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 :)

ARRI: 16mm film is 3K, 35mm film is 6K...

aliasing is good. No, wait, it's beautiful. Give me aliasing. It's so fucking romantic.
No, wait, it's not. Grain is. Why? who knows? what does it mean? film knows it.
Film sees for ours, and we are completely blind.
grain structure is not useful. It's not details, you know it's just grain. And you keep it.
35mm is 3K, we know. but sometimes it's 100K. At the same time, native 4K is not even close to itself, but rather 960x760 and soon x 0.
8mm is infinite.
4K<8mm
film is magic
Arri says the truth about film being immortal
lenses perform 200000000l/mm on film , and 20l/mm on sensor.
probably Red one does not exist. It definitely doesn't exist.
Red is film.
or whatever you want to read.
ultimately, we learn that 1920x1080 > 4096x....
the most amazing abortion is that crystal clear torrents of notions came right after the pictures that speak for themself. Kurosawas's sad ramining frames with their stunnig fisher price quality (Arri?-)scanned frames that remaind me of hand painted Laurel and hardy black and white movies from the 30's, clearly help demostrating film superior and immortal 6k unequallity.


I hope this thread have given you a small kick off to learn more
Click to expand...
you bet?
 
Looks pretty sharp to me! Sorry for the large images...

(Babel- from s16mm neg to 35 blow-up to blu-ray to jpeg):

babel6.jpg


babel5.jpg


Did you know that all of the Morocco scenes from Babel were shot on super 16 with Arri sr3's? They could have used any format they wanted, why did they choose to shoot a smaller gauge film? Because it complemented the story and the results were beautiful.

Its an incredible waste of time to get into arguments about this stuff, no matter what side youre on.
 
he he
 
hunterrichards said:
Its an incredible waste of time to get into arguments about this stuff, no matter what side youre on.
Click to expand...

Hi,

Those images look pretty filmic, S16 DOF is also beautiful, I don't understand why everybody demands as S35 Sensor or bigger. 2/3 Scarlet could be very good, time will tell.

Stephen
 
Scarlet 2/3"

Scarlet 2/3"

Stephen Williams said:
Hi,

Those images look pretty filmic, S16 DOF is also beautiful, I don't understand why everybody demands as S35 Sensor or bigger. 2/3 Scarlet could be very good, time will tell.

Stephen
Click to expand...

You can get an idea of what Scarlet will look like from the Acam dII footage,

http://www.ikonoskop.com/dii/footage/

http://www.ikonoskop.com/files/87/A-camdII_Bil_h264.mov.zip

With all the Machine Vision 5 Megapixel 2/3" lenses around, the C mount model should have a good range of focal lengths, you can use a Canon or Nikon to C mount adapter for long lenses also. Many of the Machine vision lenses are f/1.2 and some 2/3" lenses are fast like the 17mm f/0.95 from Goyo Optical,

http://www.goyooptical.com/products/industrial/highreso.html

As well as using Arri Standard lenses with a Arri to C mount adapter.

From what I can see the Acam dII looks crisper than Super16, [Hunter] can you extract frames the same size as the Bable (2006) frames to compare?

PS. Do you know a good free program [Windows] to make 1920x1080 h.264 files from TIF or BMP frames? (and to go the other way around?)
 
Of course, grain really adds to the perception of resolution...

Graeme
 
David Mullen ASC said:
Yes, I think that's partly why some film-outs of digital material to 35mm look on the soft side, the lack of any grain.
Click to expand...


But before that step you could render your digital movie in a simulation of the stock you would print/record (for example KODAK VISION3 500T Color Negative Film 5219/7219).
 
These pictures look like made by a first generation webcam and good lights.
Compact digital cameras are a lot better to me.
Also dynamic range here is more apparent then real. In photoshop there is not so much to play with curves.
The top is overexposed.
It is just almost grey image with very little color detail. details are probably reduced to 8mm equivalent, after all those passages (s16 -> 35mm -> scanning -> blue ray compression -> Jpeg compression)

I wonder how a top quality PAL tube camera at its best, digitized live, would have performed in this situation.
I have no idea why S16 over 35mm, I can't believe Brad Pitt could not put together enough 35mm film. Lighter camera? HDV (+mini35) is lighter and 4 times the the image.
It may look less professional but surely it is more.
These remind me of those color documentaries from the second world war. Big grain and little else. A bit annoying to the eyes.
 
in photoshop, I duplicated Brad, downsized it to 1280x720, then made it 1920x1080 again, and compared it to the original. Stearing at the monitor, no one could tell which was which: absolutely identical. So the original could not have more details than 1280x720. I went down to 1000x563, and there was a very tine difference in the grain, barely recognizable on the fixed picture (definition was still identical). I doubt I could be recognized with motion, since grain of course would change for every frame.
But anyway let's comfortably say the details are somewhere above 1000x563 and under 1280x720.

This is nowhere close to someone's claim - I think someone advertising in favor of Arri - of S16 having a 3K resolution.
I see there have been some passages, but then again I wonder: what it counts is image quality on the screen, not what you start with.
In the case of 16mm, since it has to be blown up (plus scanned, sometimes) anyway, with this tiny original frame, we end up at standard definition for details, and apparent resolution just a bit higher only thanks to grain, which does not contain more information about a scene, it just offers its own grain. But hey, that apparent advantage on definition is at least the same as the one we get with sharpening. So what.
I see no point in shooting S16mm for blowup. Original HDV material has lots more details than what you end up starting with S16, once projected. Or at least we can say HDV (3x1440x1080) is bad, therefore S16 is bad too.
Please be honest. You can't call that ugly image good. Ask 1 million persons if it has more details than the pictures they take with their 10MP compact cameras.
Red is 10MP too, and with bigger sensor. There 's nothing esoteric here. It's poor image vs rich images.
If you don't agree, never mind. One may say S16 is beautiful, it's fine if you like it, actually it may not be necessarily unpleasant, but saying it has the details of 3K is over 3 times ridiculous. While 3K is twice 2K, S16 does not even reach half of blue ray potential.
 
similar unaestethics with plastic lens holga chinese camera.
"lomography"
and so on.
Some people like it.
Lo-fi, dirty, vintage... it's the same fashion as with sound.
 
Who but a technician or engineer gives a flying f*** about resolution. You can make great images on a Fisher Price Pixelcam.
Click to expand...

I'm not an engineer and I f...... love high resolution.
Actually I see that almost every man shooting with decent still or video cameras do care about resolution. This does'n not mean he has to be obsessed about it.
Anyway you are free to sell your Red camera and use fisher price.
Some one once told me the same about tennis rackets. Strange, every champion had the best racket. what a surprise in hollywood they use the best cameras available and not Pixelcams. And 35mm rather then 16.
when you buy a monitor, a printer, plotter, camera, you want resolution. perfectly normal. I'd say also in tissues, fine details are more appreciated than simple ones.
Fine detail is nice. rough is not very nice. Usually. It has always been.

but please everybody compare images from cameras in fair tests, and share them.
Words have no worth when comparing images. Let's have the images speak. 2K, 3K, 4K, dynamic range, low light, lenses, bokeh, breathing... there are never enough tests.
 
This is the best article (many test photos) I've seen on this topic yet:

http://www.clarkvision.com/imagedetail/scandetail.html

Or jump right to the end:

"Results/Conclusions:

From these tests, it is my opinion that digital cameras will match Fujichrome Velvi 35mm film when they reach more than about 10 megapixels. Somewhere in the 12-16 megapixels will produce color image quality comparable to 35 mm film (this is a compromise of more intensity detail and less color detail than film). Somewhat fewer megapixels, approximately 8-10 Mpixels will match 35mm film intensity detail but at below 35mm film color detail.

Medium format film: about 50 digital camera megapixels are need to match Fujichrome Velvia in 6 x 4.5 cm.

Large format: more than 200 digital camera megapixels are need to match 4x5 Fujichrome Velvia film. How much more needs further testing. "
 
I've got this idea (grain has nothing to do with it. We already know grain does not contain details about a scene. Grain structure is a grid, though not orthogonal. Seen grain is comparable to seen pixels. Is it not true information. information is not inside a pixel nor the pixel itself, information is made of many pixels, and the more pixels = more details. This is a secret that children know, and professionals seem to not know):

Arri refers to oversampling requisites, not film's resolution. In order to not loose information from an analogue source, the signal must be sampled have at least double rate.
With CDs having a sampling rate of 44.1kHz, you can get a highest frequency of 22.05kHz in an ideal system.
So, if they say S16 is 3K, perhaps they really want to say - but do not explaing it for a good reason - that S16 should/could be scanned at 3K in order to retain 1.5K of real information. So, should they face judges, they could justify their claim, while now still blow into their own sails for the ones who take their claim too literally.
Ambiguos information exists because people - yes, the "pros" - seem to be too lazy to make precise tests and compare. Everyone relies on some one else so everyone fuck yourself in advance.

now the 16mm is better than the d20?
Click to expand...
 
Sample rate must be double the highest frequency you want to record, however, in image processing terms, frequency is lp/mm. Line pairs is obviously half as numerous as lines. So to capture 3k lines, you need 3k samples. To capture 1.5k line pairs you need 3k samples. You're mixing lines and line pairs in your above example.

The problem is that although grain is not detail, it appears in the image as detail. To accurately preserve the grain and not have it alias, you must sample at a high enough rate. That leaves us to either sample at high enough to get the picture details, and maybe get some greeblies from the grain, or to sample at high enough to capture the grain properly and then downsample with filters designed to preserve detail and not give problems with the grain.

Graeme
 
You must log in or register to reply here.
Back
Top