise.

Will use some foam core pieces per Vinny's excellent idea.
Vinny's post reminds me of a post from our resident DP (MW) to some matt boxes used in the industry (Airel?), and while a work of art, the cost was nearly astronomical. Reminds me of my Father always saying don't make a Hollywood production out of things.

The petal design would be good and appreciate that suggestion as well.
Not sure but this seems like something a 3D printer would do well.

Many Thanks for the expert help and advice.


Tim

[Drew Bedo]

CNC machining? It could also be done as a one-off with 3-D additive printing in plastic or metal.

[Jac@stafford.net]

Why would anyone want to build a rigid hood when you could have a less expensive, adjustable, perfectly functional shade?

[ic-racer]

Other things not mentioned are:
1) Mount a real compendium shade and adjust it to suit and copy the dimensions.
2) Experiment with 'good guess' designs and 3D printing then CNC the final one out of metal.

[Tim Povlick]

Other things not mentioned are:
1) Mount a real compendium shade and adjust it to suit and copy the dimensions.
2) Experiment with 'good guess' designs and 3D printing then CNC the final one out of metal.


#1.
I have an adjustable shade but it's the wrong aspect ratio. Your post got me to thinking; I can put a mask on the outside of the hood with correct aspect ratio and get the final dimensions. I was trying some experiments and this seems straightforward enough.

#2.
More than likely I won't go with metal but 'good' plastic. Then when the camera is dropped, the hood will absorb some of the impact.

Thanks!

Tim

[Daniel Stone]

Why would anyone want to build a rigid hood when you could have a less expensive, adjustable, perfectly functional shade?

Jac,
If you read Tim's original posting, he won't need an "adujstable" hood simply because he'll only be using one lens, and he's planning on making/having made one to suit the lens.
[EDIT]: after seeing Tim's response (above), it might work to simply cut that "hard mask" to fit to the outside of that adjustable(bellows, like a Lee?) hood.
If he's intending on only using image-forming light in, then a hard-mat(as Vinny mentioned) will allow for more contrast(due to less 'scattered light' bouncing around within the camera housing.

Tim,
IIRC from when I was researching a similar task myself, make sure to have your lens set/focused for ∞ when cutting/sizing your hard-mat, not closer than ∞. Even if you're using a helicoid for focusing, the extra amount of scattered light(due to lens being farther from film, so IC gets bigger) will be negligible, at best. Hope you're well! Been a while!

cheers,
Dan

[Tim Povlick]

Jac,
If you read Tim's original posting, he won't need an "adujstable" hood simply because he'll only be using one lens, and he's planning on making/having made one to suit the lens.
[EDIT]: after seeing Tim's response (above), it might work to simply cut that "hard mask" to fit to the outside of that adjustable(bellows, like a Lee?) hood.
If he's intending on only using image-forming light in, then a hard-mat(as Vinny mentioned) will allow for more contrast(due to less 'scattered light' bouncing around within the camera housing.

Tim,
IIRC from when I was researching a similar task myself, make sure to have your lens set/focused for ∞ when cutting/sizing your hard-mat, not closer than ∞. Even if you're using a helicoid for focusing, the extra amount of scattered light(due to lens being farther from film, so IC gets bigger) will be negligible, at best. Hope you're well! Been a while!

cheers,
Dan

Hi Dan,

I have the Lee hood but would rather have a fixed one one doesn't have to check the adjustments. For this camera I am trying to make it very fast shooting. I'll use the 8x10 for when time permits.

Thanks for the tip about the focuser. I was thinking the macro position woulb be the problem but you got me to thinking more about it and I can see you're right.

I have tried to reduce the amount of internal reflection with some flat blcak paint. That helped but nto enough. Some place I have some Protostar flocking material and I think that would really cut down the reflection.

In telescopes they use the protstar material on the internal walls and a series of several baffles. There is no stray light!

All is well here. The fires got very close (1/2 mile) but no problems. Busy with the day job. I hope you are doing well also.

Thanks and
Best Regards,

Tim

[Daniel Stone]

Hi Dan,

I have the Lee hood but would rather have a fixed one one doesn't have to check the adjustments. For this camera I am trying to make it very fast shooting. I'll use the 8x10 for when time permits.

Thanks for the tip about the focuser. I was thinking the macro position woulb be the problem but you got me to thinking more about it and I can see you're right.

I have tried to reduce the amount of internal reflection with some flat blcak paint. That helped but nto enough. Some place I have some Protostar flocking material and I think that would really cut down the reflection.

In telescopes they use the protstar material on the internal walls and a series of several baffles. There is no stray light!

All is well here. The fires got very close (1/2 mile) but no problems. Busy with the day job. I hope you are doing well also.

Thanks and
Best Regards,

Tim

Tim,
Glad to hear you guys weren't involved(as in property being involved) with the fires! I'm in FL now, chasing down work on the big white superyachts, so not able to shoot LF at this present moment in life ;).
I remember you talking about that Protostar material when we last met up in-person, glad it can find another use besides telescopes!
Definitely test the ∞ vs. macro situation for maximum "coverage" of stray-light reduction. My brain is pretty much fried from being in the FL sun all day today washing down two sport fishing boats, so maybe I got my memory log a bit backwards due to being tired :P. I wouldn't want to mis-quote something.

-Dan

[el french]

+1 on the ProtoStar flocking. It can really make a big difference.

Would it be correct to rephrase the formula by saying the length should equal the focal length of the lens and the width and height should equal the width and height of the film? My 200mm lens also has a diameter of 200mm, so the width and height of the compendium should be smaller than the diameter of the lens.

p.s. Is there a case where the compendium should be larger than the film size?

[Emmanuel BIGLER]

From ic-racer:

The rays enter through the front nodal point.

Ahem ... except if the lens is not symmetric: in an asymmetric lens, the rays enter through the the entrance pupil of course! And nowhere else ;)
(But hopefully for most view camera lenses, except telephotos, [and so far no retrofocus in the large format world, except an improbable 120 mm retrofocus lens aus Jena, mentioned once here] the entrance pupil is located very close to the front nodal point :cool: )

--------------

Another approach to determine the proper rectangular mask to put in front of your lens would be the following experiment.
Install the camera on a tripod in a darkroom and illuminate the open film gate from backward with a light bulb. Place a piece of translucent paper in front of the lens and see the (defocused) projection of the film gate on the translucent paper through the lens, light propagating in the reverse direction.

This will give you the shape and size of the rectangle that best fits your compendium for a given film shape and size.
Beware however if the lens is a wide-angle lens, that the light bub can actuallly send enough slanted rays to cover the whole visible angular field. May be the use of the ground glass of another piece of translucent paper in the film gate would help to actually get an illumination system with slanted rays up to the extreme angular field?
To be checked in the real world, but this experiment is easy and does not require any maths.
Whichever method you'll implement to design you compendium, e.g. with a rectangular mask like in professional compendiums, double check after having cut the mask located in front, that, as seen from the back, from the corners of the opened film gate, you cannot see the edges of the mask through the lens, looking from the rear of the camera.
Otherwise you'll get some basic good old mechanical vignetting.
But double check also with the lens stopped down at the usual working f-number: if you can see the edges of the mask with the lens wide open, this might not be a real problem if you don't see them when the diaphragm is closed down to the usual f-stop.

Again, wide-angle lenses are those for which an efficient compendium is the most difficult to fabricate without vigneting.

And if the lens is a telephoto, not only you do not have to care for the position of the nodal points (a telephoto being asymmetric, nodal points are definitely disqualified in this game, as explained), but you do not even have to care for the position of the pupils: simply look at the backward-projected image of your film gate, in front of the lens!

[Bob Salomon]

Hi Dan,

I have the Lee hood but would rather have a fixed one one doesn't have to check the adjustments. For this camera I am trying to make it very fast shooting. I'll use the 8x10 for when time permits.

....Tim

As long as you are always shooing at the same image ratio/distance.

[Tim Povlick]

Hi Emmanuel,

Many thanks for that method. It will be simple and once done can be easily verified.

I enjoy studying about the technical aspects as much as capturing images, so a math solution would have been of interest.

Turns out there are 3D printers at work so I may be able to have a hood 3D-printed w/o to much effort.

I have a piece of material similar to Protostar and it seems to really knock down the internal reflection. A piece of Protostar material is placed 'somewhere' for safe keeping. If I could find it, the project might finish.

BTW the camera is a Fotoman 6x12 and lens is a Rodenstock Apo-Sironar-S 150mm 5-5.6. Camera build quality is good just need to have for this long lens to deal with internal reflections off the cone.

_..--
TiM

[Emmanuel BIGLER]

and lens is a Rodenstock Apo-Sironar-S 150mm 5-5.6.

Hence the situation is very simple. We are dealing with a quasi-symmetrical lens design for which the pupils are located very close to the nodal or principal planes, within a few millimeters.
I do not have any data sheet for the precise position of nodal / principal planes or pupils for Rodenstock LF lenses, but if you consider that the entrance pupil il located a few mm in front of the actual, physical position of the iris, for your project, you'll be precise enough for doing a simple drawing on the back of an envelope.

The font filter mount of the apo sironar S is located about 25 mm (about 1 inch) in front of the entrance pupil / front nodal point. Hence if you attach by any means a rectangular mask of about 6 cm by 12 cm at (150 - 25) = 125 mm (about 5") in front of the filter mount, you'll get something close to an optimum compendium mask.
You simply have to keep the proportion between the mask dimensions and the distance to the entrance pupil constant i.e. in the proportion of 150/120 for the long size of the rectangular aperture.

For example if you wish to use a smaller compendium mask located closer to the filter mount, the calculation is just a simple proportionnal correction. For example, imagine that you wish to set the mask 100 mm (4") in front of the filter ring. Hence the distance to the entrance pupil will be only 125 mm (5"). The mask's opening will be 60 x 125/150 = 50 mm and 120 x 125/150 = 100 mm

Actually the real image size delivered by a given 6x12 film back depends on the back's manufacturer.
I have a Horseman 6x12 back and image size is actually 56 x 114 mm i.e. slightly smaller than 60x120 mm. 56 mm is a common value for rollfilm image width, on Hasselblad backs image width is 54 mm. If LF rollfilm backs, image length can be .. variable from one back to another ;)

But anyhow, after a rought design on the back of an envelope as explained above, you should of course make the final visual test to check for a possible mechanical vignetting.

Regarding the position of the entrance pupil, for all classical wide-angle and standard view camera lenses, it is located very close to the center of the shutter, i.e. very close to the physical iris.

But for a telephoto, things are very different.
For the 360 mm Tele-Arton, the entrance pupil is located a few mm behing the last lens vertex, i.e. deeply recessed, if you look at the image of the iris (= the entarnce pupil by definition) as seen from the front of the lens.
And even worse: the conventional perspective ray tracing similar to what we know from ancient Renaissance times in the camera obscura will fail due to the fact that the pupils are not located in the nodal planes!
Hence for the 360 mm Tele Arton (a Schneider lens) for the 4x5" format [actually the 360 Tele-Arton was designed for the 5x7" format], I would start by an hypothetical rectangular mask of size 4" by 5" located ~ 350 mm in front of the last lens vertex (where the entrance pupil is located) i.e. about 340 mm in front of the lens board.
And since this distance would be too large for any practical use, I would reduce everything by a factor 2, i.e. a 2" by 2.5" rectangular aperture located about 175 mm in front of the last lens vertex. Looking at Scheider's detailed specs It would then be easy to subtract from this distance of 175 mm the actual distance between the front and the last lens vertices and attach my rectangular mask to the front filter ring at the proper distance.

(and now, sotto voce: actually I would simply use a bellows compendium that I have in my collection, and I would stretch the compedium's bellows while checking from the rear, looking for vignetting, and that's it (no maths required ;) )

[Tim Povlick]

Hello Emmanuel,

Many thanks for the detailed and informative explanation. Initial testing will use this as a basis. I have the Lee adjustable hood and the front has two slides / slots where one could put the test mask in place. I have the Horseman 6x12 back so it's good you pointed out the actual image size is not 6x12. (I wish the industry had gone with 70mm film)

One thing that's nagging at me is the solution should also look good and not spoil the pretty camera. But that's outside the scope of this thread.

My thanks for your expertise and time!

Kindest Regards,

Tim

[Emmanuel BIGLER]

the solution should also look good and not spoil the pretty camera.

Once you have found the approximate shape, size and position for the compendium mask, you have many degrees of freedom as far as the actual look of the item can be.
A 3D printing project looks attractive: I know that some machines can fabricate parts made of a kind of a black polymer. The ones I know deliver some kind of semi-translucent, white-ish stuff, not very pretty, that would need some subsequent painting in matte black.
But you can also imagine a combination of nicely hand-crafted alumin(i)um or wooden parts assembled together. For holding the lens hood I would favor a simple clamp-style system around the filter ring in order to be able to precisely position the hood in rotation. You can of course glue any parts on a threaded ring that you can find for cheap (an empty filter holder) but you'll have to think about some means to setting the hood in the right "horizontal" position after screwing the ring to front of the lens barrel.

Have fun!