Jump to content

EXPOSURE: D log E curve


Recommended Posts

Hello!

I'm trying to understand the D log E curve beyond my rudimentary understanding of it, if there are any resources/information that can elaborate further on how to understand the DlogE in reference to exposure/dynamic range on a digital film sensor, i'm all ears. 

My understanding of the Dlog E curve is that it is a graphic representation of the dynamic range of the film stock/sensor. It consists of 3 parts the toe, shoulder and the SLP (shadow threshold, highlight threshold and everything in between respectively). Beyond this I feel like I'm missing pieces, is this something that is applicable during shooting or is it more so during processing. When is it appropriate to shoot in either the toe or the shoulder? Basically, Im trying to find an explanation that helps me fill in the gaps. 

Robert

Link to comment
Share on other sites


Hi Robert.  I'll give it a try, and probably explain a few things you already know along the way. 🙂

I think it helps to start with how it works for film, because it's easier to visualize.  If you look at the pictures in this post on another forum - https://www.minilabhelp.com/forums/topic/25897-control-strip-issues-v30-ra-kodak-chem-lorr/?tab=comments#comment-62349 - they're Kodak control strips, with calibrated pre-exposed steps.  The lightest, most see-through parts have no exposure; and the more exposure the negative received, the more "dense" (i.e. opaque) the negative is.  

Consider what it means that the straight-line portion is "linear" - that a measured amount of exposure will have a predictable and proportional change in how opaque the negative is.  The amount of change stays constant:  if you photography a grey card first without a filter, and then make a second exposure with an ND 0.3 filter in front of the lens, there will be a change of 0.3 optical density between the two exposures when you measure the negatives with a densitometer, as long as the two exposures are within the straight-line portion.  

The toe and shoulder are where the change in exposure is not equal to the change in recording. That's "tonal compression." If you have details that were exposed in the non-linear parts, and try to bring them back into the straight-line range, they probably won't look how we expect they would look.  For example, if an actor's skin is overexposed enough to be in the shoulder, there's little chance their skin tones will look right if you bring the levels down in post.

Digital sensors are fundamentally a bit different from film, and digital "density" is really "sensor voltage interpreted as a level of brightness."  (That sentence glosses over a lot of complex math and engineering.) Hypothetical digital characteristic curves are still interpreted as logarithmic graphs that simulate an optical density-like response, partly because it's how it was done in the past and we're used to working that way, and mostly because a log curve is easier for us to wrap our heads around than a linear representation. 

Looking at a log curve can be a handy way to see at a glance how a camera or film stock might respond to exposure.  The curve only tells part of the story, though, especially for digital sensors.  A camera's sensor might clip long before the log curve would indicate - the manufacturer might have given the curve a longer shoulder so they can still use the same curve with better sensors in the future, or to prevent issues if raw footage is post-processed far differently than it was exposed.  And for both film and digital, each color layer (for film) or color channel (for digital) has its own curve, with potentially different toes and shoulders for each color - meaning one color might clip or show noise before the others.

Intentionally exposing in the toe or shoulder is pretty rare - you have examples like Gordon Willis deliberately pushing and underexposing The Godfather so that the image couldn't be brightened at all in printing, locking in his exposure choices.  On the 2007 film Sunshine, to convey the intensity of the sun up close, Alwin Küchler intentionally overexposed the negative by up to 10 stops in certain scenes so that details were actually "burned out."  It would seem much more common to make small exposure changes to avoid going too far into the toe or shoulder, like underexposing slightly to retain highlight detail, or overexposing a greenscreen shot slightly to get a better key.

Generally, I think the idea is to know where the toe and shoulder start, and how that will translate into the final image via post, so that you can know on set when your exposure is in trouble, and do something about it before it's too late!

If you want to do a deeper dive into this, the first few chapters of David Stump's Digital Cinematography book go into much more detail.  It's not the easiest read, but it at least has more pictures than the wall of text I just wrote.  But I hope it helps!

  • Like 1
  • Upvote 1
Link to comment
Share on other sites

9 hours ago, Daniel Klockenkemper said:

Intentionally exposing in the toe or shoulder is pretty rare - you have examples like Gordon Willis deliberately pushing and underexposing The Godfather so that the image couldn't be brightened at all in printing, locking in his exposure choices.  On the 2007 film Sunshine, to convey the intensity of the sun up close, Alwin Küchler intentionally overexposed the negative by up to 10 stops in certain scenes so that details were actually "burned out."  It would seem much more common to make small exposure changes to avoid going too far into the toe or shoulder, like underexposing slightly to retain highlight detail, or overexposing a greenscreen shot slightly to get a better key.

Ah, this is an interesting way to put this as far as being able to "lock in exposure choices." I don't think I ever thought of using over/under exposure to that extreme.

 

9 hours ago, Daniel Klockenkemper said:

Generally, I think the idea is to know where the toe and shoulder start, and how that will translate into the final image via post, so that you can know on set when your exposure is in trouble, and do something about it before it's too late!

So if I'm tracking this right, it is a good idea to know where/what your finishing LUT's thresholds will be beforehand so adjustments can be made, during production.

Awesome, that helps a lot. Thanks Dan. Hope you're doing well dude! 

Link to comment
Share on other sites

9 hours ago, Robert Chuck said:

Awesome, that helps a lot. Thanks Dan. Hope you're doing well dude! 

Thanks, and likewise! 

I think you've got the right idea.  I'd say there's three categories - "this looks good," "this might look iffy on the monitor but I know I can deal with it in post," and "I need to do something about this before we do a take."  Watch enough movies and you'll see that even the greatest cinematographers occasionally have a shot here and there, where they got into a little trouble with the exposure... but it's only something that a real nerd like me would notice.

This interview with Gordon Willis about The Godfather Part II is great, and I think one of his statements applies just as well to the plethora of choices in digital cameras we have today:  "There’s a great deal of latitude in the Eastman color negative. A lot — an incredible amount. But you have to know where it is you’re going to put it."

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...