7D vs 16mm Kodak vision 3

PaPa

PaPa

Major Contributor
Hey guys.

So, as soon as my canon comes in, I plan on running some tests with two cameras running side by side, the Canon EOS 7D and some 16mm kodak vision 3 ( not sure what ASA quite yet ).

Basically, i really want to push the digital as far as i can to see how it compares to the film latitude, skin rendition, highlight detail, shadow detail, saturation, hue, sharpness and detail, etc...

The goal, is to do a comparison between the two to help find out how close one can come to mimicing those ( discussed above ) features from film on the 7D.

the other goal, is since we can use custom scene files, i want to see if there would be a way to compare both the digital footage and film footage together in post, do some curve adjustment until they are somewhat similar and save that change as a custom scene file to upload to the 7D so that one can work with something that mimics what one might get from raw film.

Remember, the goal isn't to get an exact match, but something similar enough that will please those of us who want to get as close as possible to the same look. (within dynamic range capabilities)

What would you guys recommend in terms of tests to perform? I don't want to buy a crazy amount of film, but maybe 100 to 400 feet should be sufficient.

Any comments, suggestions or anything else would be appreciated! Very excited to perform this test.

p.s. please no hijacking thread about the fact that digital will never be film. I am aware, and everyone is aware as it's been discussed time and time again, please keep responses relevant to the goal. Thanks.
 
Perhaps the most significant thing is -- what are you going to do with the film after it's shot? If you transfer to a regular video format or HDV or something like that, you're going to be giving up a lot of what the film has to offer. How are you planning on viewing or editing the shot film?
 
well my friend, thats where you come into play :) What transfer type do you recommend? Highest quality would be best, but i know its expensive, thus the amount of footage should hopefully be not too long.
 
Only fair competition would be a 2k transfer on a da vinci to D-5. If you are shooting 100ASA film (even old Vision 2) with great lenses, there wont be a competition. s16 still has more resolution and more latitude. Go up around 500ASA and the grain gets more distracting, you might have a better competition.
 
The problem here is a lot of the look on film is determined in the telecine. Two different colorists can produce two entirely different looks from the same stock.

Quick edit for above commenter: Vision 3 500T is incredible. The grain level is really really low [if exposed correctly]. There still won't be a comparison. Not anymore at least. Maybe with an older stock. But vision 3 holds up exceptionally well.
 
I've been wanting to try some of the 500T Vision 3. I've seen scans of the frames, but never in motion. I've shot thousands of feet of 7212 and love that stock.

You're absolutely right that it comes down to how it is telecined and the look that you and the colorist achieve.
 
well my friend, thats where you come into play :) What transfer type do you recommend?
Click to expand...
Well, first things first -- 16mm or Super16? I would, of course, recommend Super16, as the aspect ratio is closer to the 16:9 aspect ratio of HD.

Second, what camera/lens? The lens on the front of the Super16 camera can make a big difference in the recorded image sharpness.

For stock, I haven't shot Vision 3 or any film for a few years now; my last favorite stock was Vision 2 250D. I would go with daylight balanced stock because the sensor in the 7D is likely to be daylight-balanced as well; testing will confirm this and if it's actually tungsten then I'd say go with a tungsten stock. 250 is fast enough to get something decent shot without being overly grainy; 500 in 16mm starts to get wicked grainy, although being Vision3 it should help.

400' shouldn't cost too much, a roll should be about $130 US, processing shouldn't be too much, but transfer can be expensive. Find out what transfer houses are nearby and what they offer, and plead the "please, it's just a test, we'll mention your lab if you help a brother out" line and you might get a great price break. For "Essence of the Force" we got a $900/hr Spirit for, heck, I can't remember, probably around $300/hr but that's been a while... 400', properly color timed, should take about a half an hour. If you're shooting under controlled lighting circumstances and you throw a Macbeth or, better yet, a DSC Chroma Du Monde in the shots, you should be able to do a one-light transfer and that should be cheaper and faster still.

For testing purposes I wouldn't really want to do a lot of color correction at the transfer phase, because you're trying to find out how the film records colors, right? On the other hand, the color correction you can do at the transfer phase will be far more robust than anything an h.264-recorded 7D file could manage, so if you want to show the range of options, you might want to experiment heavily with CC at the transfer house. Just be aware that it's probably going to cost a pretty penny. Bono Labs in Virginia will do a telecine to uncompressed HD files on a hard disk for $175 for 15 minutes; if you shoot a Chroma Du Monde at the head of the reel and don't change the light color temperature at all, you might be able to get it all done in their 1/4 hour charge.

There's really no point in doing it if you're going to have the film telecine'd to BetaSP or DV; the minimum should be DVCPRO-HD or HDCAM, and the best way would be raw data files.
 
Barry_Green said:
Perhaps the most significant thing is -- what are you going to do with the film after it's shot? If you transfer to a regular video format or HDV or something like that, you're going to be giving up a lot of what the film has to offer.
Click to expand...
that depends on your workflow. if you do a one light you're right of course, but if you do scene by scene color correction on the dailies you can extract most of the good stuff. of course that costs money but if it's a short film not necessarily so much, i've done it several times. you can also do a technical grade best light, but going to 10 bits or at least low compression 8-bit is really a must then. it's hard to do much with hdv after the fact.
 
Thanks for that info Barry.

The local house is Technicolor. As a film student at Concordia university, we get some pretty good rates. But you're right, i want to get the highest quality transfer possible. I expect to shoot under a minute of stuff on super 16mm, and plan on using either Vision 3 500t or 250. The lenses will have to be whatever the school has available to me, but they do have some PL mount lenses so i think in comparing my nikon primes on the 7D to the PL mount lenses, i think although they may not be cinema standards, they will work just fine for the test.

I will start a thread when the process begins and will look forward to some guidance. Thank you for your input!
 
it's not just about "quality". video gamma is very different from film and the s/n ratio is lower, so you need to make some compromises, do you want to clip the highlights? compress them? do you want to keep them saturated or fall off to white, and so on, same for the blacks and everything else. you *will' be throwing away information, so you'd better decide which information is redundant for the look you want and get rid of that. better spend money on a great colorist than a great machine, if you can't have both. done correctly the film quality will be kept even if you're viewing it on a qvga cell phone.
 
You know,

I was thinking about this a bit. Why do you want to spend your money on this? I mean, it really won't prove anything. I'm not saying "you're stupid!", but I'm just interested.

Panavision did this test, BSC did this test, [Zacuto? (maybe cinvate)] did this test. And really, nothing new comes out of it. I guess I'm just curious as to why you want to do this test. To just do it? Personal interest?
 
Sure call it personal interest. I prefer than then saying I'm stupid. :)

Did they do the test with the 7D, cause thats what im interested in seeing. Especially knowing what kind of curve/saturation adjustments to make in order to match them as closely as possible.

Matt, would you care to expand a little bit about what you were saying regarding what information i would want to throw away?
 
Mattykins said:
Well they haven't run a 7D, but I can't imagine it will be much different than what we already know.
Click to expand...

Is there a picture style out there that can be downloaded to emulate a specific film stock on Canon DSLRs?

Not trying to be a smartass here -but that seemed to be PaPa's goal in all this, which I do find kind of interesting. Or did you simply mean that we already have most of the information we need in order to emulate the look of 16mm film?
 
Mattykins said:
Panavision did this test, BSC did this test, [Zacuto? (maybe cinvate)] did this test. And really, nothing new comes out of it. I guess I'm just curious as to why you want to do this test. To just do it? Personal interest?
Click to expand...

Ok, where can I see those tests?

I was curious to compare the 7D to film myself, but I was planning a test much less ambitious... My plan was to shoot side by side with a traditional SLR, and have the film developped to CD.

BTW, if you don't have enough money for a film scan or telecine, you should consider the ghetto approach: a slide scanner.

-- X
 
I only meant there is a ton of information out there. I'm not trying to down-talk his idea. I just am trying to understand why one would invest their own money (and a good amount of it) to get more information on something that's been pretty well covered by professionals in professional and controlled settings.

Almost every digital cinema system (and a good handful of the VDSLRs) has been put to the test against film before. I just assume use the information that is already out there. But that's just me.
 
xbourque said:
Ok, where can I see those tests?

I was curious to compare the 7D to film myself, but I was planning a test much less ambitious... My plan was to shoot side by side with a traditional SLR, and have the film developped to CD.

BTW, if you don't have enough money for a film scan or telecine, you should consider the ghetto approach: a slide scanner.

-- X
Click to expand...


http://synapticlight.com/2009/01/08/zacutos-great-camera-shootout-08/

http://www.bscine.com/my_documents/my_files/BSC_Evaluation_Shoot_2009_web.pdf (This link is worthless)

On the top row are 435 and 416 Cameras with 2 different film stocks each from Kodak and Fuji with all the digital cameras below which are:
Genesis
D-21
Red
Sony F 35
Silicon Imaging 2k
Sony 900R
Viper
Panasonic 3700
Sony EX3
Panasonic HVX 201
Canon 5D Stills Camera
This was the culmination of our evaluation where we shot the Daylight exterior set up simultaneously on 19 different formats. The rig was designed by Joe Dunton and was the culmination of the week. A very successful
and ambitious shoot.
Over the week we shot four different set ups on all the formats seen here, day and night interiors, close up , and day exterior and just for good measure here is a still taken of the crew that were filming the exteriors on the Friday of the shoot.
Click to expand...

I can't find the Panavision one. But both Zacuto and the BSC rocked film and digital together (including the 5D in both cases). I just can't foresee the 7D would be vastly different.

I was just curious as to why he was doing it, that's all.
 
Last edited:
Thanks for those links, Mattykins.
Very cool.

(by the way, I don't think anyone was questioning whether or not the test occurred)
 
There's some value to a test like this, only cuz a lot of the visual comparisons are with 35mm film, not 16mm. Though the more information you can provide on the 16mm end, the more value it would have to me. Some details if you can record them:

-What camera you used..
-What lenses used? I would suggest Zeiss Mark II's primes or Mk 1's if your school does not have the II's. Old angenieuxs and old zeiss zooms tend to be too soft.
-Super 16mm if possible, since it will match closer to the aspect ratio of 1920x1080.
-What did you rate the stock as, and what was your exposure?
-Where did you get the transfer done? There's huge amount of variables at any transfer house. Just because some place has some "rank cintel" doesn't mean much of anything.
-Ask for 2K scans, or some lesser resolution uncompressed to your firewire drive. It isn't as expensive as it used to be.

You don't have to shoot 400' for a test, just shoot 200' or less and short end the roll. If you can't afford a top notch transfer, I wouldn't bother to do this,
 
Last edited:
I will be going through with the test. I will record all of those things and get the footage up for everyone to play with so that they can make their own comparisons as opposed to judging this through youtube. Should be exciting :)

But i will want to get the highest possibly quality transfer (within reasonable budget constraints), so most of the money will definitely be going towards that.
 
You must log in or register to reply here.
Back
Top