Blank

800 is the native ISO of the URSA Mini Pro's so you should be shooting at 800 (not 400) anyways for optimal picture and DR.

I'm usually don't alter 800 ISO even shooting ProRes in the URSA Mini 4.6K, I will generally over expose 1 - 2 stops then push down the shadows to get an inkier look.

Why does the ISO matter when shooting in RAW? In my tests, I could barely discern a difference between lowest and highest settings in post when capturing in either extreme.

Why does the ISO matter when shooting in RAW? In my tests, I could barely discern a difference between lowest and highest settings in post when capturing in either extreme.

I'd do a few tests with the Cine Meter II App where aperture is the variable and test 180 & 360 Degree Shutter and 800 - 3200 ISO where you are intentionally under and overexposing the image by at least 2 stops to start to see where the differences are.

For example, starting in the middle, try to light the scene at 800 ISO at T5.6 and 180 Degree shutter, then brighten the light or open the aperture to T1.5 to intentionally overexpose by 2 - 4 stops.

Why does the ISO matter when shooting in RAW? In my tests, I could barely discern a difference between lowest and highest settings in post when capturing in either extreme.

What matters is the ISO you meter for -- e.g., there's a difference between the correct exposure at 400 and 800, as your light meter will tell you. But when shooting raw the actual ISO camera setting doesn't matter if the f-stop is the same in both instances. The sensor captures the same amount light regardless of whether the camera is set to 400 or 800. There will be no difference.

Why does the ISO matter when shooting in RAW? In my tests, I could barely discern a difference between lowest and highest settings in post when capturing in either extreme.

I'd do a few tests with the Cine Meter II App where aperture is the variable and test 180 & 360 Degree Shutter and 800 - 3200 ISO where you are intentionally under and overexposing the image by at least 2 stops to start to see where the differences are.

For example, starting in the middle, try to light the scene at 800 ISO at T5.6 and 180 Degree shutter, then brighten the light or open the aperture to T1.5 to intentionally overexpose by 2 - 4 stops.

Interesting. Im looking forward to seeing this. Thank you.

Why does the ISO matter when shooting in RAW? In my tests, I could barely discern a difference between lowest and highest settings in post when capturing in either extreme.

What matters is the ISO you meter for -- e.g., there's a difference between the correct exposure at 400 and 800, as your light meter will tell you. But when shooting raw the actual ISO camera setting doesn't matter if the f-stop is the same in both instances. The sensor captures the same amount light regardless of whether the camera is set to 400 or 800. There will be no difference.

Thanks John. This makes a lot of sense. Do you think there is a good choice, given this understanding, about which iso to shoot in raw? Is there one that gives you export flexibility later with regard to the aperture settings, in your experience?

There is no flexibility in aperture settings! Aperture changes DoF and you can't reverse that in post.

You can get some flexibility in exposure, though, if the scene contrast is lower than the dynamic range of your camera. Let's say you measure a difference of 12 stops with a spotmeter between the darkest and the lightest area in your scene. Let's then assume your camera has true 14 stops of DR (not only claimed DR). Now you get 1 stop of flexibility for both over- or underexposure.

Setting your meter to the native ISO of the camera should make this equal on both sides. Setting it to 400 will leave you no margin for the highlights, but 2 stops in the shadows. 1,600 ISO will reverse this, more margin before clipping highlights, but no margin before shadows drown in noise.

There is no flexibility in aperture settings! Aperture changes DoF and you can't reverse that in post.

You can get some flexibility in exposure, though, if the scene contrast is lower than the dynamic range of your camera. Let's say you measure a difference of 12 stops with a spotmeter between the darkest and the lightest area in your scene. Let's then assume your camera has true 14 stops of DR (not only claimed DR). Now you get 1 stop of flexibility for both over- or underexposure.

Setting your meter to the native ISO of the camera should make this equal on both sides. Setting it to 400 will leave you no margin for the highlights, but 2 stops in the shadows. 1,600 ISO will reverse this, more margin before clipping highlights, but no margin before shadows drown in noise.

Thank you. I suppose the more tests I do the more novice I realize I am and confused. The google chrome 4k version of the original video I upscaled to 8k looks tolerable with some basic banding that I can live with. The Safari version of the same video is utterly destroyed and unacceptable for online delivery. As well as iphone. I suppose what bothers me is that I can find examples, say, of camping videos on youtube that are lit only by a fire and candle or either or and their blacks look absolutely inky and clean and without artifacts. I have done tests with a light meter to get a close enough exposure to a part of the image but it is not solving my problem. I do not understand the ratio issue, which is perhaps the problem, however, again, I really do not understand how some people are capable of such clean clear results on the same platform with what I assume is the same or less level of light. Thank you for your patience.

For example, this video, at about 9:45. Seems she’s shooting 60fps. Granted theres no movement and perhaps she is using a low light camera and actually knows what she’s doing, but my question is, what is it, specifically that keeps her blacks so clean? Why do my blacks react so dramatically on the same platform in such a negative way? She claims her upload process is nothing special but that it is the photographer skills that matter, and while I appreciate this conceptually I am still at a loss for understanding exactly WHAT that means. It seems like I cannot find anyone to really break down for me clear and concisely what is happening and how to fix it.

Her video:

Seems she’s shooting 60fps. Granted theres no movement and perhaps she is using a low light camera...

Her camera is a Nikon D750, introduced in 2014, that shoots 1920x1080 at a maximum of 59.94fps. ISO sensitivity is "auto" plus 200 to 12,800.

Seems she’s shooting 60fps. Granted theres no movement and perhaps she is using a low light camera...

Her camera is a Nikon D750, introduced in 2014, that shoots 1920x1080 at a maximum of 59.94fps. ISO sensitivity is "auto" plus 200 to 12,800.

How do you believe she achieving her results?

There is no flexibility in aperture settings! Aperture changes DoF and you can't reverse that in post.

You can get some flexibility in exposure, though, if the scene contrast is lower than the dynamic range of your camera. Let's say you measure a difference of 12 stops with a spotmeter between the darkest and the lightest area in your scene. Let's then assume your camera has true 14 stops of DR (not only claimed DR). Now you get 1 stop of flexibility for both over- or underexposure.

Setting your meter to the native ISO of the camera should make this equal on both sides. Setting it to 400 will leave you no margin for the highlights, but 2 stops in the shadows. 1,600 ISO will reverse this, more margin before clipping highlights, but no margin before shadows drown in noise.

The only other conclusion that I can draw is that because it is underexposed, which I like visually, and because the lighting is “soft,” the are less instances of “this area is lit vs this other area is not lit” which makes the footage mostly in the “in between” levels of light, which, while it looks preferable on my render, is perhaps setting off artifacts during compression.

The google chrome 4k version of the original video I upscaled to 8k looks tolerable with some basic banding that I can live with. The Safari version of the same video is utterly destroyed and unacceptable for online delivery.

Are you able to watch it in 8K? What resolution are you previewing with?

The google chrome 4k version of the original video I upscaled to 8k looks tolerable with some basic banding that I can live with. The Safari version of the same video is utterly destroyed and unacceptable for online delivery.

Are you able to watch it in 8K? What resolution are you previewing with?

Upscaling it, to my understanding affords it a higher Youtube quality encode, even if it is 4k footage. Watch it on both Chrome and Safari at max quality and while it may not be perfect in Chrome the difference in quality is absurd.

Granted theres no movement and perhaps she is using a low light camera and actually knows what she’s doing, but my question is, what is it, specifically that keeps her blacks so clean?

The Nikon D750 is a DSLR that has a native ISO of 15800 and does a lot if signal processing in the camera itself as opposed to a Digital Cinema Camera that is designed only to capture data that will be manipulated in color grading software like Resolve. They are two very different tools.

The second issue is that most interior shots that look like they are in "the dark" are actually lit extensively but with a much higher contrast ratio, to "invoke" the look of a dark room, but if you were actually on set, the room would be rather bright. In your sample video with the construction lights in the frame and the only source of light, there is no way you can hold detail in the lights and have any separation in tone in the darker areas without lighting those darker areas "up" closer to the light bulbs in the scene. Like many people new to this area you are mistaking the role of tools versus craft. Lighting is 90% of the "cinema" look, and probably the hardest skill to learn, as you must retrain your eyes to see the way a camera sees. For some excellent, and evocative lighting, watch "The Queen's Gambit". To learn how to do it, volunteer on a real film set as a PA and watch what a grip crew does. You will learn more in a week than you will in a year in film school, and WAY more than you will learn on forums like this.

In the olden days, we used to look through a dark glass to get a better impression of what the scene would look like to the camera.

In the olden days, we used to look through a dark glass to get a better impression of what the scene would look like to the camera.

Yup, the joy of composing with an upside down, reversed image:)

Granted theres no movement and perhaps she is using a low light camera and actually knows what she’s doing, but my question is, what is it, specifically that keeps her blacks so clean?

The Nikon D750 is a DSLR that has a native ISO of 15800 and does a lot if signal processing in the camera itself as opposed to a Digital Cinema Camera that is designed only to capture data that will be manipulated in color grading software like Resolve. They are two very different tools.

The second issue is that most interior shots that look like they are in "the dark" are actually lit extensively but with a much higher contrast ratio, to "invoke" the look of a dark room, but if you were actually on set, the room would be rather bright. In your sample video with the construction lights in the frame and the only source of light, there is no way you can hold detail in the lights and have any separation in tone in the darker areas without lighting those darker areas "up" closer to the light bulbs in the scene. Like many people new to this area you are mistaking the role of tools versus craft. Lighting is 90% of the "cinema" look, and probably the hardest skill to learn, as you must retrain your eyes to see the way a camera sees. For some excellent, and evocative lighting, watch "The Queen's Gambit". To learn how to do it, volunteer on a real film set as a PA and watch what a grip crew does. You will learn more in a week than you will in a year in film school, and WAY more than you will learn on forums like this.

Thank you, there is a lot to think about here. If inky black, compared to lit areas is what the goal is, like say, the original Halloween, are you saying that contrast is king? Or that more light in the shadows will offer opportunity to go darker there while helping the highlights? I’m sorry, I know it must be like talking to a child, this is just very complicated for me, my skills, if any are in other areas. But I’m trying to understand as best I can. I think the thing I’m grappling with, is how do you achieve a look with lots of lights that looks like it’s someone’s apartment they might come home too with all that extra light being motivated by appropriate and subtle lamps? Clearly there is not enough light in my original shot, perhaps in the yellower experiments, but how do you achieve that much light without simply blasting light that fills the whole room and kills all the spooky shadows in corners and other areas, totally defeating the mood? I suppose this is why there are teams of very experienced people who do this stuff.

In the olden days, we used to look through a dark glass to get a better impression of what the scene would look like to the camera.

Haha. Would, say, sunglasses work?

Not really. They can be pretty expensive (if they belonged to a famous person):
https://auctionet.com/en/888534-contrast-glass-kodak

So, IMHO there are two main factors: texture and contrast.

If you have enough texture, like in the shot with the tent, the noise is covered to some extent and the compression has some detail to hold onto.

If you have large areas of constant color, like in your interior, you need to increase the light level without reducing the contrast. On a film set, we would often exchange the light sources in the practicals (the light-emitting objects in the scene) with stronger ones and light the background accordingly. This way you keep contrast but don't risk muddy shadows.
In post, you finally drop the intensity in the shadows to make the scene look dark. Sometimes, if the shadows are too clean, you may even add some (only some!) grain to generate detail for the compression to avoid large blotches of uniform color.

Not really. They can be pretty expensive (if they belonged to a famous person):
https://auctionet.com/en/888534-contrast-glass-kodak

So, IMHO there are two main factors: texture and contrast.

If you have enough texture, like in the shot with the tent, the noise is covered to some extent and the compression has some detail to hold onto.

If you have large areas of constant color, like in your interior, you need to increase the light level without reducing the contrast. On a film set, we would often exchange the light sources in the practicals (the light-emitting objects in the scene) with stronger ones and light the background accordingly. This way you keep contrast but don't risk muddy shadows.
In post, you finally drop the intensity in the shadows to make the scene look dark. Sometimes, if the shadows are too clean, you may even add some (only some!) grain to generate detail for the compression to avoid large blotches of uniform color.

SUPER, super helpful. Now I’m starting to understand. THANK you.

Good to know I can help. Here is a look at the waveform of the tent with the fire with extents activated.
Even if the shadows are pitch black, there is detail everywhere:

Waveform_dark_scene.jpg (207.93 KiB) Viewed 5111 times

Let me give you some explanation why this relates to compression. Our eyes can't see fine detail at low contrast. Think of a wall painted in white, but with a grainy surface. With soft lighting you see it as smooth, but with a light coming in at a very flat angle, you see every detail.

Now, image compression is killing information where you are not supposed to see it, i.e. in smooth surfaces. Instead, it'll only store the same information for a larger group of pixels. But if this area is too noisy, that larger group will change shape from frame to frame. It'll look blotchy and nervous, not smooth any more.

Seems she’s shooting 60fps. Granted theres no movement and perhaps she is using a low light camera...

Her camera is a Nikon D750, introduced in 2014, that shoots 1920x1080 at a maximum of 59.94fps. ISO sensitivity is "auto" plus 200 to 12,800.

How do you believe she achieving her results?

I believe it's all about the contrast. The light sources are properly exposed and the blacks are crushed to remove noise in the shadows. The light sources are not boring and there are some triangular forms which are strong compositionally. Also, note that the lamp is behind her head so it adds some depth and the silhouette is clearly visible.

Good to know I can help. Here is a look at the waveform of the tent with the fire with extents activated.
Even if the shadows are pitch black, there is detail everywhere:

Waveform_dark_scene.jpg

Let me give you some explanation why this relates to compression. Our eyes can't see fine detail at low contrast. Think of a wall painted in white, but with a grainy surface. With soft lighting you see it as smooth, but with a light coming in at a very flat angle, you see every detail.

Now, image compression is killing information where you are not supposed to see it, i.e. in smooth surfaces. Instead, it'll only store the same information for a larger group of pixels. But if this area is too noisy, that larger group will change shape from frame to frame. It'll look blotchy and nervous, not smooth any more.

Very slowly I’m starting to comprehend this issue, at least more than I have, and it’s largely because of your very clear, articulate and in depth breakdowns of this issue. Thank you. And thank you for your patience. So, with regard to exposure, as has been pointed to early in this thread by others more experienced than I, the lower exposure leads to less contrast between light and dark in the shot, which, in addition to soft light is creating a environment where pretty much everything is dim and smoothed out in a sort of middle ground of light level, which is essentially the opposite of what compression cooperates well with? I think one thing that still confuses me and perpetuates my denial is that the Prores render plays without any obvious noise before it is uploaded to Youtube, and seems watchable, at least on s computer, which confuses me. I will certainly continue to test, and your input has been absolutely crucial to developing a comprehension of this subject. Thank you again.

ProRes is a production codec, meant to be visually lossless. H.264 and H.265 are distribution codecs, meant to be as small as possible at the cost of some quality loss.
We once made a test with 10 generations of ProRes and H.264, i.e. decompressing, changing one pixel, and recompressing. The result in ProRes was only slightly worse than the first generation. The one in H.264 looked like an experimental Manga.

creating a environment where pretty much everything is dim and smoothed out in a sort of middle ground of light level

In your video example everything is compressed into the "lower ground" of light level, ie: if you looked at a histogram you would likely see a single "mountain" to the left (dark) end of the histogram. If you point your camera at a dark scene, and make it look dark in the viewer panel, the histogram will show a narrow (steep) mountain to the left. As you open the lens or increase the shutter angle or increase the light you will see the mountain not only move to the right, but it will expand, left to right, as it moves over. You are now using more of the camera's dynamic range to describe the same information. That will mean more separation between tones in that area.

Of course you can expand the range later in software like Resolve, but depending on what codec you shoot and the amount of underexposure, you can "break" the image by trying to do too much in software. This results in banding and accentuating noise, that is then further accentuated by the encode going to H264-H265, which is then re-encoded going into YouTube. Starting with underexposure, trying to fix it in post, then "up resing" it, then going to an end user codec like H-264, then having YouTube encode for their platform compounds errors exponentially as you go along. The best pipeline is good lighting, full exposure, darken in post, output at the shooting resolution to a quality codec like ProResHQ, or ProRes 444, and then upload. Vimeo is a superior platform to show your work over YouTube visually, and don't forget that everything you upload to YouTube becomes property of Google. That's right, your copyright goes bye bye. They own it.

awppollock's original question has, once again, been lost in the mix here. Let's agree to let the courts deal with copyright issues, and get back to Low Light Shooting. It is actually a question of "how do I invoke darkness with light?"

It is actually a question of "how do I invoke darkness with light?"

If you disregard large film sets and expensive lighting for the sake of argument, I think the 4.6K is still quite good in available light in a documentary context or with practicals in frame and a few lights off camera. It's not going to see in the dark, but it generally does have some tolerance for underexposing and still giving a rich black when graded. My main concern is finding FPN in the areas that are still visible like faces and clothing.

In low light I'm generally lighting scenes using practicals in the frame then adding in a light source out of frame that is believable enough to the viewer, not unlike Alex's experiments.

For example, if I I'm in a car at night and I want to have a face in the frame lit with passing car lights, street lights and the dashboard I would still be inclined to add a fill light, like using the brightness of an iPhone screen that could simulate another passenger using their phone while driving that gets some light on the subject's face to reduce the likelihood of fixed pattern noise. It could also be the dash light, that's where creativity comes into play when you don't have a big lighting budget.

I would still be inclined to add a fill light, like using the brightness of an iPhone screen that could simulate another passenger using their phone while driving that gets some light on the subject's face to reduce the likelihood of fixed pattern noise

Yes, what you are describing is what I called "referred lighting", where the added light is "referring" to the practicals in the frame, but reduces the light level difference between the practical and the foreground with artificial (added) light in a believable way, and gets the areas of importance up off the noise floor of whatever sensor at what ever ISO.

Another issue is if you are shooting RAW or BRAW, there is little to no noise reduction taking place in the camera, unlike a Hand-Cam. Black Magic assumes that you would prefer to dial in your own NR on a clip by clip basis as needed. I have been grading a project mostly shot on a RED 6K with available light, and it is crazy noisy. I add a touch of Neat Noise Reduction plug in, and it's amazing how great it looks, creamy mid tones, and inky smooth blacks. The problem is that it renders verrrry slooooowly.

Learning the difference between low light and low-key lighting would probably be a good place to begin gaining some understanding. https://en.wikipedia.org/wiki/Low-key_lighting

This example is low light, the results are unsurprising, the image has no contrast and the blacks are floating in space. Once the blacks are lowered there's almost nothing left and the subject is underexposed by about 4 stops. Additionally the light is all fill, there is no key light which is why she's underexposed 4 stops.

scopes.png (483.95 KiB) Viewed 4838 times

Further, if the subject had been kept camera left immediately following this grab the BG would be the lights in the room instead of the wall. She breaks the 180 rule for no reason then forces poor composition. I'd advise putting some conscious effort and thought into the lighting approach and blocking.

This on the other hand is what a proper low-key shot is supposed to look like.

Low-key_cat.jpg (51.17 KiB) Viewed 4838 times

The fill level is the same as the OP's grab with the addition of a hard key light. If the key light is coming from the direction of a practical source it is referred to in the larger industry as motivated lighting.

Nothing about BC Tipi's video is remarkable, limited dynamic range and crushed blacks, meh. This is basically day for night.

BC Tipi.png (666.2 KiB) Viewed 4838 times

Good Luck

Of course, it is. And then there is in-camera noise reduction and the edge sharpening.
But the presence of texture illustrates why the side effects of compression are less obvious.

Just to point something out that I noticed from earlier in this discussion:

ISO is not always just metadata. That is true of some but not all cameras - some of them actually apply different levels of analog amplification to the signal before it is encoded: https://www.dpreview.com/articles/54268 ... ts-complex


Also, even though I would tend to agree that the sensor can often be an acceptable substitute for a light meter when it comes to setting the exposure of the camera (particularly when combined with a false color display), it is NOT sufficient when it comes to comparing lights with each other. The light meter is sometimes helpful for setting camera exposure, but it is much more important for setting up proper lighting.

It appears to me from this thread, and another posted elsewhere on this forum about depth of field, that awppollock is beginning a new chapter in the world of digital cinema cameras, as opposed to, perhaps, consumer equipment. There are a large number of overlapping issues he appears to be working on, lighting, tonal compression and underexposure, video compression and codecs, the relationship between lens settings and DOF, and the differences in streaming platforms. He seems like a thoughtful fellow with a thirst for knowledge, but this forum is perhaps not the most efficient way for him to learn the basics of shooting and post production. Could some of the posters here recommend reading material and/or Youtube channels where he may study these topics more efficiently?

I would also recommend to awppollock that he consider volunteering as an unpaid Production Assistant on any local indie productions, and try to work his way into the camera department. Watching others with a lot of experience in lighting, setting up shots, blocking, and production design is a great way to learn quickly at no cost at first, and then working your way up to $100/day as second AC is way less expensive than film school and hundreds of thousands in student loans. Don't get me wrong and start another argument here, I went to a very good art school myself, a very long time ago, and it was worth it. However, I see a lot of people in this forum and a few others that have bought a "professional" digital cinema camera and think that alone will make their images look like what they see on Netflix. If that were true, everyone who bought a good hammer at Home Depot would be a finish carpenter. As I said earlier, don't mistake a good tool for good craft. Best of luck, to you awppollock. Keep learning!

[CaptainHook]

The discussion on copyright has been split away to the off topic section. Please try keep the discussion towards using the cameras etc.

Hi Alex, I don't mean to hijack your thread but I am learning (trying) to shoot dark shots as well. I have a BMPCC6K though and I wanted my shots to look like this:

instead.. it turned out like this: (I tried to match my video to this thumbnail from a movie)

I shot at 3200 Iso. From what I have read in these comments it seems best to shoot the scene with lots of light then in post make it darker? But I am just confused I guess on how those other shots look crystal clear with no noise while it is night outside with not a lot of light???

I shot at 3200 Iso. From what I have read in these comments it seems best to shoot the scene with lots of light then in post make it darker? But I am just confused I guess on how those other shots look crystal clear with no noise while it is night outside with not a lot of light???

Rob, having a P6K on the high gain does not make the camera a good for low light (more like low key) camera. Typical of cinema cameras is to shoot with the correct amount of lights. To answer your confusion on how those other shots look crystal clear with no noise while it is night outside, they set the camera at low gain (to avoid noise) and make decisions on what shutter angle and iris opening they want to shot at. Then they light the set/subject measuring the area they want to expose correctly and that there is enough light for the shutter angle and stops they expect to shoot at so as not to starve the sensor and underexpose the shot. That's mostly it. Then there are the cameras that are much better in low light than the P6K... a different conversation.

OK, thank you. I will try to light more and shoot at a low ISO like 400 and then in post try to make it "look" like night or dark.

OK, thank you. I will try to light more and shoot at a low ISO like 400 and then in post try to make it "look" like night or dark.

Rob, light where you need to only light and measure it for your camera setting, letting everything else fall off into black which can be okay because they are to be in the dark of night anyway. Best to you!

This has nothing to do with camera exposure at all but I think since we are talking about lighting, I think this is worth noting. I am by no means a gaffer but in projects I worked on, I always have a gaffer and depend on them to light the set given the parameters.

I have this old app called set.a.light 3D for photographic lighting simulation which sometimes how I will experiment with various throw and intensity of light through simulation before I apply it in practice. It has helped me learn about lighting and was fun to do. It's not cheap but there is a free trial to test it out. Here's the link. Note: I have not used it lately by my older version had light fixtures specific for photography. Their new version might have more lights like modeling and permanent which you might be able to simulate on a film set. You be the judge.
https://www.elixxier.com/en/products/setalight3d.php

This YT video is a trailer for a MasterClass on Lighting. For the folks who are wondering why others here like me keep on saying that you have to light your set to get the right exposure (even for dark scenes), well, you might find it interesting watching this video and seeing how much light are used to light dark and night scenes, or get that beautiful dark scene. For example, Look at 1:23 where the set is lit with the two guys seating on the table and what the shadowy/dim/dark (whatever) resulting scene looked like after. Just something to think about...

Thank you so much. I will watch this now!

Hi Alex, I don't mean to hijack your thread but I am learning (trying) to shoot dark shots as well. I have a BMPCC6K though and I wanted my shots to look like this:

I shot at 3200 Iso. From what I have read in these comments it seems best to shoot the scene with lots of light then in post make it darker? But I am just confused I guess on how those other shots look crystal clear with no noise while it is night outside with not a lot of light???

From my own experience.. Its about being bold with grading. Go bananas. Always keep in mind that what you get from BRAW files is pure RAW data without any filters. Not sexy. I always keep in mind when shooting "this is not representative to what i want". Because you have the chance to make it painterly good at your own control later on. That is, IF you have recorded enough data(Light).

Here is an example simulating "the witch" look from a shot i did this Christmas. Its shot at 3200ISO, afternoon(midday), in a dark forest at F2.8 40mm(equivalent).

Pocket4k 3200ISO

witch_Pocketgrade.jpg (994.7 KiB) Viewed 4474 times

This is a screencap from the movie "The Witch" as reference.

thewitch_screencap.jpg (895.89 KiB) Viewed 4474 times

Here is the original flat straight out of cam.

forest-original.jpg (288.91 KiB) Viewed 4474 times

Here is another example tuning down shadows/midtone at grading. The flat original is a good representation of the light that was seen that day. The sky was not that colorful as seen in the graded version. Shot at 400ISO


GRADED:

midafternoon-grade.jpg (636.43 KiB) Viewed 4468 times

BRAW - straight out of cam:

midafternoon-org.jpg (550.6 KiB) Viewed 4468 times

That! Don't starve the camera, but grade the hell out of it.

That! Don't starve the camera, but grade the hell out of it.

Agree. It was clear that we had to get it right "in camera" with the early digital cameras. That is in the past today and we have actually surpassed the organic film. What you truly want is as much light as possible for any scene keeping track of your scopes. Even the dark ones, and mold it so that you keep that crisp shadow detail (Modern look). THe bmpcc4k/6k cameras are without doubt pretty excellent in low light. That is. low light shot with light, not starved as you put it. The forest scene is a stop underexposed btw. That was a mistake from my part.

(The Witch) movie is crushed in the shadows. Most likely to get that gritty horror feel. Its an artistic choice, but the original plate is most likely pretty light and un-horrific.

Its interesting to see old movies blown up to 4k(high bitrate). I have seen many of my favorite movies with scenes that is covered with mushy noise. missed shots I didn't know about before. That was hard to see in the old cinema with a resolution at sub 800 lines of information or on a standard def tv. The technology makes it ironically unforgiving. Still.. I mostly add noise to my shots because I'm a old romantic guy that cant let go of the past.

Opto-chemical film was very different: it compressed the shadows (and the highlights) naturally because of the reciprocity law failure, or the Schwarzschild effect. If we want the same look for digital recordings, it has to be done in post.

Hi Alex, I don't mean to hijack your thread but I am learning (trying) to shoot dark shots as well. I have a BMPCC6K though and I wanted my shots to look like this:

I shot at 3200 Iso. From what I have read in these comments it seems best to shoot the scene with lots of light then in post make it darker? But I am just confused I guess on how those other shots look crystal clear with no noise while it is night outside with not a lot of light???

From my own experience.. Its about being bold with grading. Go bananas. Always keep in mind that what you get from BRAW files is pure RAW data without any filters. Not sexy. I always keep in mind when shooting "this is not representative to what i want". Because you have the chance to make it painterly good at your own control later on. That is, IF you have recorded enough data(Light).

Here is an example simulating "the witch" look from a shot i did this Christmas. Its shot at 3200ISO, afternoon(midday), in a dark forest at F2.8 40mm(equivalent).

witch_Pocketgrade.jpg

This is a screencap from the movie "The Witch" as reference.

thewitch_screencap.jpg

Here is the original flat straight out of cam.

forest-original.jpg

WOW! That looks Amazing!

So I took the advice from what was mentioned on this thread and shot in light and then in post tried to grade it to it looks like dark/night.

The problem I am having is that the footage looks blotchy.

This footage below shows three different clips. One in raw and then one graded to look like it is night. Do you guys see the same I do with the blotchiness? Any idea what I am doing wrong?
I shot these at 400 iso with the sigma 18-35 lens.. The kitchen light was on above my wife. I had two Neewer Bi-Color Video LED lights pointing at her from each side too.

Hi Robdoc,

It’s hard to say what’s causing the blotchiness, but it looks like some sort of video compression artifacts caused by the delivered format. Did you record this with the pocket6K? I don’t think any compression rate on that camera can make this effect.

Regarding grading. Personally, I prefer to “paint” the light in post when you shoot like this. The screencap of my kid in the forest is brighten up where he stands and darkened around him to make a claustrophobic feel. So that he becomes the focus and the surroundings becomes mystified and threatening. You can mould/paint the light inside davinci resolve in very powerful manner. Instead of grading evenly the overall frame. It’s a subjective thing, but you should try it if you have not.

Here is another example of how to tune the shadows and brightness in a frame. In this frame I want the audience to notice the empty chair. I have brightened up the chair and darkened its surroundings. :

The chair_grade_focus.jpg (230.54 KiB) Viewed 4311 times

Here is a frame with an overall colour grade. :

The chair_grade.jpg (267.18 KiB) Viewed 4311 times

Raw frame:

The chair_raw.jpg (183.97 KiB) Viewed 4311 times

In this frame I want the audience to notice the empty chair. I have brightened up the chair and darkened its surroundings. :

The chair_grade_focus.jpg

Here is a frame with an overall colour grade. :

The chair_grade.jpg

Raw frame:

The chair_raw.jpg

I think that this is very nicely done. In the first capture, my eyes went first to the lamp and window and then naturally to the chair. With the overall grade, one is more or less stuck at the window with nowhere to go. What I especially like is that the light on the chair comes across as plausible/natural rather than artificial.

Thanks Rob.

Yeah, the window is an eye catcher and a bit hard to tune down in a natural fashion. If the original shot was about the chair, which it was not. Then I would use some blinds om the window or frame it in another fashion. But it’s only a proof of concept.

Here is how it works in motion. Password: chair

I'm not sure what is meant by blotchiness - is it the color or tonal banding?
If so, there can be a few things that could have cause this, starting with the most obvious that we can/cannot cancel out - the compression rate at post delivery. Before that, it can be from pushing it too much. One thing I notice is some color shift on the light in the raw capture that when processed in post and pushed hard caused the bandings.

Yes, that is it. Tonal banding. I am not sure what could be doing that to the video. I even exported it from resolve as a different format and it still gives the tonal banding on the video. (I use the BMPCC6K)
The settings are:
Blackmagic RAW
Constant Bitrate
5:1
6K
Film