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 :)

Shooting in low light

Look at Matt Ryan's spaceship example -- 16k of light was probably enough for 320ISO, but then he exposed for ISO500 (if memory serves) to save that extra bit of highlights... Conversely, those shots are super clean AND were shot at 3k.

Yeah 3K, 48fps. I got a decent exposure at 320ISO and like you mention I exposed for 500ISO to keep my highlights in range knowing that I had a noise free low end.

Really, really nice work!

Thanks Robert, I appreciate it! I was pleased considering we only had a $5,000 shooting budget and were a crew of 3.
 
Night scenes, when looked at with your eye but lit for good exposure in camera...actually aren't that "dark". What makes a night sceen look and feel right is high contrast, often hard light with a majority of the scene darker/below mid gray and flirting with the shadows, with a few light sources/slashes of light nearing or past burnout. Rim and edge lights help pick out shapes and details without bringing direct light onto the subject(s) and losing the "night" feel. Motivating a light in the scene that lets you put a key key light opposite of the noseline on the talent from the camera's PoV (direct light on the side of the face opposite the camera) can be effective. Having distant lights in the background of the scene that you can let go into overexposure helps sell the shot as "night".

But there's often quite a lot of ambient light used to bring up overall exposure, so your camera/medium isn't digging into it's noise floor/grain. Electronic viewfinders and on camera LCDs let you "see in the dark" and if you try to make the scene look good in the monitor...well, we're back to the classic mistake of demanding that the monitor be a what-you-see-is-what-you-get instead of using exposure tools and ignoring that the image on the monitor when judging exposure. If it's truly dark, a monitor will look "too bright" and that generally leads to an underexposure.

Saying "but it looked great on monitor" while the histogram is crushed on the left end of the scale and the underexposure goalpost is filled doesn't get you hired back.
 
It's OK to have very underexposed areas in the frame IF you are going to leave them that dark. After all, imagine a scene in a well-lit room that is normally exposed for 400 ISO, let's say, and looks really clean... and then the character turns off the room lights at the end, leaving the room several stops underexposed in some areas and pitch black in other areas. Well, the image noise doesn't change just because the room got much darker suddenly unless you start trying to lift that image in post to see more into the darkness.
 
David - you have made a couple of recent posts with reference to the noise levels at a certain ISO and I have to admit to being confused by this. In the world of RED, it is my understanding that ISO is only metadata and that there is absolutely no gain applied in the camera with this setting. To me, this implies no change in noise whatsoever, no matter what ISO you shoot at, as long as you are correctly using the other exposure tools (goal posts, histogram, stop lights) to get proper exposure. I know that changing ISO will shift your mid-grey point but I really didn't think it changed the raw signal in any way. What am I missing?

Thanks!
While ISO is only metadata and doesn't affect the RAW data, it does affect how the data is viewed. As your ISO value gets bigger, your mid-grey point moves towards the lowlights. As it moves lower into the signal, it compresses the highs and stretches the lows - if it gets low enough that you start to lift signal from near the noise floor, you're going to get visible noise. You can counter that with a curve that crushes anything near the noise floor, but then you'll end up with crushed blacks (which you may or may not want).
 
If you're seriously contemplating spending $30k-odd on an Epic, but you're worried about low-light exposure - I would STRONGLY recommend you instead take $15k out of that $30k and spend it on lights, some grip gear and perhaps a little 3.5kw portable gennie. Then put the remainder towards a Scarlet, or F3 or whatever.

It'll allow you to do so much more than just having an Epic will.
 
David - you have made a couple of recent posts with reference to the noise levels at a certain ISO and I have to admit to being confused by this. In the world of RED, it is my understanding that ISO is only metadata and that there is absolutely no gain applied in the camera with this setting. To me, this implies no change in noise whatsoever, no matter what ISO you shoot at, as long as you are correctly using the other exposure tools (goal posts, histogram, stop lights) to get proper exposure. I know that changing ISO will shift your mid-grey point but I really didn't think it changed the raw signal in any way. What am I missing?

Thanks!

I'm using the term "gain" loosely... sure, the camera just records at its native sensitivity but if you pick any ISO higher than that, at some point, the image will have to be amplified in some way to get normal brightness, whether its the signal to the monitor or the image after you convert it in post using the iSO setting you chose... which is why higher ISO selections give you more noise. If you exposed a scene for 1600 ISO, you can't just leave it at its native ISO through post or else it will look very dark.

Personally I don't really care what the native ISO of the sensor is, I just care about how the final results look at different settings. If 800 ISO at 3200K looks acceptable and I need those settings, I will use them.

The thing to keep in mind is that at higher ISO's and lower color temp settings where noise can be a problem, you have to be honest and accurate in your exposures because you have reduced your flexibility to be wrong and fix mistakes later. An accidental 1-stop underexpose may look fine once corrected at a conventional, safe setting but become a problem when working on the edge of acceptable noise.

It is the same issue in film, you may test and find that underexposing and push-processing by 2-stops is acceptable, but you'll find that you have little room for error, you can't actually underexpose by 2.5-stops on this or that shot and fix it in post without having the grain jump out at you, whereas misexposure with film rated and developed normally can fall within an acceptable range of corrections.
 
Sure, sometimes you make a choice to shoot in some really dark space without enough light for the ISO you picked for the noise level... and thus get more noise to get the shot, but it's a choice.

There are daylight-balanced LED flashlights, by the way, and there are really bright flashlights. Plenty of movies have been shot with flashlights on 500 ISO film stocks after all, but it means picking the brightest flashlights on the market, or building your own.

It's not the flashlight per se that causes the noise, it's trying to lift the image in camera or in post to make it brighter. You could have a flashlight in a scene with the camera set to 320 ISO and have a fairly noiseless image, it's just that the flashlight isn't going to read very bright. So the trick is to get the brightest flashlights on the market, preferably daylight-balanced for the Red camera but not required, get the fastest lens on the market, set the camera to an ISO level that gives you an acceptable noise level, and if necessary, cheat some additional light into the scene.

Yes, there are now some more sensitive cameras on the market, and the Dragon sensor can be rated faster than the MX sensor with less noise increase, but people have been managing to shoot flashlight scenes in movies for a hundred years now without needing 3200 ISO cameras!

Also, with the powerful grading tools available now, one might be able to lift the lit portion of the frame to an acceptable level while leaving the darker/noisier areas mostly alone or even crush them further.

Of course what's possible to do in post and what's practical or efficient timewise can be very different.
 
A few things that might have been mentioned here. (I just skimmed trough the thread) so sorry if I repeat something.

If there is not much movement in the scene you can shoot with 360 shutter. That double your exposure time. Also if camera is fixed or you got some good post people on crew you can shoot long exposure plates. Basically you then lit your foreground action and then shoot a plate and or panorama for your background with very long exposure times. We do this a lot to get coverage in the wide shots where we can not lit the whole scene. This works particularly well if you can naturally separate your forground from the background and if not using bluescreen for this option works quite well also. Basicallty light for your talant and a small blue screen, then remove the bluescreen and shoot the backplate clean in time lapse with what ever exposure is needed.

At least it a viable option for those shots where you can not bring in the helium balloons and all that, and you can get really nice exposing night skies this way.
 
Worst case scenario, there's always Neat Video :thumbsup:


Sometimes makes me cry that we've lost sight of the craft of image making… Budgets that are too tight for projects that in the past would never have been made (and possibly still shouldn't), combined with the "shoot it and let the post guy sort it out…"
 
Sometimes makes me cry that we've lost sight of the craft of image making… Budgets that are too tight for projects that in the past would never have been made (and possibly still shouldn't), combined with the "shoot it and let the post guy sort it out…"

I don't know if the availability of another tool like Neat Video constitutes losing sight of the craft of image making. Of course the 'fix it in post' ideology is bad practise but the flexibility to shoot using a wider variety of light levels is just an evolution of the cinematographer's tool set, not an 'issue' directly linked to political issues like production budgets.
 
Worst case scenario, there's always Neat Video :thumbsup:
Got caught in a situation when shooting a rehearsal dinner for one of my nieces, not realizing it was at a venue with little or no lighting. Some of the shots had some light but others were so dark I could not recognize any of the people in them without drastically changing the gamma in Premiere Pro. Some shots I saved using Lumitre effects and combinations of other effects while others required even more drastic measures. The grade will not be even across the finished video, but at least the newlyweds will be able to recognize who all were in attendance. Thanks to Neat Video, the final Blu-Ray won't embarrass me TOO much. '-)
 
I've been following this thread with interest and have a question about when to do noise reduction for low-light footage.

If you are post / master is being done at 2K, is it better to deal with noise reduction before or after scaling to 2K? I understand that scaling down provides some inherent noise reduction.

-M
 
+1 on lost craft. There is a difference between "low light" and "no light". Its about lighting. Test, shoot, experiment, learn the limitations of the camera sensor and how to expose for every situation you might encounter. Its about ratios and you deciding what you want to see and what detail is important to the scene.
 
I've been following this thread with interest and have a question about when to do noise reduction for low-light footage.

If you are post / master is being done at 2K, is it better to deal with noise reduction before or after scaling to 2K? I understand that scaling down provides some inherent noise reduction.

-M

Seems I've read somewhere here on Reduser that it is best to do noise reduction (with Neat Video, that is) last... probably because other things you've done that may affect noise will have made less of it, or perhaps because it makes for a faster render. I've followed that advice but after thinking about it, I'm not so sure that's the best way to go. Wish someone from Neat Video would chime in because I'm sure they've tested for this.
 
+1 on lost craft. There is a difference between "low light" and "no light". Its about lighting. Test, shoot, experiment, learn the limitations of the camera sensor and how to expose for every situation you might encounter. Its about ratios and you deciding what you want to see and what detail is important to the scene.

I've got a different take... I don't see it as lost craft but instead as evolving craft. Movie making formulas are not laws of the universe... they differ according to camera choice, post expense compared to shooting expense, to personal choice of the decision makers. Lamenting that "one size fits all" no longer is the norm is not taking into account that movie making today is diverse not only in content, but in the technology as well, IMO.
 
Back
Top