The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Optimum exposure

GMB

Active member
I came across this article on LuLa about optimum exposure. Basically what it says is to expose so that the brightest point of the scene is 99.5% white in the raw file and then correct in post production by bringing down exposure. Thus, if you photograph a black cat on a black carpet you would severely overexpose to have a raw file where everything is white (just short of blowing highlights) and then bring it back to the correct exposure in post.

The idea behind is that you capture more information. (Basically it seems to be taking ETTR to the next level.)

Intuitively it seems right and I have been able to myself to bring overexposed but not blown out shots to the right exposure. But I am curious what folks who know more than me about digital capture think about this.
 

dchew

Well-known member
Georg,
There is a 10+ page debate going on in a Lula thread with some very smart people in it, and they cannot seem to agree on much at all in regards to this topic! As most things, I think it depends on camera, scene, software, and the pilot's ability to use the software.

If the scene is such that you can reliably tell from your camera's histogram that you have not blown any important highlights, then I think it makes sense to go to that right edge, with a few caveats.

I find the histogram's reliability in the highlights to be somewhat scene dependent. Maybe simply because of the JPEG compression that the histogram is based on, maybe something else, IDK. I use two cameras: An IQ180 on a technical camera and a Sony a7r. In both cases I still bracket if the scene has a wide dynamic range.

There are two ways to shift the histogram. Increase the amount of light hitting the sensor or increase the camera's ISO. The first way makes sense to me (improving the S/N ratio). The second is camera dependent and ISO dependent. In some sensors, there is a benefit to increasing the ISO to a point. That point depends on the sensor and camera.

I'm not so sure about the black cat scenario. To bring that image back down in the raw software I have to drag at least the exposure slider down, and probably others as well. These controls affect different parts of the tone curve in different ways, and each software is of course different. I find it difficult to believe that moving any software slider that far will produce the same result as an in-camera exposure that puts the black cat closer to where it "should" be. Exposing so the cat is light gray then dropping it in development may end up with a more pleasing result, or it may not! Sure, you could fiddle with a variety of controls to make it look acceptable, but now I'm spending a bunch of time, and for what benefit? So I think there is a practical limit to ETTR, regardless of whether the S/N ratio is better or worse in a case like the black cat.

I keep going back to these things are systems; from capture to print. Any "rule" has limits and the limits depend on our individual system. And I am the most inconsistent link in my system chain!

Dave
 
Last edited:

jagsiva

Active member
I read the article and while the first half sounded quite radical, his examples in the second half do not have exposure nearly to the extreme that he was recommending earlier.

I find that if I hug the right side of the histogram slightly on the IQ, most outdoor exposures are fine. If I need something more precise, I always use a spot meter, get a few readings then decide for myself.

Bottom line is that the LCD is reading out some form of converted file, and this will be different from camera to camera.
 

Shashin

Well-known member
But once you bring the tonality down to normal, you lose the data anyway (although I don't know how much data you are getting if you are simply scrunching it up to one side of the histogram anyway). Which kind of begs the question of how much data is needed to make a continuous-tone image--answer: a lot less than people realize.

And how you get back to something normal is really important or you are going to have color problems, as evidenced in the authors images. Photography is not simply a data problem, but one of tone and color reproduction for the human visual system. And ultimately, subjective.

A good photographer--and this worked in film (he is not actually saying anything new, just badly)--would control the exposure depending on the object/scene. But it is also done with rather deep knowledge of the process (what works on the coast of Maine does not work on the plateaus of Tibet).

Yes, exposure is very important. No, it is not that important. It is not simply a data problem, although "data" is the new black. There will always be zealots that "find the truth." The good ones know how to make money with that.
 
Last edited:

Wayne Fox

Workshop Member
i found the article was just rehashing old concepts that have been around more than a decade now, and I really didn't see anything that added or changed the concept of ETTR based exposure, other than trying to wrap it up with a nice wrapper/slogan.

I don't see how it takes ETTR to the next level, because the entire concept of ETTR is to expose as far to the right as you can without clipping. Using that concept the black cat on black background would do exactly what it did. As for the idea of getting to 99%, sure. Yep. Just expose to the right as far as you can without clipping. Same thing. What it measure later doesn't really matter. Frequently you can't help but clip some small parts of a scene so you have to evaluate it. But to worry about getting from 96% to 99% to make it perfect ... why? Don't think you'll see any real difference with the final data.
 

synn

New member
If you have a black cat that needs to be photographed and is currently sitting on a black carpet, grab it, place it on some white marble and shoot. ;)

I don't overthink ETTR. Be careful not to clip the highlights and you'll be able to bring everything up to your liking in post. Most modern cameras and modern post processing software are able to do that.



This was shot as as single exposure. No bracketing, no blending, no nothing. I just made sure the lights didn't blow out (too much), then brought the shadows down back in post.
 

satybhat

Member
I generally try and over-expose by half to one stop. I feel that that suffices - classical ETTR. If I split any more hairs, I might go bald... and I would rather take my compositions to the next level than the exposures.
Just saying.
 

rga

Member
I too am also wary of one rule fits all. Perhaps I'm wrong, but I do three things:
1. Unless intentional I rarely clip highlights.
2. I set the camera to show clown highlights at a value of 252 rather than 255. This acts as more of a reminder to watch the highlights, some of which my brain might not see when I am engrossed in the moment, and
3. I tend to expose digital sensor capture as I did positive film: overexposed highlights are much more difficult to recover than under exposed shadows (expose for the highlights and let the shadows fall where they may).

Kind of my starting point. Wondering if I'm missing something?
 

Wayne Fox

Workshop Member
I too am also wary of one rule fits all. Perhaps I'm wrong, but I do three things:
1. Unless intentional I rarely clip highlights.
2. I set the camera to show clown highlights at a value of 252 rather than 255. This acts as more of a reminder to watch the highlights, some of which my brain might not see when I am engrossed in the moment, and
3. I tend to expose digital sensor capture as I did positive film: overexposed highlights are much more difficult to recover than under exposed shadows (expose for the highlights and let the shadows fall where they may).

Kind of my starting point. Wondering if I'm missing something?
If your missing it then I am too, pretty much how I do it including setting the warning at 252.

I don't overthink ETTR. Be careful not to clip the highlights and you'll be able to bring everything up to your liking in post. Most modern cameras and modern post processing software are able to do that.
Yeah, I think it's easy to overthink it as well. And unless there is a ton of dynamic range in the image, not getting the histo all the way right probably won't show much difference in the final file. Trying for 99% just seems fraught with complexity and unnecessary effort that may be better spent focusing on composition and important photographic elements of the image ...
 

GrahamWelland

Subscriber & Workshop Member
I've found myself going old school and using my Pentax / Zone VI calibrated meter with zone scale and exposing for zone VIII/IX in the highlights. I know that I'm probably at 90-95% max (if that) but I can safely ignore any blinkys from any camera I use this way. Just remember to ignore specular highlights and let them blow.

Even my IQ260 nanny says I've over exposed but nope ...

Life goes full circle I think.
 
Last edited:

alvinjamur

New member
The critical thing to take note of is the state of the image ie., Linear .vs. Non Linear and what transformations are applied at each stage.

As an Astrophotographer dealing with HUGE dynamic range but compressed into very tiny linear space there are certain transformations of data u can and cannot do when the image is linear.

Here is a very good youtube video (using the BEST image processing software on the planet : Pixinsight)....

https://youtube.com/watch?v=lWXj6Pc_hog

Then ask yourself the question about why shooting right is a precursor of an optimal exposure.

aLV
Al Vinjamur Photography
 
It really depends on your camera I think... although I've started using a Pentax 645Z only recently, I've already confirmed what I read about this camera before I bought it - it doesn't seem to handle overexposure very well.

You can use live view to get a sense of when highlights will blow out since the blinkers are based on Raw data, but you may get the false impression that you're shooting severely under-exposed, if you don't keep in mind that the Z has like 5 stops of usable shadow push at ISO100.

I think this may be due to how the mid-point of the camera's DR was placed up high on the DR scale, but Pentax should really integrate a "full DR" mode that lets you see the whole range while shooting and adjust the metering and preview to compensate. As it is, I have to shoot the opposite way that I was always used to with ETTR: aim low and push in post.
 

miska

Member
Since I am using a tech cam (with copal shutter) and a digital back, I am finding exposure a bit more difficult. The shutter only has 1 stop increments on exposure time, and sometimes I am really at the limit. So either it blows out seriously highlights, or then looks under-exposed. I also have the impression that the over-exposure warning on the IQ160 are a bit conservative (or perhaps Lightroom is good at recovering blown out skies - that's usually where the over-exposure is).
So I'm now trying to learn exposure again - I think I need to not worry too much about blinking skies (or clouds). But of course blowing them out too much is not a good idea...
 

GrahamWelland

Subscriber & Workshop Member
miska,

In my experience the IQ backs are pretty conservative when it comes to blinkies (and remember you can set the level) as they don't look at the overall exposure and seem to look only at any one particular channel. I have found through tethering to a MS Surface Pro in the field running C1 Pro that I get a far more accurate view of what really is unrecoverable and when exposure warnings are displayed on the back (or in C1 Pro).

I used to be conservative and avoid any warnings unless they were obvious specular highlights which would result in unnecessarily underexposed images. I was always a slave to the histogram too. Now I've found out pretty quickly the real dynamic range and recoverability through the realtime field comparison of what the back warns of & histogram, and what C1 can fully recover or deal with minor adjustments. Thus I don't worry too much if I set the warning levels to 250 or even 252 that I partially indicate clipping in any one channel (i.e. blue) because I know what will result in a good exposure. It has been very useful also applying some of the highlight recovery or exposure compensation within C1 at capture time and comparing it to the level of warnings that the back displays - you'd be even more surprised than you think about just how much leeway is available in well captured images exposed to the right.

You are correct about cloud blinkies too - it's worth doing your own tests to identify just how much recovery is available. It's made my exposure experience a lot more satisfying recently once I learned to relax about what the back displayed. Obviously if you are seriously clipped right up against the right side on a channel, or especially two, then you're not necessarily going to be able to get a decent color neutral image recovered out of it.

I'm not a huge fan of shooting tethered out in the field but it has been a very valuable exercise in retraining my eye at capture time vs working with either under or over-exposed files when I process the images later. I don't need to do it now but it was a useful process to go through.

Every time I shoot with the Surface Pro attached to my IQ back I secretly curse Ken Doo for making me use a windows machine (and Apple for not having a tablet with USB) :ROTFL:
 

miska

Member
No ! I will not be tempted by tethered shooting in the field. I will not. I will not :)
Thanks Graham !
 
Top