Photographing Paintings

. . . When you come at something like this it is difficult to navigate a clear path through all the information, suggestions and counter opinions. . . .

Yes that takes H A R D W O R K and a disciplined W O R K E T H I C - both it appears you have, Bravo.

Good luck with you venture.

WW

On the subject CFL light sources can you tell from the histograms whether they are comprimised light source,...

No, you can't.

I think it's a pretty safe assumption that virtually all CFLs are problematic, having very spiky spectral output with very weak "fill in" between them.

The main reason, in my view, is that CFLs mainstream came to market as an energy-efficient replacement for household incandescent bulbs (the sort that can burn your fingers). (In the US, where I'm from, the incandescents were largely mandated out of the market.) Now, given that they were introduced on the basis of energy-efficiency, who, in their right mind, would buy a CFL with low efficiency, aka low lumens per Watt? In other words, if you were to manufactur a CFL with, say half the equivalent output, your CFLs would barely sell; how long would you be in business?

Now, setting aside CFLs for a moment, consider the more conventional fluorescent tubes (essentially the same thing, but in a straight line). There have been many varieties of these, including the "eco-friendly" as well as special tubes intended for color work. If you look at the lumens/Watt ratings for these, you'll find that the eco-friendly tubes are rated much higher. As I said before, they are putting out their light energy in spectral zones where the human eye is more sensitive, thus are visually much brighter. So you can actually use this as a rule-of-thumb to evaluate fluorescent tubes (including CFLs) - if the lumens\Watt ratings are high, then they are virtually guaranteed to have poor color reproduction qualities. And it's probably a pretty safe bet that virtually all CFLs will be problematic.

Thankyou Joe for the flash recomendations.

I hope is the last part in the puzzle. When you come at something like this it is difficult to navigate a clear path through all the information, suggestions and counter opinions. I have done a lot of research for this. Two of the best web sites I initially found were.

Photographing Paintings with a DSLR | Marc Dalessio and

How to Photograph Paintings and Prints with Copy Lighting

From these sites I devised my initial test setup shown in my first post, now I am looking at something quite different thanks to you contributions. The suggested lense means i really need to work horizontally for a start.

 

On the subject CFL light sources can you tell from the histograms whether they are comprimised light source, not that I intend to use them now.

 

On the subject of reflections off impasto, you definitly need some as thats what you get when you view the painting in a gallery. Although the question of how pictures are lit for viewing in galleries is a good one, as conservation of the painting is often a major consideration as well.

I just took a look at your first reference website--in my experience a $20 tripod is no bargain, especially if you have to use an elaborate tethering system to prevent camera movement that a $20-30 cable release will eliminate just as well on a good tripod and head. Also, setting a painting on an easel at a slant is guaranteed to make it harder to get the camera parallel to the painting. I have photographed a lot of artwork over the last 30 years, on a wall that I made sure was perfectly vertical. It is 8' x8' x3/4" plywood covered with homosote so that screws can be driven or pushpins used to support work. I always make sure that the work is parallel to the wall, so keeping the camera square with the work is easy. It is also a lot kinder to your back since you won't have to lean over a tripod to see what is in the viewfinder.

in my experience a $20 tripod is no bargain, especially if you have to use an elaborate tethering system to prevent camera movement that a $20-30 cable release will eliminate just as well on a good tripod and head.

- Another advantage of using flash is that any vibration is frozen by the very short exposure - perhaps effectively 1/4000th of a second at the 1/8th 'power' setting.

The instantaneous brightness of flash also overcomes moderate ambient light. So you don't have to work in dim light in order not to 'contaminate' the copy lighting.

You also don't need to worry about the flash exposure damaging or fading the painting. There's an academic assessment of the effect of flash exposure online somewhere, which comes to the conclusion that such fears are groundless, and that museum curators are over-cautious and basically don't know what they're talking about when it comes to light damage from a flash exposure.

It's a lengthy PDF. If I find the link again I'll post it.

The article is here.

I was mistaken about it being a PDF since I've copied and pasted it into an offline document for my own reference.

 

I have been and will continue to shoot raw and use custom WB. I

When you shoot raw, custom WB has no effect on the data captured in the file. It simply appends metadata that your software may read in doing its initial rendering, but whatever is in that rendering is reversible. If you are going to use a neutral card or color checker, it is a waste of time to fuss with custom WB because the when you use the image of the neutral card or color checker, you will be imposing a white balance that will override the custom WB.

- Another advantage of using flash is that any vibration is frozen by the very short exposure - perhaps effectively 1/4000th of a second at the 1/8th 'power' setting.

 

The instantaneous brightness of flash also overcomes moderate ambient light. So you don't have to work in dim light in order not to 'contaminate' the copy lighting.

 

You also don't need to worry about the flash exposure damaging or fading the painting. There's an academic assessment of the effect of flash exposure online somewhere, which comes to the conclusion that such fears are groundless, and that museum curators are over-cautious and basically don't know what they're talking about when it comes to light damage from a flash exposure.

 

It's a lengthy PDF. If I find the link again I'll post it.

I was referring to the possibility of the camera moving out of position as well as the problems of vibration. I agree that flash should obviate camera movement as a cause of unsharpness.

Thankyou all for your guidance this far. I am away for the 'festive' season and will be thinking over my next steps, probably talking to my printer in the new year and then deciding on how to proceed.

It is my intention to follow up and add to this thread about how it goes, even if its just to say it became all too daunting or expensive.

Many thanks

The article is here.

 

I was mistaken about it being a PDF since I've copied and pasted it into an offline document for my own reference.

Thanks for the interesting article.

Yes, I believe that the most important reason is to avoid distraction of other visitors.

There is also the challenge of doing it under available light. I remember some of the pictures my father

had from museums, available light with ASA 64 film. (Yes, ASA days.). And then not so many years

later, when I got to shoot some myself. Trying to be as still as possible, to shoot with the shutter

at 1/15 or even 1/8, and full aperture. (Focus carefully with the rangefinder.).

One that the article didn't mention, is light absorption by pigments, that produces electron-hole

pairs, that go to the surface, where the hole oxidizes something else.

It seems that a big problem with house paint is UV absorption by TiO2, and holes oxidizing the

paint binder. Other pigments absorb visible light, and might be pure enough for this effect.

(But again, as the paper says, the cumulative effect should be small.). Less pure chemicals

will allow recombination before the holes get to the surface.

Unlike photographic dyes, pigments are pretty though.

I do remember a museum that allowed photography with a free permit. I was then stopped

by a guard, when my coat covered up the permit. I suspect it was partly copyright reasons.

I don't remember now if that one allowed electronic flash or not.

I have also known museums that restrict certain galleries for copyright reasons.

The 300 pixel-per-inch printing 'rule' is nothing of the sort, just a guideline.

 

A 24 megapixel camera is perfectly capable of being printed to 36" x 18" with no stitching required. Especially if the subject has no very fine detail, which I suspect is true of most oil paintings.

 

If you must stitch, you need to keep the exposure exactly the same between frames, and that means using manual exposure mode. If you're letting the camera regulate the exposure, there's most of your problem.

I believe 300dpi is usual for the scanned laser optical photo printers. Inkjet printers have to dither their droplets, so even with a higher than 300dpi resolution, you will get less in an actual picture.

I don't know how big the usual scanned laser optical printers go. Maybe only 11x14 or 16x20.

And note that larger prints tend to be viewed from larger distances, so the resolution doesn't need to increase so much.

I believe 300dpi is usual for the scanned laser optical photo printers. Inkjet printers have to dither their droplets, so even with a higher than 300dpi resolution, you will get less in an actual picture.

- A digital camera produces pixels, not dots, and each pixel takes several (at least 4) printer dots to reproduce it. So there's little relationship between the Ppi of a digital camera image, and the quoted Dpi of a printer.

Besides, many printer drivers make an excellent job of effectively upscaling before spattering their resemblance of the original onto paper!

Here's an example I've posted before.

1. Full frame with cropped area outlined:

2. Pixel level view of digital image:

3. Magnified view of final print:

I find the more 'rounded' image in print far more visually acceptable than the small collection of square pixels would suggest

- A digital camera produces pixels, not dots, and each pixel takes several (at least 4) printer dots to reproduce it. So there's little relationship between the Ppi of a digital camera image, and the quoted Dpi of a printer.

Different print technologies work differently.

Scanning lasers can generate a wide range of intensities at each pixel (dot).

Dye-sublimation, which heats up dye over each pixel also should be able to generate the appropriate

shading for each pixel. That is, one dot per pixel at the specified printer resolution.

But I agree, for inkjet it isn't so easy. It is not easy to generate a wide enough variety of

drop sizes.

(As far as I know, dots are commonly used as shorthand for pixels.)

Besides, many printer drivers make an excellent job of effectively upscaling before spattering their resemblance of the original onto paper!

Rescaling is interesting. The math for resampling is well defined, but I am not sure how good different

systems are at doing it.

As far as I know, dots are commonly used as shorthand for pixels.

- That's the trouble. People, and especially advertisers, just aren't technically savvy or careful enough in their use of abbreviations, or of language in general.

It leads to all sorts of confusion, ignorance and mistaken beliefs, which of course the advertising profession loves, promotes and excels in.

- That's the trouble. People, and especially advertisers, just aren't technically savvy or careful enough in their use of abbreviations, or of language in general.

 

It leads to all sorts of confusion, ignorance and mistaken beliefs, which of course the advertising profession loves, promotes and excels in.

I suppose, but I don't find it as bad as some confusing terms.

For one, "lines" as short for "line pairs" or cycles of spatial frequency is confusing by

a factor of two.

I suppose ideally you might want nice square pixels, but most print technologies can't

make a square. Laser beams tend to be round, and probably Gaussian shaped.

Dye-sublimation uses a linear array of little heaters, which probably do come close to ideal

pixels, maybe even square. There is thermal lag, based on how fast you can heat up or

cool down the little resistors, though.

I had some 4x6 prints made at a nearby Walgreens and asked what printer they use.

It seems to be a DNP model DS40 dye sublimation printer, and which does seem

to quote 300dpi. (Yes dpi and not pixels/in).

The newer DS620 and DS820:

http://dnpphoto.com/Portals/0/Resources/DP-DS620_820_UsersManual_Ver.2.00_En.pdf

have both 300dpi and 600dpi modes.

For one, "lines" as short for "line pairs" or cycles of spatial frequency is confusing by

a factor of two.

- Again, the confusion only comes because of incorrect or imprecise use. Line-pairs per millimetre (lppmm) is perfectly understandable. One line-pair = 1 black and 1 white line of equal width side-by-side. It's the incorrect term 'lines per mm' that's confusing, because it's not clear whether a single line is 1/x mm wide, or whether lines are spaced at that pitch.

If people could be bothered to type the extra 'p' and 'm' as in 'lppmm' then it would be obvious what was meant. As opposed to the oft seen lazy 'lpm', which should really be interpreted as a meaningless lines-per-metre. That would be OK for road markings I suppose! :rolleyes:

'X cycles per mm' is perfectly clear. One sinusoidal variation of light-to-dark takes place every 1/x millimetres. No confusion to be had there at all.

"I suppose ideally you might want nice square pixels."

- I don't think so! Not in a print. The obvious half-toning of an office laserjet is bad enough.

 

(snip)

 

"I suppose ideally you might want nice square pixels."

 

- I don't think so! Not in a print. The obvious half-toning of an office laserjet is bad enough.

Well, if they are round, then they are dots.

Everyone knows what dots are, not so many, pixels, and more

important, it is shorter.

In the case of dye-sublimation, I think you can make a square heater, and maybe

not a round one. There will be some diffusion through the plastic sheet, so the spot

won't come out quite square. But if round, there would be white space around it.

On the other hand, scanning laser beams are round and have some overlap between

scan lines.

If a small basket holds four apples and you have five baskets, how many oranges do you have?

That is a simple enough question. I don't see why anyone would be confused. I mean an orange is the same size as an apple so you would have 20 oranges or apples or pieces of fruit. Simple, right?

Changing tags only makes for confusion doing the math.

If an image is 800 pixels wide and you print at 200 pixels per inch, how wide would the print be?

800 pixels / 200 pixels per inch = ?

Doing the math, the pixel tags cancel out, the inch tag remains and you end up with 4 inches. Easy and intuitive.

=======================================================

If an image is 800 pixels wide and you print at 200 dots per inch, how wide would the print be?

Doing the math, nothing cancels out and you wind up with 4 pixels/dots inches.

No wonder people have a hard time doing the math if they are first told to use dots.

The main difference is that a dot is a spot of a single solid colour ink, or of black ink. It's not a pixel.

It needs to gang together with a whole bunch of other dots to resemble a pixel.

The case of dye-sublimation is slightly different, since the blob of sublimated dye can be varied in density, and so can be translated directly from a single pixel to blobs of dye. It still takes 3 or 4 layers of sublimated dye to recreate the pixel colour though.

Hi All

I am having difficulty sourcing a Canon EF 50mm f/2.5 Compact Macro Lens, I failed to get after two good examples on ebay. I was wondering how the Canon EF-S 60mm f/2.8 Macro USM compairs performance wise as there seem to be more available?

Thanks

Hi All

I am having difficulty sourcing a Canon EF 50mm f/2.5 Compact Macro Lens, I failed to get after two good examples on ebay. I was wondering how the Canon EF-S 60mm f/2.8 Macro USM compairs performance wise as there seem to be more available?

Thanks

I have not used the EF-S 60mm f/2.8 USM. It seems to have a good reputation.

60mm is very close to 50mm, so the Subject Distances for any given sized painting will be similar, a bit farther for the 60mm.

Be aware that the EF-S lens will only mount on EF-S mount Cameras - that is all EOS digital APS-C Format Cameras, from the 20D and 300D models to the current models, that will be of no concern to you if you intend to stick with APS-C Format.

Sites like this provide a crude A/B Comparison, you can flip between the target shots: [LINK]

WW

Thank you for that information William. I looks like they are very similar beasts. The 60mm lense being a newer offering though.

Paul

... yes, the EF-S 60mm macro is a newer lens.

I think that the EF-S 60 was made so Canon had an EF-S offering, to use on APS-C Format cameras to 'equate to' their existing EF 100mm Macro used on 'full frame' cameras.

You might also have a look at the recently released EF-S 35mm f/2.8 IS STM Macro Lens - on your an APS-C Cameras, that will give you a shorter Subject Distance. I haven't used this lens but, on paper and in lab tests, it appears to be very good.

WW

Hi All

Update, I have bought a Eos 750D and am awaiting the arrival of Canon EF-S 60mm f/2.8 Macro USM. I am currently trying to buy a couple of

YongNuo YN-560 III flash units and a Yongnuo RF-602 or 603 Flash Trigger. My problem is selcting the correct model of trigger that is compatible with the 750D. Anyone know which one I need? this info would be greatfully received. I am try to buy on amazon in the uk

Thanks

Like others who have commented, use only flash. If you have a slip space in your barn doors, use a UV filter over your filters...even if your bulbs say “UV protected”. This filter will essentially eliminate a spike of purple inherent in the spectrum of 5500 lights. Which to our eyes add up to a faint blue. Most of my clients really like polarizers set up on both the lights and the camera. It is the “ Velvia “ of richness, if you remember film. Polarizers on lights eliminate the ambient bouncing of photons. But what good are they unless you have also a polarizer on your lens? Ok, after all this filtration, there will be a little bit of green cast. Not much, but enough to require a either a change in post or to add a smidgen of magenta to your white space.

If you can get your lights at more of an acute angle to eliminate speculation highlights and not use polarizers, this is fine. So there are a few specular highlights... so what? It’s a painting ...There should be some highlights!

And if there is any relief(depth) to your painting...forget photographing via sidelights..light from above! And yes, use a soft box (baffled to increase directionality)

There are a number of variables still at play: mediums of art and their surfaces, whether you’re intending to reprint art or just post digitally. Oils tend to be contrasty and rich than acrylics. Using polarizers(PLs) tend to darken dark oils. PLs on my own work of oil colored fiber photos do not work at all! I avoid PLs altogether and use ambient modeling light. Go figure!

Getting back to the specific painting you have mentioned of oil with a gold frame. Cast 90% of your light with PLs for the painting at 45 degree angle, BUT use a unfiltered light behind your camera square onto painting. The reasoning here is that PLs dumb down any reflective surface like the frame(turns gold to brown). The one light behind camera reignites the gold frame without affecting the painting.