Light meter reading wrong in live view video mode

[Jon Grimson]

I'm really stumped by a problem with the light meter reading. I'm using Leica-R lenses with Leitax adapters; great lenses. Shooting everything manual of course. In stills mode (viewfinder or live mode) the light metering works great and is accurate with these lenses. When I switch to video in live mode the metering is completely out. So my question is, how are you guys metering light shooting video in live mode? Is this happening to anybody else? Does it happens on Canon lenses? Or does it happen only to adapted prime lenses?

The only work around I can come up with is quickly going from still live mode; dialing in my settings and setting exposure and then switching to video and ignoring the light meter bar that dives from perfect exposure reading all the way down to -3. The LCD image looks the same, but the fact that the light meter goes wacky make no logical sense to me. If I try and open up in video mode and trust the light meter reading, it is completely and obviously blown out in the LCD screen. I don't get why switching from stills to video in live mode causes this?

Any ideas, or help? Or is this another workaround to contend with.

[Chad Nickle]

Take a few seconds of video and check the histogram, the light meter sucks for checking video levels on this cam anyway.

[Brian Luce]

are you saying that the LCD is not wysiwyg? That on playback what you saw in the LCD isn't what you saw live? Or are you saying that the metering only is off?

I've been snapping stills and using them as a baseline for the LCD. There don't seem to be any good solutions other than stand alone light meters.

[Peter Rhalter]

Aperture Priority has been accurate for video with Live View on my 7D. I do most of my shooting with older Takumar and Yashhica lenses so the aperture is set manually.

[Jon Grimson]

Thanks everyone. The LCD is wysiwyg. The frustartion is the light meter is not just slightly off in live mode for video; it's completely way off. I guess I expected better with all the great accolades about advanced improved metering in the marketing spiel.

The big question to me is if it works fine in live mode for stills, why would it fail to work when I switch over to video? It should be identical. You can instantly see the meter go to the floor when it had just been accurate in stills mode. That's what I don't get.

Have not tried AP. Was hoping I could operate in full manual mode and have some trust in metering. It's not practical to use a seperate light meter most of the time and frankly I expected the camera to be able to perform this really basic function. If I have a defective unit that is what I'm wondering now, but it sounds like everybody is living with a non-functioning light meter for live mode video.

Would love to hear more from anyone using adapted prime lenses and see what they experience

[John Richard]

If you half push the shutter release button, you will hear the mirror move out of the way and then the light meter takes a reading, but when you go back to shooting video the light meter is blocked and cannot take a reading.

Histogram is the way to monitor where your blacks, midtone, and highlights pixels are being distributed during video recording as was suggested. Histograms are actually better than a light meter.

[Jon Grimson]

John, if that's true then that explains the problem logically. I guess this caught me by surprise since I've never seen anyone discuss this situation i.e. Philip Bloom, etc. in any tutorials I can find or describing a method to judge exposure when shooting video.

I come from shooting video cameras where you always have zebras and can instantly see exposure problems, etc.

I agree histograms are great for tools this. I have to get used to using the histogram on a still for exposure and then flipping over to live mode video.

It would be great if that could be engaged in real time like video camera zebras but I know that is wishful thinking.

Any other tips from you guys on how you manage your shooting workflow on exposure is certainly welcome!

[Norman Pogson]

I tried metering and shooting with a Canon 50mm f1.4 I metered in Tv for a 1/60th and got f13 in photo mode, switched to video and got the same reading with the mirror up.

I switched out to a 1980's Nikkor AI 50mm f2 and repeated the process, the meter gave me f11 in still mode and just under f11 in video mode. In video mode the meter does jump around, going two stops on the slightest movement of the aperture ring.

Conclusion is the meter for the old lens seems to be just over one stop out.

I have turned down the brightness on my LCD screen on the 7D to match more closely what I see on my monitor. So I am happy to eyeball my LCD screen usually.

[Jon Grimson]

Norman that's a lot more in the ballpark than what my meters read. Maybe I just need to spend a lot more time with it and learn to trust the LCD. Not that comforting doing it that way but oh well.

So do you conclude that the Canon lenses respond more consistently than your adapted ones? 1 stop difference is one thing; but the meter reading difference between modes with my Leicas is night and day. Mind you that the LCD looks the same so my feeling is that I will have to trust it when I toggle from stills to video after checking exposure.

This seems flakey to me!

[Norman Pogson]

Hi Jon, modern Canon lenses are designed to "talk" to the camera body, putting an adapter on the camera obviously totally takes that away, meaning the camera is working with less data for exposure.

Also older lenses are smaller, my Nikkor is half the diameter of the Canon 50mm, so I think if we want to use older lenses with adapters, we will have to either calibrate the LCD brightness to our computer monitors, or use light meters.

When I want to film a scene or subject, I do slow down to a methodical pace, I'm primarily shooting stock footage, so I always start with my 50mm and meter with that. I then swap out my lens to whatever I need, but I have a base line exposure, then just fine tune it for the lens I shoot with.

[Jon Fairhurst]

I prefer zebras and false colors to histograms for video. A histogram won't help you get the face lit to the same level from shot to shot.

[Lance Watts]

Your light meter reading is screwy because you are using a Leica lens. The same is true for any lens that is unable to communicate with the 7D body. Whenever, I use Nikon lenses, the meter is always underexposed, but as soon as I throw on a Canon or EOS compatible lens, everything is dandy.

[Brian Luce]

Originally Posted by Jon Fairhurst

I prefer zebras and false colors to histograms for video. A histogram won't help you get the face lit to the same level from shot to shot.

You 5d'ers just love to flaunt your zebras. One day we'll have them, one day.

Actually I've always heard the same thing about histos, that they aren't a great way to properly expose.

[John Richard]

Another suggestion is to just use a real light meter such as a Sekonic.

If you are setting up shots of a face and using lighting, then a light meter is invaluable for matching shots.
The light meter allows you to adjust key and fill lights to have the same stop differential between key and fill as well as correct aperture exposure and match to other shots.

[Liam Hall]

However you get to your exposure setting, you should always check the histogram. Overexposure on this camera is not a good way to go - and it's very easy to screw up!

Metering a grey card and checking the histogram regularly seem to work for me. Zebras would be nice, super nice...