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 :)

2:1 or 16:9 for 1.85:1?

T

Tim Commeijne

New member
Joined
Jan 5, 2009
Messages
1
Reaction score
0
Points
0
The manual of the RED workflow for Final Cut Studio says the 2:1 aspect ratio is suited for film. You have to crop the sides in post for the 1.85:1 aspect ratio.

Now, wouldn't it be better to work in 16:9 en crop the top and bottom?

you will lose fewer pixels this way, no?

Any thoughts?

Tim
 
No you shoot 16x9 (1.78 : 1), protect the full frame, and compose for cropping to 1.85 -- then you transfer the full 16x9 image to 35mm and let the projector mask crop it the tiny bit from 1.78 to 1.85.

As for home video and broadcast, I'd make both a 16x9 full-frame and a 16x9 with a tiny 1.85 letterbox master (the letterboxing probably won't be noticed in half the 16x9 sets out there anyway).

Basically treat it as a 16x9 full-frame project and just compose with that loss of headroom in mind if it gets trimmed to 1.85 for projection. 1.78 (16x9) and 1.85 are so close that you could almost get away with just composing for 1.78.
 
David Mullen ASC said:
No you shoot 16x9 (1.78 : 1), protect the full frame, and compose for cropping to 1.85 -- then you transfer the full 16x9 image to 35mm and let the projector mask crop it the tiny bit from 1.78 to 1.85.

As for home video and broadcast, I'd make both a 16x9 full-frame and a 16x9 with a tiny 1.85 letterbox master (the letterboxing probably won't be noticed in half the 16x9 sets out there anyway).

Basically treat it as a 16x9 full-frame project and just compose with that loss of headroom in mind if it gets trimmed to 1.85 for projection. 1.78 (16x9) and 1.85 are so close that you could almost get away with just composing for 1.78.
Click to expand...

is that what you did for your last Red shoot?

and does anyone have a visual of this?

thanks.
 
David Mullen ASC said:
A visual of what? A 1.78 hard matted image on a 35mm strip of film?

The two RED features I shot were in 2:1 mode and composed for cropping to 2.40, so they will eventually be recorded to 35mm anamorphic.
Click to expand...

i was just having trouble getting through all the numbers and measurements. let me take a look again. thanks for the info, david.
 
David Mullen ASC said:
The two RED features I shot were in 2:1 mode and composed for cropping to 2.40, so they will eventually be recorded to 35mm anamorphic.
Click to expand...

Hi David,

I recently did the same thing for a short film (You can see some pictures here: http://205.234.135.241/forum/showthread.php?t=21043), shoot 2:1 and compose for 2.40, although the film is not going to distributed theatrically. My question is why you chose 2:1 over 16:9? Thanks.
 
Christopher Grant Harvey said:
Hi David,

I recently did the same thing for a short film (You can see some pictures here: http://205.234.135.241/forum/showthread.php?t=21043), shoot 2:1 and compose for 2.40, although the film is not going to distributed theatrically. My question is why you chose 2:1 over 16:9? Thanks.
Click to expand...

Well, "Manure" was my first RED movie, using Build 15, and I just heard rumors that 2:1 was more "stable" -- otherwise I would have shot 16x9. If the movie had been framed for 1.85, I would have risked 16x9 anyway, but it seemed with a 2.40 movie there was less pressure to take that chance, whether or not it really was a risk or not.
 
Absolutely true David. With build 13, which was the last release of the red software before Manure, there were lot of issues with people getting the camera to actually record without issues. In my own testing and in reading posts from reduser, there was enough issues with recording 16x9 to make it unreliable (dropped frames, codec errors, etc...). To think about it, 16x9 recording is the most taxing on the internal processor, as its the highest number of pixels the camera has to encode, downconvert, etc. In addition, there were some post concerns, as it wasnt always supported. I think the biggest post issue was the 16x9 quicktimes not rendering properly. So when build 15 came around there had not been enough of a proven track record risk recording in 16x9.

Of course, since build 16 and the newer quicktime codec, everything has changed and 16x9 is much more reliable.
 
I also want to point out that you could use the "user reticule" option in the framing menu to create a 1.85 frame inside your displayed 16x9 frame. So you wont have to guess your framing. You'll use something like:

width=1280
height=692
x = 0
y = 14

[shameless plug]
Then if you wanted the 1.85 for editing, with the upcoming version of R3D Data Manager (released this week), enter those numbers in the framing aspect menu and it will produce the 1.85 aspect in the downconverts.
[/shamless plug]
 
Yes, now I recall... the 4K 16x9 problems at the time (last summer) were more related to the Apple Quicktime codec and not the RED camera so much, though there were also a few issues there as well. Anyway, I just didn't want to deal with them at the time and stuck to 2:1.
 
David Mullen ASC said:
Anyway, I just didn't want to deal with them at the time and stuck to 2:1.
Click to expand...

Okay great to know.

So you would shoot 16:9 in future. I was just asking in terms of the number of pixels if it made a difference to the filmout process when losing some to achieve a 2.40 extraction.
 
is this fast math correct?

is this fast math correct?

is this correct?

4K 16:9 is 4096 x 2304, so the full area is 9 437 184 pixels
4K 2 : 1 is 4096 x 2048, so the full area is 8 388 608 pixels

so, the difference is 1 048 675 pixels - more than ONE MILLION PIXELS!!!

and this is just starting point BEFORE croping to 1:85. difference is even more cruel when cropped appropriately:

16:9 cropping is 4096 x 2214 (rounded) so the area is 9 068 544 pixels
2:1 cropping is 3788 (rounded) x 2048, so the area is 7 757 824 pixels

difference? 1 310 720 pixels. => 1.3 million pixels
and comparing to full available 16:9 area, you are not using 1 679 360 pixels ...

but... you decide :)

filip
 
David Mullen ASC said:
A visual of what? A 1.78 hard matted image on a 35mm strip of film?

The two RED features I shot were in 2:1 mode and composed for cropping to 2.40, so they will eventually be recorded to 35mm anamorphic.
Click to expand...

Hi David, or others with the knowledge ...

can you elaborate on the anamorphic recording?

I probably misunderstand something, you shot 2:1 and cropped to 1:2.4 now you have an image of 2048 853 pixels.
What advantages do you get by spreading it optically to the 35mm negative and shrinking it back in projection?


Thanks.
 
karilla said:
Hi David, or others with the knowledge ...

can you elaborate on the anamorphic recording?

I probably misunderstand something, you shot 2:1 and cropped to 1:2.4 now you have an image of 2048 853 pixels.
What advantages do you get by spreading it optically to the 35mm negative and shrinking it back in projection?


Thanks.
Click to expand...

We shot 4K 2:1, so when cropped to 2.39... : 1, it's about 4096 x 1713, assuming we finish to 4K (right now, that's up in the air.) If we finish in 2K, that would be about 2048 x 856. These pixel dimensions would also be true if we shot Super-35 film and finished in 4K or 2K, then cropped to 2.39... : 1 and converted to anamorphic for film-out.

As for stretching this cropped area vertically to become an anamorphic image onto 35mm 4-perf for print projection... you don't have any choice -- that's how 2.40 images are projected in movie theaters, in the 4-perf 35mm anamorphic format. You don't "shrink it back" during projection, you spread it out horizontally by using an anamorphic projector lens.

Basically a recorder like an ARRILASER, when set to record a cropped image to 35mm anamorphic, records every horizontal pixel row twice to double the height of the image, creating a 2X squeeze. So a 4096 x 1713 image or a 2048 x 856 image becomes twice as tall by repeating each horizontal pixel row during the film-out.

Now if you had shot with anamorphic camera lenses, the image would already have a 2X squeeze, so the dimensions of the digital file would be closer to 1.20 : 1. So you start out with more vertical resolution in general as opposed to a cropped spherical image getting stretched vertically by the film recorder.

For example, compared Super-35 scanned at 4K and cropped to 2.39... to a 4K scan of 35mm anamorphic photography. Super-35 is scanned at 4096 across in 4K mode, but 35mm anamorphic doesn't use the soundtrack area so the scan is 3656 x 3112. But Super-35 cropped to 2.39... is around 4096 x 1713 -- compare that to 3656 x 3112 and you'll see how much more vertical resolution there is with anamorphic photography.
 
You must log in or register to reply here.
Back
Top