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 :)

RED high lights blow out myth busted.

Mark Toia

Mark Toia

Well-known member
Joined
Jan 23, 2008
Messages
4,142
Reaction score
0
Points
0
Location
Could be anywhere in the world today.
Hello all,

We finished this TVC late last year and I thought I'd share the outcome.

http://www.zoomfilmtv.com.au/ftp/RED_test/Toia_Zoom_GGi_1280.mov

Shooting a person dressed in black, standing in snow in the middle of the day is a killer exposure test for any camera.

You need great (f-stop) range to retain detail in the snow as well the persons black wardrobe at the same time.

When I was first offered this project I did hesitate, I have shot 35mm in the snow a few times in the middle of the day and it came out with hot whites, great shots none the less but the sun hitting the snow is so white its hard to not have blown out whites. RED is no better than FILM with exposure latitude and to be honest I thought I'd end up with the same problem, if not worse.
I took the plunge and said to everyone, Lets go for it and shoot this thing on RED.

What we came back with was amazing, We had detail everywhere, So much so the we even crushed the finished grade a little bit to give it a bit of grit and extra contrast and still the whites and blacks hung in there.

The cameras did wig out when we were at 12,000 feet doing hand held aerial shots of the stunty jumping out of the heli. The hard drives were locking up momentarily at -20 degrees c , when we got to 0 c degrees all was okay again. Other than that the cameras worked flawlessly.

Once again what amazed me most was the detail in the snow and the blacks the actor was wearing. In our final 2k master you can literally see the actual snow crystals on the ground.
I went back to and old ad we shot on 35mm a couple of years ago and there was NO detail in the snow at all. Same conditions.

With the MX chip coming out, wow! seriously, like do we need it! yes...... we do.:smiley:

Have a nice day all.

PS. Please excuse this quality of this small compressed H264 quicktime movie, detail is not there compared to the final master.
Id like to post the 4gig 2k version to show you all the amazing detail, but the download time would be ridiculous.

enjoy

http://dl.dropbox.com/u/1363279/Toia_Zoom_GGi_1280.mov
 
At what ISO did you rate the camera ?
 
I was sure you were going to say 500 ISO or something since that seems to be the general advice when shooting outdoors with lots of light and trying to protect the highlights.
Looks pretty awesome nevertheless.
 
Mike N said:
I was sure you were going to say 500 ISO or something since that seems to be the general advice when shooting outdoors with lots of light and trying to protect the highlights.
Looks pretty awesome nevertheless.
Click to expand...

I never move off the base ISO, I either filter or lens to suit the f stop i want to shoot at, in this case I shot every thing at f2.8

I shoot exactly the same way I shot film, nothing changed for me.
I just pretend that I have a roll of 320 in the mag.
 
Nice Mark. How was this posted? Are you still using the proxy method in FCP? Was it graded in FCP, Color, other tool(s)?

Nice work!
 
Steve Sherrick said:
Nice Mark. How was this posted? Are you still using the proxy method in FCP? Was it graded in FCP, Color, other tool(s)?

Nice work!
Click to expand...

If I told you, Some serious Post houses would send someone around to break my legs.
 
monitors

monitors

One thing people need to watch out for is looking at the images 1:1 on their monitor and adjusting the "sharpen" until the images look like good DSLR still images, in general that is way too much sharpen.

In a movie theatre you are sitting back so that the width of the monitor is 1/2 to 1/3 the distance you are to it. If you sit too close to your monitor you need to be back more to judge the sharpness. If you look at a 4K image 1:1 on a 2K monitor, you will need to be back 6 to 9 feet to see it like it will be in the theatre.

I'm bringing that up since I think people are way overexposing their cameras, more so for people with a "video" background. Video tape had a very limited range, and new sensors are MUCH better than the limits of older video tape.

If you combine the ability to underexpose the sensor with the greater dynamic range of R3D over video tape, you can back off on the white level.

A 90% white card should record between 25% and 50% linear sensor signal in the green sensors for Digital Cinema use.

If you over sharpen you will see some noise at normal viewing distance, so DON'T over sharpen. Then you can expose well for the 1280x720 resolution of a projected release print, over-sharpening just adds un-needed noise and brings out some of the wavelet induced artifacts.

Its DIGITAL CINEMA not digital video, so make it look like cinema not video, don't clip...
 
There is no sharpening going on here, It is just straight out of the camera.

No under exposing was used on the day.

I'm not sure what your problem is, I have posted this just to show people the DR.

I don't shoot video cameras. so I don't no what your talking about.


Dan Hudgins said:
One thing people need to watch out for is looking at the images 1:1 on their monitor and adjusting the "sharpen" until the images look like good DSLR still images, in general that is way too much sharpen.

In a movie theatre you are sitting back so that the width of the monitor is 1/2 to 1/3 the distance you are to it. If you sit too close to your monitor you need to be back more to judge the sharpness. If you look at a 4K image 1:1 on a 2K monitor, you will need to be back 6 to 9 feet to see it like it will be in the theatre.

I'm bringing that up since I think people are way overexposing their cameras, more so for people with a "video" background. Video tape had a very limited range, and new sensors are MUCH better than the limits of older video tape.

If you combine the ability to underexpose the sensor with the greater dynamic range of R3D over video tape, you can back off on the white level.

A 90% white card should record between 25% and 50% linear sensor signal in the green sensors for Digital Cinema use.

If you over sharpen you will see some noise at normal viewing distance, so DON'T over sharpen. Then you can expose well for the 1280x720 resolution of a projected release print, over-sharpening just adds un-needed noise and brings out some of the wavelet induced artifacts.

Its DIGITAL CINEMA not digital video, so make it look like cinema not video, don't clip...
Click to expand...
 
light meters

light meters

Mark Toia said:
... so I don't no what your talking about.
Click to expand...

Sorry, my comments were not directed at your example, but rather to the on-going discussion of the 320 speed and people setting their light meters to lower index values to minimize noise.

EI 320 in fact should be putting the red and blue sensor exposure for a 90% white card at lower levels than would be used for video shooting, since we don't have any true RAW data to look at I can guess from the green bias of the Kodak sensors that 320 would place the red or blue exposure between 50 and 25% linear ADC signal, green may be a bit higher.

The DRX setting can be a help for things like snow that don't have much chroma, since you can get back a stop or so from the red/blue vs. green sensor unbalance.

With the M-X sensor being said to have the same sensitivity but lower noise and 14bits, it will be lower down in the "under-expose" part of the sensor, that is relative to the way video is shot with 90% white being closer to the top of the waveform. Shooting M-X at 800 would put the 90% white down a little over a stop from where it is at 320 on the M sensor. So the recomended M-X exposure is where I have been saying it should be for Digital Cinema use to hold enough highlights over 90% white.

Everyone knows that the EI does not tell you much since the way the light readings are taken is important.

In the case of snow, you can take a reflected reading off the snow from the camera's point of view, then increase the exposure by 2.5 stops to set the snow at about the white clip, or increase the exposure 1.5 stops to set the snow at about 50% sensor signal.

Light meters read for 18%, and 90% white is about 2.5 stops brighter, so if you open 1.5 stops you have one stop of extra head room. A little more since 320 has some head room to start with, so with two stops of head room 90% white is 25% sensor signal or so (green may be clipped more than red and blue because that is the way the sensors are made).

Anyway, people were talking about their lightmeters needing to be set to numbers off of what the camera is set to before, under-exposure is anything that does not clip green. Someone contacted me about making a white card that is closer to 100%, but even 100% is not bright enough to simulate specular reflections, like from snow, to get those you want "white" to be a stop or two under green clip.

Using EI 800 on the M-X sensor should make it easer for people to avoid green clipping because even if their light meter is a half stop off they will have more sensor "under-exposure" than using 320...

Reading snow may give a brighter reflection than a 90% white card because of the specular reflections, and so you automaticly reduce the exposure so that the white or gray card speed is no longer 320 if you did the compensation between the snow and card readings. Since the EI/ISO numbers are based on the 18% gray reading, if the histograms are used you may be exposing at a shifted ISO without being aware the 18% gray card peak is moved off the place it would be if the light reading was taken of it and centered for right spot for the ISO selected.

The comments about sharpen are because people were exposing high to avoid noise, and if you sharpen less you see less noise so can expose lower to hold the highlights better, even half a stop of "under-exposure" can help hold enough highlights to get a more cinematic look.

What method did you use to set the exposure?

BTW, if you want to hold the highlights in 35mm ECN, ask the lab to pull process you can get much more highlight detail that way. Many labs use old bleach that does not get all the silver out and that gives blown out highlights. Some labs do not have enough speed control to pull process two stops so you may need another lab. Also when the scans are made the processing of the scans should not clip the highlights and the scan should have "white" at less than clip in the uncompressed 4K TIF frames so you can grade the white levels later.
 
Dan -

Can you explain the reggae beat to me? I want to know why people like it so much when they're stoned.
 
You must log in or register to reply here.
Back
Top