10-31-2018 12:31 PM
LED lighting can create a different flicker effect than florescent or high intensity discharge lights (i.e. mercury vapor or high pressure sodium) because the driver module cuts the actual LEDs on and off very quickly with a near square wave light output pattern from the emitters which go from no to full output and back in nanoseconds. Most of our other artificial light sources provide some significant non-binary transition between on and off states where there is still significant light output over the entire time between peak and minimum; for example even when all excitation is removed you will see a brief glow from the phosphor of a florescent tube and the "off" period is far more brief during the operating cycle compared to what you see after turning the lamp off.
More complex LED arrays use a sophisticated driver module that pulses the LED junctions at a rate far faster than the human eye can perceive and the resultant duty cycle (percent of time it is on) provides the observed light level. But depending upon when the camera shutter is open the camera may see a different number of pulsed light events for each exposure because the shutter isn't synchronized to the driver module. The effective LED array light output is based upon duty cycle which depends upon the width (period of time) of the on pulse and the width of the off period. There are a nearly infinite number of combinations of on and off width and frequency leading to the same perceived light output but with very different captured image effects.
Think of taking a pair of equally timed long exposure images and firing a flash manually where one exposure has three flash events of equal intensity and the other has five so if one is correctly exposed the other will be either over or under exposed. With LED lighting you stand a good chance of having this happen especially with a more sophisticated lighting source that is controlled by an advanced PWM (pulse width modulation) controller instead of a simple line frequency fired module which tends to have a slower modulation rate with extended on time. You could simulate the time lag effect of other types of bulbs with the LED driver but then you would run into a problem of color shift because instead of being modulated between full brightness and off the LED would be driven in a more linear fashion producing some output most of the time but at a lesser intensity and resulting different color temperature given the way LEDs produce visible light.
Some arrays attempting to achieve very high specific light output will pulse the LEDs at a higher current than recommended by the manufacturer providing extreme brightness during the brief on event but protecting the LED from excessive heat by running it at a duty cycle that still provides reasonable, albeit shortened, LED life. These types are likely to be the most problematic for photographic purposes but none of them are going to advertise their method of achieving very high light output at a competitive price. A completely flicker free LED array would be DC driven with a 100% duty cycle but this is going to result in a bigger, heavier, and far more expensive array for the same light output so such a device would tend to be sold at a very high price to specialized users.
For a simple test of the camera simulate the scene using incandescent lighting and see if the over-exposure issue disappears. Incandescent lights powered by AC also have a slight flicker but because of the tremendous thermal lag of the superheated filament it is not visible to humans or a practical camera.
LED lighting has some huge advantages over other light sources but like most new technologies it brings additional problems and unwanted side-effects. One area receiving significant study now is the impact of this differnet flicker environent on human health and behavior including increased potential for epileptic events in children. As we learn more, LED driver protocol will change and some of those changes are going to have various impacts on photography that we will have to learn about and adapt.
11-01-2018 09:42 AM
"I'll test with some incandescent lighting since it's continuos."
You can try that but it will not help. Your problem is not light flicker. You have a correct exposure and then you have an over exposure, by several stops. If light flicker is happening the light flickers off, not brighter. Which is what you would need to get an over exposure. A two second SS will see the same amount of flickers anyway!
Since you are shooting at f32, I will bet the lens is not closing all the way down to f32 on the over exposure. This could be due to a faulty lens and is almost the only way your problem can happen.
11-04-2018 06:20 AM
Thank you ALL for all the input.
"ebiggs1", you are correct, it must be the lens. i tested same senario with my 50 mm and didnt have any issues even when shooting 36 continuous shots. So THANKS again
11-04-2018 06:49 AM
I'm have exactly the same problem. I posted about it earlier today 'problems with exposure'.
The exact same settings, same field of view/subject/background and one turns out fine then completely over-exposed.
I'll be following this thread to see if there's a fix, because I'm throwing away a ton of photos at the moment.
11-04-2018 06:54 AM
11-04-2018 09:37 AM
"... I don't think flicker is the problem"
The flicker answer was and is silly. You are getting an overexposure because you are getting more light. Not less light. That is where you need to concentrate your efforts.
11-04-2018 09:44 AM
I've been having the same issue but outdoors (no chance of flicker). Two shots, same background, subject matter, composition, settings . . . everything seemingly the same and one looks okay, another completely over exposed and blown out. New Canon 5D mark IV and really upsetting
With 2 second exposures, I do not think it could be flicker.
But, i would not completely rule it out, either. Unlike analog light sources that merely dim during their “OFF” portion of their flickering, a digital, solid state light source will go completely dark. The human eye just cannot see it go dark. If you fire the shutter at the moment when the solid state light source is dark, then you may wind up with an over exposure.
It is very easy to eliminate this as a source of the trouble by simply enabling “Light Flicker Compensation”, and see if the problem just simply goes away. If it does not, then you know the problem is not flicker. If you are shooting with artificial lighting, then it is probably a best practice to enable LFC, anyway.
11-04-2018 10:04 AM
With more sophisticated LED arrays, the anti-flicker setting in the camera won't have any effect. It is set up to operate at an exact multiple of either of the commonly used 50 or 60 hz line frequencies (line frequency depends upon country/region) and if the flicker rate isn't an exact and recurring multiple of that then the camera anti-flicker algorithm cannot work.
Historically common light sources (florescent strips, high intensity discharge mercury or sodium stadium lights, etc.) will flicker in sync with the applied line frequency. However some LED arrays, in particular those that can operate from an internal or external battery pack or those using a more advanced array, have a driver that is not tied in sync to these commonly used line frequencies so the camera won't detect and correct for the problem. Canon adds this line frequency restriction as a footnote in the manual (at least for the 1DX family).