Jump to content

RED production schedule


Carl Brighton

Recommended Posts

  • Premium Member
But, because Dalsa choose to employ the same sort of resolution creative accountancy as RED, that automatically qualifies it as a "real" cinematography camera, even though its native resolution is less than other cameras that don't quality.

Jim Murdoch, is this you again?

Link to comment
Share on other sites

  • Replies 495
  • Created
  • Last Reply

Top Posters In This Topic

  • Premium Member
They are not the same thing, no matter what verbiage one wants to use to indicate otherwise.

 

There is no such thing as "2K HD." HD is 1920x1080. 2K is 2048x1556, at least in its "full frame" guise. The native color space is also different, as is the sampling rate. In a video system, there is a maximum sampling rate (based on today's systems) of 4:4:4. The sampling rates of most scanners exceed that, because they're not locked in to a video subcarrier rate as their base.

 

I'm not necessarily claiming that the end results are vastly different, but they are different. 1920 is not 2048, at least not on this planet. And lenses, recording methods, and industry/commerical success does not magically turn something into something it's not. There are enough inflated claims and general misunderstandings already. Let's not add to them. The current generation Genesis is an HD camera. Even Panavision doesn't dispute that, regardless of how it's used.

 

michael this is somewhat confusing as hd is high definition not 1920 x 1080, it also refers to 720 and the 4:4:4 colour space is not restricted to 1920 x 1080 either, further more there are many other combinations of output in terms of resolution and colour space. the current genesis is an hd camera and is arguable that any future digital cameras, be they 2k, 4k or 10k, are hd cameras as they will all be of 'higher' definition than a SD camera. i also fail to see that because film is scanned at a possibly higher sample rate that a 2k scan has any greater claim over the words 2k as a 2k data output from a digital camera. further more whilst 1920 is clearly less than 2000 by 80 pixels and a whopping 128 pixels less than the 'recognised' figure 2048, we all know that many '2k scans' are in fact frequently much lower than that and are often in their non 'uprezzed/ sampled' form less than 1920.

 

keith

Edited by Keith Mottram
Link to comment
Share on other sites

Hi Carl, I'm sure I'm bugging the hell out of you with my constant, no doubt dumb replies, unfortunately I'm a Philosophy graduate and get a kick out of arguing with people - suffice to say I enjoy and learn a lot from the process, regardless of the outcome from said argument. So here I go again...

 

your brain is able to make accurate assumptions about the level of detail of the areas it doesn't sample. If the camera and transmission system had some way of knowing what part of the scene the fovea was looking at at any particular instant, the data rate could be massively reduced, as you would only need high resolution in those places.

 

Dude, I wasn't trying to make a point about the relative merits of the human vs machine image sampling system - I was simply making a basic point that your brains makes intelligent estimates to fill in information it lacks and I'm sure that works fine for you, so as a methodology it's legitimate to use it. After that the only criterior to be worried about is accuracy.

 

The luminance signal from a true RGB camera is made by adding together distinct ratios of the full-bandwidth Red, Green and Blue signals, and these can be regenerated with a high degree of accuracy.

 

The luminance signal from a Bayer Mask does not contain such information, only an estimation of it.

 

In my defence, I was making no such mistake -I acknowledge that in my previous post, see here:

 

[/i]

resolution as a usefull term (in this discussion) [is usefull] only insofar as the 'processed' image of a certain size (4K Bayer) is more or less better [at] 'accurately' reproducing reality than another image [formed] from composites of three colour specific alligned monochrome images [of a smaller size] (2k 3chip system). My argument was that some form of intelligent processing can re-produce reality better in that case, not as accurately as 3 chip 4K, but better than 3chip 2K I suspect.

 

To reiterate - taking two sets of same size projected images - one from a 4K Bayer mask image file (with de-bayer processing), and one from a 2K 3chip image - and all else being equal - the 4K bayer image will represent a closer approximation to the original image - when compared to say a hypothetical 'pin-hole type' optical projection of the same scene. Again I bet a 4K 3chip process would be more accurate still, but that is not in dispute. By accurate I mean the degree to which, for each discreet section of the original image, the luminance and chrominance information recorded by each system, matches the original (whatever that means) information for said section.

 

Discuss.

 

On completely different note, i've clearly misunderstood something along the way, otherwise I wouldn't have been confused by this:

 

In a video system, there is a maximum sampling rate (based on today's systems) of 4:4:4. The sampling rates of most scanners exceed that, because they're not locked in to a video subcarrier rate as their base.

 

I was under the impression that 4:4:4 colour sampling simply meant that for each recorded segment/pixel the full and seperate luminance and chrominance information for it's Red, Green and Blue aspects was recorded/sampled (to whatever degree of accuracy - 10bit, 12 bit, 14bit, 14million bit - whatever (hmmm, probably should have said Bit or something, letting on my degree of ignorance there) - so, um, how can most scanners exceed a sampling rate of sampling everything? Does they sample more than just the RGB a chrominance aspects of a pixel, or, er, what? Or are you simply talking about cameras recording 4:4:4 YCbCr (i.e. colour difference subsampling for each pixel) vs 4:4:4 RGB (i.e. no subsampling) because I was under the impression, could be very wrong, that the D20 and possibly other cameras can provide RAW files containing the 4:4:4 RGB colour info (or at least the Bayered info - which I suppose is not technically true 4:4:4 RGB) - I'm not trying to argue, I'm just wondering if there is a gap in my knowledge somewhere that Michael could hopefully fill.

 

Colin Elves

Link to comment
Share on other sites

  • Premium Member

Hi,

 

> i also fail to see that because film is scanned at a possibly higher sample rate that a 2k scan has any greater

> claim over the words 2k as a 2k data output from a digital camera

 

I don't think anyone's really trying to say that; the problem is that there's currently only one camera which is capable of doing it (Dalsa), and it isn't usually dealt with in the same way as a normal 2K workflow in any case.

 

Phil

Link to comment
Share on other sites

There is no such thing as "2K HD." HD is 1920x1080. 2K is 2048x1556, at least in its "full frame" guise.

 

Sure, but how many movies are shot 4:3 these days? And while you could theoretically use anamorphic lenses to get widescreen with a 2048x1556 sensor, in practice what people do is crop vertical pixels. 16:9 2K is 2048x1152, which is pretty close to 1920x1080. Anyway, the DCI digital cinema exhibition spec specifies only a 2048x1080 container for 2K.

 

The native color space is also different, as is the sampling rate. In a video system, there is a maximum sampling rate (based on today's systems) of 4:4:4. The sampling rates of most scanners exceed that, because they're not locked in to a video subcarrier rate as their base.

 

Err... this makes no sense. 4:X:X numbers represent the relative resolution of luma and chroma channels. 4:4:4 just means that the chroma channels have the same resolution, both horizontally and vertically, as the luma channel. (This is necessarily true in any RGB image, since the luma channel doesn't exist independently of the chroma channels.)

 

Higher numbers wouldn't mean anything. The second two numbers are relative to the first, and '4' is basically chosen for the first number because it's conveniently divisible. I suppose something like 4:8:8 would mean that the chroma channels were twice the resolution of the luma channel both horizontally and vertically. But nobody would ever do that with footage designed for consumption by the human visual processing system, which, of course, is much more aware of brightness than color.

 

If you mean that a scanner can scan at 4K and scale the resulting image to 2K, creating a better image than a native 2K capture device, this is true. But this isn't best described in terms of the 4:X:X notation. And this is something which video acquisition systems are perfectly capable of doing. Any time you shoot SD on an HD consumer camera, you're performing such scaling, though obviously with fewer pixels at the moment. There is no barrier imposed by being "locked in to a video subcarrier rate" -- it's all just data these days.

Link to comment
Share on other sites

I'm not necessarily claiming that the end results are vastly different, but they are different. 1920 is not 2048, at least not on this planet. And lenses, recording methods, and industry/commerical success does not magically turn something into something it's not. There are enough inflated claims and general misunderstandings already. Let's not add to them. The current generation Genesis is an HD camera. Even Panavision doesn't dispute that, regardless of how it's used.

I don't have any particular problem either way whether a camera should be designated "HD" or "Digital Cinematography", it's just that if the Genesis isn't, then the Dalsa and the Red aren't either. The Genesis produces native 1920 x 1080 RGB. The Dalsa and the RED produce synthesized RGB which is then re-scaled to to match the so-called 2K spec of 2048 horizontally.

 

If I take the DV output of my handycam and put that into some sort of upscaler that converts the 720 x 576 to 2048 by whatever it is, does then make it a "Digital Cinematography" camera too?

 

If not, then the only camera that could be truly described as "2K" would be a Genesis with a 6134 pixel sensor.

Link to comment
Share on other sites

Err... this makes no sense. 4:X:X numbers represent the relative resolution of luma and chroma channels. 4:4:4 just means that the chroma channels have the same resolution, both horizontally and vertically, as the luma channel. (This is necessarily true in any RGB image, since the luma channel doesn't exist independently of the chroma channels.)

 

Higher numbers wouldn't mean anything.

 

Incorrect.

 

The numbers themselves do mean something quite significant, they're not just relative to each other. In the case of "4:4:4," it means that the sampling rate is 4x the subcarrier frequency. "2", as in "4:2:2", represents twice the subcarrier. That specifies a very specific number of samples. The numbers aren't just pulled out of thin air, they're based on the video system and how you deal with images in that system. With a scanner, you can have higher sampling rates because you're not limited to video "rules" and frequencies. You can design a scanner to sample at a higher rate than video systems, thereby yielding digital information that's more accurate to the original image. Once again, in most cases, the 4x video subcarrier rate is sufficient to yield a good result, but also once again, it's not the same as what you can do with a scanner, which is not based on video.

 

I don't have any particular problem either way whether a camera should be designated "HD" or "Digital Cinematography", it's just that if the Genesis isn't, then the Dalsa and the Red aren't either. The Genesis produces native 1920 x 1080 RGB. The Dalsa and the RED produce synthesized RGB which is then re-scaled to to match the so-called 2K spec of 2048 horizontally.

 

If I take the DV output of my handycam and put that into some sort of upscaler that converts the 720 x 576 to 2048 by whatever it is, does then make it a "Digital Cinematography" camera too?

 

The term "digital cinematography" has nothing to do with resolution, nor does it imply a specific resolution. "Digital Cinematography" indicates a form factor and design criteria that make it appropriate for use in "traditional" cinematography applications, such as dramatic productions, that is different than cameras more suited for, say, ENG or home video usage. This would mean things like matte boxes, high quality lenses, high quality recording system, follow focus, and other features that allow higher quality imagery to be captured using more precision tools. The question of resolution is a different, albeit somewhat related issue. The original post I was responding to claimed that the Genesis and other devices are "2K" cameras, and that somewhere there exists a "2K HD" system. They are not, and there is not.

 

If you take your handycam and outfit it with interchangeable lenses (not really possible, but this is theoretical), a matte box, a follow focus, better weighting, and a better recording system, then yes, it could reasonably be considered a "digital cinematography" device, although it's not one I would likely want to use for that purpose.

Link to comment
Share on other sites

Incorrect.

 

The numbers themselves do mean something quite significant, they're not just relative to each other. In the case of "4:4:4," it means that the sampling rate is 4x the subcarrier frequency. "2", as in "4:2:2", represents twice the subcarrier. That specifies a very specific number of samples. The numbers aren't just pulled out of thin air, they're based on the video system and how you deal with images in that system. With a scanner, you can have higher sampling rates because you're not limited to video "rules" and frequencies. You can design a scanner to sample at a higher rate than video systems, thereby yielding digital information that's more accurate to the original image. Once again, in most cases, the 4x video subcarrier rate is sufficient to yield a good result, but also once again, it's not the same as what you can do with a scanner, which is not based on video.

 

Well, that was a blast from the past. Thanks for the refresher on analog capture. None of that has any applicability to digital acquisition. When a Red captures off the 4K sensor and digitally scales to 1080p before recording, or when a consumer HD camcorder captures off a 1440x1080 sensor and records SD, it's doing essentially the same thing a film scanner is doing when it scans at 3K for a 2K image. There is no "video subcarrier" anywhere in the equation, and the imaging system can function at whatever resolution it wants to.

 

The term "digital cinematography" has nothing to do with resolution, nor does it imply a specific resolution. "Digital Cinematography" indicates a form factor and design criteria that make it appropriate for use in "traditional" cinematography applications, such as dramatic productions, that is different than cameras more suited for, say, ENG or home video usage. This would mean things like matte boxes, high quality lenses, high quality recording system, follow focus, and other features that allow higher quality imagery to be captured using more precision tools.

 

Yes, this is a much better way of defining things than quibbling over technical specs. Though, to quibble a bit myself... I'd say the definition you've given describes what a "digital cinematography" camera is. I'd say digital cinematography itself is any use of a digital acquisition device in a scenario where film has traditionally been used, even if the particular device wasn't actually designed with that in mind. That is, if you're using a consumer camcorder to shoot a movie, you're doing "digital cinematography". You're just not doing it with a digital cinematography camera.

Link to comment
Share on other sites

Incorrect.

 

The numbers themselves do mean something quite significant, they're not just relative to each other. In the case of "4:4:4," it means that the sampling rate is 4x the subcarrier frequency. "2", as in "4:2:2", represents twice the subcarrier. That specifies a very specific number of samples. The numbers aren't just pulled out of thin air, they're based on the video system and how you deal with images in that system. With a scanner, you can have higher sampling rates because you're not limited to video "rules" and frequencies.

 

I thought the sampling for SD component was usually done at 13.5MHz, as that is an exact multiple of both the PAL and NTSC line frequencies. Anyway, didn't the 4:2:2 notation coincide with the introduction of component video, which doesn't have subcarrier frequencies?

 

 

The term "digital cinematography" has nothing to do with resolution, nor does it imply a specific resolution. "Digital Cinematography" indicates a form factor and design criteria that make it appropriate for use in "traditional" cinematography applications, such as dramatic productions, that is different than cameras more suited for, say, ENG or home video usage. This would mean things like matte boxes, high quality lenses, high quality recording system, follow focus, and other features that allow higher quality imagery to be captured using more precision tools.

 

If you take your handycam and outfit it with interchangeable lenses (not really possible, but this is theoretical), a matte box, a follow focus, better weighting, and a better recording system, then yes, it could reasonably be considered a "digital cinematography" device, although it's not one I would likely want to use for that purpose.

 

So, according to you the ONLY thing that stops the Genesis (the closest thing I have ever seen to a traditional film camera, that is actually a production model and in regular commercial use), from being regarded as a "digital cinematography" camera is a lousy 128 pixels horizontally and some airy-fairy tripe about sampling rates and "colour space"?

Link to comment
Share on other sites

  • Premium Member

Mike,

 

You'd be best to give up bothering with this thread. It's clear from a few of the posters that they are these DIY filmmakers who have little clue of how a television picture, digital or analog is made. It degrades the good conversation that those who are obviously experienced in the field are trying to have, and pains me to see how much ignorance is past off as experience. The signal to noise ratio is often too low on such threads as it is here. It's one reason why these gossip threads are nothing but destructive. They answer no questions and feed ignorance. Imagine if many of these folks didn't have the web to "learn" all they know about video? Imagine trying to say that 4:2:2 is somehow an analog term? The RED folks are smart trying to create buzz about their camera by disguising these viral threads they put all over the web as some of sort of public involvement in the development of their camera. It's simply a form of viral marketing that gets their brand in peoples heads. Imagine a million dollars in free marketing and the sad part is that these sites allow it without any compensation. But as I said early on, they will find these threads actually will hurt them more in the end than help them as most all viral and Astroturf marketing plans do.

Link to comment
Share on other sites

Mike,

 

You'd be best to give up bothering with this thread. It's clear from a few of the posters that they are these DIY filmmakers who have little clue of how a television picture, digital or analog is made.

 

Maybe you can explain how subcarrier frequencies make it impossible for digital acquisition systems to oversample?

 

Imagine trying to say that 4:2:2 is somehow an analog term?

 

I think you'll find I didn't say any such thing. I said Mike's explanation for why video cameras couldn't oversample the way film scanners do was not applicable to digital acquisition systems.

 

I also said that talking about oversampling as being better than 4:4:4 was meaningless. 4:4:4 means that choma and luma information have the same resolution. With digital image data, that's all it means; there's no subcarrier frequency that things are being measured relative to, or whatever. It doesn't imply anything about whether the image in question was generated via oversampling, or anything else. Generating a 2K RGB image from an 8K scan might get you a better 2K image than generating one from a 2K scan (or from direct 2K capture... but it's still a 4:4:4 image.

Edited by Chris Kenny
Link to comment
Share on other sites

  • Premium Member
So, according to you the ONLY thing that stops the Genesis (the closest thing I have ever seen to a traditional film camera, that is actually a production model and in regular commercial use), from being regarded as a "digital cinematography" camera is a lousy 128 pixels horizontally and some airy-fairy tripe about sampling rates and "colour space"?

Carl's post is the best example that Walter is absolutely right, some people here are only trying to create controversy. In this particular case Carl quotes Mike as saying that digital cinematography has nothing to do with resolution and then has the temerity to accuse him of of not counting the Genesis among the digital cinematography camera beause of a lack or resolution...

 

Thank god for the 'ignore' button! Bye Bye Carl...

Link to comment
Share on other sites

Max Jacoby:

"Carl quotes Mike as saying that digital cinematography has nothing to do with resolution and then has the temerity to accuse him of of not counting the Genesis among the digital cinematography camera beause of a lack or resolution..."

 

What?! Not so much a "lack of resolution" as "significant lack of resolution"

 

'Twas ever thus...

Call me an ignoramus, call me anything under the sun, but don't you ever ever answer the question!

(By "answer the question" I of course mean "provide a meaningful answer to the question!")

 

So, getting back to my original question: When is anybody (outside the RED Skunk Works, that is) going to see a real RED, in the aluminium flesh, with all its processing happening on-board in real time, running off a battery, and shooting actual pictures which can be seen on an actual monitor screen?

Link to comment
Share on other sites

So, getting back to my original question: When is anybody (outside the RED Skunk Works, that is) going to see a real RED, in the aluminium flesh, with all its processing happening on-board in real time, running off a battery, and shooting actual pictures which can be seen on an actual monitor screen?

 

I'd guess they're shooting for NAB. Whether they'll make it... I doubt anyone outside of Red knows. Possibly nobody inside of Red knows.

Link to comment
Share on other sites

  • Premium Member
So, getting back to my original question: When is anybody (outside the RED Skunk Works, that is) going to see a real RED, in the aluminium flesh, with all its processing happening on-board in real time, running off a battery, and shooting actual pictures which can be seen on an actual monitor screen?

 

Carl, isn't it obvious that the camera is not yet finished, that it is still in development?

 

I don't get the gist of your question. Why do you need to know a specific date for completion? Do you have an impending project that you are thinking of buying the RED camera for? What difference does it make to you whether it's finished by NAB in the spring -- or not until the fall anyway?

Link to comment
Share on other sites

Gentlemen:

 

There is considerable misapprehension here about the meaning of the terms "4:2:2", "4:4:4" and so on.

 

These terms are not specific to any particular resolution, nor necessarily to any sort of digital sampling process. The original expressions were devised nearly 20 years ago, when the D-1 digital component tape format was first devised, although not specifically for that either.

 

With a traditional top-quality PAL signal, the luminance normally had a 5MHz bandwidth, and for normal television transmission a chrominance bandwidth of around 1.3MHz was considered the maximum that could be appreciated on an ordinary interlace scan TV. Thus, the chrominace bandwidth was about one-quarter of the luminance bandwidth, and much the same for NTSC. So ordinary colour TV as broadcasted could be defined as 4:1:1 , the second "1" indicated that the chrominance information was supplied for every scanning line.

 

The original D-1 digital component standard specified a chrominance bandwidth of half the luminance figure, since this would be beneficial for chroma keying and similar "work" functions, however half of this was discarded when the time came to transmit it. This gives the figure "4:2:2"

 

In reality, since the chrominance bandwidth is reduced horizontally, there is no real need for the chrominance information to be supplied on every line and so in some compact tape systems (Mini DV etc) a "4:1:0" system is used, where the chrominance is only supplied for every second line.

 

The original D-1 standard also specified a 13.5MHz luminance sampling rate since 13.5MHz is an exact multiple of both the PAL and NTSC line rates. Although this is not much used much anymore in professional equipment it is the basis of the domestic DVD format. In modern players as single 27MHz crystal (2 x 13.5MHz) is used as the master clock for everyting including the colour subcarriers, for both PAL and NTSC players.

 

Digital composite VTRs such as the D-2 format did the digital sampling at multiples of the colour subcarrier frequency to minimize interference, however component based systems did not need to use any particular sampling frequency, as long as it was at least twice the maximum luminance frequency.

 

The curent 1920 x 1080 HDTV standard was derived from the old NHK 1125-line analog HDTV system. Because it was designed on vacuum-tube based cameras, 45 scanning lines were set aside for the vertical retrace leaving 1080 active picture lines. Since retrace is not meaningful with CCD cameras that leaves 1080 lines. Scaling this to the 16 x 9 shape gives 1920 pixels horizontally. The figure of 2048 mentioned for filmscanners is simply the binary number 11111111111 and is the result of an 11-part chain of binary counters being easier for a computer to implement. Otherwise the difference between 1920 and 2048 is insignificant.

 

Since they human eye's ability to resolve colour detail is a fixed percentage of its ability to see luminance detail, the same relationship exists whatever the luminance resolution, so the terms "4:2:2" and so on are constant regradless of the resolution

Link to comment
Share on other sites

Since they human eye's ability to resolve colour detail is a fixed percentage of its ability to see luminance detail, the same relationship exists whatever the luminance resolution, so the terms "4:2:2" and so on are constant regradless of the resolution

 

Werner, thanks. You clearly know much more than I do about the origin of this notation, but what you're saying, if I'm understanding correctly, is basically what I've been trying to say. This notation describes the resolution of chroma information horizontally and vertically for a format relative to the resolution of the luma information for that format, and that's basically all it describes.

 

The fact that the numbers are relative to each other (rather than being defined relative to a specific sampling rate or whatever) makes it clear that the notion of something being better than 4:4:4 doesn't really make any sense.

Edited by Chris Kenny
Link to comment
Share on other sites

Sure, but how many movies are shot 4:3 these days? And while you could theoretically use anamorphic lenses to get widescreen with a 2048x1556 sensor, in practice what people do is crop vertical pixels. 16:9 2K is 2048x1152, which is pretty close to 1920x1080.

 

Lot's of films are still shot and scanned using the full gate. The final image normally ends up being a wider aspect ratio though.

Link to comment
Share on other sites

Carl, isn't it obvious that the camera is not yet finished, that it is still in development?

 

I don't get the gist of your question.... Why do you need to know a specific date for completion?What difference does it make to you whether it's finished by NAB in the spring -- or not until the fall anyway?

You're right, you totally misunderstand.

 

Do you have an impending project that you are thinking of buying the RED camera for?

 

No, but I really, truly want the RED to succeed.

I want Jim Jannard to hit the market with 1,500 or so cameras and I want him to do it as soon as humanly possible. I'll come in and help him assemble them if he wants, even just packing them into cartons and shipping them.

 

I just want to be sure of the schedule. RED keep repeating that they've never missed a deadline, so what is the deadline?

Link to comment
Share on other sites

  • Premium Member
No, but I really, truly want the RED to succeed.

I want Jim Jannard to hit the market with 1,500 or so cameras and I want him to do it as soon as humanly possible. I'll come in and help him assemble them if he wants, even just packing them into cartons and shipping them.

 

Why? It smacks of desperation, why you are so eager for RED to succeed? Are the other cameras on the market not available to you, or are they limiting your creativity in some way? Is "4k" going to revolutionize your whole image-making experience?

 

I'm not cheering against them, but c'mon.

Link to comment
Share on other sites

I think Mr Mullen's point is what really counts considering the name of this thread.

 

When are we going to be able to actually go out a buy a RED system, or rent one and try out the workflow for a production?

 

I'm as excited as the next guy at the promise of affordable 2K let alone 4K, but before I'd commit my hard earned cash I want to get my paws on one and put it through the grinder for couple days, then see how the footage performs in post and filmed out.

 

Just my 2 cents

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.

Forum Sponsors

BOKEH RENTALS

Film Gears

Metropolis Post

New Pro Video - New and Used Equipment

Visual Products

Gamma Ray Digital Inc

Broadcast Solutions Inc

CineLab

CINELEASE

Cinematography Books and Gear



×
×
  • Create New...