Demos comparing 35mm quality to different digital cameras

Forum covering all aspects of small gauge cinematography! This is the main discussion forum.

Moderator: Andreas Wideroe

Daniel
Posts: 391
Joined: Wed Feb 05, 2003 12:17 am
Location: Chile

Demos comparing 35mm quality to different digital cameras

Post by Daniel »

Hello,

The company where I work since two years ago, did produce in 2013 a demo that compares 35mm (Vision3) scanned at 4K RGB 16 bits log/flat (TIFF files) via a DFT Scanity with Arri Alexa (recording in Pro Res 4x4) and Red Epic-X MX (R3D files).
Same glass, ISO settings, T stop and fps. The actress performed the 'same' action three times. In the fourth shot (daylight), the 35mm footage still had more information in the highlighs that we could have recovered through a luminance key, but we wanted that the color grading to only be a primary color grading with in total only one node (per each take) for the color grading (that was performed through a Da Vinci Resolve).

https://vimeo.com/105584682

Serious analyse of image quality through Vimeo is not possible, I agree, but some differences between film and the two digital cameras are still noticeable.
We do have a 35mm print of this test demo, that was done at Weta Digital and Park Road Post in 2013.
The Vimeo version has the 3D LUT burn-in (that we used to emulate the film-out via Arri Laser 4K to Fuji 4511 / and contact print to Kodak 2383)

We produced another demo, this year, that will be released soon.
This year we compared 35mm (Kodak Vision3) scanned at 2K RGB 10bit log (DPX files) and three digital cameras recording RAW Arri Alexa (ARRIRAW), Sony F65 and Red Epic-X MX.
Once it is released I will post the link.

Of course in both cases (in my opinion) film looks better!

All the best,
Daniel
Tscan
Posts: 548
Joined: Fri Sep 03, 2010 6:44 pm
Real name: Anthony Schilling
Contact:

Re: Demos comparing 35mm quality to different digital camera

Post by Tscan »

The RED camera looks pretty bad out of the three. The Alexa comes close, but the film image is still more pleasing... all technical factors aside. The most noticeable difference in my eyes is that film has a more matted texture which i find more pleasing, and digital is more glossy and less pleasant to the eyes. The difference reminds me of digital audio, with too much treble and not enough bass.
Reborn member since Sept 2003
JeremyC
Posts: 153
Joined: Fri Jun 01, 2012 2:51 pm
Real name: Jeremy Cavanagh

Re: Demos comparing 35mm quality to different digital camera

Post by JeremyC »

The RED's flesh tones looked pink almost as though it needed a white balance and I noticed it couldn't handle the highlights on the lounge/ocean background set up. The Alex looked slightly plasticly in its flesh tones and I was surprised at the lack of grain in the 500T example.

However, I am looking at these images on a lap top - not anywhere near a critical viewing device and with both the Alex and the RED I would have to know how each of them have been set up/adjusted, something I am very familiar with broadcast cameras but not with sensors set up for cine work.

But heartening to know I could just throw my 20 year old S16 camera at these set ups and not have to worry about tweaking anything once the lighting is set up.
User avatar
S8 Booster
Posts: 5857
Joined: Mon May 06, 2002 11:49 pm
Real name: Super Octa Booster
Location: Yeah, it IS the real thing not the Fooleywood Crapitfied Wannabe Copy..

Re: Demos comparing 35mm quality to different digital camera

Post by S8 Booster »

the film looks much better and seem to cover a wider range of high and low light. the look is by far more aesthetic. the digital/digital cams imagery looks more or less constructed/designed and look un-natural by comparisation.

surely there may be more to get through digital/digital cam post processing but as presented here the film seem more natural and balanced.

on the other hand i suppose none of us would be too turned off at a cinema seeing the cuts projected since there were no comparisation.
..tnx for reminding me Michael Lehnert.... or Santo or.... cinematography.com super8 - the forum of Rednex, Wannabees and Pretenders...
carllooper
Senior member
Posts: 1206
Joined: Wed Nov 03, 2010 1:00 am
Real name: Carl Looper

Re: Demos comparing 35mm quality to different digital camera

Post by carllooper »

If the various variables involved (ISO/ASA, aperture, shutter, scene contrast, etc) completely characterised each of the camera/chip/film tests, AND the values of each reference variable were all the same reference values, then there would be no difference at all between the images.

The fact that there is a difference means the reference variables employed may not (or can not) completely characterise each of the tests. And/or the systems in use are not yet calibrated with respect to such reference variables.

There is a fundamental flaw in this "comparison". China girl tests are used to calibrate imaging systems. To use the same tests to compare imaging systems is flawed (or silly). For example, if I find from a test I'm getting blown highlights on a shot (or not getting them as the case may be) where I otherwise wanted something else, then I would use the test to re-calibrate the system, ie. alter some setting in the lighting, camera, light meter, chip post-filtering algorithm, post-production process, script, direction, concept, etc.

For example, I test a Super8 camera and find in the result that it's underexposed due to not taking into account light lost to the viewfinder. Should I change cameras as a result of this test? Well sure, but I could also calibrate the system to ensure the result isn't underexposed the next time around. Or I could change the concept of the film to one that requires underexposed shots. Or I could take into account the light lost to the viewfinder. Either way each of these is what the test is designed to reveal, ie. that which needs adjustment.

Or lets suppose I'm getting more red in an image than I should be according to a reference test (and what I'm otherwise after). That means the camera/film system being tested requires some tweaking in order to remove that bias. That's the purpose of the test. It's called calibration. If you don't do the calibration then the tests have no meaning. To what end is the test? This is the important question. And if you do the calibration (for which such tests are designed), then the comparison has no meaning.

For it is only when all the images become exactly the same that each of the systems can then be said to be fully calibrated, ie. with respect to an ideal reference system. Which is why it is conceptually silly to use such tests to compare systems because in the technically correct end case (ie. ideal full calibration) the outcome will be no difference between the images at all. Nothing to compare. And in the non-ideal case (real world systems) there will remain differences for which there is no common reference value, ( = no way of making the images identical). The list of reference variables in the provided tests are only those in common (ASA/ISO, aperture etc), ie. the list hides those that are not. And for obvious reason because those not in common are incompatible. The amount of cheese used in each test is not capable of occupying the same formula as that which accounts for the amount of chalk used.

By way of example of the differences, there are tests one can do to see how long film should be in the developer. Applying the same test to a camera chip we find that no matter how long we soak the chip in the developer it still doesn't work. We could argue from this test that film is therefore better. Indeed I'd be a lot happier arguing the case for film on these surreal grounds, than anything in these China girl tests. Or after a long day on the word processor, we might find ourselves at night, doing some drawings on a piece of paper, and upon making a mistake, we inadvertently find ourselves looking for the undo button ... from which we might conclude (rightly or wrongly) that drawing on a computer is better.

So all we're left to appreciate in these empty tests is a lack of any calibration, and/or imagination, and the purpose (or repurposing) of which results in yet another phony film/digital comparison.

What is lost is an appreciation of the underlying systems and the genius/ingenuity of each system. The madness under the surface and the practice they require. That which the artist/magician entertains. Biting the bullet rather than standing endlessly paused before a test, wondering what to do or say next.

C
Carl Looper
http://artistfilmworkshop.org/
User avatar
MovieStuff
Posts: 6135
Joined: Wed May 01, 2002 1:07 am
Real name: Roger Evans
Location: Kerrville, Texas
Contact:

Re: Demos comparing 35mm quality to different digital camera

Post by MovieStuff »

You are exactly correct, Carl.

Qualitative "raw" image tests that apply no calibration are rather pointless with the possible exception of resolution, which is a quantitative measurement. The only real value of having access to a variety of raw captures from different sources is to document what was required to get each of them to look "normal". This presumes, of course, there is a standardization for what everyone considers normal. For instance, people that shoot 8mm film a lot might like the way Kodachrome looks and would like to see raw digital have the same appearance. Those that shoot color negative a lot might think that was a terrible idea and embrace a raw digital image that looked flatter and more like color negative. If someone came out with a digital format that looked like neither in a raw format but could look like either after appropriate calibration, does that mean that digital format isn't as good as either Kodachrome or color neg? The truth is that color negative looks simply terrible when transferred unless someone comes along and grades it properly. The same holds true for Kodachrome. That's why calibration is required. Therefore, to compare raw images of different formats to make a qualitative judgment is a waste time.

Roger
Daniel
Posts: 391
Joined: Wed Feb 05, 2003 12:17 am
Location: Chile

Re: Demos comparing 35mm quality to different digital camera

Post by Daniel »

BTW, ProRes 4x4 is a video format not a RAW format.

This is not a test trying to show the best imagery you can get from each camera for each shot, then yes you could change the lighting or place ND in the windows etc. This first test is based in that there are no lighting modifications, lens modifications etc. To get the same stop in the lens, we decided to align the ISO of the digital cameras to the film stocks (500T or 250D depending on the shot). Then we lighted and worked as we do in 35mm, and then just replaced the film camera by two different digital cameras.

Concerning the debayering parameters of the Red Epic media, we used REDlog / REDColor2.
The debayering was done through da Vinci Resolve (version 9x) using Resolve debayering not RR board, as we found it cleaner.

In post-production, we limited the color correction to primary color grading, specifically primary color grading + overall saturation using one node per each take. The colorist used the 35mm shots as reference, so that each take could have at least the same grading intention. Another test could have been to do a specific, completely different grading for each take, using secondaries, but this was not the intention of this test.

We then made a film-out of this test to finally watch the 35mm print. The web version has the 3D LUT we used for film-out emulation, burn't-in.

This test is pretty simple, the method we choosed for this test worked for us to compare results, knowing some results could have been different if other methodology was taken.
On another hand this test is just another one, that shows the quality of 35mm film in terms of highlight handling, flesh tones and that pleasant feel.
carllooper
Senior member
Posts: 1206
Joined: Wed Nov 03, 2010 1:00 am
Real name: Carl Looper

Re: Demos comparing 35mm quality to different digital camera

Post by carllooper »

Daniel wrote:BTW, ProRes 4x4 is a video format not a RAW format.

This is not a test trying to show the best imagery you can get from each camera for each shot, then yes you could change the lighting or place ND in the windows etc. This first test is based in that there are no lighting modifications, lens modifications etc. To get the same stop in the lens, we decided to align the ISO of the digital cameras to the film stocks (500T or 250D depending on the shot). Then we lighted and worked as we do in 35mm, and then just replaced the film camera by two different digital cameras.

Concerning the debayering parameters of the Red Epic media, we used REDlog / REDColor2.
The debayering was done through da Vinci Resolve (version 9x) using Resolve debayering not RR board, as we found it cleaner.

In post-production, we limited the color correction to primary color grading, specifically primary color grading + overall saturation using one node per each take. The colorist used the 35mm shots as reference, so that each take could have at least the same grading intention. Another test could have been to do a specific, completely different grading for each take, using secondaries, but this was not the intention of this test.

We then made a film-out of this test to finally watch the 35mm print. The web version has the 3D LUT we used for film-out emulation, burn't-in.

This test is pretty simple, the method we choosed for this test worked for us to compare results, knowing some results could have been different if other methodology was taken.
On another hand this test is just another one, that shows the quality of 35mm film in terms of highlight handling, flesh tones and that pleasant feel.
Hi Daniel,

the tests may be simple but what is the purpose of the test? To what end? What is the intention? This is the single most important question. Without such an intention the tests have no meaning. You say "This is not a test trying to show the best imagery you can get from each camera for each shot"

Ok. So what is the test for?

You say: "On another hand this test is just another one, that shows the quality of 35mm film in terms of highlight handling, flesh tones and that pleasant feel."

On the other hand? To what? You haven't stated any other reason for the test.

When the workers at Bletchley Park, during World War 2, performed tests on enigma machines, they too would have said the tests they carried out were not tests trying to show the best encryption you can get from each machine for each message. But in their case, they actually had an alternative reason for the tests. The tests they did were designed to determine what could be adjusted in the machines in order to transform gobbledegook back into a message.

And then those adjustments were made

It is this follow through act, implicit in any correlation test, that is the purpose of such tests. Without such an intention (the follow through act), all you are left looking at is just an expression of the problem: the gobbledegook.

I should say that I love the look of film. But digital can't be measured using the same yardsticks. It is designed with the concept of change in mind. For example, if the gradients on a face look unpleasant to you, you can alter it. You don't have any where near the same kind of freedom with film. Film is the result of countless tests having already having been done towards some notion of what the ideal film image should look like. What we might appreciate in film is the result of a 100 years of such tests and calibrations. All the work has already been done. But digital is a lot more relaxed in the sense that it doesn't have to conform to notions of what an ideal image should look like ahead of time. The end result can be deferred and re-twigged in the pipeline. As a result it is a lot more intention orientated. And it is designed with that in mind. It is not as "hardwired" as the film pipeline. Film, for it's part, insofar as it has worked within the framework of producing (as close as possible) a ready-made result, decides in advance what an ideal image should look like. And locks down a lot in the process. There are conceptual and philosophical differences between film and digital, which, if ignored, just lead to arbitrary comparisons, at arbitrary junction points, in the production pipelines of each.

C
Last edited by carllooper on Wed Sep 17, 2014 2:47 am, edited 1 time in total.
Carl Looper
http://artistfilmworkshop.org/
Daniel
Posts: 391
Joined: Wed Feb 05, 2003 12:17 am
Location: Chile

Re: Demos comparing 35mm quality to different digital camera

Post by Daniel »

Carl,

The primary intention is just to compare the different results one would get, using film or digital devices like the mentionned ones, when used in the same scene setup, with the same glass, T stop, and lighting conditions through a specific post-production workflow. As it is the same scene shot three times, one have some basis to compare A, B or C.
Why comparing ? Because we think this is of interest of D.o.P. Directors, Producers etc. In fact when we projected the 35mm print, last year, an interesting conversation followed it. To some extent, the results of the test, do make tangible the (great) properties of film, that seems to be forgotten by many people in our industry.

You may have another method for comparing, as you have suggested, or another intention when preparing your own tests, and to my opinion that's just fine.

Daniel
carllooper
Senior member
Posts: 1206
Joined: Wed Nov 03, 2010 1:00 am
Real name: Carl Looper

Re: Demos comparing 35mm quality to different digital camera

Post by carllooper »

Daniel wrote:Carl,

The primary intention is just to compare the different results one would get, using film or digital devices like the mentionned ones, when used in the same scene setup, with the same glass, T stop, and lighting conditions through a specific post-production workflow. As it is the same scene shot three times, one have some basis to compare A, B or C.
Why comparing ? Because we think this is of interest of D.o.P. Directors, Producers etc. In fact when we projected the 35mm print, last year, an interesting conversation followed it. To some extent, the results of the test, do make tangible the (great) properties of film, that seems to be forgotten by many people in our industry.

You may have another method for comparing, as you have suggested, or another intention when preparing your own tests, and to my opinion that's just fine.

Daniel
I added this to my previous post, not realising you had made a new post:

I should say that I love the look of film. But digital can't be measured using the same yardsticks. It is designed with the concept of change in mind. For example, if the gradients on a face look unpleasant to you, you can alter it. You don't have any where near the same kind of freedom with film. Film is the result of countless tests having already having been done towards some notion of what the ideal film image should look like. What we might appreciate in film is the result of a 100 years of such tests and calibrations. All the work has already been done. But digital is a lot more relaxed in the sense that it doesn't have to conform to notions of what an ideal image should look like ahead of time. The end result can be deferred and re-twigged in the pipeline. As a result it is a lot more intention orientated. And it is designed with that in mind. It is not as "hardwired" as the film pipeline. Film, for it's part, insofar as it has worked within the framework of producing (as close as possible) a ready-made result, decides in advance what an ideal image should look like. And locks down a lot in the process. There are conceptual and philosophical differences between film and digital, which, if ignored, just lead to arbitrary comparisons, at arbitrary junction points, in the production pipelines of each.

Certainly selling the merits of film vs digital to "industry clients" is another use of a China Girl test. But it's a phony one because it hides the intention (to sell the merits of film vs digital), behind the facade of a technical test otherwise used to make adjustments to the very systems supposedly in question. It is a misuse or abuse of those tests. One should declare in advance the purpose of the test, rather than add it in at the end, as if it were just a happy (or sad) side effect of the test.

A camera/film system is just one component in the entire imaging system, which, when isolated in this way can easily lead to arbitrary conclusions such as film is better. Or digital is better. Neither are better or worse than the other. There is no real technical test that can decide that. There is only what you want to do with each or either. It is in relation to that which you would use such tests. (Or towards creative ends which I've seen). Any other use is just bullshit. But then that's what a lot of industry clients might only be capable of understanding.

I'm trying to think of another analogy. Suppose we're comparing two power supplies, both labelled 30 volts. But when we measure one we see that it's only 20 volts where the other is 30. Does that mean the 30 volt supply is better? Should we take this discovery to a meeting of industry heads on which power supply is better? Do we put up a power point show with graphs showing that power supply X is underpowered? No. It just means we change the label on the 20 volt supply to 20 volts. Or we increase the voltage it supplies to 30 volts. We bring the descriptions in line with the objects or vice versa. Towards identity between objects or between objects and their description. This is the end result - nothing left to distinguish them. The tests are towards such an end. Not the expression of difference but the expression of identity - in which there is no difference. But difference is incredibly important - but it's not with identity converging tests that such can be expressed. Not in the end. One can see that end game in advance. It's a dead end for any expression of difference.

Hope that helps to make sense of where I'm coming from. It's from the edge of the universe. Rather than a boardroom discussing which equipment to purchase.

C
Last edited by carllooper on Wed Sep 17, 2014 4:02 am, edited 1 time in total.
Carl Looper
http://artistfilmworkshop.org/
User avatar
MovieStuff
Posts: 6135
Joined: Wed May 01, 2002 1:07 am
Real name: Roger Evans
Location: Kerrville, Texas
Contact:

Re: Demos comparing 35mm quality to different digital camera

Post by MovieStuff »

Well, to me, the only point of a presenting comparative original imagery is if none of the imagery can be adjusted. It would be like shooting Kodachrome, Agfachrome and Ektachrome slides and putting them all in a slide projector to see how each handled the same subject. That would be a legit test because you are looking at the original frame directly with no adjustments intended or needed.

But whether you are shooting color negative or shooting digital, there is zero intention of showing any audience the original captured neg or digital file so what's the point of presenting the originals side by side for an aesthetic comparison? The proper way would be to have the same subject and, as Carl has pointed out, calibrate or adjust the originals until they all have the same desired look, if possible. To whatever degree the same look is achieved -or falls short- the only usable information is a record of what efforts it took to achieve that look from each original. Otherwise, it's like looking at a kitchen counter full of ingredients for two different recipes of chicken soup and asking which recipe tastes better. Until it's finished, there's no way to tell.

Roger
carllooper
Senior member
Posts: 1206
Joined: Wed Nov 03, 2010 1:00 am
Real name: Carl Looper

Re: Demos comparing 35mm quality to different digital camera

Post by carllooper »

MovieStuff wrote:Otherwise, it's like looking at a kitchen counter full of ingredients for two different recipes of chicken soup and asking which recipe tastes better. Until it's finished, there's no way to tell.Roger
That's a perfect analogy.

C
Carl Looper
http://artistfilmworkshop.org/
Daniel
Posts: 391
Joined: Wed Feb 05, 2003 12:17 am
Location: Chile

Re: Demos comparing 35mm quality to different digital camera

Post by Daniel »

Carl,

This test is not designed to choose a camera to purchase nor to sell you or anybody anything. And the purpose of sharing it, is to present it to people that do shoot film as well as digital, as simple as that, just professionals, film students, or filmmakers like in this board. People that share a genuine interest in filmmaking.

The interpretation of the results is personal and depends on what one is looking for as technical parameter or aesthetical consideration.

The test is as much neutral as we could; and there was not a special treatment for film over digital.

The exposed/undeveloped Vision3 stock was sent to be processed via courier in another country to emulate a 'real' workflow in case a Lab is not in town, so in other words, film stock wasn't developed same day, but about 5 days later. Granted, we developed on a first quality lab (FotoKem).
Concerning the debayering of the Red Epic media, we used highest decoding quality available on the Da Vinci Resolve software.

All in all, the results are what they are, once again, for the choosen methodology. Besides the inherent limitations of the uploaded vimeo version (very compressed) each image system seems to offer a type of image that has its own 'identity' - however beyond that, to me it is very clear that two of them (Alexa and Epic) share some artificial looking specially in the flesh tones compared to the film imagery.

Concerning what you wrote :
''if the gradients on a face look unpleasant to you, you can alter it. You don't have any where near the same kind of freedom with film''
Well in my experience this is just exactly the contrary. I can attest that there is much more freedom during the color grading (assuming digital intermediate post production) starting with a proper exposed and developed modern color negative stock (like any Kodak Vision3 for instance) scanned at 16bits than with any digital camera I know. Have you tested lately Vision3 scanned at 16bits ? I think it is worth a look.


Roger,

The good news is that some ungraded frames could be uploaded (by next week), if anyone is interested.


Daniel
carllooper
Senior member
Posts: 1206
Joined: Wed Nov 03, 2010 1:00 am
Real name: Carl Looper

Re: Demos comparing 35mm quality to different digital camera

Post by carllooper »

Daniel wrote:Carl,

This test is not designed to choose a camera to purchase nor to sell you or anybody anything. And the purpose of sharing it, is to present it to people that do shoot film as well as digital, as simple as that, just professionals, film students, or filmmakers like in this board. People that share a genuine interest in filmmaking.

The interpretation of the results is personal and depends on what one is looking for as technical parameter or aesthetical consideration.

The test is as much neutral as we could; and there was not a special treatment for film over digital.

The exposed/undeveloped Vision3 stock was sent to be processed via courier in another country to emulate a 'real' workflow in case a Lab is not in town, so in other words, film stock wasn't developed same day, but about 5 days later. Granted, we developed on a first quality lab (FotoKem).
Concerning the debayering of the Red Epic media, we used highest decoding quality available on the Da Vinci Resolve software.

All in all, the results are what they are, once again, for the choosen methodology. Besides the inherent limitations of the uploaded vimeo version (very compressed) each image system seems to offer a type of image that has its own 'identity' - however beyond that, to me it is very clear that two of them (Alexa and Epic) share some artificial looking specially in the flesh tones compared to the film imagery.

Concerning what you wrote :
''if the gradients on a face look unpleasant to you, you can alter it. You don't have any where near the same kind of freedom with film''
Well in my experience this is just exactly the contrary. I can attest that there is much more freedom during the color grading (assuming digital intermediate post production) starting with a proper exposed and developed modern color negative stock (like any Kodak Vision3 for instance) scanned at 16bits than with any digital camera I know. Have you tested lately Vision3 scanned at 16bits ? I think it is worth a look.


Roger,

The good news is that some ungraded frames could be uploaded (by next week), if anyone is interested.


Daniel
Hi Daniel,

I'm not the one suggesting you are trying to sell anything. You are. You are selling the idea that the test is neutral and from such we might conclude that film looks better. Film may very well look better but this test does not make any contribution to that conclusion in any way whatsoever. And if it's not buying into the film/digital debate then what is the test doing? That's the big question. If the test is not to sell some notion of quality between different systems, what is the test? It has no other purpose other than the one and only one that you yourself have suggested.

But such a suggestion completely misunderstands the purpose of such tests. The tests are not to suggest which system is better or worse at rendering some notional concept of quality. Although, of course, that's how you are using such tests. But such a use is bogus. The history of tests like this is calibration. It is not some framework for the assessment of which camera does a better job. The purpose of such tests (and I'm thinking of the colour chart in particular) is to determine a reference for how one might then adjust the camera and/or a range of other points along the entire imaging pipeline, in order to obtain or maintain the reference values. If you don't do this, or even try, the test is meaningless.

What is the point of the colour chart in there? If you've done the work correctly all the colours should look exactly the same. It's like you've set up the test but you haven't followed through on the test. You haven't solved anything. You are still looking at the problem and not the solution. But worse than this, you are using the problem to suggest something, when it doesn't suggest anything other than you have more work to do. You need to do that work. Until you do the work you can't make any statements about film vs digital. Don't worry about biases. Bias is exactly what you are have to introduce. Bias each image towards the reference value. Check the pixel values against your refernece values and grade the image. Do the HDR toning on your digital files. Do everything that is done in each domain. The machines themselves are designed for such. One day they might think for themselves but for the time being we end up doing the thinking for them. And thans goodness I say. Consider an aperture. What is an aperture if not a way of biasing the amount of light falling on the film in favour of a particular type of result? Now some bright spark came up with the idea of auto-exposure - the idea that machines could do some thinking for us. A built-in light meter that adjusts the aperture for you. Ha ha. But it was nevertheless a great idea. Auto-exposure has come a very long way since. But it still requires some genius to think up such things and make changes as required. And that includes us as end-users as well. We can't just sit back and just assume the machine should do everything out of the box. Or we could but we'd have a hard time calling ourselves artists or "professionals" if we did that.

Grading. This is more difficult in film. You say that, on the contrary, there is more freedom in film - but then you go on to assume a digital intermediate. THAT'S DIGITAL. Try finding the same freedom on a contact printer or an optical printer. THAT'S FILM. And it's not possible, using such equipment ,to make anywhere near the range of adjustments you might otherwise entertain in digital. On the other hand, most of the time, you don't have to. The point I'm making is that digital is very plastic and open ended. Open to whatever notions of quality you might like to entertain. If you don't like the look of film you can go for something else. And if you do like the look of film well you can always try and achieve that. I'm not suggesting digital is better in this regard. Only that it has a different philosophy. If you don't like the particular look the presets in a camera has given you, you can change it. The fortunate thing with film is that for many of us (myself included) I don't want to change much about it at all. It already looks good to me. I love film. But I'm also very much a digital age guy as well. I walk both sides arguing for both.

To what end are your tests? If I'm misunderstanding then elaborate. Because if it's not to buy into the film/digital debate (where it fails), nor is to calibrate the cameras and workflows in question (which it hasn't but which is something you could do), then what on earth are the tests for?

C
Last edited by carllooper on Wed Sep 17, 2014 4:56 pm, edited 1 time in total.
Carl Looper
http://artistfilmworkshop.org/
Daniel
Posts: 391
Joined: Wed Feb 05, 2003 12:17 am
Location: Chile

Re: Demos comparing 35mm quality to different digital camera

Post by Daniel »

Hi Carl,
What' is the point of the colour chart in there?
The X-Rite Colorchecker Digital SG and Colorcheker color charts were used as some reference of color and gray scale, just to have some known patches in the pictures.
During the grading these patches were used for the primary color grading, but also the flesh tones of the actress. The color grading was a compromise between the color charts and the flesh tones.

Aside from this answer, I don't see any need to elaborate anything else. I just respect that anyone could find this test without any interest or empty/wrong purpose. That's just fine.
On the other hand, if one read the very first 2 or 3 feedbacks of this thread, it seems to me that whatever the methology choosen, the test communicates some differences and/or similarities between the three image system employed.

Thanks anyway for your feedback!
Daniel
Post Reply