5D MKIV - HELP ?! metering is overexposing & quick review with 16-35 F4 is

Hi, I finally got a mkIV after 5 years of my MkIII.
I’ll do a bit of a review at end of this post.

But first shoots with it , I’m very impressed. Much nicer files , not the smirry smudgey look of the mkIII.
And much less colour noise in shadows , possibly more luminance noise as iso goes up. But much better , and the resolution bump is decent.
Processing raw images in capture one , seems mor comparable to D800 or A7rii files, rather than 5d mkiiis.
Which is great.

One problem though is the meter seems to be giving me over exposure readings especially in evaluative mode.
- I shoot manual all the time , and am used to metering from different sections of a scene then working out the ‘correct’ exposure from my various readings of the scene.
It seems nearly 2 stops different from my mkIII meter..

Say I’m metering for dark areas in trees in shade or a dark road or stone building . - normally on my mkIII in evaluative I would set the exposure to 2 stops under .
Actually I’ve been using this technique for 20 years, and has worked well for me on various film and 5D series cameras.
If I do this , these dark areas are almost a stop over exposed. So I found I have to set exposure to 3-4 stops under the meter from dark areas.
Everything i meter for , and say it’s an average tone and constrast scene , it’ will be 1 stop over if go on the meters exposure.

I’ve tried center weighted and this seems similar to readings I was getting from my mkiii.
Partial seems very hot and miss .

Any thoughts ?
 
Mar 25, 2011
16,848
1,835
The 1 series cameras meter the AF spot, but the 5 series cameras do not.

It sounds like you think you are metering different areas of the scene, but you are not. This could be the cause of the miscalculation. Try letting the camera meter the shot.

However, in liveview,using the touch screen to focus on a area seems to also meter on that area. I haven't researched this, I notice that the lighting changes according to where I focus though.
 
Upvote 0
BasXcanon said:
Yeah thats the negative side of increased DR, in digital you overexpose 1-2 stops on the skies instead of the shades.

Thanks for reply
But really ... increased DR attribute! I don’t think so.
It’s the meter , trust me . I’ve been metering manually all my life on various different cameras and metering modes .
And while yes this is a method more commonly taught for exposing B&W negs . It does translate very well to digital , if you want great exposure and tonal range in your shadows and mids.
And it holds in most normal lighter to highlight tones but often not things like backlit clouds. In which case I’ll shoot another exposure for that section if I see it overexposing, or adjust exposure.

- I know what you recommend exposing 1-2 stops over the highlights , I do sometimes but this often leads to trees and shadow areas being too dark.

Honestly even on a flattish scene- say a street with stone or red brick houses and a mix of cars and trees on a cloudy day - taking in minimal sky in the frame , which evaluative should nail fine ( as it has on previous 5D series) it’s over exposing.

I suspect they’ve changed the alogorithm for evaluative , or simply over all .
To help avoid under exposed scenes. Or mines is mis calibrated
 
Upvote 0
Oct 26, 2013
1,140
426
If it is not too late, I would return it and get another. I have found metering to be very variable within camera lines. When I first bought a 6D, it was overexposing in evaluative a bit (not much, but more than I was used to). I returned it and the 2nd model underexposed. The third try was a keeper. Similar story with my Sony A7 - it overexposed by a good 1.5 stops. Returned it and the next one only overexposed by 1 stop! So, there seems to be quite a variation or tolerance when it comes to exposure - at least in my limited experience.
 
Upvote 0
dak723 said:
If it is not too late, I would return it and get another. I have found metering to be very variable within camera lines. When I first bought a 6D, it was overexposing in evaluative a bit (not much, but more than I was used to). I returned it and the 2nd model underexposed. The third try was a keeper. Similar story with my Sony A7 - it overexposed by a good 1.5 stops. Returned it and the next one only overexposed by 1 stop! So, there seems to be quite a variation or tolerance when it comes to exposure - at least in my limited experience.
Thanks , could be possible. And I’ve been lucky in the past .

Reminds me when at college we did a test to get our cameras true exposure reading And we found everyone’s meters were different when metering grey cards . Maybe I’ll go back to this test to see if it is that.
 
Upvote 0

SecureGSM

2 x 5D IV
Feb 26, 2017
2,360
1,231
klickflip,
I own 5D IV and can confirm that 5D IV metering is in fact more accurate than the same of 5D III. let me explain what is really going on here and you will understand why your +/-2 stop rule is no longer works as it should for you.

your old 5D III metering system used to treat all objects as if it was metering a mid grey colour regardless, be that white or black. therefore darks, blacks and deep shadows were overexposed to appear on photo as ... mid grey. and your whites were underexposed to appear on photo as .. mid grey... hence you had to compensate your exposure by -2 or +2 stops accordingly.
OK, but.. why 2 stops, not 3 or 4 or 5? lets see..
here is why.. useable photographic dynamic range of 5D III sensor is 5.4 EV (stops) only. yes, 5.4 stops that is full DR less safety margins. safety margins are areas with LEVEL values dangerously close to the sides of the curve. those areas cannot be printed as 0-10 value is too much white to print and 250 is too much black to be distinguishable from the crashed black.
so, the entire usable area of 5D III histogram is 5.4 EV left to right. Your mid greys are located right in the middle of the histogram. That's 5.4EV/2 =2.7 EV stops mid grey to extreme shadows and extreme highlights.
so, when you spot meter with 5D III for extreme shadows, your camera would overexpose to shift the extreme shadows area right to the middle of the histogram overexposing by approx. 2.5 EV and underexposing spot metered extreme highlights by approximately 2.5 EV.
These rules are changed for 5D IV quite a bit. firstly, usable photographic dynamic range of the camera has increased dramatically and measured to be approximately 6.7 Ev (stops). 6.7/2 = 3.35 EV mid grey to extreme highlights or extreme shadows.
this is what you have observed as well: with your new camera you have to apply stronger exposure compensation for your spot metered shadows. yes, 3.5 EV would be correct. give or take.
with extreme highlights situation is a bit different in that camera will shift the curve to the right a bit, Automatic +0.66EV to 1.66EV EV ETTR depending on the scene if you will :)

p..s. evaluative mode camera reading value for the test shot of the X-Rite ColorChecker Card (reflected light): ISO 100, F5.6, T=0.4s, Sekonic 478D incident light reading for the same: ISO 100, T=0.4s, F5.6

I have attached -3EV, 0EV and +3EV exposed images (JPEG, Lightroom, no processing applied, 100% quality) and screenshot of the LR screen with Histogram right on the money for me. Both Sekonic 478D and 5D IV evaluative metering mode produced exposure values that perfectly agree.



klickflip said:
BasXcanon said:
Yeah thats the negative side of increased DR, in digital you overexpose 1-2 stops on the skies instead of the shades.

Thanks for reply
But really ... increased DR attribute! I don’t think so.
It’s the meter , trust me . I’ve been metering manually all my life on various different cameras and metering modes .
And while yes this is a method more commonly taught for exposing B&W negs . It does translate very well to digital , if you want great exposure and tonal range in your shadows and mids.
And it holds in most normal lighter to highlight tones but often not things like backlit clouds. In which case I’ll shoot another exposure for that section if I see it overexposing, or adjust exposure.

- I know what you recommend exposing 1-2 stops over the highlights , I do sometimes but this often leads to trees and shadow areas being too dark.

Honestly even on a flattish scene- say a street with stone or red brick houses and a mix of cars and trees on a cloudy day - taking in minimal sky in the frame , which evaluative should nail fine ( as it has on previous 5D series) it’s over exposing.

I suspect they’ve changed the alogorithm for evaluative , or simply over all .
To help avoid under exposed scenes. Or mines is mis calibrated
 

Attachments

  • canon 5D IV Clipping Point Range and DR.jpg
    canon 5D IV Clipping Point Range and DR.jpg
    143.2 KB · Views: 118
  • 5D4_0009.jpg
    5D4_0009.jpg
    1.5 MB · Views: 115
  • 5D4_0010.jpg
    5D4_0010.jpg
    910.7 KB · Views: 117
  • 5D4_0011.jpg
    5D4_0011.jpg
    1.4 MB · Views: 117
  • 5D4_0009_LR.jpg
    5D4_0009_LR.jpg
    318.7 KB · Views: 124
Upvote 0
Thanks GSM for that explanation.
I’m kinda with you , but maybe I’m visualising how the expanded DR affects tones in the recorded image differently to your explanation.
- do you have any tone charts that show this to help?

Not having a go with anything I say , but genuinely wanting to learn and discuss more about any differences in extra DR vs metering .

I get that we have more available EV on the mkIV. But I don’t see how that affects meter reading.
- I’ve been shooting manual and asssing the the readings within a scene for over 20 years , from b&w & colour neg to Trany to digital. So know the differences in DR . Eg I use to over expose negative file by 1/2 -1 stop and underexpose tranny by 1/3rd (then push develop +1/3rd)
I know in some films and some digital camera I’ll lose or keep more highlights and shadows , but I’ve never seen that alter my base exposure readings.
Digital i overexpose a bit if I can ( depending on highlights) to keep S/N ratio down and more flexibility in raw processing .

I know well fine the meter sees mid grey 18% , and base all my scene reading and calculations on that.
As I mentioned I was taught to use the zone system.

So why when I meter for a mid grey tone section and set my camera to that exposure does it record that as a light tone? Whereas every other camera I’ve had recorded that similar tone section as a mid tone (we’ll ignore colour to keep it simple) produces that tone as expected eg mid tone .

I kinda get what you are saying - in that what was dark grey/ near shadow with detail , on a mkiii` was typically 2 stops under , but now in MKIV is 3.5 stops under .
I get that you are saying the EV is expanded. But surely this means that say on mkiii all tones were crushed to black at -4 , yet on a mkiv we still have some tone.

The centre point should stilll be the same just the EV is expanded out (maybe not quite equally on mkiv) on either side from the centre.

Surely 18% grey metered and exposed should be recorded as that regardless of how much DR a camera has ?
Then depending on a cameras DR I can choose my exposure to get the best exposure for my needs .
I’ll shoot some white grey and black cards over the next couple of days and get a mkiii as well to test the differences, in meters and what they record.
Thanks GSM for that explanation.
I’m kinda with you , but maybe I’m visualising how the expanded DR affects tones in the recorded image differently to your explanation.
- do you have any tone charts that show this to help?

Not having a go with anything I say , but genuinely wanting to learn and discuss more about any differences in extra DR vs metering .

I get that we have more available EV on the mkIV. But I don’t see how that affects meter reading.
- I’ve been shooting manual and asssing the the readings within a scene for over 20 years , from b&w & colour neg to Trany to digital. So know the differences in DR . Eg I use to over expose negative file by 1/2 -1 stop and underexpose tranny by 1/3rd (then push develop +1/3rd)
I know in some films and some digital camera I’ll lose or keep more highlights and shadows , but I’ve never seen that alter my base exposure readings.
Digital i overexpose a bit if I can ( depending on highlights) to keep S/N ratio down and more flexibility in raw processing .

I know well fine the meter sees mid grey 18% , and base all my scene reading and calculations on that.
As I mentioned I was taught to use the zone system.

So why when I meter for a mid grey tone section and set my camera to that exposure does it record that as a light tone? Whereas every other camera I’ve had recorded that similar tone section as a mid tone (we’ll ignore colour to keep it simple) produces that tone as expected eg mid tone .

I kinda get what you are saying - in that what was dark grey/ near shadow with detail , on a mkiii` was typically 2 stops under , but now in MKIV is 3.5 stops under .
I get that you are saying the EV is expanded. But surely this means that say on mkiii all tones were crushed to black at -4 , yet on a mkiv we still have some tone.

The centre point should stilll be the same just the EV is expanded out (maybe not quite equally on mkiv) on either side from the centre.

Surely 18% grey metered and exposed should be recorded as that regardless of how much DR a camera has ?
Then depending on a cameras DR I can choose my exposure to get the best exposure for my needs .
I’ll shoot some white grey and black cards over the next couple of days and get a mkiii as well to test the differences, in meters and what they record.


So now to record a very dark tone properly on mkiv I now need to underexpose from the spot meter reading by 3.5 stops , but surely the centre point of 18% when that tone is exposed properly should still be the same when viewing the raw file with the processors standard canon mkiv profile , before any adjustments?

One thing is puzzling me , why does the mkIV’s meter only show -/+3 then , should it not be -/+5 with this expanded DR now?

Cheers!
 
Upvote 0

SecureGSM

2 x 5D IV
Feb 26, 2017
2,360
1,231
I have updated my post above with images, diagrams to demonstrate what is really going on with 5D IV DR and clipping range.

to address your question regarding "... But surely this means that say on mkiii all tones were crushed to black at -4 , yet on a mkiv we still have some tone...."

lets take a look at my 6D chart attached. its DR and clipping range is a little bit wider than the same of 5D III. 0.3 EV only.
Please note that DR boundary values are set at 10 to 250 and not all the way from 0 to 255.
now, let's look at the second chart. if you extend boundary of measured DR all the way from 0 to 255, then we get much wider DR value.
incidentally, we get exactly 4EV to play with extreme blacks!!!! there are _SOME_ tones there left as dropper would indicate values very close to either 0 or 255, but.... BUT... BUTBUT... we can see this value but in print this will come out as crushed black or clipped highlight. we have to apply safety margins to preserve highlights and shadows, don't we..

regarding the mid grey metering: "... Surely 18% grey metered and exposed should be recorded as that regardless of how much DR a camera has ?..."

that's 118 value. and as per my post above and both Sekonic incident light metering and in camera evaluative metering perfectly agree! identical reading: ISO 100, F5.6, T=0.40s.
if your camera does not produce accurate result with similar setup, then there is something is not right with it. do you have an access to a lightmeter to confirm your camera metering?

p.s. "...I get that we have more available EV on the mkIV. But I don’t see how that affects meter reading..."

the simple explanation is that spot metered extreme shadows and extreme highlights end up sitting right in the middle of the histogram : 118 value. That't no good :) in order to place extreme highlights back at the right end of the histogram, I have to compensate 5D IV spot metering reading by +3.6EV - lets make it +3.5 EV for simplicity.. And in order to place extreme shadow back at the left end of the histogram, I have to compensate camera readings by -3.1 EV - lets make it -3.0EV for simplicity. Please note that there is more headroom to play with highlights (0 to +3.6 EV) then with shadows (0 to -3.1EV)
 

Attachments

  • canon 6D Clipping Point Range and DR to the  MAXXXX.jpg
    canon 6D Clipping Point Range and DR to the MAXXXX.jpg
    137.2 KB · Views: 115
  • canon 6D Clipping Point Range and DR.jpg
    canon 6D Clipping Point Range and DR.jpg
    143.9 KB · Views: 118
Upvote 0
Thanks GSM,

Really helpful seeing those charts.

I did d a quick test against a white and grey backdrop in partial and centre weighted .
And partial seems to be giving a reading of 1/3rd over centre weighted . Both were recorded as just over mid grey .
So that’s no bad thing.

And I was out shooting contrasty scenes today consisting of lots of trees and a small river. Some lit in harsh sunlight , some in to the sun . So quite challenging scenes.

Definitely seeing the -3 coming into play to get the darker areas exposed correctly. And very impressed how highlights in backlit clouds were still retained and recoverable.
Also the fact that I can push shadows in processing to recover detail where the MkIII would have been horrible and full of colour noise and the grey silvery large noise.

I do now think that the meter is a lot more sensitive , as some reading were giving me exposures that were off.
Particularly where in sunlight . Again generally having to underexpose more that I normally would.
Which i’ll Get used too very soon.

Plus the 16-35 L f4 IS is very very sharp and minimal flare shooting into the sun through trees , with a nice highlight star flare .
 
Upvote 0

SecureGSM

2 x 5D IV
Feb 26, 2017
2,360
1,231
I am glad it works out for you well. Yes, Canon RAW highlight recoverability and colour magic is what keeps me in the Canon camp. Speaking of Highlights and 5D IV:

I am sure that you are aware that there is a brand new Dual Pixel RAW option available on 5D IV. it can be used for a slight focus shift,etc. Hardly an exciting stuff.. but..

Smart people dug into the Dual Pixel Raw file quite deep and found out that an extra full stop of highlights contains and can be recovered from the subframe! this feature cannot be accessed with Canon DPP but Current version of RawDigger will export separate TIFF files to work with! what it means is that for the Dual Pixel RAW enabled files (this option has to be enabled in menu, note: this will double your RAW file size), one can recover up to 1 stop of hard clipped highlights.
Essentially, Dual Pixel RAW is a combination of two frames: Main Subrame and Auxiliary Subframe.
Main subframe contains: A +B information and auxiliary subframe: only A

" ...dual-pixel CR2 file is effectively 15 bit data without much change in the format :)... Iliah Borg, RAWDigger

this can be very useful when shooting very contrasty scenes and bracketing is not an option i.e. run and gun situations, back lit subjects on bright background, sports, wildlife, BIF.... extra stop of DR at the right end of histogram.. yes, please!

once again: Canon DPP does not utilise DPRAW highlight recovery capabilities. RAWDigger TIFF export of both subframes is a nice workaround that may rescue that hero shot for you and me one day.

I strongly recommend the following two write ups on the issue. some comment on the pages are also quite revealing.

http://www.imaging-resource.com/news/2016/09/01/dual-pixel-raw-redux-forget-focus-tweaks-canons-new-tech-can-yield-an-addit

https://www.rawdigger.com/howtouse/canon-dual-pixel-mode-highlights-are-there


p.s. Canon 16-35 F4 IS L is a beauty of a lens. Very contrasty, punchy colours,.. superb! I have attached an image taken with my F4 IS lens just to demonstrate what I am talking about.
I recently stepped up to Canon 16-35 F2.8 III L lens as I need every bit of that extra 1 stop brighter aperture and regretfully had to let my F4 IS lens go.


klickflip said:
Thanks GSM,

Really helpful seeing those charts.

I did d a quick test against a white and grey backdrop in partial and centre weighted .
And partial seems to be giving a reading of 1/3rd over centre weighted . Both were recorded as just over mid grey .
So that’s no bad thing.

And I was out shooting contrasty scenes today consisting of lots of trees and a small river. Some lit in harsh sunlight , some in to the sun . So quite challenging scenes.

Definitely seeing the -3 coming into play to get the darker areas exposed correctly. And very impressed how highlights in backlit clouds were still retained and recoverable.
Also the fact that I can push shadows in processing to recover detail where the MkIII would have been horrible and full of colour noise and the grey silvery large noise.

I do now think that the meter is a lot more sensitive , as some reading were giving me exposures that were off.
Particularly where in sunlight . Again generally having to underexpose more that I normally would.
Which i’ll Get used too very soon.

Plus the 16-35 L f4 IS is very very sharp and minimal flare shooting into the sun through trees , with a nice highlight star flare .
 

Attachments

  • IMG_2947_DxO.jpg
    IMG_2947_DxO.jpg
    1.5 MB · Views: 113
Upvote 0
Looks like theere could be great potential in the b frame of dual pixel , really intesersting. Hope Capture one might intergrate it soon if other manufacturers start it as standard too. Increase in files size is negligee in the bigger picture.

I did some some very quick tests of metering modes comparing mkIII and mkIV today .
Seems very similar when looking at a flattish subject like a wall / backdrop. But when looking at a constrasty subject matter , like a check shirt the mkiv was over exposing 1/2 -1/2 .
But comparing identically shot raws , the shadows are lighter and the AWB is much superior in mkiv in a room with mixed daylight and tungsten.

Yes the 16-36 is super contrasty.
Here’s one i shot yesterday . Not bothered about sun highlight burning out was metered for mids. , as bringing it back too much leads to a heavy HDR look which I dispise. And I can always overlay a slight touch of tone/ colour if I wish.

But very very sharp and punchy, even at 5.6 , the fore ground is sharp when it focused on mid background.
 

Attachments

  • bells brae scenic 744.jpg
    bells brae scenic 744.jpg
    2.6 MB · Views: 113
Upvote 0