Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Basic rules of exposure

David Mullen ASC said:
But there are plenty of examples where some minor clipping is allowed and even necessary in order to not expose important detail in the bottom noise floor at the other end.
Click to expand...

Obviously, fully agree. In this sense the idea of raw workflow is to retain maximal amount of data "almost everywhere" as the mathematicians tend to say. (But this is now getting more complex than Bob asked for.) It's going to be interesting to see the implications of the 16 stops DNR of Monstro sensor on all this.
 
Lauri Kettunen said:
It's going to be interesting to see the implications of the 16 stops DNR of Monstro sensor on all this.
Click to expand...

Well, to be blunt, but what it will mean is that we will finally get something like what we've already had for decades shooting color negative film...

It will also mean more flexibility in misexposure, i.e. more latitude.
 
Sidney thought with RED expose to preserve highlights (generally speaking) because information is better stored in shadows. In post drag up shadows and bump up contrast to make a nice roll off in the highlights - much nicer than if sensor had clipped highlights.

but now wise people say - expose to the right because sensor uses half of resolution on brightest highlight...

In film its opposite - more information hidden in highlights. less in shadows, Sidney learned from sitting in telecine next to colorist pulling down blown out skies and making them blue (and noisy - but blue none the less). therefor "expose to the right" makes sense for Sidney when shooting film...

but now wise people say - expose to the right on RED because sensor uses half of tonality (resolution?) on brightest highlight...... now Sidney is confused...

Sidney knows that some times blowing out highlights is fine and all that... but where to center the exposure - left or right on histogram?

Or should Sidney light as flat as possible?

stills from David´s film "Manure" looked like he aimed as close as possible for final look (or contrast ratio) in camera?


best,

Sidney
 
I am not tempted at all by the technique of overexposing to lower the noise level.

Doesn't look natural or "right" to my eyes.

Above all, i think the most important thing about exposure is to use a chosen "look" (LUTs) for monitoring (EVF and LCD), and test the camera in most common and extreme situations and see what happens in your Post monitor and draw your own conclusions depending on what you want. (Then test again...etc)

It's the best unscientific way i found to achieve predictable and constant desired results.

I am not sure i'd go the light meter route because RED CMOS doesn't react like film at all at first, but with specific LUTs, the final result can look superb.

Antoine
 
Graeme Nattress said:
The skill of exposure is knowing which highlights to protect, and which to let go. If there's no detail in the specular itself, all we need to do is ensure any roll-off is pleasant. Aiming your exposure to what you think is the most important part of the scene is the key to making every camera work. Although we're alway making cameras better to reduce exposure constraints, I don't see that the main skill of cinematography - lighting - and how it works together with exposure, is going to go away soon.

Graeme
Click to expand...

well said... especially from someone so technical...
 
Thanks Stuart. I'm only technical when I need to be. I much prefer to be artistic. But as art illuminates science, science informs art.

Graeme
 
David Mullen ASC said:
Well, to be blunt, but what it will mean is that we will finally get something like what we've already had for decades shooting color negative film...
Click to expand...

:-) Indeed, did not thought of that as never had resources to work with film. I've climbed the other way around and struggled with the technical limitations of video camcorders and thus tried to learn to push the equipment to their limits.
 
Fantastic thanks, thanks to everyone who is contributing. I tink tgere are many people on the boards who are wrangling with RED exposure. I say keep it simple because i know there are some great and technical minds out there, when the info is condensed down and the message is simple, its great.

Asking a basic question about exposure is also quite exposing, if you know what i mean. I get the feeling there are plenty of people out there who would have liked to ask the question, but wouldn't want to for one reason or another. :biggrin:



Lauri Kettunen said:
Bob,let's keep it simple and say you need only to make sure you store maximal amount of data and focus on colors and tones only afterwards. For this, set the raw view on and set the exposure just below clipping, which is to say just below the point the first traffic light turns on. This is the main idea of raw workflow. Some people say you may let one channel to clip a bit and the image is still ok. Although one channel clips, this way you get more data stored on the others. So, it's a compromise and matter of taste.

Sometimes one or two channels may dominate. If there's enough light you may want to use filters to get a better balance. This is to say, get a proper amount of (digitized) data stored for all channels.

While you shoot you probably want to check colors or how dark are the dark parts of the image. For this set either the redspace or rec709 on. Still I find it better to exploit the raw view to set exposure. For, then I have more accurate meters to control exposure just below clipping.

Summing up, the whole point is to store as much data as possible. Eye's response is not directly proportional (that is linear) to the amount of light. Films are manufactured to respond somewhat the same way as an eye, whereas CMOS sensors do respond linearly to light. That's why one has to exploit a gamma curve at some point to transform the image obtained from a CMOS sensor to match with those captured by the eye. The raw workflow means you need not to fix the gamma curve while you shoot but instead you may leave that to the post edit. And you are in best position in edit if you have maximal amount of data available when you start to post process the data. This is why the rule of thumb is to shoot on the right.
Click to expand...
 
Musings of a frustrated, non CF card carrying, shooter.

Musings of a frustrated, non CF card carrying, shooter.

We've had our Red now for only a couple of weeks and, as we have no CF cards yet, I'm unable to test my exposure theories as they apply to the camera. This thread has been extremely interesting and my theory of meter for ratio and use the histogram for stops looks fairly solid - untested by me though it is.

As I started in stills before moving to cine and video 25 years ago, I'm not fazed by tight DR. Shooting transparencies sorta makes anything else look easy. Many years ago I spent a great deal of time getting my head around moving my Zone System knowledge to roll film and I learned the first thing you have to accept is that, in order to get the exposure you want, some things simply have to fall off the edge of the world. It really clarified my thinking and, I must say, changed my framing as the decision of exactly what I wanted exposed and what I could live without helped me see that, often, much of what was in my frame was extraneous.

Anyone who ever shot Kodachrome 25 knows that blacks and whites do not have to be bad (man, were those blacks fantastic!). It depends on the quality of white you get and (again untested) I can't help but think this is a post issue (or opportunity?) I once had a Schneider 90mm wide angle and the whites were unbelievably creamy and luscious on transparency stock.

I am yet to fully understand how "good" or "bad" the clipped highlights will look from this camera (the purple sun thing bothers me, why can't that go white instead?) but I agree with Graeme; if there's nothing in there you want to see, why is it a problem if it's white? And blacks shouldn't be noisy so long as you're not trying to rescue detail that you haven't exposed for. Expose for black and get black. Or am I wrong?

The one thing it seems I'm going to have to change from my film/Zone System thinking is the concept that there is no such thing as "correct" exposure. It seems from reading this thread that, with the Red's sensor and compression algorithms, there is indeed a point of exposure one could call "correct", with zone 9 whites near the right of the histogram, leaving enough room for roll off to a nice shoulder in post.

Great discussion, loads of great info guys. Can't wait to get those cards and actually shoot something!
 
Ok so, after extensive noodeling and research on various threads and my own experimenting. Here are some conclusions i have come to with regards exposure on the RED and the tools the camera offers and how they serve us.

1. The histogram, stop lights and stacked color chart are very useful. The traffic lights will tell us when we are clipping. Set your 1 button to RAW view and flick between that and your chosen gamma settings i.e REC709 etc.

2. With regards RAW, well here's the rub. There is a slight gamma curve applied to it, so it's not strictly RAW, the gamma curve serves to give color in the RAW image and it shouldn't affect overall exposure too much.

3. False color, use it but be aware when using false color and viewing a RAW image that it is not a true representation, i.e it is not calibrated to the RAW image. So the best option is to use the stacked colors ( it's colors equate to the false color chart but they are acurate ) they will tell you when you are near clipping a RAW image.

4. The histogram is a true representation of the RAW image.

Any thoughts?
 
I like the way they light Boston Legal. I think they shoot on film. The whites looked washed out in many shots but that doesn't really bother me. I guess there is no 'proper' exposure.
 
problem with noisy blacks is in post production. or noisy green - it affects the edges thereby causing problems for compositors.

noisy blacks is an issue if you underexpose a face little to save the sky from burning out for example. then in post you pull up face to "normal" exposure and you get noisy face... expose face right blow out sky - nothing you can do in post to make it come back - much nicer to blow out sky in color grading.

ofcourse if you have full budget and crew you just add the fill light needed...

so you could say there is a "right" exposure for a given scene.
 
A shoulder to fry on

A shoulder to fry on

First of all, for the most part, light for story/mood. However, there is a specific caveat concerning skin tones off key: the inherent noise characteristics of the RedOne manifest mostly in Zones 2 & 3, therefore, IF, and only if, I need to read attractive skin tones in the shadows (lead actress for instance), I will add an extra stop of fill even if it "feels wrong" for the shot. Especially in a closer shot of the face, the noise can be distracting.

To directly address examples on this thread and exposure issues in general when setting a stop on the RedOne: I agree that considered choices about what you are blowing out to allow ETTR can be give you a better "neg" in grading, just be aware that the transition from the clip to genuine detail can be steeper than film negative so beware of things like a face against a clipped window or sky.

If you have uncontrolled lighting in the field and a 14 stop range to capture there is only so much you can do. Don't forget to be creative, its what we get paid for. If you're really screwed go back to the basics - place and fall - if the biscuit is the skin tones then get them in the fat part of the exposure and live with whatever goes away. Getting crappy skin tones because you're trying to avoid a clip or a noisy bottom end is rarely the right move.

One thing I have noticed more and more lately in watching the work of DPs I really respect is that they often use silhouetted subjects to draw attention to body language and dialog rather than sell out the setting of the scene to hold skin tones. Obviously that would be a poor strategy for some scenes, but worth considering if your lighting/budget/time is limited and the compromise exposure leaves you with a less than a pleasing image.

Just my .02

Cheers-
 
good tips all....
very illuminating(pun intended)


it seems as though the "protect the highlights" slogan was a little misleading

being that half the tonal values are represented in the brightest 1 stop of overall exposure,and a lot of the shadow noise comes from lack of tonal values represented in the darker areas of our shots

it seems protecting the highlights...or the rolloff of them at least, is already done for us by the nature of the linear color space.
 
thanks for the thread, very interesting. Also sorry for hijacking it.

As for my monitoring, I definitively love to switch between Redspace and RAW, and see how my exposure falls. My question is just in terms of shortcuts and key buttons. I do use the the option with the red edge for focusing purposes. But I can't seem to be able to use the keymap option to have both buttons assigned to both the Colorspace (the one with all the colors for seeing where you clip) and another button to have the the red edge?

Is there any way to have the three EVF buttons to be assigned to 1X, Red edge, and color clipping mode? I love those three modes so much, but it is so annoying to have to go into the menus just to change the mode.

Thanks
 
But Jeezey... not much to complain about nowadays with video. Use to be that if you got four stops of latitude you were doing good. Now it's up there with film stock though not the same. You can light badly and only get three or four stops out of high speed film stock still. No subsitute for proper shooting.
 
muffinny said:
thanks for the thread, very interesting. Also sorry for hijacking it.

As for my monitoring, I definitively love to switch between Redspace and RAW, and see how my exposure falls. My question is just in terms of shortcuts and key buttons. I do use the the option with the red edge for focusing purposes. But I can't seem to be able to use the keymap option to have both buttons assigned to both the Colorspace (the one with all the colors for seeing where you clip) and another button to have the the red edge?

Is there any way to have the three EVF buttons to be assigned to 1X, Red edge, and color clipping mode? I love those three modes so much, but it is so annoying to have to go into the menus just to change the mode.

Thanks
Click to expand...

Not sure don't use the RED edge myself, the falsecolor monitor which i think you mean when you say colorspace is pretty good, but it's a bit redundant as it's not calibrated for RAW. You can use it in redspace to get an idea of the hottest areas and then work out from the bar graph which areas are peaking, but it's a bit of a hassle and it would be nice if the false color meter was a direct reading from the RAW image when in RAW mode.
 
chuck colburn said:
But Jeezey... not much to complain about nowadays with video. Use to be that if you got four stops of latitude you were doing good. Now it's up there with film stock though not the same. You can light badly and only get three or four stops out of high speed film stock still. No subsitute for proper shooting.
Click to expand...

I'd still shoot film every time if i could though,eh chuck?
 
You must log in or register to reply here.
Back
Top