I found another post asking why the horizontal bands appeared. I'd like to know how to spec. the wright LED strip lighting for my custom built-in soft boxes to avoid this?

A little more: I'm building custom diffused light boxes into the walls and ceiling of a table top height photo booth for fast and consistent product photography. Though I am not well versed in the reasoning, I am well aware of the nasty horizontal stripe phenomenon that is ruining our images.

Can you help me to specify what to avoid when specifying LED strip lighting for my project? Also, is LED strip light even capable of what I'm attempting to do? I Don't have a lot of space in the wall to recess the spiral fluorescent lamps.

Stats: Camera - Usually using my iPhone7 Lighting - I am in testing mode and my first attempt proved to be an eppic fail. :(

Please help! test light image test light image 2

~Much appreciated Angie

  • \$\begingroup\$ Also consider USING the gradients to your advantage :) \$\endgroup\$ Commented Apr 6, 2019 at 17:58
  • \$\begingroup\$ photographylife.com/light-frequency-issue has some nice background on light frequency / pwm issues, also suggests shutter speeds that are one full or one half cycle of the light (though doesn't mention that some lights choose their own pwm instead of following the 50 or 60 Hz mains frequency). Some cameras, e.g. cpn.canon-europe.com/nl/content/education/tipsandtricks/5152.do have "Anti-flicker shoot" which will delay the shutter to match the max light frequency \$\endgroup\$
    – unhammer
    Commented Aug 2, 2019 at 18:18

3 Answers 3


If I understand the question correctly (and I may well NOT be understanding it, in which case apologies for the useless answer! ;)):

This is usually caused by interference between the frequency at which the light oscillates and the speed of the rolling shutter in the camera. These days, most LED light systems are controlled by a PWM (pulse width modulation) controller, which cycles power on and off at a relatively fast rate to simulate variable brightness levels. In some cases the frequency of the cycling gets close to the frequency of the rolling shutter, which can cause problems.

The brighter the LEDs are with a PWM, the more time they spend "on" and the less time they spend "off". The dimmer the LEDs are with a PWM, the more time they spend "off" and the less time they spend "on". You will usually find that the interference patterns get worse with moderate to low frequencies.

There are two possible solutions. The first is to use a linear voltage regulator rather than a PWM to vary power to the LEDs. A linear voltage regulator is less efficient as it burns off the excess energy as heat, however it will not result in oscillations in the light intensity as a PWM does. Therefor, dimmer light is dimmer light, strait up.

The other option is to use an LC-filtered PWM. A filtered PWM will use an inductor (L) and capacitor (C) to maintain voltage and current in the moments when the PWM cycles off. You can usually find PWM controllers that are pre-built with filtration, however sometimes it's not specified. It is usually easier to find a linear regulator instead. Just make sure you use one that has a nice, large heatsink. ;)

I thought of a third option. Use a PWM that operates at a very high frequency. Something that operates at 15khz should avoid any problems with shutter interference.

  • \$\begingroup\$ Wow, welcome back! It must be the moon that brings you out :-P \$\endgroup\$
    – dpollitt
    Commented Nov 15, 2016 at 3:04
  • 1
    \$\begingroup\$ Hah! Well...actually, it's the moon that keeps me in! :P It's a megamoon tonight...too bright to even image with my narrow band filters. So, I'm fiddling around online, rather than acquiring data. ;) I might get out there a little later here once the megamoon clears the trees, and image it just for the record. ;) I want to be able to compare it to normal and apogee (mini) moons as well, so I can demonstrate the size differences. \$\endgroup\$
    – jrista
    Commented Nov 15, 2016 at 3:10
  • 1
    \$\begingroup\$ @jrista LC-filtering would average voltage and current resulting in a very poor dimming curve. LEDs should be driven with constant current. Thus the PWM in the first place. \$\endgroup\$
    – bogl
    Commented Nov 15, 2016 at 12:25
  • \$\begingroup\$ It's true that the dimming curve isn't ideal. LEDs tend to stutter out at lower settings with a filtered PWM. Similar issues with a linear regulator though... Best solution then would be to use a high frequency PWM so that the PWM cycles don't interfere with the shutter rate. \$\endgroup\$
    – jrista
    Commented Nov 15, 2016 at 20:59
  • \$\begingroup\$ This might be best answered on electronics.stackexchange.com but I think you were close - what you need is a linear current regulated power supply, or a high-frequency PWM one, both these things are easily found thee days for quite cheap. \$\endgroup\$
    – John U
    Commented Jul 4, 2018 at 11:05

If you have manual control of shutter time, aperture, and ISO you can always shoot manually using a shutter time that is long enough to cover a full cycle of the lights' oscillation.


It is my understanding that a shutter speed of 1/30 or 1/60 in the United States will solve the problem. 1/30 is a full cycle, and 1/60 is a half cycle.

  • \$\begingroup\$ Only if the lights are operating at mains-frequency or a multiple or it, most PWM controllers will just have some cheap internal oscillator at some convenient frequency with low stability. A slower shutter speed will help though. \$\endgroup\$
    – John U
    Commented Jul 4, 2018 at 11:06

Your Answer

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

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