90d: Photos over exposed (double iso) when shooting in non live view mode

[vizualequalizer]

I have a Canon eos 90d and the photos are over exposed when shooting in non live view mode (when switching back-and-forth between live view and viewfinder modes; the ISO doubles), or another words looking through the viewfinder, the ISO will double. There's no exposure addition or subtraction on exposure stops. I am confused please help!

[jrhoffman75]

In case no one looked at the link I posted above, here is some information posted by our forum member ray-uk on the DPReview site

"There has been discussion lately about viewfinder and live view metering givin different results, a look at the cameras specification goes some way to explaining this:

80D Viewfinder metering

metering from 7,560 pixel 63 zone meter sensor
Evaluative metering linked to all AF points
Partial metering, 6% of viewfinder centre
Spot metering, 3.8% of viewfinder centre
Centre weighted average metering

80D Live view metering

metering from 24.2 megapixels image sensor
Evaluative metering 315 zones
Partial metering 6.1% of live view screen
Spot metering 2.6% of live view screen
Centre weighted average metering

-- hide signature --

7D Mk II Viewfinder metering

metering from150,000 pixel 252 zone meter sensor
Evaluative metering linked to all AF points
Partial metering, 6% of viewfinder centre
Spot metering, 1.8% of viewfinder centre
Centre weighted average metering

7D Mk II Live view metering

metering from 20.2 megapixels image sensor
Evaluative metering 315 zones
Partial metering 10% of live view screen
Spot metering 2.6% of live view screen
Centre weighted average metering

--------------------------------------------------------

When photographing a scene that has various dark and light spots it stand to reason that the 2 systems may well give different results depending on where the dark and bright spots are. It looks from the above information that evaluative metering could be more error prone in this respect and centre weighted could give the more consistent results.

I think this just highlights the fact that automated metering, although good in most circumstances still requires a little intelligent input from the user."

[ebiggs1]

"I did do outdoor testing and the same issue. I should have thought of this earlier, but I tested it with my Rebel 1Ti and have the same issue although it is much less exagerated, only about 1/3 - 2/3 stop difference."

OK fine, if you did the outdoor test and saw a 1/3 stop, very hard to see BTW, I would consider this normal.  If you saw a 2 or 3 stop difference that is not normal and I would suspect the camera or lens is faulty. (After doing the full reset and the basic settings I outlined?)

The next step, without changing any of those basic settings, you need to try a different lens. Do the daylight test again.

[Waddizzle]

"When photographing a scene that has various dark and light spots it stand to reason that the 2 systems may well give different results depending on where the dark and bright spots are. It looks from the above information that evaluative metering could be more error prone in this respect and centre weighted could give the more consistent results.

I think this just highlights the fact that automated metering, although good in most circumstances still requires a little intelligent input from the user."

The cited "above information" are specs.  That is an extraodinary conclusion to reach just by looking at specs alone.  

Personally, I have found Evaluative Metering to be very quite good at metering a subject or a scene.  Canon has selected it to be the default metering mode because it works, and it works very well.  It is not "error prone" as the article suggests.

[vizualequalizer]

I'm not sure I understand your last post @waddizzle. Are you saying make sure the mirror box is dust free, and use a rocket blower?

[Waddizzle]

---

@vizualequalizer wrote:

I'm not sure I understand your last post @waddizzle. Are you saying make sure the mirror box is dust free, and use a rocket blower?

---

Yes, I am.  Make sure the image sensor is clean for Live View metering.  The OVF metering sensor is probably inaccessible because it is located at the top of the mirror box [viewfinder].

Make sure your locked AF point is the same for both modes.  Using a tripod works best.

[vizualequalizer]

Okay. I don't have a rocket blower or the like. Regarding your AF points being the same, I was just about to upload pics for this thread, when in fact, the focus points are different, which I'm pretty sure like you are saying influences the meter!

As soon as I can I will reshoot and nail down the constants and post the results.

Thank you kindly for your time and attention, viz

[Waddizzle]

---

@vizualequalizer wrote:

Okay. I don't have a rocket blower or the like. Regarding your AF points being the same, I was just about to upload pics for this thread, when in fact, the focus points are different, which I'm pretty sure like you are saying influences the meter!

As soon as I can I will reshoot and nail down the constants and post the results.

Thank you kindly for your time and attention, viz

---

Use "1-pt AF" point selection for both OVF and LV modes.  LV does not default to 1-pt AF.  Also, use One Shot AF mode, so that the metering locks when focus locks.

According to my light meter, all of my Canon bodies overexpose about almost 1/3 of stop in the OVF.  I have never tried the same tests with the LV mode.  I think this is done to capture more details in shadows.

[jrhoffman75]

I guess it would depend on a definition of error if you believe your stand alone meter is correct.

"According to my light meter, all of my Canon bodies overexpose about almost 1/3 of stop in the OVF".

Unlike the other metering modes, Evaluative Metering has no set pattern. It is an intelligent mode that prioritizes its evaluation based on where the focus point is, since it assumes that is the location of interest.

The greater the number of metering zones the more precisely the camera can determine the conditions surrounding the focus point. I believe that can contribute to a different exposure reading between OVF and Live View, particulary for complex lighting situations like the original images that were posted. Throw in Auto AF point selection and the potential for differences grows more. Add in the fact that both methods use different sensors for measuring exposure. I would be more surprised if they gave the same result, just like two cameras side by side could give different readings. 

One could call it an error if a particular mode is deemed correct and the other mode is producing a different result.

[Waddizzle]

---

@jrhoffman75 wrote:

I guess it would depend on a definition of error if you believe your stand alone meter is correct.

"According to my light meter, all of my Canon bodies overexpose about almost 1/3 of stop in the OVF".

Unlike the other metering modes, Evaluative Metering has no set pattern. It is an intelligent mode that prioritizes its evaluation based on where the focus point is, since it assumes that is the location of interest.

---

I am using a Sekonic L-858D.  I measure both reflected light and incident light.  I use 1 pt AF using the center AF point.  I use Partial Metering and Evaluative Metering, and I get identical exposure settings in the cameras..

Histograms in the cameras confirm the slight overexposure, as do both LR and DPP.  It is a pretty consistent result across six different DSLR models, and my M3.  All of the histograms "expose to the right", so to speak.

I never said it as an error.  It just seems to be how Canon "bakes" their images.

[ebiggs1]

"I was just about to upload pics for this thread, when in fact, the focus points are different, ..."

This means you did not do the test as outlined.  You must follow these instructions precisely.

Here they are again...

"... reset the 90D.  Set it to P mode, fix ISO to 200. Use just the center focus point and One shot. WB set to daylight. Now go outside on a sunny day a take several general shots.  Make sure you have some vertical lines in some of them. Like trees, a fence or buildings, etc.  It the 90D makes the great photos it is capable of the camera is working correctly.  This means you set something incorrectly for the photos in question."

"I don't have a rocket blower or the like..."

It's not a bad idea to have one but this is not your problem right now.