How to get bright white backgrounds?

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

iconoclastica
Posts: 486
Joined: Sat Jun 25, 2016 12:34 pm
Location: Wageningen, Gelderland

Re: How to get bright white backgrounds?

Post by iconoclastica »

I have been shooting plants against white backgrounds for a few years intensively. In my experience it is far easier to get good results if the background is lighted and not a light itself. Use independent light sources for foreground and bg, one a flash unit, the other continuous light. I prefer continuous light for the fg, so I can see what I am doing. The flash output is balanced so, that the peak of the bg light hits the right side of the histogram. It may start disappearing, but should not completely fall over the right edge. When the bg is overexposed the light 'creeps' over the edge of the object.

Distance between object and bg is another invaluable help. Even for macro I use a large white panel, some two meters away.

In the raw file, it never gets whiter than 254 (Canon). In post I select the bg with as few mouse clicks as possible and create a masked 255-white layer to avoid a grey shade in print. Sometimes I only need a single mouse click, but only when the object has hard edges. You may be more succesful for that.

Sometimes I made a double stack, one without fg lighting, to get a better defined contour. This wat especially necessary when white hairs overlapped the white bg. For solid objects I don't see the point.

My method has not been tested with metal objects though. In an earlier life phase I have photographed metal too. To get that right the key was to realize that I was looking at a complex mirror. If you can figure out where the reflections come from, then you can mangae what exactly will be reflected where. A classic example is the wine glass with wel-defined, black contour in front of a white bg. In your case I think your bg shoud be black or grey everywhere where your camera can't see it, but the reflecting edges can. If you're selecting and redefining the bg in photoshop you'll only need a narrow ring of white bg around your object.

A more informed way of tracing the rays that enter your camera is by reversing the light. Dim the light in your studio and shine a bright light through the view finder. The look aound to see where the reflections go.
--- felix filicis ---

soldevilla
Posts: 684
Joined: Thu Dec 16, 2010 2:49 pm
Location: Barcelona, more or less

Re: How to get bright white backgrounds?

Post by soldevilla »

I have been photographing archaeological pieces for the last few years that need to be removed from the background and made transparent (or not...) My approach is different:

I raise the piece to blur the background, and the background is always a piece of black velvet. This makes it very easy to select and remove the background with a threshold, but it obscures the edges of the piece, so I lay out pieces of white paper on top of the background so that in the image I see a narrow strip of velvet around the piece. , but there is white paper that sends light to the edges of the piece.

Then in the processing it is easy to convert everything to black or select it and have a transparent png
Attachments
IMG_7685.jpg

kaleun96
Posts: 270
Joined: Sat Oct 20, 2018 3:47 pm
Location: Stockholm, Sweden

Re: How to get bright white backgrounds?

Post by kaleun96 »

iconoclastica wrote:
Tue Nov 29, 2022 2:58 pm
I have been shooting plants against white backgrounds for a few years intensively. In my experience it is far easier to get good results if the background is lighted and not a light itself. Use independent light sources for foreground and bg, one a flash unit, the other continuous light. I prefer continuous light for the fg, so I can see what I am doing. The flash output is balanced so, that the peak of the bg light hits the right side of the histogram. It may start disappearing, but should not completely fall over the right edge. When the bg is overexposed the light 'creeps' over the edge of the object.

Distance between object and bg is another invaluable help. Even for macro I use a large white panel, some two meters away.

In the raw file, it never gets whiter than 254 (Canon). In post I select the bg with as few mouse clicks as possible and create a masked 255-white layer to avoid a grey shade in print. Sometimes I only need a single mouse click, but only when the object has hard edges. You may be more succesful for that.

Sometimes I made a double stack, one without fg lighting, to get a better defined contour. This wat especially necessary when white hairs overlapped the white bg. For solid objects I don't see the point.

My method has not been tested with metal objects though. In an earlier life phase I have photographed metal too. To get that right the key was to realize that I was looking at a complex mirror. If you can figure out where the reflections come from, then you can mangae what exactly will be reflected where. A classic example is the wine glass with wel-defined, black contour in front of a white bg. In your case I think your bg shoud be black or grey everywhere where your camera can't see it, but the reflecting edges can. If you're selecting and redefining the bg in photoshop you'll only need a narrow ring of white bg around your object.

A more informed way of tracing the rays that enter your camera is by reversing the light. Dim the light in your studio and shine a bright light through the view finder. The look aound to see where the reflections go.
Thanks for the comments. I guess in my case, with the limited working distance I have, illuminating the background rather than it being a light itself is a bit trickier due to the diffused light hitting the edge of the coin. I will try giving the coin another 10-20mm of height from the background to see if that makes much difference though.

I like what you say about realising you're looking at a "complex mirror". While the problem I've highlighted in this thread can seem kind of insignificant, it's made me go down a rabbit hole of trying to better understand light, different types of reflections, and how to control those reflections. At least to a degree I hadn't considered before in my earlier macro work.

Agree with what you and others have said about only needing a narrow ring of white background around the object - the difficulty is that the size of this ring changes from coin to coin and needs to be quite close to the actual diameter of the coin to not result in any reflections. So the trick is then how to make this white background variable in size, rather than cutting out dozens of black felt "masks" of various diameters and shapes. I can't think of any easy solution to that so I think the best approach is to aim for parallel light rays either reflecting off the background itself or being emitted from it (as a light source).
- Cam

iconoclastica
Posts: 486
Joined: Sat Jun 25, 2016 12:34 pm
Location: Wageningen, Gelderland

Re: How to get bright white backgrounds?

Post by iconoclastica »

You will need but a single gobo ring if you make it so that you can reposition it:
shade.png
--- felix filicis ---

kaleun96
Posts: 270
Joined: Sat Oct 20, 2018 3:47 pm
Location: Stockholm, Sweden

Re: How to get bright white backgrounds?

Post by kaleun96 »

iconoclastica wrote:
Wed Nov 30, 2022 9:13 am
You will need but a single gobo ring if you make it so that you can reposition it:
shade.png
Interesting idea, though I mentioned a bit earlier that I had tried a variable "mask" for the background using an 3D printed iris, the issue that prevented me from continuing with that is that the coins I photograph are not always round. Some are quite oblong, like the one I posted at the beginning of the thread. Perhaps combined with more distance between the coin and background a circular mask could still work but in my tests with a separation of about 30mm, a circular-shaped mask didn't work well enough. I think it's also a bit more difficult with a vertical setup as I have a limited working distance and also still need some form of mount to hold diffusers around the coin as well as something to catch the coin if it falls off the pedestal.

I'll probably still need to employ some form of variable mask but more for doing a rough "masking" of the background and then rely on parallel light / reflections as the means of preventing any remaining stray light that may reflect off the edges. For example, a Fresnel lens with an LED at the bottom of the setup and above that an iris to limit how much of the fresnel lens is in the frame to only what is necessary. So most of the light is stopped from reaching the coin by the iris, limiting the angles at which stray light can hit the coin's edge.
- Cam

ray_parkhurst
Posts: 3417
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: How to get bright white backgrounds?

Post by ray_parkhurst »

The usual method for doing white backgrounds with coins is to use a white backlit panel. I understand that you tried this but it resulted in undesired reflection off the coin edge, presumably due to placing the coin directly onto the illuminated surface. You could eliminate this problem by setting the coin on a pedestal to reduce the edge reflection. You could even vary the height of the pedestal to vary the amount of reflection if you wanted to have a finite amount of edge illumination.

Or was there some additional reason the backlit panel didn't work?

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: How to get bright white backgrounds?

Post by rjlittlefield »

I have this crazy vision of using an imaging display for the backlight -- take one photo with full backlight, threshold to get an approximate mask, dilate to produce a blob that is slightly larger than the coin, then invert and display that blob as a backlight image that would automatically be bright in just a narrow strip around the coin. I expect this would give a nice result, but it could be an unrealistically large amount of development unless you need to do this a zillion times. I still prefer the ideas based on collimated light.

By the way, for your application the backlight should not be perfectly collimated in the sense of parallel rays. If the background illumination rays are parallel, then most of them will not make pass through the lens aperture so as to be visible in the image. Instead, the lens would see the backlight only in the center of the image, an area the same size and shape as the lens entrance pupil. What you really need is more like the condenser system in a slide projector, in which illumination rays from all across the field converge to an approximate focus that is small and positioned to pass through the lens pupil.

--Rik

kaleun96
Posts: 270
Joined: Sat Oct 20, 2018 3:47 pm
Location: Stockholm, Sweden

Re: How to get bright white backgrounds?

Post by kaleun96 »

ray_parkhurst wrote:
Wed Nov 30, 2022 10:50 am
The usual method for doing white backgrounds with coins is to use a white backlit panel. I understand that you tried this but it resulted in undesired reflection off the coin edge, presumably due to placing the coin directly onto the illuminated surface. You could eliminate this problem by setting the coin on a pedestal to reduce the edge reflection. You could even vary the height of the pedestal to vary the amount of reflection if you wanted to have a finite amount of edge illumination.

Or was there some additional reason the backlit panel didn't work?
Hey Ray, yeah I always use a pedestal that's about 20-30mm off the background, depending on the coin. I could go a bit higher but not a huge amount without risking the coin falling and bouncing off a surface onto metal plates and rails. But even with a pedestal I still seemed to get stray reflections when using a backlight, though by this I mean a 3W LED COB or similar, about 30-40mm in diameter. I had tried a number of setups with that light source, such as diffusion gel, 3D printed domes or concave shapes (with thin walls), 3D printed flat panels, diffusion modifiers from my flashes or small LED panel (not bright enough on its own as a backlight), etc. It's been a few months since I've done that testing so may have forgotten other things I've tried. The only thing that really prevented it entirely was using a shroud around the coin, which should be effectively the same as masking off everything of the backlight except that which is necessary to produce a thin outline around the coin.

I should add that this is, relatively speaking, a miniscule amount of stray light causing problems on the edges. Often small enough that I don't notice it from my 7" monitor, and even zooming-in on the edges from a still photo doesn't always show it unless I'm focused on the very edge, so it can be a bit of a pain as I may not even notice the issue until after the stacking process. I've attached a photo below that might better show the amount of reflection I'm sometimes getting. If I remember correctly, this is using a backlight but with some diffusion. Not ideal, I know, but at this point I had tried a direct backlight and it wasn't working any better so was then experimenting with various things.
ex5.jpg
rjlittlefield wrote:
Wed Nov 30, 2022 11:22 am
I have this crazy vision of using an imaging display for the backlight -- take one photo with full backlight, threshold to get an approximate mask, dilate to produce a blob that is slightly larger than the coin, then invert and display that blob as a backlight image that would automatically be bright in just a narrow strip around the coin. I expect this would give a nice result, but it could be an unrealistically large amount of development unless you need to do this a zillion times. I still prefer the ideas based on collimated light.

By the way, for your application the backlight should not be perfectly collimated in the sense of parallel rays. If the background illumination rays are parallel, then most of them will not make pass through the lens aperture so as to be visible in the image. Instead, the lens would see the backlight only in the center of the image, an area the same size and shape as the lens entrance pupil. What you really need is more like the condenser system in a slide projector, in which illumination rays from all across the field converge to an approximate focus that is small and positioned to pass through the lens pupil.

--Rik
That would definitely be a fun project to try but would a display be bright enough and not be washed out by the key light from the ring flash?

Glad you mentioned the thing about not wanting perfectly collimated rays. I had a suspicion with the retroreflector idea that in a perfect world it may fail for this exact reason, which is why I tried 3D printing my own, figuring the imperfections in the printing and material may improve it slightly. On this topic, in terms of avoiding haze/a soft image from the backlight, is it only necessary to prevent the light from the background from being diffused off of surfaces close to the lens? In other words, direct light into the lens should be fine - the worse that will happen is overblown highlights in the area the light is coming from (or potentially lens flare if the light is coming from an angle)?
- Cam

ray_parkhurst
Posts: 3417
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: How to get bright white backgrounds?

Post by ray_parkhurst »

The backlit panels that most folks use are more like what Rik suggested, but the reflections will still happen unless you can get the coin high enough above the panel. To help I'd suggest masking an area just larger than your field of view at minimum magnification, but you'll still likely need to go higher than 30mm to kill the reflections.

iconoclastica
Posts: 486
Joined: Sat Jun 25, 2016 12:34 pm
Location: Wageningen, Gelderland

Re: How to get bright white backgrounds?

Post by iconoclastica »

I wouldn't call it stray light. You're probably looking at the walls, ceiling, or at the extent of the too large backdrop. This is what happens:
scan_30112022.jpg
If you only move a black piece of cardboard from the side (behind the coin) towards the centre, then at some moment you will notice that the reflections disappear. Or ask someone to do it for you once. From there make it practical.

BTW, that's a very interesting coin - Viking money?
--- felix filicis ---

kaleun96
Posts: 270
Joined: Sat Oct 20, 2018 3:47 pm
Location: Stockholm, Sweden

Re: How to get bright white backgrounds?

Post by kaleun96 »

I'll see if I can do some testing this weekend and try rule out some other factors. I think if I can leverage the hole inside the rotation stage and place an LED at the bottom of it, I can effectively gain more height between the background and subject, though not as straight-forward as that if I then have something between the light source for the background and the coin (e.g. a lens).
iconoclastica wrote:
Wed Nov 30, 2022 2:49 pm
BTW, that's a very interesting coin - Viking money?
Thanks! It's a bit older than that, actually from around the time of Alexander the Great. Minted in Tarsos (modern day Turkey) between 361-334 BC, under the Persian satrap Mazaios. The coinage from this time from this city can be particularly interesting, in both the iconography and the history of what was happening at the time (Satrapal revolt, Persian dominance waning, Alexander conquering Asia Minor, etc).
- Cam

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: How to get bright white backgrounds?

Post by rjlittlefield »

kaleun96 wrote:
Wed Nov 30, 2022 11:54 am
would a display be bright enough and not be washed out by the key light from the ring flash?
This issue is common whenever there's a mix of continuous and flash illumination. One solution is to adjust the exposure time so the amount of light from the continuous source, summed over the whole exposure, is properly balanced with the amount of light from the flash during its single pulse. If the continuous contribution is not bright enough, then use a longer exposure time.

... in terms of avoiding haze/a soft image from the backlight, is it only necessary to prevent the light from the background from being diffused off of surfaces close to the lens? In other words, direct light into the lens should be fine - the worse that will happen is overblown highlights in the area the light is coming from (or potentially lens flare if the light is coming from an angle)?
Light that enters the lens from anywhere in the subject field should be fine. Most of that will either go through the aperture to form the image you want, or will be harmlessly blocked at the aperture, or will be harmlessly reflected back into the environment. A very small portion will bounce around inside the lens to create veiling glare, but modern coatings are pretty good at preventing that.

The stray light that you need to worry about comes from outside the subject area. Some of that will get through the aperture and then bounce off the edges of lens element or the sides of the lens barrel, mounting adapters, extension tubes and so on, all of which will happily create visible haze if the reflections are bright enough.

--Rik

ray_parkhurst
Posts: 3417
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: How to get bright white backgrounds?

Post by ray_parkhurst »

rjlittlefield wrote:
Wed Nov 30, 2022 4:20 pm
kaleun96 wrote:
Wed Nov 30, 2022 11:54 am
would a display be bright enough and not be washed out by the key light from the ring flash?
This issue is common whenever there's a mix of continuous and flash illumination. One solution is to adjust the exposure time so the amount of light from the continuous source, summed over the whole exposure, is properly balanced with the amount of light from the flash during its single pulse. If the continuous contribution is not bright enough, then use a longer exposure time.
There is an excellent and surprisingly high-end You Tube video that shows how to do this calibration. Unfortunately I looked but could not find it. I'll look again but you might try searching as well for axial lighting in coin photography with backlight. I'm terrible at finding such things so I'll bet you can find it before I can...

J_Rogers
Posts: 87
Joined: Mon Aug 09, 2021 2:29 pm
Location: North Carolina, USA

Re: How to get bright white backgrounds?

Post by J_Rogers »

ray_parkhurst wrote:
Wed Nov 30, 2022 5:16 pm
There is an excellent and surprisingly high-end You Tube video that shows how to do this calibration. Unfortunately I looked but could not find it. I'll look again but you might try searching as well for axial lighting in coin photography with backlight. I'm terrible at finding such things so I'll bet you can find it before I can...
Is this it?
https://www.youtube.com/watch?v=jvPiMaLy2RE

ray_parkhurst
Posts: 3417
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: How to get bright white backgrounds?

Post by ray_parkhurst »

J_Rogers wrote:
Fri Dec 02, 2022 9:50 am
ray_parkhurst wrote:
Wed Nov 30, 2022 5:16 pm
There is an excellent and surprisingly high-end You Tube video that shows how to do this calibration. Unfortunately I looked but could not find it. I'll look again but you might try searching as well for axial lighting in coin photography with backlight. I'm terrible at finding such things so I'll bet you can find it before I can...
Is this it?
https://www.youtube.com/watch?v=jvPiMaLy2RE
Yes indeed, thanks for finding it!

Post Reply Previous topicNext topic