Black flickering bars

[Willen]

Hi there!

 

I am getting this effect:

 

 

When shooting pics with my phone camera (not many settings there) under the light of my bright 6500K desklamp (that I use for painting). The image on the phone screen shows rolling bars like when filming a TV screen.

 

I am aware of the effect of different scans on screens. I don't know how to avoid it :( Any pointers? Some pictures do not show this effect but I am at a loss on why:

 

[Doug Sundseth]

Looks like a moire pattern to me. They're most commonly caused by an interaction between the grid pattern of photosites on your camera sensor and the grid pattern of a subject (for instance, printed backgrounds will normally have an underlying grid from the dot pattern of the printer.

 

If that's not the cause, I suppose it might be an interaction between the flicker frequency of a fluorescent light and the scan frequency and pattern of the camera. I think that pretty unlikely, though.

[Hibou]

While a cameras sensor syncing up wth a florescent light source is pretty rare. I had it happen with my DSLR a few years ago. It was in a convention hall lit by hundreds of florecent tubes. At first i though it was an banding issue with my camera's sensor, probably caused by too little or too much electrical current. But i managed to reproduce it on both my cell and a friends camera. Ultimately after a bit of research and a call to my camera manufacturer's tech reps i pretty much narrowed it down to the camera syncing up with all the florecent lighting.

 

I would try using a different light and shooting a few test pictures with a different background. Then by process of elimination you can figure out which of the three factors: the phone, the paper back ground, or the light is your issue.

[Willen]

Thanks.

 

Background is plain white paper, actually. Exact some setup on both pictures, except position of the lights. I would believe it has something to do with syncing with the lights, since I found out the effect was accentuated if a third fluorescent light was used straight on the mini.

 

I could not remove it completely thou. And I have no incandescent sources of light that are not warm-orange, if that even exists :(

 

This leads me to believe (the engineer in me) that the auto-white balancing of the phone camera is actually doing something to adjust for the different perceived light temperature or something (I am no expert here, I know nothing about photography but some about EM) that results in an accentuated effect. I didn't know it was called Moire, will look that up.

[Hibou]

All camera talk aside i really like the paint job you did on your mini.

[Willen]

Thanks

 

I am sharing the WIP here: http://www.reapermini.com/forum/index.php?/topic/49228-wip-dancer-girl/

[Wren]

I got an effect like that once taking pictures with an iPad against a certain background (which was just the bottom of a white cardboard box, I was trying to take quick pictures).

[Willen]

All I found so far online is a mention to some arcane feature some cameras have that is called "Antialias Filter". This eliminates the effect, to the cost of blurring the image a bit (it reduces sharpness of focus). However, no further explanation to the actual phenomenon, how can it be diminished by camera settings, nada :(

 

I hate when I do not find things to _understand_. I can learn about anything... if there is a good explanation. I hate it when stuff is like "magic"

[Doug Sundseth]

An anti-aliasing filter won't help much if that's moire. They're very useful for high-frequency moire, but essentially useless for low-frequency moire. Plus your camera almost certainly has one.

 

Moire is caused when you overlay two regular patterns of similar but different frequencies. For example, if you overlay two sheets of window screen and move them slowly about, you'll see patterns of light and dark appear and move about.

 

In the case of cameras, the patterns are caused by the interaction of a pattern on the subject (threads in cloth or dots from a printer, for instance) with the gridded array of individual pixel sensors (photosites) in the camera. In some cases, the dark parts of the pattern will fall onto the photosites while a tenth of a millimeter away, the light parts of the pattern will fall onto the photosites. This can be exacerbated by the interaction of the final pattern in the photo with the grid of dots in a screen, resulting in a moire pattern that changes as you zoom in and out.

 

Since you said you're shooting on white paper, I think it's pretty unlikely that you're seeing moire here. Try shooting with an incandescent bulb or in daylight (which does not flicker on and off like a fluorescent). If the pattern goes away, you'll know that it's bulb flicker that's causing the problem.

[Willen]

Thanks Doug for the great explanation!

 

If it is, indeed, bulb flicker, do you care to propose an explanation about how different spatial configurations can make the effect stronger or diminish?

 

Also, why would bulb frequency produce this effect, simple because of close interaction of bulb frequency with camera scanning frequency? What I am trying to understand is why, being those two things constant (my electric feed should be pretty constant, and I don't believe the camera scanning freq would change, can it?) sometimes I get the effect or not.

[Doug Sundseth]

When you change the distance between the lights and the subject (or add new lights), you change the total light level. You camera reacts to that by changing its shutter speed or aperture. If it's the shutter speed that changes, you could get a difference in scan rate so that it comes close to the flicker rate.

[Willen]

I wish I could like your post more. And send you beer or the beverage of your choice for a virtual toast and a huge thanks for your insight.

 

Much experimenting shall be done this weekend

[Spike]

I've had the same thing happen with my DSLR taking photos at a hockey practice rink (white walls and "ice", plus dozens of overhead flourescents). I thought it was a camera thing too until I messed around with different shutter speeds. It seemed to be most likely to occur at very high speeds (1/1000 sec and higher) and at wider apertures. It also seemed to be more intense on certain surface angles.

For what it's worth, your particular effect looks kind of cool.

[Willen]

As evidenced in my WIP thread, I did not have this issue with a more "pro" camera borrowed from my brother.

 

I am lying. I saw a start of it, and come to a theory:

 

When I put the lamps very very close to the model, the camera display started to show a hint of this moving bars effect. I could not find the setting for shutter speed in the camera (I am a noob!), so I assume that it automatically compensated for the amount of light by increasing the light (with more light, I am assuming it requires less time to capture the image, thus a higher shutter speed).

 

But in my rig, closing up et all, with the lamps about 30cm from the mini, it worked great. I need to learn about white balance and ISO, but the results were satisfactory

[MonkeySloth]

I just got this with my phone's camera when I added a new 100w incandescent light to the mix. My other lights are LED and an small incandescent and my photo background is special plastic\paper that's used as a defuser for lighting (the only background my phone likes to focus on) so I'm getting it for the shot being to bright I guess because if I move the 100w light back a few inches it goes away.