Eos 5D Mk IV--End of ETTR?

[YuengLinger]

I'm finding even +1/3 is blowing more highlights than I like.

Sort of found some confirmation on Art Morris's site where he finds images are exposed correctly even though he isn't pushing the histogram up to the right side as on previous models. I'm also still getting used to this.

(He also claims that the 5D IV has better IQ than the 1DX II!!!):

"As compared to the EOS-1D X Mark II, Canon’s blazingly fast flagship body, the 5D IV offers lighter weight and lower cost. Not to mention far superior image quality...

The high ISO performance of the 5D IV at least rivals that of the vaunted high ISO king, the EOS-1D X Mark II. Control of fixed pattern noise with the 5D Mark IV is outstanding."

http://www.birdsasart-blog.com/2016/10/12/why-i-love-my-5d-mark-iv-the-beyond-incredible-imaging-sensor-and-so-much-more/

Otherwise, not finding much info about metering, exposure tips for the 5D IV. Thoughts and/or links would be appreciated!

 

[Mt Spokane Photography]

Individual cameras respond differently, so yours may or may not be typical. The ISO sensitivity is set during manufacturing, just as autofocus is set. There is a tolerance. Generally, tolerances are set to allow boosting the exposure at least a little.

 

[YuengLinger]

Individual cameras respond differently, so yours may or may not be typical. The ISO sensitivity is set during manufacturing, just as autofocus is set. There is a tolerance. Generally, tolerances are set to allow boosting the exposure at least a little.

Have you used the 5DIV?

 

[Sporgon]

I'm finding even +1/3 is blowing more highlights than I like.

Welcome to low noise and on-chip ADC ! The answer to your question is..........Yes !

 

[YuengLinger]

I'm finding even +1/3 is blowing more highlights than I like.

Welcome to low noise and on-chip ADC ! The answer to your question is..........Yes !

At least there is an explanation! From the 20D to the 60D to the 5DIII, that was a lot of shots ETTR, sort of training the brain to expect it would go on that way forever. Taking a little time to adjust.

 

[[email protected]]

I find that exposing to the right with the 5D4 is usually causing more problems than it's solving, especially as I don't sit there and think about changing settings between each frame, as I'm usually rushing to keep up with my subjects.

This is actually pretty nice. No fuss, no muss. But it can be confusing when I'm operating with two bodies. I often use the 7D2 at the same time, and I find myself either ETTR on both or neither, and one or the other is suboptimal, depending on the body. I bet the 7D3 exposes much more like the 5D4 and the 1DX2, so things will become consistent again.

 

[stefang]

I'm finding even +1/3 is blowing more highlights than I like.

Sort of found some confirmation on Art Morris's site where he finds images are exposed correctly even though he isn't pushing the histogram up to the right side as on previous models. I'm also still getting used to this.

Doesn't this mean that the machine does ETTR all by itself? So no end of ETTR, but end of manual ETTR.

 

[SteveM]

I found +1/3 on the mklll and zero on the MklV

 

[YuengLinger]

Thanks for the replies so far.

Anybody also notice that the histogram needs to be interpreted a little differently too? In my case, I'm used to previous bodies slightly "exaggerating" what was happening with the RAW file. What I mean is I could take shots with the highlights pressed right up to the right side of the histogram, but still have only the specular highlights blown out. With the 5DIV, if the highlights are even slightly touching (like a point the size of a grain of sand!), or if a fair amount of midtones are to the right of middle, I'm looking at an overexposed image.

 

[Mikehit]

That is interesting.
The histogram traditionally works off the jpg and that in turn depends on what jpeg you have dialled in. Note that even if you switch to raw only, it still uses the last jpeg style you dialled in (the same jpeg is displayed on the camera LCD).
I recall that about 18 months ago someone developed a way of showing a histogram based on the raw data - I wonder if the 5DIV does this?

 

[Mikehit]

One other question - is the image overexposed on the computer after downloading or on the LCD? Have you been able to recover highlights in post processing?

 

[Loswr]

I recall that about 18 months ago someone developed a way of showing a histogram based on the raw data - I wonder if the 5DIV does this?

Do you mean an ML module? Sounds like you might be referring to universal white balance (google UniWB), but that's much older than 18 months.

 

[Luds34]

That is interesting.
The histogram traditionally works off the jpg and that in turn depends on what jpeg you have dialled in. Note that even if you switch to raw only, it still uses the last jpeg style you dialled in (the same jpeg is displayed on the camera LCD).
I recall that about 18 months ago someone developed a way of showing a histogram based on the raw data - I wonder if the 5DIV does this?

That's my understanding. I don't spend a lot of time using the histogram on the camera (since I am shooting RAW) but pay more attention to it during post. With the Canon bodies I've used I've decent success recovering "blown" highlights, within reason of course.

 

[LDS]

I'm finding even +1/3 is blowing more highlights than I like.

AFAIK ETTR is image-dependent, it's up to you to decide how much you can push the highlights (for a given image) and how much clipping is acceptable (again, for a given image).

What kind of metering are you using? What is the setting of "highlight tone priority"? If highlights are already placed enough to the right, overexposure will inevitably clip them.

For ETTR I usually use spot metering and "highlight tone priority" off (obviously), and I would really like the 5D had the exposure level indicator the 1D and 7DII have, to quickly assess an image dynamic range and understand how much I can push highlights.