Locus of Focus

[arith]

Glenn, thanks for the link to the field curvature example ... indeed my fears would have been confirmed in this case, that the "locus of focus", i.e., the "field of focus" has a "field curvature". What I got wrong was that I expected the curvature to be in the other direction, towards the camera.

This field curvature does not seem to be significant with the prime lenses mentioned for photographing 2D artwork, at least at the single focal distance used for photographing the 120 x 80 cm grid in the Photozone reviews (unfortunately they only did this for the one distance, so there may still be concerns ... will the lenses work well closer up?)

I wish reviews of lenses would publish diagrams like the one in Photozone, but go further and shade in a cross section of all the region that would be "in focus", what I like to call the "locus of focus", instead of just a curved line, since there will be a depth of focus all along the curved line. That way one could tell if focusing on the corners (as suggested in the Photozone article) would put the center, and hence the entire scene, into the field of focus. In the hypothetical example pictured here (based on the diagram from Photozone)



one has such a shallow DOF along the field that focusing at the corners will definitely put the center out of focus ... it seems that there is no way, with the given camera settings (and hence with the given field of focus), to put everything in the plane into focus.

Stan

This is really interesting and shows how lazy I've become by not considering it. I've read somewhere it is always best to use the central registration point for focusing but as I normally fill the frame like to use a registration point near a third.

Surprising I didn't know the focus arced as above, but more as if the camera is the centre of a sphere, with appropriate depth of field.

But here they talk about the plane of focus;

http://en.wikipedia.org/wiki/Scheimpflug_principle

Also here, so I suspect I was wrong and the points of sharpest focus lie on a plane.

https://www.cambridgeincolour.com/tu...ft-lenses2.htm

Of course TnT lenses might be useful for art that isn't so easy to get at, but I sure don't understand the locus of focus.

[pnodrog]

If you focus near a third it actually helps as you end up with the field of focus moving forward and bisecting ideal plane of focus therefore optimising the use of depth of field in trying to bring everything into focus. In practice with a flat surface most people will focus in the centre which may result in the corners being out of focus.

[ajohnw]

John,
Thanks for the pointer to reproduction lenses ... I am not familiar with them. It may well be that a high quality planar field of focus is beyond my modest means... but nonetheless I will be happier if I know that I have done a decent job of finding out what works best for me, and how it compares to a first-rate system. I will sit down later this evening and write up a summary of my findings, and what concerns me about them, and post them here ... I have recently purchased a grand old DSLR, the 8 Megapixel Canon 1D, Mark II, and I want to make reasonably sure that I choose a nearly best possible lens (modulo my budget) for photographing artwork with it. Perhaps I will need to use my venerable 4 Megapixel CoolPix 4500 for small pieces ... it seems to have a respectable macro capability.

Stasch

Your cheapest option is likely to be a portrait lens. It too is likely to have been computed for modest distances and the resolution of these tends to be on the hight side as well. One thing you will have to watch for is light fall of towards the corners on full frame even with full frame lenses. Some manufacturers correct for this in the camera and for distortion. Looking at test reports on their 85mm f1.8 it's best performance crops up at F4 / F5.6. At that point fall off, ca and distortion are negligible. The 50mm is more or less the same. Sigma primes are also fairly good but unfortunately have rather wide apertures which pushes the price up. Another option is major make early macro or portrait lenses plus adapter. It's better to go for ones that were aimed at AF film cameras and early digital if they have an aperture ring. Coatings are likely to be better. AF confirm adapters are available. Some I believe can actually be set up to tell the camera what focal length is actually on the camera. The others tell it that it's a 50mm F1.4. In general most lenses loose problems at F5.6 some at F8. Macro lenses have to have low distortion and high resolution to do their job and portrait lenses are much the same. Part of that is flat field as far as the sensor is concerned. Of the general makes Tokina are probably the most reliable. SD indicates modern glasses. ATX too I believe but SD's do for certain. Zooms invariably have distortion but this can be removed with software as can camera alignment.. Resolution is never as good as primes. One aspect of that is that a 3:1 range zoom is usually a lot better than one with a longer range. Some 4:1's can be good. A lot depends on the final viewed image size. :-) I'm a pixel peeper - if I can focus to the pixel level I do. Only problem then is the appalling depth of field at that size of circle of confusion.

I would forget fake white lighting if I were you - tubes for instance. Most of them simulate white by increasing blue. Cameras expect daylight so dynamic range can be reduced by excess blue. Hot filament lighting is a better option really or flash. The flash can be angled just like any other light source and with some the sensor put on the camera. 2 flashes could be used. Best option in many ways are older auto flashguns. Much cheaper than ettl etc. All you have to watch for is that there is a suitable ISO / Aperture rating available. I picked up a Hanimex hammer head recently very very cheaply on ebay. 99 pence as it needs a minor repair. Many old flashes have high trigger voltages but adapters are available with sync sockets and flash gun sockets that isolate the camera. Googeling flash trigger voltage brings up more info. No auto on the flash gun - well the sums are easy and there is often a calculator on the flash. Not really a problem in fixed type set ups.

If you do go for a manual lens I have tried to use an AF confirm adapter on a microscope. Hopeless but it;s a different world when there is a need to focus to 0.001mm of better. A right angle finder with a 2x mag setting was no better. Either may work out for your use. I found that my best bet was an eye cup and the viewfinder. A web search for a make and model of lens often brings up user test reports of one sort or another. Best look for one that shows near full resolution shots.

John

[Stasch]

Manfred, you have as always brought up some excellent points. It should be a routine calculation, using the law of cosines, to find the deviation of the corners of a flat painting caused by the camera lens axis not being perpendicular to the plane of the painting. If you could give me an guesstimate of an average angle error that a photographer with just a tripod mounted camera would be likely to make, then I will be happy to crank out some numbers, say for a 100 mm f/2 lens. Another factor to take into consideration is that the photographer likely will not be able to point the camera at the exact center of the painting, so a guesstimate on the angle error when trying to point at the center could be tossed into the calculations as well. As for the irregularities in the flatness of the painting, a guesstimate on the average deviation could be tossed in as well....

You will no doubt find my goals very modest compared to my great fascination with the minute details ... I have 9 grandnieces and 6 grandnephews, all still children at home, and part of the fun of my recent years has been to try to show them what a clever uncle they have. One big effort in this direction was spending 2.5 years studying silversmithing ... absolutely fascinating subject ... Now it is painting and photography! Napoleon Bonaparte had the great French mathematician Laplace working in his administration; at one point he complained that Laplace had unfortunately brought the spirit of the infinitesimal to his administrative work. I would no doubt be inclined to follow in Laplace's footsteps in a similar situation .... it is in my DNA!

Consequently I would like to know what the requirements are---for the photography end---for Giclee quality prints ... if I can achieve that then it would increase my bragging rights when I make prints of my (still under development) artwork for them.

Your question as to why there are so many good reproductions of paintings out there is one that I would definitely like to know the answer to. When working on the math poster that I had mentioned early in this thread, it was necessary to make some acquaintance with copyright law for pictures. One of the fascinating stories that I ran across was the Bridgeman Art Library vs Corel
lawsuit of 1999. Bridgeman had received permission from major museums to photograph works of the old European masters, and had invested a considerable sum of money into perfecting photographic technology to reproduce the paintings. I do not know any details of what aspects of photography they thought needed to be improved, nor do I know the standards that the great art museums set for reproduction ... whatever they are, they are surely the gold standard. My question is: how close can hobby amateurs come to producing the gold standard? As for Giclee quality printout, I am pretty sure that the large format printers from Epson (that fit on a desktop) suffice for say 17" x 21".

It would be nice to know which consumer lenses and cameras can produce Giclee quality files, and under which conditions .... if I couldn't afford it, at least it would make great coffee table conversation! (like that Leica lens that cost over 2 million dollars...)

[Stasch]

Paul, my reply to Manfred, just posted, probably answers your query about what I am looking for.... Thanks for the info on your setup.

[Manfred M]

I remember being in either an art gallery or a church in Europe a few years ago, near closing time and watched a couple off people in white lab coats setting up to photograph a large painting. Banks of softboxes and other tools; with a camera set up on a platform large enough to support the camera and the photographer. The camera and lens were set up, and it was a DSLR with a lens I recognized; certainly nothing exotic, and I don't recall the brand. What did strike me was the effort in lighting (they were using continuous lighting, not strobes) and setting things up properly, rather than the camera / lens combination being used to take the picture.

Now Glicee, that is pure marketing gibberish. Gicleur is the French word for nozzle, and Glicee is a nice way of saying a high end ink-jet print, no more no less. The term was coined in the early 1990's as a way of obscufating the fact that the the fine art prints were made using a computer and a high quality inkjet printer, and the negative connotations of that technology.

[ajohnw]

There is some info on the equipment the high end people use on here. It matches a program that was on in the UK some time ago on a similar subject, I like look of the monitor but it seems to be over 2000 quid.

http://www.petermennim.com/fine-art-...cambridge.html

Tongue in cheek really - if you have a camera capable of capturing something perfectly or at some other level of perfection a monitor and printer are also needed that will match it. In some ways it's worse at the publishing end of things. I know some one who regularly has work taken with a 5D MK111 or MK11 published and they always want to know where the rest of the files is. Often when it's the first time they have handled this type of work they wonder if they can actually do anything with it.

John

[Stasch]

Steve, the "plane of focus" seems to be the practical view of how a lens focuses, but one can certainly find articles on-line that challenge that assumption.

A Pentax forum from 2009 has quotes such as:

6. The optical designs of most 50mm lenses are very similar, and nearly all show field curvature that is easily misinterpreted as a lens being "soft". I tested eight different 50mm lenses from Nikon, Canon, Zeiss, Olympus, and Sigma, and found that all had field curvature that became obvious with the appropriate subject matter. I did so because in my testing of the Zeiss ZE 50mm f1.4 Planar, I detected strange variations in image sharpness across the frame, even at f5.6. At first i though this was a bad lens, but subsequent testing of the eight 50mm lenses showed this to be common.

One conclusion was that people may be mistakenly assuming their lens has soft corners when they are actually seeing the effects of field curvature ... the lens is focusing sharply, just not on the presumed plane of focus.

[Stasch]

Paul, The 1/3 rule certainly fits in with the assumption of a curved field of focus.

[Stasch]

John, thanks for the detailed recommendations ... it will take me some time to digest them. The local camera shop convinced me to try the continuous fluorescent lighting because it was cool and long-lived.

[xpatUSA]

Stan,

Have you thought about flipping the Canon for an APS-C sensor'd camera? Manfred touched on it earlier when he said your camera has a APS-H size. (very convenient for Canon, it is the largest size that was economical re: chip yield from a slice of silicon, not so convenient for you).

Then you can use a full-frame lens on a camera with a smaller sensor = use of the middle of the lens and a smaller angle of view. A Nikon D90 might suit your purpose quite well (good resolution, < 6um pixel pitch) and they are becoming available quite cheap. Or a different Canon if you're locked into their lenses.

Apologies to anyone that has already mentioned this!

[jcuknz]

I would caution you Stan to take the 1/3 rule with a grain of salt because if we are talking about the same thing it not a constant and changes depending on subject to camera distance .... with for instance a 50mm on full frame the 1/3 starts to move towards 1/1 at around 25ft, getting there at around six feet.

I rarely bother about DoF prefering to select the point I want to be the point of focus and accepting that some things beyond and nearer will be soft.

[xpatUSA]

John, thanks for the detailed recommendations ... it will take me some time to digest them. The local camera shop convinced me to try the continuous fluorescent lighting because it was cool and long-lived.

Have you considered LED lighting? I really bought some cheap-ish LED floodlights, 5000K color temp, with a >90% CRI (which, of course we take with a pinch of salt ;-).

There is a pronounced blue spike in the published spectrum which doesn't really show up using my camera (WB is set to sunlight) but is discernible if looking at a white card. If I bought them again I would probably get the 4000K model which has a much lower blue spike. The nice thing with LED's is that the rest of spectrum is quite smooth which might, or might not, be suitable for your camera and paintings (your call). Continuous cool LED lighting makes my setting up nice and easy too.

Single LED flood, lit from left. RAW image was white balanced on patch #22 gray.

[ajohnw]

One other thing I have seen on lighting pictures for photography and copying etc is the use of 2 halogen desk lamps. I spend some time trying to get to grips with taking photo's through microscopes and white balance is a serious problem. One thing I have noticed is that cameras tend to latch onto the correct value when a microscope is fitted with a halogen bulb. Like old fashioned photographic daylight bulbs tungsten bulbs need over driving to achieve a temperature that the camera can cope with. This doesn't seem to be the case with halogen. Flash - most studios use flash often with an umbrella because they want diffuse light and don't want colour balance problems.

The excess blue comment came about because of experience with high powered LED's on microscope. The highest powered versions often put out twice as much blue as green to make the colour temperature higher. Strip lights - my monitor calibrator comes up with entirely different numbers to what is stated on the tube. This is probably down to a similar problem. They put out various combinations of spikes of coloured light. I've had mixed results with fluorescent light balance in cameras so would wonder if you would get accurate colouring. In some circumstance it is likely to be impossible.

John

[Stasch]

A few years back I attended an art show in Toronto where one of the artists was selling limited edition prints of their work, appx. 17" x 21", for about $1 K each, and found out that they were printed using a top end desk top Epson printer. But at that time I did not think to ask about the camera equipment.

My goal is quite limited ... paintings with a diagonal measurement of 6 inches to 6 ft. So I would hope that two lights would be enough.

I looked up the XRite color checker chart on Amazon.ca, and saw that it was offered for $122.99. Do these things deteriorate with age? I recall that Pantone puts out sets of color strips which are supposed to be replaced every year or two.

[Manfred M]

I looked up the XRite color checker chart on Amazon.ca, and saw that it was offered for $122.99. Do these things deteriorate with age? I recall that Pantone puts out sets of color strips which are supposed to be replaced every year or two.

If you check a bit deeper, you will find that x-Rite owns Pantone, and yes they do recommend replacing the swatches every year or two. I find if you store them in a cool, dark place there is no significant colour shift for a few years, but if you use them in a daily production environment, I would expect the shelf-life to be considerablly shorter.

I had a Pantone swatch book in a previous job and remember unearthing it about 4-1/2 years later when I left the company; I had stored it in a cabinet that was not opened that often. I compared mine to a new one that the Quality Control department had, under their calibrated MacBeth light source (also now owned by x-Rite) and I could not see any difference in the colours.

[pnodrog]

I would caution you Stan to take the 1/3 rule with a grain of salt because if we are talking about the same thing it not a constant and changes depending on subject to camera distance .... with for instance a 50mm on full frame the 1/3 starts to move towards 1/1 at around 25ft, getting there at around six feet.

I rarely bother about DoF preferring to select the point I want to be the point of focus and accepting that some things beyond and nearer will be soft.

Yep, theory vs practice. However it was not the rule of thirds that I was referring to but focusing off centre on the flat object. Ultimately you can worry about theory all you like but when focusing I tend to believe my eyes. In another thread someone questioning the need for a depth of field preview button and I think I responded saying that I seldom use it but I would not buy a camera without it.

So jcuknz I think we basically agree.

[Stasch]

John, Wow, what a setup they have! ... Apo-Digital Schneider lenses, Phase One digital back, Apple Cinema monitor ... definitely out of my price range, but nice to know about. And it is all in such a simple setting ...

I had a crash course in the complexities of matching monitor view, my home printer output, and the print shop output a few years back when I worked on a math poster project ... it began to feel like a black art. There was no photography involved (just working with Photoshop); it is only now I am getting my initiation into thinking about the photography aspect. As Manfred and others have suggested, I will probably find out that I have been worrying about all the wrong things .... I suspect that is the price every novice has to pay...

[Stasch]

Ted, that is something that never crossed my mind ... somehow I always thought "bigger is better", and would have gone for a full frame camera if there were cheap ones floating around on the used market. I will certainly keep your suggestion in mind ...
Best,
Stan

[Stasch]

It looks like the 1/3 rule, like so many things in photography, will need a bit of experimenting to determine how to use it.