Here's the obvious test to do to get a handle on whether it is lighting or shutter:
Shoot outdoors at the same shutter speeds. In fact, shoot outdoors at all the shutter speeds, in steady light.
Set up a tripod, set the ISO to 50. Invoking the Sunny f/16 rule (which in practice is often f/14 or f/11), between 10 am and 2 pm with the sun behind you, the shutter speed is 1-over-ISO or 1/50 at f/16. So shoot a series of shots:
1/25 f/22, 1/50 f1/6, 1/100 f/11, 1/200 f/8, 1/400 f/5.6, 1/800 f/4, 1/1600 f/2.8, 1/3200 f/2, 1/6400 f/1.4, 1/8000 f/1.2 Say! You get to use the f/1.2 lens at f/1.2!
The exposures should all be very close if your shutter is accurate. Shutter accuracy is usually allowed to be +/- 15% except above 1/4000 where it can be +/- 30%. Shoot the series of a normal outdoor scene, ideally with a person close enough to the camera to show skin without blocking the view. You can also shoot at 1/3 stop increments if you like, perhaps especially in the higher shutter speeds.
Shoot it again with an evenly illuminate blank wall or a medium gray sheet of cardboard or even a plain brown sheet of cardboard. It has to be flat and have the sun shining directly on it, not sideways.
This will test how even the shutter is. If you see some bands then the shutter has variation during the slit travel. It is probably best to make 3 shots or more at each shutter speed for this test, in case there is intermittent variation.
I predict the camera will pass the tests just fine.
snapsy wrote:
If the dark band was the result of cycling LEDs then the orientation of the band wouldn't follow the orientation of the shutter curtains between the photos taken at landscape vs portrait.
I thought the same way at first and shot down my own idea, but I reconsidered it, and the orientation will not disturb the effect since the LEDs are probably blinking on all at the same time.
jorgegarcia wrote:
ISO 800 Exposure 1/1600 sec Aperture 1.2
In this one, the LEDs blinked on as the slit got to the right side. Or alternatively if the slit travels the opposite way, they blinked off as the slit got a little bit away from the right side:
Notice how at the top part of the frame, the camera has captured some much dimmer ambient light (yellow sodium vapour) reflecting off the penguins. So the shutter is working fine.
Plausible points Monito, thanks. I'm still thinking that it might be the shutter, esp. when I see the same effect on both sets of photos/lighting. It'll be interesting to see how this turns out. Your ideas for the experiments are good as well, assuming whatever the issue is can be reproduced somewhat readily. OP, please keep us informed.
I contact canon and sent them the images and they believe something is wrong and offered to pay for shipping to send it in for service. i am still within b&h's return/exchange period so i have also contact them with regards to this to see what they think. in the mean time, i will try to test it out using Monito's methods and see what happens.
The penguins are very likely due to LED refresh rates. LED Christmas lights work more like fluorescent than incandescent in that they switch on and off rapidly to fool our eyes into thinking they stay on all the time. If you shoot at a shutter speed faster than 1/60 you are in danger of getting this effect.
For the 1st shot it's hard to tell for sure without further testing.
For the face half in shadow, if you were focusing on the dark area it could be that there just wasn't enough light. f/1.2, 1/80, ISO 10,000 is dark, and faces are often challenging to focus on in the dark.
It's say the other forum members were 100% correct by saying it's the cycling of lights. It all makes perfect sense. The penguin shots are a great example -- especially the horizontal one that shows some ambient light up top. You haven't noticed this with previous cameras simply because you probably haven't had a camera that could reliably shoot at ISO 10,000 with a f/1.2 lens. I don't think sending the camera in for "repair" is a wise idea. There isn't anything wrong with the camera, and if you send it away for "repair" all that will happen it will be knocked around in transit. Now THAT is something that might break a camera.
If you don't doubt that LED's cycle very fast, the easy way to tell is to turn your head real fast while looking at them. You'll see a trail of dotted lights. I notice this a lot with LED tail lights in certain cars --- mostly Cadillac. If you turn your head while driving, you'll see a very strange broken light trail. Some LEDs are better at this than others, and I'm pretty sure that an inexpensive made in China christmas decoration isn't going to use the highest grade LED lights.
jorgegarcia wrote:
Let's not get side tracked here with why I chose x settings. These pics are for shits and giggles. Thank you to those trying to help me out.
Since these pics have no meaning and are just for sh!$s and giggles.... Why not go outside and take a few pics so that everyone can eliminate the ideas about the rate the lights are cycling?
I agree with most others with that high of a shutter speed youre just capturing the lights refresh rate. I won't get into the exposure choices, they wouldn't have been mine, and do seem quite silly, but some daylight shots with similar shutter speeds would certainly rule out shutter issues. If canon wants to take a look at it let them, maybe even see if they will calibrate some of your lenses as well so the trip is not totally useless.
I just got few images today with a yellow band, didn't know anything about this issue, will be ready next time. I was at a high shutter speed on every image and under fluorescent lights.
EB-1 wrote:
I think we are seeing the interference of the light cycling and the portion of the slit open (a movement across the short dimension of the frame). If there were a mechanical problem with the shutter it would not occur at different places in different shots.
EBH
Why not? My old EOS650 had a shutter go bad and it occurred in different spots shot to shot.
the shutter speeds are so high it could be catching the artificial light bet
