Reflection, Refraction, Specular Light and Shiny Subjects

Images taken in a controlled environment or with a posed subject. All subject types.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Reflection, Refraction, Specular Light and Shiny Subjects

Post by steveminchington »

Following on from a previous post which got into some interesting discussion about strange things happening with light when photographing reflective subjects, like the aperture being decreased and DOF increased with subsequent degrading of resolution.

This got me thinking about the effect of having a chrome plated objective stuck between the subject and the light source. If we have a large highly reflective object in close proximity to the subject then some of that light must get reflected onto the subject. And if the subject is also reflective then that light will get reflected back again.

So I made a cover for my Nikon CFN Plan 10x 0.3, from thick black paper complete with a nose cone with a 7mm aperture (which is the size of the front element). I did a comparison test with the objective uncovered then covered, and it seems there is a significant amount of light bouncing back and forth between the subject and the objective casing. So much so that when the objective was covered, the drop in reflected light caused a 1 stop reduction in exposure. That's 1 stop of unwanted light causing flare and some of these strange artifacts we have seen.

This first image is of a five pence coin at 10x with "bare" objective and a "soft" LED ring light. Working distance = 9mm.

Image

Second image is the same set up and exposure but with the black cover on the objective.

Image

Third image is a crop from the centre of No1 at 100%

Image

And the final image is a crop from the centre of No2 at 100%

Image

I don't know if anyone has tried this before, but from now on my Nikon CFN Plan is going to be wearing his little black jacket!

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: Reflection, Refraction, Specular Light and Shiny Subject

Post by rjlittlefield »

Good experiment -- many thanks for posting.
That's 1 stop of unwanted light causing flare and some of these strange artifacts we have seen.
These aspects need some more investigation.

Certainly a mask on the end of the objective will prevent the misleading "iris" appearance that you get when photographing a shiny convex surface such as a spider's eye, HERE.

But the flare issue is much less clear. Unwanted light causes flare when it enters the lens from outside the subject area and then somehow bounces around so as to hit the sensor anyway. I don't see how masking the end of the objective will affect that.

Likewise for any other artifacts. Masking the end of the objective will reduce the amount of light that strikes the subject from angles near the axis. Essentially the mask is making the illumination more oblique. It's unclear whether this change will make other artifacts better, worse, or just different.

--Rik

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Re: Reflection, Refraction, Specular Light and Shiny Subject

Post by steveminchington »

rjlittlefield wrote:Essentially the mask is making the illumination more oblique.
I'm just including a picture here so you can see what I have done. The cover doesn't affect the angle of illumination at all. The cone is actually glued to the end of the objective with some light glue. I don't know the technicalities of what is happening with the light, I'm just showing what I have seen, and the only difference between the two shots is that the first is with a chrome objective and the second it is black. The first image has a lot of flare around the design and in the second one this is greatly reduced. Also it looks to me like there is more blurring in the first shot in the OOF areas.

Image

ChrisR
Site Admin
Posts: 8662
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I've made similar things, out of black paper and aluminium tape, but more as a lens hood.
Some objectives - I can't remember which ones (I've played with a few, without keeping records... there are pictures here somewhere) aren't immune from out-of-view lighting causing flare. You'd probably get a fair bit of out-of-view reflected light from a coin.

I can picture several influences at work here; at the very least you have more light, coming from more directions when not using the snoot/mask/hood. I also can't particularly see what I'd be quick to call Flare. If your near-axis extra illumination is filling the shadows, then your contrast would go down. It would with flare too, but that doesn't mean it's the same.
At the least, I'd like to see a stack with everything in focus. Most of what I notice to be different, is out of focus.
Coins and other shiny things are something of a special case. Reflections are particularly important so in the same way that the ideal studio has black walls, avoiding random reflections from the equipment is good for control, so yes, I can see it as useful inthat context.

NB, the plan apo 2x and 4x, plus any of the BD series with the outer parts removed, is black, some are skinny too. I found one in my desk drawer today which I forgot I had - it'll do for illustration. The non ELWD ones are smaller diameter glass of course.
Image

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: Reflection, Refraction, Specular Light and Shiny Subject

Post by rjlittlefield »

steveminchington wrote:
rjlittlefield wrote:Essentially the mask is making the illumination more oblique.
I'm just including a picture here so you can see what I have done. The cover doesn't affect the angle of illumination at all.
Let me illustrate what I meant.

Here is a picture of my CFN 10X NA 0.30 focused on a penny. I've annotated the picture to show three "families of angles", to use the notation of the Light Science & Magic book.

Image

The gray section labeled "Excluded by aperture" is the region where no significant light can come from, because that's the glass that leads to the black hole of the camera.

The green section labeled "Always allowed" is the region where illumination can reach the subject without bouncing off the objective.

Neither of those is altered by your black paper.

What's affected by the black paper is the region I've labeled "Excluded by mask". That's the area from which light can reach the subject only by being reflected off the shiny end of the objective.

You've told us that adding the black paper reduces the exposure by 1 stop. That means about half the light is coming by reflection from the red region.

I hope it's obvious that the green region alone is narrower and farther off axis than the green+red regions combined, and that's what I mean when I say that the mask "is making the illumination more oblique".
ChrisR wrote:Most of what I notice to be different, is out of focus.
Not only that, but they're differently out of focus. Check out what's in focus versus not, in the crops.

--Rik

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Post by steveminchington »

Rik - Thanks for the illustration and explanation of what you meant. I understand what you were getting at now. Light, Science and Magic is a good book although it's a while since I read it now.

Yes, the focus is slightly different - I couldn't avoid this as I had to slide the subject stage away to get the cover on the objective, and then slide it back and re-focus. Which ever way I had done this, I would never have got the focus exactly where it was before.

I agree with ChrisR that the only way to get a proper comparison would be to run off a couple of stacks. I'll have a go when I get a moment.

ChrisR - Thanks for the pic of your objective - I wish I could find things like that in my drawer that I had forgotten about! I wonder if the CFN plan comes apart like that.

My guess is that this extra illumination from the red zone is not a problem in most cases since many people get excellent results using a shiny objective. But in the odd instances when this light is a problem (as in the shot I am trying to do at the moment of a hard drive read write head which has shiny gold wires) we need to mask it.

ChrisR
Site Admin
Posts: 8662
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I wonder if the CFN plan comes apart like that
It doesn't, quite! You can see on the "BD" the annulus for the incident light, so there's no equivalent cover on the CFN. You may be able to unscrew the outermost shell, as the one far right, which has the numbers on. Some are very tight. It won't gain you much, though may make it easier to fiddle with, while adding your black paper.

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Post by steveminchington »

Ok so I have tried to be as accurate as possible with my little experiment now and have run two stacks on this coin back to back under exactly the same conditions, except the first stack is with the bare (shiny) objective and the second stack is with a black 'mask' on the objective.

With the unmasked objective, light reflecting off the end of the objective in the 'Red zone' is increasing the flare around the letters on the coin. This is causing some of the detail adjacent to the flare to appear out of focus. But if we look at the crop from the masked objective, the flare is reduced and the detail is resolved.

Well that's my uneducated guess at what is happening - of course, I would welcome the views of the experts.

One would probably not try to photograph a coin under these conditions but I am just trying to see if masking the objective makes a difference. On a more fitting example e.g. the hard drive read / write head image I posted, the difference was not so pronounced, but the unmasked objective did cause some undesirable highlights and tended to flatten the image by filling in some of the shadow detail.

As before, image 1 no mask on objective

Image

Image 2 mask on objective

Image

Crop at 100% without mask

Image

Crop at 100% with mask

Image

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Steve, thank you for the additional information.

I've pulled your two images into Photoshop, overlaid and aligned them, and made a little animation that shows several aspects that I think are important.

The animation is 2 seconds each, of 2 images (one marked up with colored arrows).

Image

The aspects I think are important are marked with different colored arrows.

Green arrow marks a long bright line that appears without the paper cover. This is easy to understand as a reflection of a reflection: a long edge of the coin reflecting the light source that is reflected from the objective.

Blue arrow marks a similar effect at much smaller scale, a small bit of the coin catching some light that it does not when the objective is covered.

Red marks some catastrophic loss of sharpness that I find very hard to understand. It almost looks like the extra light is causing the stacking software to lose its way and show us unfocused frames.

How are you doing the stacking? Can you show us crops from a couple of correspondingly focused frames showing the bright blobs indicated by the red arrows?

--Rik

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Post by steveminchington »

rjlittlefield wrote: Can you show us crops from a couple of correspondingly focused frames showing the bright blobs indicated by the red arrows? --Rik
Rik - I have checked all of the stack images and the blobs you have marked with the red arrows are not quite in focus in the stack with no mask. That frame is the last in the stack so I should have gone past that point. It's really hard to see such fine detail through the viewfinder. Also I was using 10 micron increments which I have discovered is not really fine enough to catch this detail. Not only that, I had left the vellum paper ring in the ringlight I was using as a diffuser for something else.
Apologies for being such a rubbish research student, but I can safely say that test is invalid. I will do it again at 5 micron steps which I can just about do on my rig, WITHOUT the diffuser paper, and post again.

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Post by steveminchington »

Ok, so I hope I have got it right this time...
I have run two more stacks of 23 frames at 5 micron steps, starting and ending at the same place. I did this by checking all the frames side by side and matching the focus and pulling both sets from a bigger stack which I over-ran each end.

So unmasked first

Image

Now with mask

Image

Crop at 100% from same area as before - unmasked

Image

Crop at 100% with mask

Image

Perhaps you can inspect these now Rik and see what you think.

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

This last test looks like a good comparison.

Most of what I'm seeing is local expansion of the bright reflections, which makes sense in terms of more near-axial light with the uncovered objective. There are also some dark areas especially left of center that pick up a bit of brightness. I'm guessing that those are slightly sloped regions that are either picking up the near-axial illumination directly, or possibly picking it up indirectly as a reflection from that big ridge that runs through the lower right corner.

Here's an animation of almost the whole frame, reduced to 50% to make the gif small enough.

Image

--Rik

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Post by steveminchington »

Thanks for taking the time to look at this Rik. I must learn how to do GIF animations as they are a great way to compare images - much easier than looking from one to the other.

So nothing disastrous happening - no loss of detail as I first thought, just a difference in illumination. Depending on the subject, this could either be of benefit or not as the case may be, but something to be aware of when lighting the subject.

So far for me, the extra illumination hasn't been beneficial. The patch left of centre seems to be picking up light from the right which as you say may be a small hump being illuminated by the bright edge.

Thanks again - I'm glad that one is sorted.

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

steveminchington wrote:I must learn how to do GIF animations as they are a great way to compare images - much easier than looking from one to the other.
In the current Photoshop CC, the recipe goes like this:
1. Create a layered image containing one layer for each frame you want to show.
2. If necessary, select all layers and do Edit > Auto-align Layers.
3. Open the Timeline window and select Create Frame Animation. (In earlier versions of Photoshop, I believe the Timeline window was named Animation.)
4. In the upper right corner of the Timeline window, click on the little pulldown menu box and select Make Frames From Layers.
5. Select all frames in the animation, then hover over the "0 sec." label, right-click, and select the delay time (number of seconds that each image is to be shown). I generally prefer a slow animation, say 2 seconds per frame, so that the viewer has enough time to do a little thinking while each frame is shown.
6. In the toolbar at bottom of the Timeline window, change Once to be Forever.
7. Test the animation by pressing the play button (right-pointing triangle, tooltip "Plays animation").
8. Adjust timing and frame contents as necessary.
9. In the upper right corner of the Timeline window, click on the little pulldown menu box and select Optimize Animation. Be sure that both boxes are selected and click OK.
10. In the main Photoshop menus, select File > Save for Web. Select GIF and be sure the Looping Options are set to Forever, then press "Save..." and place the file.

Note that with GIF there is nothing like the JPEG compression quality to adjust file size. To fit within 300 KB (or whatever) it is necessary to resize the image. In the current Photoshop this can be done "on the fly" as part of the Save for Web dialog.

I hope this helps!

--Rik

steveminchington
Posts: 215
Joined: Sun Jul 22, 2012 2:30 pm
Location: Bedford UK

Post by steveminchington »

Thanks very much for the instructions Rik... unfortunately I don't use Photoshop, I use GIMP. I found a tutorial for GIF animations for GIMP which is similar to what you described using layers, but obvoiously there are differences to PS. I have been using Linux, Mac and Windows, but since I have retired, I have gone over to using Linux exclusively, since I can no longer afford to buy commercial software. Actually that was the deciding factor to use Zerene, not only because it works extremely well, but because you have a Linux version. So I am hoping that you will continue to support Linux. There are so many high quality open source applications for Linux now that I see no reason to use Windows or Mac and to be honest, I don't miss the virus and malware hassles at all!
Perhaps you could post your instructions on the FAQ page for PS users - I'm sure they will find it useful.
--Steve

Post Reply Previous topicNext topic