Photographing item with dual tone (e.g. black and pink pistol)

[Scatterbrained]

Just to show you what I'm talking about, here's my dayglow, military spec safety vest under the gun, along with the histogram:





Notice the spike to the right? It's brighter, but it's nowhere near clipping, and that's after my tone curve has been applied. With the default, flat tone curve it wasn't quite as far out to the right.

 

[Gavjenks]

You're trying really hard here. You do know that going up a stop doubles the amount of light right?

Yes, see previous example

Reflectance in my earlier example
0.5% doubles to [same stuff as at 2%, but inside of a groove or decoration or overlap of metal]
1% doubles to [black velvet would be around here]
2% doubles to [really matte really black coatings like rubberized stuff, or paint used inside lens barrels, etc.]
4% [normal flat black paint]
8%
16% [black matte basic plastic like the coffee machine handle is probably around here at 5 stops]
32% [brushed nickel probably around here at 6 stops maybe 1 higher]
64%
[highly polished chrome highlights around here somewhere at 7.5 stops]
128%
[fluorescent could be around here somewhere at 8.5 stops]
256%

As a matter of fact, I just took my florescent safety vest that I wear when riding my motorcycle at night, and put it under the gun (in place of the tiles), imagine my surprise when the image came out just fine.

Yeah, again... you probably have a much nicer camera than a 50D and you're probably using it optimally at ISO 100. So something that might just go beyond the capabilities of a 50D used at high ISOs (as it very well might be since OP is using continuous lighting and may not know that ISO matters) is just a slam dunk for you, potentially.

For example, the dynamic range of a 50D used at high ISOs can be about 6 stops lower than the capable range of a D800 used at its lowest ISO. So yeah, depending on what you have, you might be able to photograph NASA's best space-age white and black materials at 12 stops apart or something, but that doesn't mean he/she can do the same things you are doing.



That's why I said way earlier "Is there an edge that can't be captured no matter what lighting? No for a D800, probably yes for a 50D." A difference of 8 or 9 stops might cause an older smaller sensor camera to fail no matter what, whereas a new full frame eats it for lunch. Even more extremely, an early generation micro 4/3, or maybe a cell phone camera might UTTERLY FAIL to be able to possibly capture the shoes that you captured quite easily (even though those are probably more like 4 stops)

 

[Scatterbrained]

Once again you're thinking about this the wrong way. If you are supplying the light, DR doesn't matter. Period. You control the range of the scene. That's the whole purpose of using a light meter and shooting in a studio. BTW, that shot was a 5DII, same generation of sensor as the 50D, but it would work just fine with a camera phone. The DR of that scene is only a few stops. The lighting on that scene doesn't vary by more than 3 stops across the whole frame. People have been achieving shots like this since long before the D800 came out. Dynamic range is a luminance issue, not a color or tone issue. Something that is black, when properly exposed, will still be black. Something that is white, when properly exposed, will still be white. Your argument is predicated on the assumption that you're trying to illuminate everything to the same tone, which would simply render a grey image. Have you ever used a handheld lightmeter? A lightmeter takes a reading of the light falling on the scene and tells you what camera settings to use to render the scene correctly. An accurate reading will give you camera setting that will render black as black, white as white, and grey as grey. That's the point. If a subject has a matte surface, that surface will render matte under the light. With highly reflective, polished items you're shooting for the reflection; but again, DR doesn't matter, light placement does.

 

[Scatterbrained]

Okay, just went and shot it with my little M P&S (same sensor tech as the 50D BTW). Shot came out just fine. I also tested the Dynamic Range of the scene with my light meter. 1 stop. I can't imagine there's a camera in the world right now that can't handle that. From f/5.6 +.9 to f/8 +.9.



Dynamic range is about light, which is why it's always discussed in terms of highlights (bright areas receiving a lot of light) and shadows(areas receiving little relative light).

 

[Gavjenks]

Once again you're thinking about this the wrong way. If you are supplying the light, DR doesn't matter. Period. You control the range of the scene. That's the whole purpose of using a light meter and shooting in a studio.

This is simply incorrect. I'm not really sure how else to explain it. I will try once more:

Imagine just a rectangular rod sitting on a table. It is split half and half, one half is a material that reflects 4x as much light (2 stops more) as the other side. There is a sharp boundary in the middle.

You can use multiple lights or whatever to light the two ends of the rod differently, yes, but the boundary right at the very middle where it switches over, it goes from dark to light in just a single pixel's worth of your sensor. Those two pixels will have pretty much the exact same amount of light hitting the amount of the rod that each one represents.
No matter how fancy your lights are, this is going to be true for those two pixels.
No matter how bright your lights are, this is going to be true for those two pixels.

Let's say the number of photons that hits each of those pixels during an exposure is 100 photons (speaking in terms of ones that would hit your camera if not absorbed). The black side pixel will reflect 10 of those photons perhaps, the white side pixel will reflect 40 of them. A 1:4 ratio

This same ratio holds no matter what the hell you do with your lighting setup. Use 15 lights, use 1 light. Use flags, scrims, dots, softboxes, umbrellsa, CFC bulbs, candles, bonfires, bioluminescent bacteria in a box. Whatever. NONE of them are going to have much of a significant difference in luminance from one of those pixels to the other pixel. The only effect you could possibly have would be the amount of feathering of a light that occurs over a single pixel's distance, which is of course negligible.

Also, the same ratio holds no matter what strength lights you use. Expose 200 photons instead? It becomes 20 reflected and 80 reflected (still 1:4 ratio). Expose 5,000,000 photons instead? Becomes 500,000 and 2,000,000 reflected (STILL 1:4 ratio).

So that one edge there in the middle of the bar guarantees that the dynamic range of the scene will be at least a ratio of 1:4, or 2 stops in that example. Completely regardless of any lighting decisions you make at all. It might be higher than that (if, say, the black side is in shadow and the white side is more lit)! And to the extent it is, you could reduce it by changing your lighting setup to more even lighting, say, or even reverse lighting. But all you can ever do by that is reduce it back down to 2 stops minimum, if an edge like that exists in the scene.

Here is a diagram:



Initial light in middle.
Cranking up the general ambient lighting by 2x does nothing to the ratio.
Cranking down the general ambient light by 2x does nothing to the ratio.
Shining a light on just the middle area does nothing to the ratio at the critical edge, and actually increases total scene DR.
Shining a light in a vain attempt to light only the dark side still bleeds over a tiny bit and does nothing to the ratio at the critical edge.

NO MATTER WHAT you do with lighting, it will be guaranteed at least 2 stops DR in the final image due to that edge being 2 stops of reflectance difference. (a perfectly sharp cutoff of light on one side versus the other is not possible, because in real life, it's not a straight single line. It's crazy stuff like shoe patterns, that you have no chance of matching).

And if the difference at the edge were 9 stops reflectance, the scene would be guaranteed to be at least 9 stops DR no matter what.

(the 1.98 stops for example is due to the feathering being enough to slightly change light intensity from one side to the next. This is exagerrated in the diagram though, since it's only a few hundred pixels wide, not 5,000 or so like most DSLRs. In a photo, this effect would thus be cut down by a lot, and it would be closer to just 2, period.)





In fact, I'd go so far as to say it is a general rule that the minimum boundary of the dynamic range of any scene is going to be mostly defined by the number of stops' difference between reflectance values for the highest contrast single surface edge in the scene.

 

[Scatterbrained]

Until you find this mythical "thing" it's just mental masturbation. I've posted examples that have met all you objections and received nothing back but more conjecture. Find this mythical rod and I'll photograph it for you. I've given you a scene with both polished stainless steel, matte black, dayglow, gloss black, etc, and it all fit nicely in the jpeg histogram with room to spare. I have photographed for you things you said couldn't be done. Now it's your turn to put the camera to work.

 

[Gavjenks]

Look, I was simply hoping to have an interesting hypothetical discussion about what is a pretty cool logical consequence of the physics of light and materials. You've made it clear though that either you don't understand what I'm talking about (whether it be a failure of my describing it or otherwise), or you would understand but aren't reading closely, or you're just being belligerent. In particular, listing "matte black" in the same list as stainless steel as if those were equivalent or similarly relevant terms for this discussion...

Regardless, I guess I don't really care anymore, because one way or the other, it's obvious that the discussion isn't going to be had here, so I'll go ask at the next vision researcher reading group meeting at the office instead, where I know people will care about imaging theory for theory's sake.

Thanks for sticking with me up til here.

 

[Scatterbrained]

I know exactly what you're saying. Basically, that something reflective will create a highlight, that highlight will exceed the DR of the camera and therefore can't be photographed. You're assuming that an image with a wide tonal range will require a camera with more DR. What I have demonstrated for you is that it can still be photographed, with any camera, so long as you know how to light it. Sure, if you shine a light into a mirror you're going to get 100% of that light back, that's why you don't shine the light straight into the mirror What you don't seem to be wrapping your mind around is that when something is highly reflective you don't shine your light into it, you choose what will be reflected in it instead.