0
\$\begingroup\$

I have a Nikon D3300 camera. I am doing a project in which I have to project an image of uniform intensity value onto a white screen and take photo of it. This is actually the calibration part. But the problem I am having is that even when I go into manual mode, with fixed iso, shutter speed and aperture, the photo I take give me unpredictable intensity of the screen image. If I put 150 as the greyscale value, I have gotten values from 170 to 210. If I do 0, I get above 37 and below 97. I have captured repeated images for a single greyscale value, but never have I gotten a single output intensity for all the images.

my current settings are ISO:100, F14 and shutter speed 1/40. I have put ec to zero and white balence to AUTO. I am taking RAW image of the .NEF format, then converting them into .ppm by rawkit python module and then taking the intensity values from PIL(another python module. But I don't think that is relevent). I am triggering the camera from the laptop using gphoto, but I don't think it changes these settings that I have put manually in the camera. No matter what the abovementioned settings however, I am not getting the constant output. I have also set the zoom of the lens such that the screen covers the entire image and the background light fluctuations are not great. Also, the focus is on manual and flash is disabled.

Is there some setting that I am missing? Pls help...

Is there some settin

\$\endgroup\$

4 Answers 4

3
\$\begingroup\$

What type of projector are you using? It may be worth trying a much longer exposure time, especially if you're using a DLP projector that uses a rotating filter wheel, or a CRT based projector. Either will have a relatively slow refresh rate, so your 1/40th second exposure may be capturing only a few refresh cycles including a partial one- which may give you inconsistent results.

In the case of DLP projectors, the projector uses a rotating filter wheel with different coloured wedges, and the tiny mirrors on the DLP chip adjust the light intensity at each pixel as the different colours go past as the filter wheel spins. Persistance of vision gives you a full colour result. If your exposure time isn't an exact multiple of the filter wheel rotation time, one or more of the colours will be overexposed relative to the others. This will show up more where the exposure time only covers a few rotations of the filter wheel.

With a CRT based projector, you get similar problems - but this time relative to how many time the projector paints the raster during the exposure. Problems here (with a 3 CRT projector) will usually show up as a different exposure between bands, where parts of the screen are painted n times and others n+1. I'd expect the 3 colour channels are likely to be painted in sync with each other, so this would probably show up as mostly a brightness variation.

Also, if you're not in a blacked out room, the room lighting may affect results. 60Hz electric lights will usually give you a 120Hz flicker, (50Hz/100Hz in the UK) which can give you different base lighting intensity depending on where your exposure falls in the mains voltage cycle. Daylight will also give you a base lighting intensiry, but tends to change fairly slower (barring clouds across the sun).

And one thing I forgot - your projector bulb may have brightness variations for the same reason as the room lighting. With an incandescent filament rather than a fluourescent tube, the flickering's likely to be less pronounced since the filament is going to stay hot, but there may still be a small variation as the voltage/current changes during the AC mains cycle. (from a quick check on the web, opinion seems to be divided on this - some people seem to think the thermal inertia of the filament is enough to prevent flickering, others think there still may be some present - see https://physics.stackexchange.com/questions/13400/what-invisible-flicker-do-different-types-of-light-bulbs-have)

\$\endgroup\$
1
\$\begingroup\$

My D800 had a bad aperture control unit that gave inconsistent exposures at the same settings all the time. If possible, try shooting wide open and see if the problem goes away. If it does, you have a faulty aperture control unit.

\$\endgroup\$
1
  • \$\begingroup\$ It's not just a faulty unit. It is the entire design of the mechanical aperture linkage Nikon uses. It is notorious for being inconsistent from shot-to-shot. That's why most other manufacturers abandoned mechanical aperture control for electronic aperture communication between camera and lens a long time ago. Nikon has finally started to put electronically controlled apertures on some of their most recent high end lenses. \$\endgroup\$
    – Michael C
    Jul 2, 2017 at 15:39
1
\$\begingroup\$

Your basic issue is that you are using lower end consumer grade gear designed to do artistic photography and expecting to get lab grade results from it. That's not going to happen.

Neither aperture position nor shutter times are precisely controlled enough with consumer cameras (even the pro models such as the Nikon D5 or the Canon 1D X Mark II) to get lab grade results. To do what you want you need laboratory grade equipment. Such equipment is very expensive.

The mechanical aperture linkage Nikon has been using since the 1950s is well known to be inconsistent from shot-to-shot. Other manufacturers have been transitioning to electronic communication between the camera and lens regarding control of the aperture over the last three decades. Only relatively recently has Nikon begun to include electronic aperture control in mostly higher end lenses.

Electronically controlled apertures are usually much more consistent than mechanically controlled ones, but even those aren't accurate enough for scientific measurements. The same is true of electronically controlled shutters now found in most digital cameras. They are far more accurate that their spring driven predecessors, but they still aren't lab grade.

The projector you are using could also be introducing a great deal of variability to the equation. Most current projectors are based on DLP technology that uses a rotating filter wheel that creates flicker.

There are a few things you can try to reduce the variability between frames:

  • Shoot with your lens wide open. That removes aperture variability from the equation if the aperture blades are completely outside the optical path when the lens is wide open. This is the case with most, but not all, consumer grade lenses.
  • Use a much longer exposure to reduce the influence of flicker from your projector.
  • In order to maintain an exposure that does not result in total saturation (blown highlights) you'll probably need to use a neutral density filter to reduce the brightness of the target.

For more regarding the acceptable limits of accuracy with photographic cameras, please see this question and its answers and pay particular attention to some of the comments to the answers:
Is there a sane reason why ¹⁄₁₂₅ is not, instead, exactly half of ¹⁄₆₀?

\$\endgroup\$
0
\$\begingroup\$

I remember reading about a stop-motion animation project that resorted to using manual Nikon lenses on Canon bodies--so no aperture linkage at all. The aperture doesn't move between shots. 100% consistency.

But it needs to be an old manual lens that lets you set the aperture on the lens. And it needs to be on a body that doesn't have a way to manipulate the aperture of that particular lens.

\$\endgroup\$
1
  • \$\begingroup\$ You can do the same thing with Canon lenses on Canon bodies. Just set the aperture manually (via the camera), press the DoF preview button, and while holding the DoF preview button press the lens release button and twist the lens a few degrees. \$\endgroup\$
    – Michael C
    Jul 4, 2017 at 4:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.