Welcome to our community

Be a part of something great, join today!

  • 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 :)

Exposing correctly

There is certainly more to be found in this image... a little noise is not the end of the world. Everything is a balance. Take you foot of the brake a bit. If you grade to the look you want and find that you have a bit of noise because of your exposure... use some NR. There are a lot of tools in the bag. Don't be afraid to use them. Using NR is not a sin. If this is the look you want... then ignore everything else. Your call...

Jim

For $99, you can use Neat Video. That will do wonders for a lot of shots, just have to find the right balance of noise vs. plastic look. I agree with Jim, as long as the noise is not objectionable, i.e. the pattern doesn't become too apparent, then you may be fine. Everyone has different thresholds for this kind of stuff. Some want ultra clean, some don't mind some noise. But ultimately, you want to be in full control of your image if possible. Know when you'll get noise, and when you won't. That will help a ton.
 
Another recommendation for Neat Video. I had a shot that I thought I would have to throw away. I ran it through Neat Video and it has become an "oh wow" moment from everyone who sees it. Best $99 I ever spent!

As for the topic, I only use in-camera exposure tools and they work very well. Whenever I get an overexposure warning I look for something in the shot that's causing it. As David says, if it's a light or small window, let it go, better to preserve your overall shot. But if it's a highlight on the face of your talent, stop down!

And as Jim says, sometimes a shot will require you to stop down more than you would like, but that's the deal. At least you will only have a few challenging shots in post rather than every one.
 
Hi Jamie,

Looking at your histogram in the photo, you are underexposed my friend, my guess is around 2 stops. 2 stops under would probably be an issue with just about any format.

But my glass is always half full, because you are a way better shooter and Red tech now than you were that day! Welcome to Cinemotography - where all of the best lessons learned are from f-ing up on shoots! I have learned more screwing up on shoots than I did at 4 years of film school.

Your feature will go much better!
 
Thanks for all the advice guys. I've purchased Neat Video.

I went over to my collegues house tonight and saw this thread on his IPS panel based monitor after trying to grade on my TN based one. Safe to say I'm getting a shiny new HP display for Christmas.

I'll pick up this thread when I've got my new monitor, because this monitor is just useless for grading. There's no detail, and everything has a really plastic artificial feel.

Jim, David, Steve, Rich, Xioasu, Joel, and Liam, thankyou for all your help. The shoot in February is very important to us, and your help is invaluable.
 
Jamie... this is not intended to be a color grade, just a guess of what we think you can find in the image without needing much (if any) NR. An R3D is pretty flexible... the joy of shooting RAW.

Getting to know what you can do in post certainly helps one make better exposure decisions. Shoot... grade. Shoot... grade.

Best on your project.

Jim

1261406800.jpg
 
Last edited:
The before and afters should be displayed together. There are so many new objects appearing in the second one that I'm getting dizzy switching back and forth between the two posts.
 
Jamie... this is not intended to be a color grade, just a guess of what we think you can find in the image without needing much (if any) NR. An R3D is pretty flexible... the joy of shooting RAW.

Getting to know what you can do in post certainly helps one make better exposure decisions. Shoot... grade. Shoot... grade. Best on your project.

Jim

1261406800.jpg

Thank you Jim. I'm just about to have another crack and that's tremendously helpful. It is incredible how much detail can be pulled up. Like you say (it couldn't really be put better), it's about keeping the balance between nicely preserved highlights and staying away from noise.

Everything is dependent on the scene, what would work best, and trade offs must be made in order to 'hit' the desired exposure. In this particular shot, I certainly would of blown out the window (had I been able to!) and kept things futher away from the noise floor. It has certainly been interesting though, and I'm very glad now that I did preserve all the highlights and started this thread.

Thanks for the kind words about the film, I'll keep people posted if you'd like to see how it goes =]
 
As Jim and others have mentioned - testing is the key. Of course there are real world limitations on testing methodologies so making good guesses in a hurry is often required. Below I will describe one approach, keep in mind that with experience you will learn to reference many exposure indicators in making your choices.

Ignoring format (RedOne, F23, HVX200, whatever) use basic lighting and cinematography practices to stage your elements and set your lighting.

Frame up the shot, switch on false color metering (in view RAW) and spin the iris up and down while looking at where the primary green false color contour lands (this is nominal 18% gray / 50% exposure). Think about your expectations for where elements in the image "should" be in the exposure curve. For example if you have a person with medium skin tone you might want to open up until some of the face reads false color green, and some false color pink. This is your first guess for stop.

Now scan around that false color version of your scene and see what is near, at or over clipping (yellow, orange & red contours). As David and others have mentioned, some items should clip, other elements need to hold detail. If you have placed the most critical elements at the exposure level you desire and nothing you care about is clipping, you're halfway there.

The second check is shadow levels. There is a -3 stop false color contour that looks kinda blue. Note your projected stop and then spin the iris to see what is above and below that contour as your ride exposure up and down. AS a general rule (lots of creative exceptions are valid, this is just a rule of thumb) anything below that -3 blue contour is at risk of being noisy. My usual strategy is to add fill to any elements that would look lousy if noisy. This can result in a flatter look on set but a simple tweak of the curve in RedAlert, Scratch, etc can easily restore the intended contrast ratios.

While I consider the Mysterium sensor with Build 20+ a 9 stop+ DR camera, I try and keep elements of interest in a 6 stop window (3 over and 3 under 50% exposure) so there is plenty of noise free detail and pure chroma for grading. YMMV.

Cheers - #19
 
a future reference bit of advice...

I find that this is a very light-hungry camera - (not a complaint but an observation)...

...and you'll find that the kicker, rim light, hair light - whatever you want to call - is your friend. Don't leave home without it. Even in quick in-and-out, fast set-ups, it helps a lot to have a litepanels micro or roscoe lightpad or best of all, one of Richard's handy, portable Coollights that can run off a RED battery if you buy the V-lock mount - that's my favorite. It doesn't add a lot of weight, is flexible, and easy to set up, even without a crew.

Documentary-style situations that I may not have lit in the past, when shooting RED - I light them. That separation between the foreground and background helps tremendously, and your lighting setup doesn't have to be heavy or cumbersome. But it's pretty indispensable.

When people with dark hair are underexposed, it's easy to lose detail in the hair - makes 'em look like they're wearing a helmet, not hair. A bit of backlight will cure this. I would say that I almost never shoot humans indoors without it. It just looks good.
 
I'm about to debut Obscure Knowledge Test Lab. Now before anyone gets too excited, let me say it's going to start off slow, but basically I will provide some R3D test files for people to take a look at under different conditions. I've got a test subject who is very cooperative and I just upgraded my web site hosting account to unlimited bandwith so people can download the R3Ds. With each R3D will be information about lens used, camera settings, F-stop info, filters, type of light, etc. People can then play around with the files, get hands on with the processing. Hopefully this will continue once Epic and Scarlet comes along. The place to find these files and info will be here.

http://www.obscureknowledge.com/MEDIA.html

Look for a test file to appear soon and an official announcement of the site.
 
Steve that's fantastic, it will be so helpful to us. I hope you manage to get some up in the coming weeks. Much respect.

Hopefully in the next few days, maybe sooner. At least a test file should be available very soon.
 
Hopefully in the next few days, maybe sooner. At least a test file should be available very soon.

Excellent, it would be an invaluable resource for those who are coming from a video/film background like us. It's very difficult to get hold of a good range of R3D's unless you go out and shoot yourself which is expensive (at least in England) to do.

You should set up a way to donate towards server costs maybe if things really get going.
 
Excellent, it would be an invaluable resource for those who are coming from a video/film background like us. It's very difficult to get hold of a good range of R3D's unless you go out and shoot yourself which is expensive (at least in England) to do.

You should set up a way to donate towards server costs maybe if things really get going.

There used to be more sites with R3D files but they seem to have gone by the wayside lately. I know Jason H has a site, but otherwise you have to dig deep to find files. What I'm hoping to do is give folks some files with reference information so they can have a chance to play with actual footage without the expense of renting the camera. The key is the reference info. Knowing what stop, what lens, how the lighting was setup (early tests may be crude by the way, I'm not a DP but I know some good ones who will eventually help with these tests) and other information that will be valuable in judging the images.

If I can do a small part to help people who are trying to figure this stuff out I'm happy to do so. If it gets costly or expands to the point where I need to offset the cost somehow, I'll explore donations but for now I'm just trying to help. Having actual R3D files is far more informative than JPEGs or even TIFFs. You need to see what's actually recorded.
 
Would something like this give you guys enough enough about each test shot.
Test-Sample-2.jpg
 
Would something like this give you guys enough enough about each test shot.
Test-Sample-2.jpg

That is looking good- thanks for doing this. :)

One thing that would be cool to add, if it wouldn't be to much trouble- could you provide Foot Candle read outs of the key light?
 
Added Key Light Foot Candles. Anything else you guys would like to see?

Test-Sample-3.jpg
 
WOW that was quick! Thanks. :)

BTW- that screen grab you've posted- is it a program you've created? If so, is it something that can be made available? This would be extremely helpful for the test shoots that I do on a regular basis. Right now I'm just using a text file, which gets rather tedious ...
 
Filemaker is the program. Very useful for making data entry templates/records.
 
Brilliant, thanks. :)
 
Back
Top