Making HD video look like black and white film - Version 3

[Scotness]

This is my 3rd version and the best so far. I've uploaded it in 1080P - and you need to see it at that resolution to see the grain effect properly.

Anyway changes are: the mts->avi conversion was done with Freemake which did not add any frame blending (unlike the other one on #1 and #2) - and also the brightness and contrast of the grainmap was lifted which gave a nicer result.

http://www.youtube.com/watch?v=9ZfGNPMRh-8

Scot

[Will2]

Have you tried shooting the same scene in film and comparing? That might give you additional insight.

The issue for me seem to be the burned out highlights on the cars.

And what type if film are you going for? Super 8, 16 or 35?

[Scotness]

Have you tried shooting the same scene in film and comparing? That might give you additional insight.

That's a great idea - I might try that -- I'm going for Super 8 or 16 - haven't really decided - just something rough -- good point on the highlights too -- I think the compression on YT has made them a bit worse, but definitely something to watch

Thanks
Scot

[Will2]

That's where the ultra-flat curve of Arri-raw and RED raw files come in. That would translate into fake-film much better than the average DSLR or video camera. Of course it becomes a strange round about situation if you have the money to use those cameras why you'd want to emulate poor film. I think there is a 5D curve setting from Technolcolor that might help you achieve a flatter look.

Depending on how far you want to go, maybe you could consider just a little tiny bit of projector flutter. Not enough to make it look like it was projected but just a tiny bit to subconsciously make it seem more like film.

[carllooper]

The trend in digital filmmaking for the cinema is, thankfully, away from simulating a film look, away from pretending to be film. Looking at Avengers and Prometheus, for example, the look is a very pure signal. Video like but in a definition so high it's like looking at 'reality'. The special effects of those films now aim to match this pure signal rather than the other way around. And it's quite beautiful in it's own way. Gone are the fake grain and god awful "arty" colour tinting (as in Sherlock Holmes or Sucker Punch).

Each medium should be respected (methinks) for what they achieve in their own particular way rather than how well they might be persuaded to simulate the look of some other medium. The worst of early photography, apart from a kind of novelty value, were those photographs that tried to look as if they were paintings. Today they are just ludicrous. But worse, they just look kind of unhealthy. Ill. Bedridden. Gathering cobwebs. Mouldy. Fake. Empty.

Carl

[Scotness]

I kind of agree with what you're saying there Carl, but I think my approach is valid too - because if you do most of the film look processes to video you don't get film - you get the best looking video you can. You only have to look at the Miami Vice reboot to see why that is so. But anyway needless to say if I could afford to do this particular project on film I would

Scot

[hirudin]

Sorry in advance (thread hijack)

Carl, just to confirm... whether it's film or 4K digital, there's grain; there's no system that doesn't have noise. Case in point, I recently received plates that were billed as noise free 4K capture, and I gotta say, grain definitely had to be added to the visual effects to match the plates, particularly in the blue channel... from in terms of visual effects, workflow doesn't differ at all in the approach with film and hasn't changed: whatever the origination, we match to the plates. And I think you'll find the drastic colour timing will remain, digital or film...

http://theabyssgazes.blogspot.com/2010/ ... -stop.html

Thanks!

-Jesse-

[carllooper]

I'm not against processing film, or video/digital beyond what the camera/film/data contributes. Indeed a lot of what I do myself is precisely about enhancing the camera original (mainly film) beyond what it otherwise gives, be it through techniques on location (Iighting, framing, etc) or in the studio (colour correction, stabilisation, software, etc).

It's the nature of that enhancement that is interesting. Is it towards fooling the mind, or enlightening it? Is it a forward step, into some relatively fresh or new idea, or just conforming/reinforcing some old idea, or convention. Is it just business logic speaking or is there creativity in there also. Is it an idea that has it's roots in photography or in painting/graphic art? Does it matter? And so on.

Carl

[hirudin]

In terms of the effects, it's to give the appearance that they and the original occupy the same space... without that there is no suspension of disbelief. Beyond that, if the director is doing their job, it's in service of the story and contributes in a constructive way.

But that's not always the case...

[carllooper]

In terms of the effects, it's to give the appearance that they and the original occupy the same space... without that there is no suspension of disbelief. Beyond that, if the director is doing their job, it's in service of the story and contributes in a constructive way.

But that's not always the case...

Special effects are really interesting. What is a special effect? Is an unembellished photograph itself a special effect? Is a special effect everything you do to a photograph or only some of what you do?

A lot of contemporary theory is confused in this regard. It treats photography and what you do to it, or add to it, as all of the same thing. There are no special effects anymore because everything is an effect - there are no effects which are special.

A lot of this attitude is born of the "information age", in which most images, irregardless of how they were created, end up as digital data: one's and zeros. Within this digital domain it all seems to be the same thing. So much so that theorists now think this sameness extends back throughout the entire system - that if there is no observable difference at the level of information (it's all one's and zeros) then there is no real difference at any higher conceptual levels. The smallest unit comes to define the whole.

I like to make distinctions between the photographic and special effects. I call photographic images a special defect. This refers to how photographic figure looks when cut in to a painting or a graphic. The photographic figure looks like some sort of defective image. To avoid this defective look, special effects artists then change the backdrop, or props to look more photographic, (photorealism).

But I like the defective look of photographs when they are cut out and collaged into a graphic. It shows up how different a photographic image is from that of a graphic artist, painter, etc. Even if it is just all zeros and ones at it's base. It is an observable thing. There is this visible difference.

Now some times it's interesting to actually erase this difference. Whether to create a suspension of disbelief or for some other reason. One other reason is that by successfully erasing the difference (as an apparition) you are actually demonstrating intimate knowledge of the difference. This is a much better reason than "suspension of disbelief". It is not towards some cheap (or not so cheap) trick that one is creating a particular effect but towards some sort of insight. Towards some sort of philosophy.

Carl

[carllooper]

I kind of agree with what you're saying there Carl, but I think my approach is valid too - because if you do most of the film look processes to video you don't get film - you get the best looking video you can. You only have to look at the Miami Vice reboot to see why that is so. But anyway needless to say if I could afford to do this particular project on film I would

Scot

I think I get what you are saying now. It's not a film look you're after per se, but that target they can be regarded as having in common. The photographic image as something almost (if not actually), independant of both film and digital. That I can relate to. I'm doing something like that myself, but with film.

cheers
Carl

[carllooper]

Sorry in advance (thread hijack)

Carl, just to confirm... whether it's film or 4K digital, there's grain; there's no system that doesn't have noise. Case in point, I recently received plates that were billed as noise free 4K capture, and I gotta say, grain definitely had to be added to the visual effects to match the plates, particularly in the blue channel... from in terms of visual effects, workflow doesn't differ at all in the approach with film and hasn't changed: whatever the origination, we match to the plates. And I think you'll find the drastic colour timing will remain, digital or film...

http://theabyssgazes.blogspot.com/2010/ ... -stop.html

Thanks!

-Jesse-

At the level of the mediums themselves, yes, there is noise. I didn't mean to suggest otherwise. What I was arguing against was adding noise (fake grain) in order to get a film look.

In relation to matching grain, that's another discussion.

As for that teal-and-orange thing. I've seen that before. That's something I think will also change. Otherwise it should change and I say that because I can, whether the teal-and-orange industry gives a shit or not.

What I found in Prometheus (god awful film) or The Avengers (a great film) is that neither tried to look like film. There was no fake grain. If there was no noise it was because they removed it (rather than it didn't have any). And I like that. Video/digital noise is ugly. But I find film noise (or grain) is quite beautiful. They are very different. I know because I've been studying grain/noise for the last two years. There are fundamental differences between the two types of noise.

Now it's not the look of film grain, per se, that I think is beautiful. It's the fundamental nature of actual film grain that I find beautiful. I don't find fake film grain beautiful at all. I find it ugly. More ugly than video/digital noise (which is ugly enough).

The structure of film grain/noise extends right down to the atomic scale whereas video/digital noise sits only at single scale/frequency. It's flat. In and of itself this doesn't make either more or less beautiful than each other. It could very well be cultural training that determines my appreciation of the film grain. Be that as it may it is the difference between the two that stands as an otherwise objective fact. But more importantly noise (whether film or digital) is something I believe that you either have there or live with (because of the intrinsic nature of the medium) or you remove/suppress it. The idea of adding noise makes no sense to me at all to me. I know the reason why others add it in (film look, grain matching for effects, flat noise suppression, artifact masking, stair casing suppression, dynamic range enhancement) but none of which I find a particularly convincing reason. Particularly "film look" emulation. A lot of the reasons are for graphic purposes rather than photographic purposes.

But that doesn't mean you can't do it - it all depends on why for me. These things should not be done (methinks) according to some rule that it must be done this way or that, but rather, because particular circumstances, exceptions require it. There should be a specific reason for it rather than just a general rule-of-thumb approach to these things. Thats just me of course. We're all free to pursue and promote other approaches to this.

My position emerges out of a kind of difficulty with consistency. Consistency unnerves me. I like difference. Shocks. Change. Not always. Just most of the time. In some domains consistency is the very fabric of the space in which you are working. For example, in mathematics there is no such thing as inconsistency. If something is inconsistent then it's an error. Mathematics is the language of consistency and I love mathematics for that reason. But when that same consistency, which is proper to mathematics, is taken out of that context and used within the domain of the inconsistent it unnerves me. There is a another language, proper to the inconsistent, and it's not mathematics. Math can help but it can't intrinsically solve it - it reacts badly to the inconsistent - it always wants to restore consistency rather than interoperate with it.

The inconsistent is not solved by elliminating it but by working creatively with it - making or allowing the inconsistent to work.

Carl