Where did My Numbers Go ???

[Steaphany]

Ok, I know I do not sit on the bleeding edge of photographic technology, even my SD14 was only purchased after Sigma had it out for a couple years. I still feel at home with manual cameras, film based imaging technology, and first generation image stabilization. ( A tripod )

From the posts I've made, I'm sure most CiC members realize I delve into the numbers, calculating hyperfocal distances, depth of fields, exposure values, and generally working in photography as an engineering and scientific discipline.

So, in my usual preparation for some astronomical star field shots, part of experiments in film preexposure to allow the film record an image with less incident light, I wanted to set the focus between a calculated hyperfocal distance and ∞ so that I know I'd have an optimal focus and not have to rely on setting the focus by eye. A Note to Newbies - Auto Focus will not work with low light or point sources, in this case it has to be manual.

First off, all my "modern" lenses focus beyond ∞, which is why I decided to pursue the hyperfocal distance route. That's no big deal, a quick and dirty bunch of calculations in a spread sheet has that done.

Now the problem comes from my "modern" lenses, No Numbers !! At least not where I need them to be ! The "on lens" focal scale for my normal to wide angle lenses runs from 1m or 2m straight to ∞ ! The best "modern" lens is my old Sigma 70-300 Macro Zoom ( it was a kit lens with my SA9 ) that at least goes from 8m to ∞.

First Question: How do I set the focus of my 28mm lens set at f/1.8 between the hyperfocal distance of 87m and ∞ ? Especially when the lens can focus beyond ∞ and I know the scale itself may not be entirely accurate. If the lens focal adjust ring stopped right at ∞, this would be a non-issue.

I do not accept settling for f/22 just to bring the hyperfocal distance down to 7m.

Note: These hyperfocal distances come from calculating for a Circle of Confusion of 5um, for my needs the standard 30um is too large. For a star field astronomical photo, a smaller Circle of Confusion means the film can image fainter stars. Equivalent to going from ISO 200 to over ISO 1000 with no increase in noise.

A tool like dSLR Focus will not help here, I'm using my Sigma SA9, a film SLR, but the lens is all digital age technology.

Second Question: When did lens design and lens manufacturers so extremely compress the ∞ end of the focal scale ? I know the scale is logarithmic, but 1m to ∞ is a bit much.

( I know - Another weird post )

[arith]

First Question: How do I set the focus of my 28mm lens set at f/1.8 between the hyperfocal distance of 87m and ∞ ? Especially when the lens can focus beyond ∞ and I know the scale itself may not be entirely accurate. If the lens focal adjust ring stopped right at ∞, this would be a non-issue.

I know this may be a stupid answer, so just ignore it if it is: have you tried focusing on something say 100 metres away using AF, possibly using a bright light or in daylight, then possibly taking note of the position or putting a sticky label on or.... I said it was stupid.

[Steaphany]

I know this may be a stupid answer, so just ignore it if it is: have you tried focusing on something say 100 metres away using AF, possibly using a bright light or in daylight, then possibly taking note of the position or putting a sticky label on or.... I said it was stupid.

Actually, It is not a stupid answer, though the approach I thought of along this line was to focus on something a bit further - The Moon.

During the Moon's Waxing Crescent phase, 20% to 30% illuminated, when the Moon sets in the early evening, it could serve as a reasonable target with good contrast distinctions from a hard edge ( an auto focus' dream ) and a perfect distance of 384,000km. Since it sets early, there would be plenty of Moonless dark sky until dawn.

The effectiveness of the Moon is a matter of focal length. Being only 0.51° in diameter, if I use a 300mm lens the Moon would only occupy 11% of the height of the frame, but when shooting wide angle with a 28mm lens, the Moon ends up being 1% of the height of the frame and too small for the auto focus to lock on to.

Compounding the issue of focal length is exposure time. I do not have a permanently installed and precision aligned equatorial mount for tracking the sky. ( Just try align a goto mount by the view through an SLR with no illuminated reticle. ) To ensure that every star remains a point in the final image, I was figuring on using a shutter set short enough that a star would remain within 2 pixels or a film Line Pair width. Stars with a Declination of 0°, over the Earth's Equator, would cross through a fame faster than stars at other positions and a 300mm lens would need a shutter under ½ second. The choice for a 28mm lens would allow the same clarity with a 5 second shutter and 10 times the light

Alternately, I could set up before Sun set and lock the auto focus onto a view of the horizon, with the terrain here I'd easily get a clear view out to 3km to 5km, or even more, but now that means sitting out in a field until dark with the camera and lens exposed to insects and the wonderful Texas dust. The road that passes my ranch is gravel and dirt with little traffic, but when anyone does drive by they leave a dust plume that will drift across the fields. None on my lenses have any means to focal lock the focus ring's position and I dare not even put the lens cap on due to the potential of offsetting the focal position.

This is why I was aiming for a go by the numbers approach.

( I've been scratching my head on this puzzle for a while now )

[Steaphany]

Well, I've made some progress on this one. I posed this query on three forums and even sent it off to Sigma to see what their lens designers had to say. ( Since I'm using a SA9, I suspect that I may not get an answer due to their priority is their new SD1, which I say is over priced )

From APUG, I received a great explaination to my second question:

II: The compressed distance markings are usually tied to a short focusing throw. With a short throw (say, 90° instead of 270°), the lens can be focused more quickly, though perhaps not as accurately.

Which makes a lot of sense due to marketing pressures for lenses having a fast focus. To halve the focus time, you can halve the throw, everything else can remain the same. This applies equally well for auto and manual focus photographers, but does not bode well for precision photographers or photographers who are able to take an hour or more to set up for a shot. A problem here comes back to marketing, I have never seen focus throw as a specification on a lens. There is no way to filter lenses to find those with a throw of 270° or greater. Even Macro Lenses are so general purpose that they amount to being a general purpose lens first with close focus lengths added on. ( Two of my lenses have a switch to enable Macro mode. ) Photographers end up having to be stuck with design trade offs to achieve a lens that everyone would want.

If lens designers pulled themselves back from designing general purpose lenses and put in the effort to creating application specific designs, I'm sure many photographers would jump at the opportunity.

This has brought me to conclude, If I can not buy an application specific lens, I'll find the books to design what I want and make my own lenses from COTS optical and mechanical components. If I do need components machined, I know a photographic machinist that I can turn to.

Well, I found Lens Design Fundamentals, Second Edition by Rudolf Kingslake and R. Barry Johnson. Rudolf Kingslake is regarded as the American father of lens design; his book is viewed as a classic in the field. This 570 page text provides the skills and knowledge to move into the exciting world of contemporary lens design and develop practical lenses needed for the great variety of 21st-century applications. ( Fun reading, Lots of formulas )

If anyone wants a copy, just do a google search.

Now, If I want a manual focus with a 1,800° focal throw or to have a precision hard stop at ∞, I can put that into the design. Since I'm not into mass production, I do not have to address issues of reproducibility or high volume manufacturing nor would I have some marketing specialists nagging me over "Not every photographer would want that".

Infrastructure wise, I already have the software tools to permit me to ray trace and model wavelength specific lens performance.

[rpcrowe]

Steaphany,

Why would you want to use hyperfocal distances in focusing on an infinity subject?

Hyperfocal distance is, at best, a compromise, providing the widest depth of field however, not necessarily the sharpest focus at either end of the DOF.

The hyperfocal distance, when focused on, will provide "acceptable" focus from half that distance to infinity.

The degree of "acceptable" focus is determined by the circle of confusion which is, in turn determined by the original image size and how much that image needs to be enlarged.

[Steaphany]

Richard,

My reason for considering the hyperfocal distance to aid in an ∞ focus setting is due to none of my lenses have a hard stop at ∞.

The properties of the hyperfocal distance are:

1. If the lens is set to a focus position of the hyperfocal distance, the depth of field will be one half the hyperfocal distance all the way out to ∞.
2. If the lens is set to a focus position for ∞, the depth of field will be from the hyperfocal distance to ∞.

Now, when considering that the focal scale on a lens may not be entirely accurate, absolute positions can no longer be relied upon, but if you know two values which would both achieve a specific goal, then you will know that a mid point position will do what is needed, compensating for scale errors.

So, if ∞ is essential, any lens focus setting between the ∞ position and the hyperfocal distance position will permit the depth of field to include ∞, at the accuracy of the circle of confusion used to determine the hyperfocal distance. To ensure that the focus is actually at a precision that I desire, I calculate the hyperfocal distance not by a value determined by the frame size, as is traditional with the circle of confusion, by off the imager or film resolution.

The lens I referenced in my original post is 28mm, using a film with a resolution feature size of 5um, I take this value as the acceptable circle of confusion and calculate a hyperfocal distance of 87m at f/1.8. So, in stead of a hard stop at ∞, a focal setting anywhere between the 87m position and ∞ will be JUST as good.

[Steaphany]

To help you understand how I'm using the hyperfocal distance, here is a graph I just calculated that shows how the hyperfocal distance will vary depending on the value used for the Circle of Confusion:

This graph holds the lens focal length at 28mm and the aperture at f/1.8.

So, if I chose to accept a circle of confusion of 30um, ∞ would be focused at least at this level of clarity if the positioning of the lens focus setting falls between 15m and ∞. For the clarity I want to aim for, 5um, I'm correspondingly forced to use a significantly greater hyperfocal distance, 87m.

Just remember that several factors come into play with the hyperfocal distance, focal length, aperture, and clarity of the focused image as defined by the circle of confusion, these can be juggled about or applied how ever You need to capture an image. It's just a matter of understanding that hyperfocal distance can be leveraged in other areas than just depth of field.

[JonathanC]

Well Steaphany, here's where we can go "old school" with a knife edge focus setup like this:-
http://www.astropix.com/HTML/I_ASTRO...METHODS.HTM#KE
For those who don't want to know too much, the method involves placing a razor blade type object on the film rails and looking past the edge at the focussed star. When the star image is extremely small, any movement of the eye leads to an on/off of the star image. The more out of focus the star image is, the more gentle the transition from light to dark.
I've tried the Stiletto focus system, but found it inadequate. Maybe I didn't play with it long enough or try to adjust it, but I got poor results.
WRT focusing lenses at infinity, I found masking tape wound round the lens after daylight focusing works OK.

BTW drift alignment works OK on a goto mount if you have a digital SLR, even if its a bit tedious.

HTH