Fairly new piece of text by Bob DiNatale -> http://www.luminous-landscape.com/essays/the_optimum_digital_exposure.shtml
Tags: Exposure, ETTR, Metering, M.A.P., Noise
Tags: Exposure, ETTR, Metering, M.A.P., Noise
Article to dig into at Luminous Landscape - Optimum exposure
- Thread starter DominoDude
- Start date
You're welcome.AcutancePhotography said:
Yes, I've had LL bookmarked for a long time, and revisit it quite often. On a grey and rainy day it's good to brush up on things I think I know, until I get them to a point where they should be.
Upvote
0
Seems like a long-winded way of saying "expose high, process low." I already use this method when I have time, and I have one quibble with his explanation: the goal, for me at least, is not to put the brightest areas at 99+% exposure, but to put the brightest areas I care about at 99+% exposure. In some cases I'm willing to allow some areas to blow-out, either because they're too small to affect the final image or because it's a trade-off I'm willing to make for the rest of the image.
And let me see if I can avoid derailment of the thread: sensor DR irrelevant here because no matter how much DR you have (current sensor tech) you can still apply these principles to optimize your exposure. Can we please not argue about sensor DR in this thread?
And let me see if I can avoid derailment of the thread: sensor DR irrelevant here because no matter how much DR you have (current sensor tech) you can still apply these principles to optimize your exposure. Can we please not argue about sensor DR in this thread?
Upvote
0
Orangutan said:
Agree on all counts.
Though I have found that well over exposed very subtle evening skies cannot be lowered to accurately reflect the tonality of a more traditionally exposed shot, I have never learnt why.
wtlloyd said:
Which bit? I am always up for a good
Upvote
0
Orangutan said:
Fixed the tagging. I included DR first because it's about making the best possible use of the dynamic range at hand/captured.
Author explains a method, and I won't go into nitpicking and semantics, but I automatically understood it correctly. It always have to be about what I consider to be optimum - for example should specular highlights be blown, sun rays reflecting in water should be blown. It's a method and a way of thinking. We have the responsibility to apply that method wisely and to think about what we're doing, and why we are doing so.
Upvote
0
Photography is a complex, creative process. To proclaim axioms as if everyone has the same creative goal is more than a little <please insert appropriate term because I'm a loss for words>.
Upvote
0
StudentOfLight said:
The suggestion isn't about everybody achieving the same end goal, it is about utilizing the hardware to give the most potential to realise that personal artistic goal. Capturing the maximum potential is to the benefit of everybody that uses RAW. Who cares about jpeg shooters!
Besides, if you don't want to do it, don't, it doesn't alter the fact that an 'optimum exposure', as talked about in the article, is one that retains the most post processing potential.
Upvote
0
Orangutan said:
I think the author says something similar in this quote from the article:
"There can be some red highlight warning in an optimum exposure if the photo includes sunlit bright snow, specular reflections, etc.. In the example above you might choose the middle +0.6 EV exposure, however, the red 'Highlight Clipping' on the roof in the first exposure, +1.3 EV, could also be used. The bright snow covered roofs in bright sunlight would look natural being 100% white with no texture. Remember the expression of the Artist’s Vision is what matters."
In other words, your expression of what matters is what matters.
Upvote
0
privatebydesign said:
That is what I find, and I believe the reason that that we are losing accuracy with that level of over exposure, especially saturation.
I remember the original article produced by LL in 2002 regarding over exposure to achieve an improved raw file, and in those days they were dealing with( no surprise), digital tech from 2002, where tonal transition was poor compared with film, especially with CMOS, which was (IMO) inferior to CCD at that time. They were referring to the fact that if you did not have data recorded in the top one fifth of the histogram you were losing fifty percent of the potential information., which is correct in theory, but, doesn't deal with lose of accuracy and saturation.
It is now generally accepted that to maximise the IQ potential in a scene that includes bright and dark you want to pull the recorded data across to the right as much as much as possible in order to preserve information in the dark areas, allowing specular highlights to blow etc; it's all sound practice.
There is a benefit from over exposing a relatively small amount and then bringing the raw file back down in post. Look at the effect of pulled ISOs on SNR on some cameras.
However in the recent article, suggesting that the optimum exposure for the black cat is the one on the right is just a load of bull. You might get away with it when the majority of the tones are black, but that doesn't mean it is sound technique.
I took a couple of comparison shots to try and show what I mean, and have incorporated blue as it helps with the difference.
The first picture (ending 5) is exposed using settings determined by an incident light meter. The matrix metering of the camera was wanting to overexpose by two thirds of a stop more.
The second picture (ending in 6) is over exposed from the 'correct' exposure by two full stops. This was then reduced by two stops before converting the raw to 16 bit TIFF. Even with the pictures reduced for web display you can ( hopefully, I haven;t seen them yet of course) see the difference in the blue.
In the second two images, both at 50%, I have applied the appropriate curves and saturation, and then balanced the lesser exposed one in levels to equal them up. The loss of saturation in the blue bucket is clear, and I cannot get it back to being identical.
The final two images are at 100% of the black buckets and there is no difference in 'IQ' or noise. Overall the correctly exposed image is the better one.
Attachments
Upvote
0
Sporgon said:
Hmmmm, interesting... (Which means that I'm thinking seriously about your findings, the consequences and how I would like to apply the result to my shooting style.)
There are times when I want somewhat punchy colours (at concerts, or in sunset shots for example), and underexpose deliberately already in camera. Experimenting with Magic Lantern, ETTR and this LL-article, because I also want to have the best possibility to change my processing mind afterwards in post.
My wallet and other circumstances has put restrictions on the tools I can use in post, so processing is done in DPP and that means all or nothing; no layers, no masks, no selective editing. I fiddle with all data I have in my RAWs at a time.
Need to find a new natural balance in my shooting style that works in the future as well as now.
Upvote
0
Hi Sporgon
Thanks for the effort of posting those photos. I have the same experience that "over-exposure" or ETTR tends to compromise tonality in skies, well illustrated with your blue bucket (I use LR5). You comment that you have gained little in the black bucket. That bucket is providing you with sufficient shadow details for the image, because this is quite a low contrast scene. If, however, the blue bucket was in the sun and the black bucket in the shade, more exposure would have allowed you to capture more shadow detail. And in my case with a 5DII, reduced banding.
So its a case of deciding what is the priority (as you will already know!).
Nevertheless I'm not sure why tonality is lost in skies when exposing to the right as there should be more information at the right side of the histogram.
Thanks for the effort of posting those photos. I have the same experience that "over-exposure" or ETTR tends to compromise tonality in skies, well illustrated with your blue bucket (I use LR5). You comment that you have gained little in the black bucket. That bucket is providing you with sufficient shadow details for the image, because this is quite a low contrast scene. If, however, the blue bucket was in the sun and the black bucket in the shade, more exposure would have allowed you to capture more shadow detail. And in my case with a 5DII, reduced banding.
So its a case of deciding what is the priority (as you will already know!).
Nevertheless I'm not sure why tonality is lost in skies when exposing to the right as there should be more information at the right side of the histogram.
Upvote
0
Sporgon great work.
I think there is either less detail in highlights than claimed or that detail is so close to maxing out that we are seeing blowing out of the color even though the software can not recognize it. It would appear that blues may have a lower threshold of data to reach blowout zones?
I think there is either less detail in highlights than claimed or that detail is so close to maxing out that we are seeing blowing out of the color even though the software can not recognize it. It would appear that blues may have a lower threshold of data to reach blowout zones?
Upvote
0
Thanks Sporgon for illustrating my point, I did have a couple of early morning sky shots I had bracketed that I have used before to illustrate the difference, but your blue bucket works better!
Having read about it I think it is just the saturation levels, not the tonal values, but to have to go into the RAW file and apply tonal range based saturation curves seems a bit excessive seeing as how the black bucket didn't actually benefit from the exercise!
I find it interesting that there seem to be a small group of posters here that get their opinions from their own empirical results, we might not have the specific equation or technicality that causes the phenomena down pat, but we know what we see (though that often leads to accusations of fanboyism). It is a shame there is a much larger and more vocal group that rely almost exclusively on equations and theories to expound their opinions, generally until they own the relevant equipment and do an empirical test, and then their equations and opinions suddenly come more into line with the observations of the actual picture takers.
Having read about it I think it is just the saturation levels, not the tonal values, but to have to go into the RAW file and apply tonal range based saturation curves seems a bit excessive seeing as how the black bucket didn't actually benefit from the exercise!
I find it interesting that there seem to be a small group of posters here that get their opinions from their own empirical results, we might not have the specific equation or technicality that causes the phenomena down pat, but we know what we see (though that often leads to accusations of fanboyism). It is a shame there is a much larger and more vocal group that rely almost exclusively on equations and theories to expound their opinions, generally until they own the relevant equipment and do an empirical test, and then their equations and opinions suddenly come more into line with the observations of the actual picture takers.
Upvote
0
Belated thanks from me also, Sporgon. Your shots were very useful.
I hope I don't have to feel like I belong in any special camp - other than the one trying to learn more, and the ones trying to find knowledge and methods to share.
I love my Canon gear like I love my left arm - it may itch at times, but I prefer to scratch that irritating itch over cutting the entire arm off.
I hope I don't have to feel like I belong in any special camp - other than the one trying to learn more, and the ones trying to find knowledge and methods to share.
I love my Canon gear like I love my left arm - it may itch at times, but I prefer to scratch that irritating itch over cutting the entire arm off.
Upvote
0
