The end of the traditional low end cinema camera company is nigh!

[Wayne Morellini]

I've seen plenty of "cinema" from these cameras but thanks for reafirming that some RED owners continue to be douchbags when it comes to this old argument. Seriously thought these days were over. Sigh.

Don't worry, you are right. I'm not going knock an innocent like you just because you got "1" letter wrong by using the alternative Americanised version of douchebag. I can understand what you mean.

Ash, welcome, thanks for your positive truthful feed back. Cinema caneras were not meant to be only the bastardised less accurate Bayer versions, they are anything that can deliver a cinema image setup to do cinema shooting, as well as whatever other shooting. If you are talking "Prime" quality cinema camera, than that is another story where Red and a dew others fit unnthe Digital Cinema Camera realm.

There has been a real problem around here, my posts used to be a quarter the size, but having to answer and explain things simply to some around here has blown out NY writing style. Don't matter how correct what you are saying is, they cannot accept it to understand it and just have to keep grinding away as the true ... they are.

 

[Karim D. Ghantous]

I'm not going to get into the above arguement (or should that be 'argument'? ha ha) but every community thinks that it is the only one to contain allegedly immature members. Not so.

Self-flagellating (and usually virtue-signalling and hypocritical) photographers think that photography is the only profession whose adherents obsess over technical details, and that the only acceptable attitude is to not talk positively about one's favourite pieces of equipment.

Musicians really, really care about their instruments, because they train for hours every week. All that training would come undone if even small changes or inconsistencies were to occur. Maybe they should shut up about their hardware and just play, because as all professionals know, the equipment doesn't matter and all that matters is telling people how little you care about your equipment or your style.

Sometimes all you need is a jug, a box and some rubber bands (but good quality rubber bands, and that's not a joke), but sometimes you need a good violin.

Chefs and cooks would be about the same. Painters and sculptors, too. Tennis players have some concerns about their equipment and their clothes and their shoes. And yes, it does matter whether your use 15mm rods or 19mm ones, just as it matters whether your use an LED panel or an 18K incandescent studio light. It is acceptable and compulsory and responsible to obsess about these details - as long as you do your obsessing before you start shooting, and not in between takes.

What emerges when something is democratised is that it becomes fashionable to look down on people who care about what goes on behind the curtain. Hence the DSLR kidz who think a POS 5D is somehow just as good as a proper cinema camera just because it has "more depth-of-field". And yes, I have seen people say that. IIRC a fashion was emerging, though stalled, which proposed to eliminate lights.

I'm all for natural lighting but you can't have NO LIGHTS on a serious production unless you're really patient and/or anal and/or experienced, and you know how to use scrims and reflectors and negative fill (not rocket science, I grant you). You also should have some understanding of exposure that goes beyond pointing an incident meter at a camera ("Wait, so you're telling me you can point this thing at the key light? Holy shit."). And hey, for no-budget movies, that's fine and even commendable, because any creativity is better than none at all. But to make this a virtue, well, no, sorry.

 

[Wayne Morellini]

...Sometimes all you need is a jug, a box and some rubber bands (but good quality rubber bands, and that's not a joke), but sometimes you need a good violin...

Ahh, the delta blues, love that stuff.

Sorry went to doctors without phone and couldn't fix my post.

 

[Wayne Morellini]

Karim, with the technology I'm pointing to you can use natural light (even starlight to make a lit color scene, but that defeats the purpose of looking natural). You can eventually set virtual lighting through computational photography on set to what you like, and reset what you like in post, or not light in the first place. Flexibility, but also convenience, as one day it could auto light your scene, and according to your style or style request, and then fine tune to taste. On low end you are looking for massive savings per quality level. At the moment though, the thread is only about basic filming against single lens things like mini Ursa and below. If one has a production that needs more rent a camera. There are a lot of low end guys that will seldom need more than this level.

 

[Jeffery Anderson]

Personally, I don't believe any of the camera tech you've discussed would help a Director of Photography shoot a movie any better

 

[Wayne Morellini]

Well, that's an impossibility, and pretty irrelevant, as we are talking about low end, who often don't have the skill levels mentioned in this low end discussion.

Is it not a case of denying it's true. Have a search on adobe's site for their very old camera.

 

[Wayne Morellini]

OK, the following is not really what I'm talking about, that would require a serious targeted attack:

http://www.43rumors.com/ft5-panasonic-gh5-6k-photo-mode-works/

They are just trying to kick the but of the lower end professional market (mini Ursa 4k and below). 4kp50 4:2:2 10 bit maybe. I suspect maybe 200mb/s or 150mb/s (the not really buyable option). Not literally kick those cameras around, but just take enough sales away to kick competitors below a 4.6k Ursa mini around. Sakes mean everything, if you can't get enough sales, you can't cover costs at your present pricing.
http://www.newsshooter.com/2016/05/...angeable-lenses-and-raw-capture-in-the-works/

A raw option coming. Presently I can't get access to information, but the morr consumer E1 does 16mp 15fps.


http://www.panasonic.com/uk/consumer/cameras-camcorders/camcorders/hd-camcorders/hc-x1e.html

http://www.panasonic.com/uk/consumer/cameras-camcorders/camcorders/hd-camcorders/hc-x1e.specs.html

How, did that get there. Seriously, Jarred, hurry up with the new cameras before I have to buy something.

Now, seriously Panasonic needs to get with it. No 6k recording on GH5, only 10 bits, likely only 200 or 150mb/s codec. No new super sensor. Looking at the hc-x1, I can't find recorded pixel format or bit depth, and we see the Mos sensor, 1 inch, but look at the low light characteristics. These cameras needed a higher grade spec. All BM has to do to fend off an attack from the industry is get a better 4k sensor, better prosumer/Eng level software and control, maybe lower prices, and bring out a 4k pocket replacement. People would nearly be ignorant, challenged or blind to buy these two cameras instead. BM and Red have been too easy on the industry.

 

[Dan Kanes]

I've seen plenty of "cinema" from these cameras but thanks for reafirming that some RED owners continue to be douchbags when it comes to this old argument. Seriously thought these days were over. Sigh.

Seen plenty... But have you shot any ?

I used to be a canon guy back in 2008 and 2009.

Shot the first feature length film on the 5Dmk2.

Shot the first feature length film on the 7D.

These didn't really win any awards or garner any attention, but they were movies. "Cinema" if you will.

So, I get it...

but after putting in many years and hours behind the lens (I have 15 years experience in motion pictures)...

I realized... Why the hell would I want to dick around with something that doesn't work right for the purpose and every moment is a compromise.

Intelligent design dictates that you make something with a purpose to be used a certain way. I love hacking and tweaking, but once it starts compromising your end results it's more of a Kludge than a Hack.

Main point is - c300 (Mark I especially)is designed for run and gun handheld and reality shooting, and it sold well and did well - but it was capable of SO MUCH MORE THAN WHAT IT ACHIEVED.

I was was there at Paramount Pictures on c300 launch day. I also went to scarlet launch day right after. There was cinema shown at the c300 event.

But you know what pissed me off? I knew canon was capable of MORE than they were giving with that camera body, just like how they have always been capable of giving more but always chose to GIVE US LESS.

If it was purpose designed for "cinema" they would have asked professional cinematographers what they wanted - not videographers. C700 represents the expression of what Canon's interpretation of a professional cinematographer's needs are.

In conclusion - if you know, then you know - because you've shot movies, commercials, title sequences, fashion films -- or you've built a product, ran a team or been part of one, and know it's not just saying "just add software" - believe me I understand that the pipelines and technology that Wayne is talking about IS CAPABLE OF SO MUCH MORE THAN WHAT IT IS USED FOR. And that's disappointing.

It's human inability and failure - not the failure of technology.

Last time I need to get involved in these semantic diarrhea festivals.

 

[Dan Kanes]

The word Douche is derived from the french word for Shower.

"the more you know"

 

[Wayne Morellini]

He, he, you should have seen what I cut out of my message on that. I was going to make s point, the hc-x1 above is what I call a video camera, and likely barely suitable for cinema. A camcorder, as somebody mentioned, is even less than this and usually fits in a hand.

 

[Marc Wielage]

Karim, with the technology I'm pointing to you can use natural light (even starlight to make a lit color scene, but that defeats the purpose of looking natural). You can eventually set virtual lighting through computational photography on set to what you like, and reset what you like in post, or not light in the first place.

You forget four things:

1) there's an organic quality to real lighting done on the set at that moment, rather than trying to fudge it in post, and it's something the actors respond to in their performance

2) if you don't light it that way initially, then the editor and director will not immediately see the end result in dailies and during the edit

3) it costs a surprisingly large amount of money and takes quite a bit of time to relight in post

4) the end result of relighting in post does not look as good (to me) as it does when the cinematographer actually does it on the set. There's a palpable difference in quality and reality to the final effect.

Bear in mind that I've done about 20 films in the last three years where I've done hundreds (more like thousands) of little bits of relighting, simply in adding a little more fill to an actor's face or bring down the wall in back of them, essentially adding an "electronic flag" to the image. Sometimes, we're practically using a simulated Leko light to illuminate just their face and nothing else. But this adds considerably to the time and effort involved. Not every production can afford the additional 20-30-40 hours needed of post work.

As Jackson Browne once sang, "Nobody rides for free. Nobody gets it like they want it to be. Nobody hands you any guarantee." And that's especially true for post. Don't ever assume that post can make up for bad lighting (or no lighting) on the set.

 

[Karim D. Ghantous]

Well, I'm open minded on the issue simply because I don't yet understand the limits of the proposed technology. Computational re-lighting is demonstrable. But is it any good? Here are all the useful things you could have in a camera:


- 8K
- Light field (it works okay for still photography)
- Multiple lenses (and the sensors that go with them)
- Non-Bayered colour sensors
- Wide DR (20 stops or more)


Some of those have minimal limitations. Some have huge limitations. Some limitations are temporary, others might be logical and therefore impossible to overcome. But we don’t know.

 

[Marc Wielage]

Well, I'm open minded on the issue simply because I don't yet understand the limits of the proposed technology. Computational re-lighting is demonstrable. But is it any good?

If it takes a day to re-light one 10-minute scene, vs. an hour to actually light it on a real set, you're just transferring work done in one place to another. And again, it's not going to look as real done digitally.

To me, the biggest problem is that the director and editor won't be able to cut the project together because they won't be able to see the actual lighting until they get to post. If you light it for real on the set, then what they see is what they get. Ideally, what you've captured is already at least 2/3 of the way there (or more), and then you get the rest of the way in the final color session.

 

[Karim D. Ghantous]

But if we have computational re-lighting, will it take a day to re-light a scene? It would, hypothetically, take no more time than correcting a colour cast or whatever.

If we were to shoot a scene knowing that we would re-light it on a computer, we would probably blanket-light everything to make it easy. Yes, you would not be able to look at any shot in context until you apply the post-light. But it would be like a blank slate, so that you could at least imagine it.

I bet that if computational re-lighting is workable, you could do it with either two flat sensors or one light field sensor. I actually hope not, because it would make movies boring and actually not worth making at all. But you can't repress technology. It's going to come out one way or another, if it can be done. I'd like to wish away digital, but... it was inevitable.

 

[David Peterson]

If it takes a day to re-light one 10-minute scene, vs. an hour to actually light it on a real set, you're just transferring work done in one place to another. And again, it's not going to look as real done digitally. .

If there are twenty people on set, then paying for a whole day for one person in post would make a **LOT** more sense than paying for one more hour from twenty people!

To me, the biggest problem is that the director and editor won't be able to cut the project together because they won't be able to see the actual lighting until they get to post. If you light it for real on the set, then what they see is what they get. Ideally, what you've captured is already at least 2/3 of the way there (or more), and then you get the rest of the way in the final color session

People likely said the same about color grading, but now you get to see LUTs applied on set which help give you a general idea.

I bet in ten or twenty years from now we'll have similar relighting templates you can apply directly to the on set monitor to get an idea of what it will look like in post.

 

[Satsuki Murashige]

If we were to shoot a scene knowing that we would re-light it on a computer, we would probably blanket-light everything to make it easy.

Except that "blanket lighting" a scene (let's assume a house day interior for argument's sake) can be much more time consuming and thus costly from a production standpoint than lighting in a naturalistic way. If you require completely shadow-less, source-less, ambient lighting for every scene, then you really need to shoot on a stage and have heavy art direction to have light tones and pastel colors. Real locations have ceilings and walls that don't move, and natural sunlight that needs to be diffused and flagged. Lots of space lights and large diffusion frames, which means a rigging day prior to set build. Possibly two days for the set build, and then a day of lighting and shooting. Probably a G&E crew of 4x4 for the rigging day and shooting day, not to mention all the prep for art.

Then you run into the issue of reflections. If you have any glossy curved objects like white tile, a chrome kettle or a glazed ceramic mug, it will pick up reflections from any bright objects and light sources in the vicinity. And since we have introduced an immense amount of ambient fill from all sides of the frame you will have to deal with these reflections - flag them off, paint them out in post, etc.

So it's really not making things any easier, cheaper, or more efficient to re-light in post, you are simply delaying the decisions that have to be made at some point and dumping them onto someone further down the post-production line. If you commit to the look you want beforehand, then you scout for the right location or just build as much set as you need, only bring the equipment and crew that you need, light it, shoot it, and hand-off something to post that is at least 70% finished. Which means they can concentrate on polishing the last 30% in the limited time they have allotted. Everyone does better work this way.

 

[Satsuki Murashige]

If there are twenty people on set, then paying for a whole day for one person in post would make a **LOT** more sense than paying for one more hour from twenty people!

Crew typically work on a day rate, so as long as that hour is budgeted into the day without going into overtime, it wouldn't cost any more to light the shot properly. Things can go pretty fast when you have highly skilled crew who are led by someone who knows what they want and what they are doing. If not... well, then they're probably not getting paid much or getting any overtime anyway!

 

[David Peterson]

Yes, I agree if we're just talking about one shot, but when discussing hundreds of shots... all those small pieces of extra time do add up into entire whole more days of shooting. If that can be saved and shifted over to post, then producers will choose to do that.

That time isn't quite here yet, but in a few more years/decades, then I bet we'll see it much more prevalent.

 

[Craig W. Bickerstaff]

It depends what kind of lighting you're talking about surely. I mean if you are just talking about throwing a power window on someones face so you can see their eyes and face better or you want to bring out some detail that's getting lost by boosting the contrast then sure post it okay. If you want to throw in a really nice backlight to bring out their hair or you've decided you want hard shadows or something then probably post is a bit late to do such things.

 

[Wayne Morellini]

OK, memory of what happened is a bit shot due to health issues, plus I've got a couple of noisy women speaking Spanish next to me (literally I go to check if I need to reply to this and they sit down next to Mr, happens all the time, now there is three of them, thankfully they are going into their session). So I don't know how this will be because I forget what I just read. But here we go.

You forget four things:

1) there's an organic quality to real lighting done on the set at that moment, rather than trying to fudge it in post, and it's something the actors respond to in their performance

2) if you don't light it that way initially, then the editor and director will not immediately see the end result in dailies and during the edit

3) it costs a surprisingly large amount of money and takes quite a bit of time to relight in post

4) the end result of relighting in post does not look as good (to me) as it does when the cinematographer actually does it on the set. There's a palpable difference in quality and reality to the final effect.

Bear in mind that I've done about 20 films in the last three years where I've done hundreds (more like thousands) of little bits of relighting, simply in adding a little more fill to an actor's face or bring down the wall in back of them, essentially adding an "electronic flag" to the image. Sometimes, we're practically using a simulated Leko light to illuminate just their face and nothing else. But this adds considerably to the time and effort involved. Not every production can afford the additional 20-30-40 hours needed of post work.

As Jackson Browne once sang, "Nobody rides for free. Nobody gets it like they want it to be. Nobody hands you any guarantee." And that's especially true for post. Don't ever assume that post can make up for bad lighting (or no lighting) on the set.

If it takes a day to re-light one 10-minute scene, vs. an hour to actually light it on a real set, you're just transferring work done in one place to another. And again, it's not going to look as real done digitally.

To me, the biggest problem is that the director and editor won't be able to cut the project together because they won't be able to see the actual lighting until they get to post. If you light it for real on the set, then what they see is what they get. Ideally, what you've captured is already at least 2/3 of the way there (or more), and then you get the rest of the way in the final color session.

Marc, I'm a technologist, even privately called a genius among top engineers I know. I see what others can t yet, and have come up with very advanced ultimate post computational photography techniques. Believe me, you can do this live in camera on set with a suitable custom circuitry. So believe me it is coming in a way extremely close to real lighting by just moving virtual lights around on set that can control basic real lighting arrays. It does not matter if there is lighting on set, or mostly what it is like, as long as the camera can pick up all the information to do any adjustments needed in post.

Now, work wise, a extremely small crew can do an indie and relight in post live, in seconds. On a big project, you want to try it for real, and make it look better in post. We might he heading towards AI that picks the lighting for you, just pick the "AI Profile" you desire (probably a completely new term you have seen here first). So, you can puck profiles that match famous lighting directors and directors, films etc. So in 50 years this might he the case, and in 100, computers could do most of it in a virtual setting without people, a story telling computer. Have been researching and designing AI ideas since primary school. The Japanese 20 year AI project of the 80's was over reach. You need people that understand the mistakes at least to do that. But I think 50-100 years is a pretty good time frame. 2050-2100 even better. However, they have to get real about it. I was thinking of an open source AI project and wiki the over night, but open source tends to produce great results in volume rarely.

So you can eliminate nearly all crew. Do filming a lot faster, spend some time on set setting up placement and look, and a but in post improving it. The thing is that with ultra dynamic range, color, sensitivity and levels you can access all the information for post calculation. This is probably far beyond what most, or all other, scientists are envisionaging, but it is the natural practical progression of the technology.

Now real limitations. In complex structures, such as the eye, can trip up the results, because they produce different results based on light placement, color and intensity. The translucence of these structures is a complicating factor. But it dies end there. Hair, skin, cloths, etc also present these challenges. Now, if you 3D map from an infinite point cloud grid arc, you can probably achieve good results. Hence my set wide kens design of 12 tears ago, and dome lens system. However there are some technique to get closer, and you can simply set the computer to fame it or use fake structures or scan the actresses eyes (let's face it). So, it can look completely normal to the viewer. From what I'm hearing, at the moment the present techniques just aren't good enough.

To achieve these things you have to see it in your head and manipulate it like a a crash in reverse.

I can't remember what else I read here. But on the live calculations I'm working on new processing models that are looking at a 1000x performance advantage maybe, on new techniques. But, apart from what can be achieved with today's processing production technology, the techniques to look out for are Quantum Dot Cellar Automata (I prefer the magnetic type) spintronics, and a few others, as well as more advanced forms of optical and quantum computing eventually. I have been working through a number of these areas. QDCA, promise extremely low powered processing arrays. The issue has been the relatively low speed never catching up to silicon. But at such low power you can simply put more in instead. I have been working on proposals to get around this issue. QDCA is one if the promising technologies. I'm also keen to work on a FPGA solution and magnetic computing.