Is ff metering more aggressive than crop?

[Marsu42]

I've recently bought the 6d and expected the metering to be just the same as on the 60d - but that doesn't seem to be the case(?), though both use the same 63-zone metering system.

To me, the ff metering system seems to be more aggressive, and I'm more often running into blown highlights than with my 60d - esp. since the amount of highlights "hidden" in the raw and only accessible in postprocessing is much smaller on the 6d.

This is an important difference, because esp. with flash & ettl you're really depending on the camera metering system, and I see me using -ec, spot/partial metering, exposure & flash lock more often than with the 60d where eval metering and "shoot away" usually worked.

Question: Is my impression correct and ff/crop metering is different, or have I've gotten it wrong during my rather short time with the 60d?

 

[Ewinter]

I got the same feeling from the 5DIII. I don't shoot in evaluative any more.
It's the same on the 1DX, which has af linked spot metering.
I think that's the only real advantage the d800 actually has over the 5dIII, personally

 

[privatebydesign]

It is quite possible the cameras assessment in Evaluative metering, is different in different models. The "issue" is that there is a component of "guessing" what you are taking a picture of, and this is proprietary, Canon don't explain what the camera is assessing, what look up information it has, or what adjustments it is making to "standard" metering because of those assessments.

To work out how your two cameras are metering compared to each other the easiest way to do it is mount a 70-200 on a tripod, then put each camera on the lens, zoom to achieve the same framing and then compare meter readings. Spot, Center Weighted, and Average should all read the same, if there is a difference then it should be consistent, 1/3 stop or so either way. But the Evaluative could very well show marked differences for no apparent reason.

This is one reason when Evaluative was introduced many didn't trust it, I never met a prolific slide shooter who used Evaluative mode because you just don't know what it is guessing you are taking a picture of.

 

[Marsu42]

It's the same on the 1DX, which has af linked spot metering.

But on 1dx it should be better due to the rgb metering? At least that what what I was guessing when I got blown highlights with 6d shots and very greenish nature scenes - or am I mistaken and the 6d "grey" camera view cannot be the cause for the too aggressive metering vs 1dx or Nikon?

This is one reason when Evaluative was introduced many didn't trust it, I never met a prolific slide shooter who used Evaluative mode because you just don't know what it is guessing you are taking a picture of.

True, but since it covers the complete shot it's good at quickly detecting very bright spots off center and adapting accordingly, at least on my 60d. The 6d doesn't seem to put that much weight into preventing blown highlights, maybe it's a descendant of the "pro" 5d3 and is for people who should know what they're doing.

 

[Ewinter]

While it's true the RGB metering helps a lot, it's still no good at guessing what I want to be exposed precisely.
With AF spot I can meter precisely every time.

I also wonder if they did with the exposure chip like they did with the AF chip between the 60D and 6D- kept it the same size while scaling up everything else like the mirror and sensor.
For example, the 9 point af covers most of a crop vf. On the FF, a 9 point sensor mostly fills the middle. I wonder if it's the same with the metering chip

 

[Skulker]

I don't expect every camera to meter in exactly the same way. There are incredibly complex algorithms involved.

I treat it as an excuse to get out there, take some shots, and get used to how the camera works.

Some of the best advice I have had was "Choose a mode and stick to it"

 

[Mt Spokane Photography]

I've owned many cameras almost every Canon DSLR and several Nikon as well. Each one has its differences. I thought my 1D MK II had exceptionally good metering, my MK II and MK IV were very good, but the MK II was outstanding.
My 7D and all three 5D Models were good, but not up to the 1 series.

 

[Loswr]

The 6d doesn't seem to put that much weight into preventing blown highlights, maybe it's a descendant of the "pro" 5d3 and is for people who should know what they're doing.

I'm not sure the logic holds. Lots of people (including me) have observed that the 1D X (and other 1-series bodies), evaluative metering is 1/2-stop or so darker than other bodies.

 

[duydaniel]

Here are some facts I have learned:
Only 1Dx has 100,000 rgb metering sensor.
The rest of Canon lineup has 63 dual layer zone.

BUT
their 63 dual layer zone between systems are not identical.
Like the 5D3 and 7D meter slightly different. I think it is due to
newer "code" or version 1.0 vs 1.1

 

[candyman]

The 6d doesn't seem to put that much weight into preventing blown highlights, maybe it's a descendant of the "pro" 5d3 and is for people who should know what they're doing.

I'm not sure the logic holds. Lots of people (including me) have observed that the 1D X (and other 1-series bodies), evaluative metering is 1/2-stop or so darker than other bodies.

True for my 5D MKIII. I noticed it last year when I bought it next to my 7D and compared it with the 7D

 

[JPAZ]

I did note a difference when moving from the 50d to the 5diii and now I keep my EV +1/3 to the right. Actually, I forgot that I set it this way a while back but it definitely gets better exposures than when the setting was at 0.

 

[Sella174]

... I never met a prolific slide shooter who used Evaluative mode because you just don't know what it is guessing you are taking a picture of.

Yip! I only use centre-weighted metering.

I also wonder if they did with the exposure chip like they did with the AF chip between the 60D and 6D- kept it the same size while scaling up everything else like the mirror and sensor.

Good question ... and I think herein lies the answer.

 

[Marsu42]

For example, the 9 point af covers most of a crop vf. On the FF, a 9 point sensor mostly fills the middle. I wonder if it's the same with the metering chip

I agree this could be the answer - the 6d is the only ff camera with the 9/11 pt system and the 63z metering, so the metering system is not a straight copy from the 5d3.

And I know for a fact that Canon did a botch job on the 6D firmware - it's not visible to the user, but as I'm developing for Magic Lantern there are a couple of firmware calls that just don't work correctly and are bugged ...

... thus it might be also that the metering system wasn't tested that much, I recon that's what you get a 1d for (if you happen to win the lottery ). Alas, Canon is unlikely do any real 6d fw updates that fixes it, they'll probably only do a feature update in 2014/2015 to boost sales a bit (like raw hdr).

 

[iron-t]

Interesting thread. I recently moved from a 60D to 5D3 and noticed the evaluative metering is completely different on the 5D3. By comparison the 60D was extremely reliable in nearly all situations, particularly with flash. The 5D3 is rather unpredictable--again, particularly with flash--and seems to underexpose by ~1 EV as often as it gets the exposure right. This would seem to support the idea that different bodies just handle evaluative metering differently. Frankly, although cranking up the exposure on many shots in post is annoying, I'd be far more concerned if the camera were frequently overexposing and destroying highlight detail. Either way, I'm not impressed with evaluative metering on the 5D3 as I was with the 60D and have switched to spot for most shooting.

 

[Loswr]

...as I'm developing for Magic Lantern...

That explains much, Grasshopper... 8)

 

[pensive tomato]

For example, the 9 point af covers most of a crop vf. On the FF, a 9 point sensor mostly fills the middle. I wonder if it's the same with the metering chip

I agree this could be the answer - the 6d is the only ff camera with the 9/11 pt system and the 63z metering, so the metering system is not a straight copy from the 5d3.

And I know for a fact that Canon did a botch job on the 6D firmware - it's not visible to the user, but as I'm developing for Magic Lantern there are a couple of firmware calls that just don't work correctly and are bugged ...

... thus it might be also that the metering system wasn't tested that much

Very interesting... I thought the metering sensor was located between the pentaprism and the viewfinder, but I couldn't find any confirmation of this. I'd think that the top of the camera has more parts and design in common with the 5D series than the 60D. Of course, nothing prevents Canon from being cheap and using the 60D's sensor over the 5D3's.

Now I think you hit the nail in the head regarding firmware, as the root of all this may be in the algorithms used for metering. As you have documented, programming work on the 6D was likely done -to put it mildly- on a limited budget.

By the way, I've also noticed metering working differently for each of the Canon DSLRs that I've used (T1i, 7D, 6D). I never knew how much it was my skills changing over time and me adjusting to each camera's behavior as I do my EV compensations (maybe Canon was at fault too).

 

[Marsu42]

...as I'm developing for Magic Lantern...

That explains much, Grasshopper... 8)

Indeed it explains why I'm not using Nikon, because as I was raving mad over the limiting auto iso of Canon on my 6D I sat down and coded a working version (ec on m, no more iso 400 flash lock, unlimited min. shutter in av mode, selects only full iso stops for raw shooting)

 

[MonteGraham]

...as I'm developing for Magic Lantern...

That explains much, Grasshopper... 8)

Indeed it explains why I'm not using Nikon, because as I was raving mad over the limiting auto iso of Canon on my 6D I sat down and coded a working version (ec on m, no more iso 400 flash lock, unlimited min. shutter in av mode, selects only full iso stops for raw shooting)

+1

 

[pensive tomato]

...as I'm developing for Magic Lantern...

That explains much, Grasshopper... 8)

Indeed it explains why I'm not using Nikon, because as I was raving mad over the limiting auto iso of Canon on my 6D I sat down and coded a working version (ec on m, no more iso 400 flash lock, unlimited min. shutter in av mode, selects only full iso stops for raw shooting)

Thanks for your work with Magic Lantern! This also brings to mind the silly limitations on the behavior of the Set button that I thought you tackled as well.

 

[Marsu42]

Thanks for your work with Magic Lantern! This also brings to mind the silly limitations on the behavior of the Set button that I thought you tackled as well.

Yes, SET has the option to quickly set the af to the center, and it's easier to reach the points with the multicontroller. This feature is already merged into the ml 6d dev builds...

... unfortunately the "focus patterns" feature is one of the cases of broken Canon firmware, you can only select one af point at the time and not for example the three right ones (like 60d and all other 9pt) which would have been great since on the 6d they're not cross. On the other hand, crippling this could have been deliberate on Canon's part.