cinematography The cinematography of Drive

http://www.youtube.com/watch?v=KBiOF3y1W0Y

http://www.youtube.com/watch?v=Kq2a7MWbmJU

http://www.youtube.com/watch?v=_z5LXyWn3-w

I really like the cinematography with the sunlight scenes, and trying to get right in my camera. It seems the best way to do it is too shoot with the color temperature around 12000 to 15000, depending on the time of day. In that scene in the second video they may have changed it for clouds too. The chase scene looks like it was done lower, more around 9000 but trying to get that one right too. I've been trying to get it right, but something feels off.

http://youtu.be/jVK4ESOV-bk

Does anybody know what they did to get those colors?

Thanks.
 
Last edited:
K I can use a tripod. I also want to try to make handheld look good too, since sometimes I am forced to go with it, such on city streets, where the cops are okay with handheld but not tripods, and dollies so much. Some movies have quick panned handheld, like Paul Greengrass for example, and he makes it look good, so I was thinking of getting better at that style maybe.
Yeah, you and me both, brother.

But Paul et al are using these EasyRigs for their "handhelds": http://www.cameraworks.co.nz/rentals/easyrig-rentals/

Not quite the same as you and me with our DIY FigRigs and cage rigs. ;)
 
K I can use a tripod. I also want to try to make handheld look good too, since sometimes I am forced to go with it, such on city streets, where the cops are okay with handheld but not tripods, and dollies so much. Some movies have quick panned handheld, like Paul Greengrass for example, and he makes it look good, so I was thinking of getting better at that style maybe.

Sure I get that but at the minute ... You're trying to figure out the camera settings... Once you know yr settings THEN practice the handheld .... one thing at a time dude!!!
 
I think the second video looks better: it's closer to the eye. When you are in a parking with orange lights, it looks like that: orange/yellow.
Besides that: orange lights are nasty to WB.
But if you want that 'cool' look: use it.

You should get yourself something colorful: just an object with 'all the colors' (or a nice girl with rainbow stripes sweather ;) ) so you can see the effect of the white balance and the lights beter. Grey concrete doesn't give you a clue about how other colors look.

BTW,
using the filename as title gives really great related videos with the same name!!!
Really funny :-p
 
Sure I get that but at the minute ... You're trying to figure out the camera settings... Once you know yr settings THEN practice the handheld .... one thing at a time dude!!!

This.

You're doing look development right now. Drop the camera on some sticks and take notes making a single adjustment at a time and review the results on the monitor you trust the most while referring to your notes.

Practice the operating another time.
 
K I can use a tripod. I also want to try to make handheld look good too, since sometimes I am forced to go with it, such on city streets, where the cops are okay with handheld but not tripods, and dollies so much. Some movies have quick panned handheld, like Paul Greengrass for example, and he makes it look good, so I was thinking of getting better at that style maybe.

It's very difficult to make handheld look good. Try watching your test footage on a large screen. The handheld look may look okay on your camera's 4" (?) screen but on a 60" screen or theatre screen it will be near impossible to watch - particularly if that 'look' goes on for more than a brief period.

A lot of people complain about Paul Greengrass's shooting style (I like it) - and as another poster has said he actually uses an expensive rig to achieve his 'handheld' shots.
 
It's very difficult to make handheld look good. Try watching your test footage on a large screen. The handheld look may look okay on your camera's 4" (?) screen but on a 60" screen or theatre screen it will be near impossible to watch - particularly if that 'look' goes on for more than a brief period.

A lot of people complain about Paul Greengrass's shooting style (I like it) - and as another poster has said he actually uses an expensive rig to achieve his 'handheld' shots.

In my opinion, it's not the rig that makes his handheld shots so good. Yes, it plays a part in the whole formula, but if you shoot with a 35mm film camera or Arri Alexa for example on the shoulder, you'll get an entirely different result than a T2i, even on a shoulder rig.

Weight is a good thing for handheld shots. It makes them a lot more stable and the camera doesn't just bounce around. I've shot with Alexa's on the shoulder and it's a lot easier to achieve good looking handhelds on that camera. The easyrig just makes the process a lot easier. If you have a global shutter on a camera and are shooting handheld, you're a lot better off too!
 
Yeah I can use a steadicam or a shoulder cam rig. I have a DIY steadicam so far, and works better than handheld. I just didn't bring it with me that day, since I was concentrating on color looks.

Here's the same shot as the first video. It's the one where the orange lights are white, but I changed them to red:

http://youtu.be/c5DUos2G_Fk

I may have more room to play with color by shooting with the orange lights white balanced, but it doesn't look that way because I need to get better at color grading probably.

Here's the second video with the mercury vapor lights white balanced, and the orange lights orange. Only I sucked the orange out and made them white:

http://www.youtube.com/watch?v=RuCINfgz7TM&feature=youtu.be

Their is a lot of noise as you can see. So maybe shooting with a realistic white balance to start as was also suggested, is a mistake?
 
What am I supposed to garner from a short whip pan (second video) and what can only be described as the camera operator sarcastically imitating a handheld style that he wants to make fun of (first video).

Put the damn thing on a pair of sticks so that when we pause the image to examine it we aren't fighting against motion blur or rolling shutter distortion. FFS.

Do you have me on ignore or something? It's fine if you do, I don't really care, but I'll stop wasting my time posting in your threads if that is the case.
 
Nope, I've just been busy with work, and didn't reply yet. Not meaning to ignore anyone. Okay not all of the second video uploaded to youtube, and youtube, only took 3 seconds of it. I will re-upload it.
 
Nope, I've just been busy with work, and didn't reply yet. Not meaning to ignore anyone. Okay not all of the second video uploaded to youtube, and youtube, only took 3 seconds of it. I will re-upload it.

Never heard of YouTube uploading only a part of a video.
Since when does it have a threshold that uses quality as parameter? ;)
 
Nope, I've just been busy with work, and didn't reply yet. Not meaning to ignore anyone. Okay not all of the second video uploaded to youtube, and youtube, only took 3 seconds of it. I will re-upload it.

It's not about replies, it's that you are not listening, and I'm far from the first person to get frustrated with that. :no:

At least 4 or 5 people in this thread alone have asked/suggested that you put the camera down and leave the shot still while doing this sort of look development. There are tons of reasons for this, not the least of which being you can compare the exact same frame under different look parameters to ensure that you are truly seeing what changes.

So, best of luck with that, really.
 
Well I wanted to move the camera around since their are different lights in the room to observe, but I will use a tripod/fluid head and go back.

That's fine. Just take multiple static shots of the combinations in the room rather than panning around. Trust me buddy, the movement in the shot is working counter to your goal of examining the look. Take copious notes of what the settings are for each shot so that you can accurately determine what changing each image parameter does to the image.

Did you not study the scientific method in school? Change one variable at a time otherwise you cannot be certain, when you see the result, which changed variable was responsible for said result.

Worry about your handheld practice later. If you want to test camera movement under this lighting condition (a valid test) do it after you have determined your iso/color temp/taking stop/look settings. Use those settings for all of your motion tests.

The point I am trying to make is that you have to approach this with more diligence and discipline in order for the testing to really be of any value - both to yourself educationally and to your project creatively.
 
... you have to approach this with more diligence and discipline in order for the testing to really be of any value - both to yourself educationally and to your project creatively.

Yes, this. Test Shoot, not a test shot... Here's an example from my site: http://yafiunderground.com/index.php?page=lighting_distances

Note that for each distance from subject to light, I did a full range of subject to diffuser. Changing one variable at a time every time, labeling my slate so I could see the settings for each. Straight up scientific method. This is how you learn the effects that changes have on your shot. This method applies to ABSOLUTELY everything we do, not just lighting and diffusion tests.
 
Test & practice.
Test & practice.
Test & practice.
Test & practice.
http://www.youtube.com/user/fowleggs/videos

What do you think I'm doing on my little nature outings? Taking pretty pictures?
http://www.youtube.com/user/placesofnature/videos
Pfft.
No.
I'm learning lighting, movement, framing & composition, f-stops, shutter speeds, frame rates, ambient noises, intentional noises, audio editing, and Lord knows what else.
Pretty pictures, my aunt's buttered biscuits! Ha!
Pfft.
Hardly.

I'm still trying to figure out why my camera will record some blacks BLACK! and other blacks a noisy speckled mess. :grrr:

https://www.youtube.com/watch?v=jyaKjfz1SqM
 
Ray,

I can't really tell because this might be caused by the YT - but the noise looks worst in the very dark green areas where there is just a little bit of detail. Like in the lower frame where the tree line and field of something are. You might be getting some extra noise in that channel? Perhaps the cleaner blacks have less green, and the dirtier blacks have more green detail? Is de-noising by color channel even a thing?

I might also be completely off base. :D

idk what kit you are using, but you could probably find a decent ND grad filter on Amazon for ~$50 ish that was the right ring size for your lens. Would really help those horizon shots like that, give you more detail down there in the foreground.
 
I'm still trying to figure out why my camera will record some blacks BLACK! and other blacks a noisy speckled mess. :grrr:

Generally, the noise goes up as the ISO goes up. Seeing as we're limited to 48fps for "normal" shooting, that means we need a wider aperture or more light. In digital terms, the ISO is the gain you put on the imaging sensor, and just like audio gain, the higher you go the more the noise floor becomes audible/visible.

You work with a video camera, right? So it probably lists gain, not ISO. Every 6 dB (IIRC) is ~1 stop, so if you can open the aperture 1 stop, you can drop 6 dB of gain, and noise. Or add twice as much light.

I gotta get out and do more practising also, lotsa shooting coming up this summer!

CraigL
 
Thank you, David.

Your guess is more informed than mine is, by far. Much appreciated.
I don't even have the slightest idea about how to go about de-noising by color channels and something I'll investigate further. TY for pointing me in the right direction.

I'm using a crappy cam with a rinky dink little 1/2.9 chip aggravated with H.264 compression.
As long as the sensor MOSTLY gets plenty of light the image doesn't get grainy.
But in evening-like lighting the grainy noise gets pretty noisy.

Yeah, I need to get a variable ND filter (looking for a Hoya).
I know I need it in the bright daytime shots to slow down my shutter speed to 30 or 60.
Hadn't really considered using it in low light situations. Cool! I'll give it a whirl when I get it! :yes:
 
Thank you, Craig.

Yeah, I use a video camera and I generally have better luck keeping the ISO on "AUTO".
Looks like I need to quit being so lazy. :lol:

On that shoot I honestly didn't even consider locking the ISO to 100 or 200.
I was fooling around with kicking down the shutter speed to 4secs (not 1/4, but a full 4secs!)
I don't think it made any better of an image than at 1/60 or 1/30.

And that shot was at the lens' full f1.8.
THAT part I understood. ;) It just still didn't help. Pfft.

EDIT: Argh! I forgot, since computer display frame rate is 29.97fps I almost ALWAYS record in the closest thing to that, 30fps.
I've no need to shoot at 60 unless I plan on slowing anything down.
And even shooting at 60 and comparing that to 30fps I don't notice a big whup difference for what I've generally been shooting.
https://www.youtube.com/watch?v=7bwrEIbigwk
https://www.youtube.com/watch?v=QB-H7Qy5ciA

'Spearmentin'. :)



And FWIW, my next batch of video experiments will be on color grading.
Taking an average shot and tinting it more blue, then green, then goldeny brown.
Should be fun.

I just had my Tascam DR40 delivered last week and haven't really had time to experiment with that, comparing its audio collection with that of both my cellphone and video camera.
It's my understanding the on board mics aren't really great, and that to get full benefit it needs an external mic, so... until then I just wanna collect better audio than what I'm getting from my video camera which even I can hear is poor.
I get tired of running my audio through hi/lo pass filters to cut out recording hiss.
So, experiments with that are next next. :)
 
Last edited:
To denoise by color channel, I use shake (you can use your compositor of choice, not sure the precise tools though). I separate the individual channels into separate images, then you can combine them again to deal with the channels separately.

Here's adobe stuff:
https://www.youtube.com/watch?v=Efyi21EUvbo
https://www.youtube.com/watch?v=pFhrTQJg6MI
 
Back
Top