Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

General Demosaicing Questions

Christopher Probst said:
Graeme,

Ironically, I'm knee deep in this very sort of topic for a highly technical Digital Imagers article I'm putting together for American Cinematographer. I know I've asked several times to both you, Jim and Jarred for Red to contribute to this article and been declind (very politely, of course), but what you're talking about in this thread is exactly the sort of commentary I'm looking for. I wish again Red would reconsider their position in having a voice in American Cinematographer. Believe it or not, AC (and specifically ME in this instance) is not an enemy!

Perhaps I'll have to do an Epic article as a means of amends... But of course, that stipulates I get my hands on one!

Still waiting.
Click to expand...

The problem with doing an interview for an article is that you will get too many "can't answer that one"...

Jim
 
This article doesn't go into proprietary info on individual manufacturers. Rather it discusses the history of digital imagers, the types there are, how they function, the factors that comprise their functioning (CCD, CMOS, pixels, sampling, MTF, digitization, image processing, ect...) but in a general sense. Not nit-picking a specific sensor on the market.

What is interesting is hearing different opinions on sensor design, image processing, compression schemes, output... Basically I'm trying to create a definite primer in talking about digital imagers, and by default, digital cinema cameras so that all future discussions (in print) have a foundation to stand on. In creating this hopefully definitive reference, I DO feel it's a shame NOT to have Red as a presence in the article.

Thanks for at least commenting on my post.
 
Graeme Nattress said:
I find it rather amusing that one arm of Sony used to trash Bayer, while another arm of Sony make some very successful DSLRs which use Bayer patterns! Similarly the Canon / Panavision talk trashes Bayer, but Canon make a rather successful line of DSLRs all using Bayer patterns. You'd think they'd talk to each other to get their story straight.
Click to expand...

Graeme, I've viewed that presentation several times, and I fail to find where Larry Thorpe trashes Bayer... he does call it a very clever method to get color info out of a single sensor. John Galt goes on to agree that the idea Larry stated about "effective" diagonal sampling was a way to explain why with some Bayer sensors, some manufacturers were able to exceed the theoretical Nyquist limits of what the sensor should be able to produce in terms of resolution.

So I'm trying to understand what you find fault or error in, so I don't put a proverbial foot in my mouth on my article.

Care to educate me?
 
"Bayer sensors, some manufacturers were able to exceed the theoretical Nyquist limits of what the sensor should be able to produce in terms of resolution." - no they don't. Sampling theory doesn't say anything about what resolution you can get from a sensor, just what can be achieved without aliasing, and because of the nature of optical filtering, no camera I know of has enough filtering to entirely eliminate aliasing - all compromise to some degree. That's why we say "negligible aliasing" which is true, rather than "no aliasing" which some other camera manufacturers state, which is demonstrably false.

I think I've written enough on this subject on this and other forums to write a book :-)

Graeme
 
Sorry Graeme, I did not verbatim quote them and should have said apparent exceeding Nyquist... This was an expressed sentiment they made in praise of Bayer, but not a hard-line factual statement supported by data. However, I wasn't necessarily looking to dissect MY wording of the question, rather I'm trying to understand what you had issue with in their presentation.

If you have a chance and can review that, I'd love to know where you feel they were in error on the topic. I'm trying to understand all perspectives.
 
I watched it so long ago, enjoyed the MTF stuff but was very put of by the presentation on Bayer patterns. If you look back in time to my comments I made, I'm sure they'd explain my objections to the video.

The mistake about Nyquist is that sampling theorem doesn't talk about resolution, but about reconstruction of an input signal with respect to the appearance of aliasing or not, in that if you bandwidth limit according to the theorem you can completely reconstruct the sampled waveform, and of course, aliasing would preclude accurate reconstruction.

The other objection was that Bayer patterns were criticized for aliasing and chroma moire, whereas in practise (as shown amply by RED cameras) such problems generally don't occur, and if they are shown in test charts they are negligible in nature, whereas the sensor pattern of the F35 exhibits both strong luma aliasing and chroma moire:

F35chroma.png


indeed strong enough to make it visible on broadcast shows shot on the F35, not just test charts, along with a sub-1080p measured resolution, in part due to such strong aliasing. Indeed, pretty much most of the "criticism" levelled at Bayer pattern CFA sensors is MORE visible what they suggest as a superior alternative.

That's pretty much all I've got to say on the issue as I mentioned above, I've spent way too long explaining the issues first, second and third time around....

Graeme
 
I agree about the striped array as well. Naturally, it too is not an ideal and even though Panavision spearheads the idea of their "macro cells" as a super sampling scheme to derive a 1080 image, in my mind, it's still a CFA that doesn't have a discreet RGB value derived at each photosite, so therefore, a compromise. Now I'm not knocking either method (though I was wasn't thrilled with the vertical streaks that can happen with hot pin-sources in a Striped CFA frame...) it IS just the nature of the beast with any CFA that some tradeoff has to be accepted.

And before even that comment gets criticized, my general feeling on CFAs now is somewhat placated by the fact that it has been shown, especially here on these boards, just what you can get out of a finely-tuned Bayer sensor and smart processing of the Bayer data. I don't think about obtaining 5K RGB out of an Epic. But getting a good, clean 4K? Sure! I'm anxious to start playing. And "super-sampled" 2K? No worries there at all.

On that note, I find it ironic with all of the focus on resolution, resolution, MTF and whatnot, that many people are choosing to put older and older glass on Reds. To me, it's just a validation of Otto Schade's work that we humans don't need high frequency detail. We need bulk lower- and mid-band resolution with good contrast reproduction to feel an image is sharp. In fact, I recently found Camtec an UNUSED set of 1950s-era Kowa spherical lenses. They can't keep them in the house! People like Matty Libatique keep taking them out to put in front of R1s. Do we think that is going to deliver superior MTF for the 4K signal? Certainly not. Does it somehow look more interesting. Definitely.

Therefore with the Epic now at 5K, and in my mind delivering a very solid usable 4K resolution, I think we're more than covered for the foreseeable future resolution-wise. And it's not like humanity is really, going to need that much resolution anyway (Unless post production turns into only shooting wide shots and pushing into the image for all subsequent sizes!) What I AM looking forward to is the increased data rate and color information the lower compression schemes Epic is offering at 5K.

And how much measured resolution do we really need? Out of curiosity. have you guys tested when you dialed-in your OLPFs to see if we really even need true 3K of resolution at the end delivery? Many people seem happy enough with 2K DI workflows and presentation. However, I certainly wouldn't cry if the industry adopted a full 4K DI workflow and presentation... It certainly couldn't hurt.

One thought that I keep pondering is I feel the next frontier in sensor improvements may not be in ever increasing pixel counts, but rather in decreasing the size of the other components on the CMOS chip around the photosite to help gain back as much pixel area as possible while maintaining the high-enough 5K pixel count. If Moore's law continues, hopefully there are gains to be made on that front, only increasing the preformance of CMOS photosites, dynamic range and lessening the fill factor if you get my meaning.

Okay I'm rambling.
 
The best way to increase the MTF curve at lower resolutions is to increase it at higher resolutions. By going for a 4k (or beyond) sensor we get vastly superior MTF at 2k than a 2k camera, for instance. (and at 1k etc.) Indeed, the only way to get high MTF at lower resolutions without nasty aliasing is to aim for, capture, and project at much higher resolutions.

Graeme
 
Christopher Probst said:
I agree about the striped array as well. Naturally, it too is not an ideal and even though Panavision spearheads the idea of their "macro cells" as a super sampling scheme to derive a 1080 image, in my mind, it's still a CFA that doesn't have a discreet RGB value derived at each photosite, so therefore, a compromise. Now I'm not knocking either method (though I was wasn't thrilled with the vertical streaks that can happen with hot pin-sources in a Striped CFA frame...) it IS just the nature of the beast with any CFA that some tradeoff has to be accepted.

And before even that comment gets criticized, my general feeling on CFAs now is somewhat placated by the fact that it has been shown, especially here on these boards, just what you can get out of a finely-tuned Bayer sensor and smart processing of the Bayer data. I don't think about obtaining 5K RGB out of an Epic. But getting a good, clean 4K? Sure! I'm anxious to start playing. And "super-sampled" 2K? No worries there at all.

On that note, I find it ironic with all of the focus on resolution, resolution, MTF and whatnot, that many people are choosing to put older and older glass on Reds. To me, it's just a validation of Otto Schade's work that we humans don't need high frequency detail. We need bulk lower- and mid-band resolution with good contrast reproduction to feel an image is sharp. In fact, I recently found Camtec an UNUSED set of 1950s-era Kowa spherical lenses. They can't keep them in the house! People like Matty Libatique keep taking them out to put in front of R1s. Do we think that is going to deliver superior MTF for the 4K signal? Certainly not. Does it somehow look more interesting. Definitely.

Therefore with the Epic now at 5K, and in my mind delivering a very solid usable 4K resolution, I think we're more than covered for the foreseeable future resolution-wise. And it's not like humanity is really, going to need that much resolution anyway (Unless post production turns into only shooting wide shots and pushing into the image for all subsequent sizes!) What I AM looking forward to is the increased data rate and color information the lower compression schemes Epic is offering at 5K.

And how much measured resolution do we really need? Out of curiosity. have you guys tested when you dialed-in your OLPFs to see if we really even need true 3K of resolution at the end delivery? Many people seem happy enough with 2K DI workflows and presentation. However, I certainly wouldn't cry if the industry adopted a full 4K DI workflow and presentation... It certainly couldn't hurt.

One thought that I keep pondering is I feel the next frontier in sensor improvements may not be in ever increasing pixel counts, but rather in decreasing the size of the other components on the CMOS chip around the photosite to help gain back as much pixel area as possible while maintaining the high-enough 5K pixel count. If Moore's law continues, hopefully there are gains to be made on that front, only increasing the preformance of CMOS photosites, dynamic range and lessening the fill factor if you get my meaning.

Okay I'm rambling.
Click to expand...

I am not directly affiliated with RED but I am exceedingly grateful that they exist. If I was truly religious I would say my prayers had been answered, as I work on developing some pretty deep technologies in the areas of 3d and S-3d.


The article that you have to write is a total mine field, and although you will be covering as you say the “historic development” of these types of imagers it’s almost impossible not to get sucked into saying something about the current offerings from various manufacturers.

Before you put pen to paper “in anger” as it were, I’d like to recommend the following book.


“Modulation Transfer Function in Optical and Electro-Optical Systems” by Glen D. Boreman, SPIE Press, ISBN 0-8194-4143-0. It talks a lot about the raw physics of MTF in optical systems, in Electro-Optical Systems, other MTF contributions, Measurement methods and practical measurement issues. It’s a very accessible book and characterizes pretty well the types of aliasing artifacts you get in relation to MTF.


I would also recommend that you spend at least three weeks trawling through the patent literature just to see all of the variants and possibilities that exist with implementing so called “standard technologies”. At least make sure that you have read about 60-100 key patents concerning the technologies you wish to present.


Then after that I think you get the idea that one has to tread VERY carefully indeed as everything hinges on the implementation, not nominally what the technology claims to be. An engine + a transmission + drive shaft, axles and four wheels does not a Bugatti Veyron make; (I think you get the idea!).


In engineering, this is what fascinates me so much as that now, with cheap powerful computers and highly “evolved” software tools, that small teams of focused engineers that are good at wearing more than one hat, can accomplish amazing things that twenty years ago would have required very large teams of discrete departments, (where the main challenge would be just to communicate requirements between different departments). My impression of RED team’s work (and its only an impression as I don't really know) is that there is a fairly enlightened approach that looked at the problem between hardware, signal processing, coding schemes and software as a whole entity right from the start. This integrated approach yields far more elegant and effective implementations than an artificially divided set of problems and skill sets. In a conventional larger company one would say, OK here’s the chip, now give it to the software guys, so in effect “up-stream” feedback is almost impossible, and optimal solutions are very rarely arrived at.


I think in “RED’s” case that they need at some point to hire (if they don’t have one already) a really good (textural) editor. [I’m very lucky as I am married to a superb editor that has saved my bacon too many times to recount.]. I think it is possible to explain in fairly simple language some of the technical advances and benefits of what RED “imagery” produces without giving away the recipe for the “secret” sauce. It’s not enough to say our images are really “nice” or really scrummy yummy… People need a little bit more than that, but if you provide hard data, people generally don’t understand it either, and it usually confuses the issue or becomes mis-applied ( a bit like SONY’s “mythic” 100% green!). It can take months and months to wrestle with the language to deliver a concise and clear explanation as to why and how something really works so that people go “Aha!”.. “Now I get it!!!” and yet you have not given too much away to your would be competitors. It’s an art and skill, something that very few scientists or technologists posses. A good technical editor is definitely worth their weight in gold.


I think the other thing that should be respected is that if you develop a new technology, it’s absolutely heart breaking if someone from the press is the first one to publically “explain” how it works. If you invent something new, in essence the owners and inventors of the technology should be the ones to FIRST present how their technologies actually work to the degree that they are comfortable with; academic/industrial white paper and so on. Things become more complicated when the technologies are already in the hands of your customers and yet a formal explanation as to the “goodness” of the system have not been comprehensively explained other than it does what it says it does on the tin, and to most people’s eye, they really love what it does!


The fact that Jim has not given any interviews regarding your topic area, to me, is completely understandable.


Just out of interest have you approached other manufacturers for technical advice on your article? [For example… SONY and others], and you also indicate a series of articles, so I guess that when the general history of the technology is over, other articles will follow that address current systems?

Is that right?


Cheers,

Eric
 
Christopher Probst said:
Graeme,

Ironically, I'm knee deep in this very sort of topic for a highly technical Digital Imagers article I'm putting together for American Cinematographer. I know I've asked several times to both you, Jim and Jarred for Red to contribute to this article and been declind (very politely, of course), but what you're talking about in this thread is exactly the sort of commentary I'm looking for. I wish again Red would reconsider their position in having a voice in American Cinematographer. Believe it or not, AC (and specifically ME in this instance) is not an enemy!

Perhaps I'll have to do an Epic article as a means of amends... But of course, that stipulates I get my hands on one!

Still waiting.
Click to expand...


Christopher

Dude….

I read your article on the RED ONE in AC.

Ohhhhh man….


You have a beautiful eye, the Vegas images are stunning; I hope one day you get to do some very serious and good s-3d one day (if you could stand to do it).

But man what you wrote.

I know the article is focused on “tips” for using the REDONE; however it came across as an unending nag session for eleven pages for everything that could possibly go wrong with the camera (I know you will say that wasn’t the half of it). In each paragraph there is virtually no positive counterpoint to effectively an unending stream of negative statements. I know that the approach that you employ is one of presenting useful truths to your peers; however I’m not really sure who the article really serves. If I was an old school cinematographer, and had read your article I would just say “screw it, …just shoot film” this digital stuff is way too risky and too much hassle.


If products that we released had the same treatment as your article, then I would be paying you huge amount s of money to NOT to use our equipment. Jim obviously has a thick skin and is a very kind and magnanimous sort of bloke, it seems. The imagery that you produce IS excellent.


Maybe the next article you write you should also use some very smart editor just as a sanity check on the tone and effect that you are trying to create? [Just as a friendly recommendation]


I honestly can’t imagine in some small part, why you are so surprised that they will not at this stage offer you an interview? Perhaps some bridge building and a very different tack might be needed, as well as to let some time pass? I don't really know... only you can be the judge of that.


Good luck with your projects.


Eric
 
Yes, old glass can look "interesting", or great, or whatever; but that is a personal decision for each individual to make, and should not set a standards frame of reference - no matter how fast these lenses fly off shelves. I don't see medium format still's guys running out to buy "old" lenses to put a lid on "high frequency detail". It's not all about high frequency detail. And we have yet to see anything in cinema that makes more resolution moot, as far as I'm concerned. With all of the options available to tame resolution, to sculpt it infinitely, that is what I'm talking about. We must move beyond broad effect.
 
Having enough potential alias-free resolution, sensitivity, and DR in the camera so that the optics become the limiting creative factor would seem to be ideal to me. Then you can make your own creative choices instead of compensating for inadequate capture.
 
Hi Eric

Thanks for your thoughts on my past article. For that piece I was asked to write a first-person account on my experiences working with the camera. AND the article was published pre-MX chip. I have read it again and I still stand behind what I wrote for the time-frame I wrote it. I have had dozens of cameramen thank me for the article because it gave pointers on how to shoot the camera, and what to look out for. None of them complained about their subsequent experience with the R1.

This article was debated both pro and con here on the boards previously; some wanting to take out a hit on me, and others not seeing the big deal. A lot has changed since then. Even with the R1 since the MX sensor came out. And now with the Epic, the gains and pains made with the R1 are being implemented into a whole new system. So I'm very excited by that. The R1 still has its quirks and annoyances. Those that bother me have all been addressed with the Epic. SOME of them were corrected with the MX chip. Don't get me wrong, I still like the camera. But it's also still not an ideal for me. It's a compromise I've learned how to use to my ultimate satisfaction on more that 100 professional projects now. BUT I still stand behind everything I said.

This new article is non-manufacturer specific and it is NOT first person. The tone is entirely different. I have been writing technical articles for AC for over 16 years now and I work tirelessly on them when I do one. In the past I've done massive stories on special lab processes, telecine and colorist techniques, even how to cut an effect reel. I work like a nut and still try to find time to edit the magazine every month.

I don't take anything I do lightly. I have done several months of research on the new topic at hand and I have been interviewing several different companies associated with digital imaging devices. The fact that I'm openly inviting Red to contribute to the story is that I value their place in the digital arena and feel its a shame not to at least ask. Its my loss and a loss for the readers of AC mag that they aren't part of the story I'm working on. And when people come at me and complain that I didn't include Red but still have the other "BIG WIG" companies represented, well I have not much to say about that.

I could write the entire article just around quotes pulled from Graeme from these message boards, but Graeme has also declined to speak in print. My only aim is to as objectively as possible, lay a ground-floor understanding on the development, design parameters and implementation considerations of digital imagers so that in the future any story discussing a digital cinema camera can have something to stand on. Who wants to have to read the same remedial explanations of what a CCD or a CMOS imager is, or why MTF is important (and what the fuck does it even really mean) in every future article talking about a camera's performance?

I KNOW people who are on this forum LOVE Red. And I too am a member of this community for a reason. I also KNEW I'd catch flack when my article came out, but I wrote what I felt and that was the assignment. It is our job to be critical of the manufacturers that supply the industry and encourage and reward progress. Likewise, it is also a right to demand reliable performance out of the products we use. I know of MANY cinematographers that were frustrated with the Red when it came out. Red has admitted openly here that they were learning as they went. And that's fine, but it doesn't make one happy on a multi-million dollar project when you're the DP and the camera system fails. Whether it's the DPs fault or not, all eyes are on him, so we DEMAND our camera systems perform under pressure.

Okay, I'll crawl back under my rock now.

Eric Lange said:
Christopher

Dude….

I read your article on the RED ONE in AC.

Ohhhhh man….


You have a beautiful eye, the Vegas images are stunning; I hope one day you get to do some very serious and good s-3d one day (if you could stand to do it).

But man what you wrote.

I know the article is focused on “tips” for using the REDONE; however it came across as an unending nag session for eleven pages for everything that could possibly go wrong with the camera (I know you will say that wasn’t the half of it). In each paragraph there is virtually no positive counterpoint to effectively an unending stream of negative statements. I know that the approach that you employ is one of presenting useful truths to your peers; however I’m not really sure who the article really serves. If I was an old school cinematographer, and had read your article I would just say “screw it, …just shoot film” this digital stuff is way too risky and too much hassle.


If products that we released had the same treatment as your article, then I would be paying you huge amount s of money to NOT to use our equipment. Jim obviously has a thick skin and is a very kind and magnanimous sort of bloke, it seems. The imagery that you produce IS excellent.


Maybe the next article you write you should also use some very smart editor just as a sanity check on the tone and effect that you are trying to create? [Just as a friendly recommendation]


I honestly can’t imagine in some small part, why you are so surprised that they will not at this stage offer you an interview? Perhaps some bridge building and a very different tack might be needed, as well as to let some time pass? I don't really know... only you can be the judge of that.


Good luck with your projects.


Eric
Click to expand...
 
Graeme Nattress said:
The other objection was that Bayer patterns were criticized for aliasing and chroma moire, whereas in practise (as shown amply by RED cameras) such problems generally don't occur, and if they are shown in test charts they are negligible in nature, whereas the sensor pattern of the F35 exhibits both strong luma aliasing and chroma moire:

indeed strong enough to make it visible on broadcast shows shot on the F35, not just test charts, along with a sub-1080p measured resolution, in part due to such strong aliasing. Indeed, pretty much most of the "criticism" levelled at Bayer pattern CFA sensors is MORE visible what they suggest as a superior alternative.

Graeme
Click to expand...

Okay I understand your argument. The way I will handle this idea is pretty no-favored-nations. I have heard several times in different situations Red felt it was being singled out and criticised but the other manufactures were not placing the same criteria on themselves (read Q67 double resolution fiasco here) and not comparing apples to apples or admitting their designs have similar or even moreso faults (the aliasing example you mention above.) Rest assured I won't be scrutinizing Bayer but not Striped CFA or even 3CCD for that matter.

In your example above, rather, I feel the best approach it to explain why aliasing/moire can occur and what in general is done to combat that. NOT going... "Well Canon moire's like a motherfucker so they suck and Panavision rules..." That's not the idea. I hope you get my meaning and realize that it is your calling out of these situations that will help me create a more balance piece.

Can we go back to discussing Demosaicing now?
 
Graeme Nattress said:
The best way to increase the MTF curve at lower resolutions is to increase it at higher resolutions. By going for a 4k (or beyond) sensor we get vastly superior MTF at 2k than a 2k camera, for instance. (and at 1k etc.) Indeed, the only way to get high MTF at lower resolutions without nasty aliasing is to aim for, capture, and project at much higher resolutions.

Graeme
Click to expand...

Exactly what I was saying I was excited about with the 5K Epic. However, at some point if we keep shrinking pixels to fit more on a fixed Super35-sized imager, we're going to lose sensitivity and dynamic range (under current CMOS technology building methods). The fact that you guys were able to make that drastic leap with the MX sensor and get 5K across, is great. I'm totally confident we can get exquisite, full RGB 2K for a DI with very little aliasing, and very eager to see what DCI deliverable 4K the Epic will produce.
 
Christopher Probst said:
Hi Eric

Thanks for your thoughts on my past article. For that piece I was asked to write a first-person account on my experiences working with the camera. AND the article was published pre-MX chip. I have read it again and I still stand behind what I wrote for the time-frame I wrote it. I have had dozens of cameramen thank me for the article because it gave pointers on how to shoot the camera, and what to look out for. None of them complained about their subsequent experience with the R1.

This article was debated both pro and con here on the boards previously; some wanting to take out a hit on me, and others not seeing the big deal. A lot has changed since then. Even with the R1 since the MX sensor came out. And now with the Epic, the gains and pains made with the R1 are being implemented into a whole new system. So I'm very excited by that. The R1 still has its quirks and annoyances. Those that bother me have all been addressed with the Epic. SOME of them were corrected with the MX chip. Don't get me wrong, I still like the camera. But it's also still not an ideal for me. It's a compromise I've learned how to use to my ultimate satisfaction on more that 100 professional projects now. BUT I still stand behind everything I said.

This new article is non-manufacturer specific and it is NOT first person. The tone is entirely different. I have been writing technical articles for AC for over 16 years now and I work tirelessly on them when I do one. In the past I've done massive stories on special lab processes, telecine and colorist techniques, even how to cut an effect reel. I work like a nut and still try to find time to edit the magazine every month.

I don't take anything I do lightly. I have done several months of research on the new topic at hand and I have been interviewing several different companies associated with digital imaging devices. The fact that I'm openly inviting Red to contribute to the story is that I value their place in the digital arena and feel its a shame not to at least ask. Its my loss and a loss for the readers of AC mag that they aren't part of the story I'm working on. And when people come at me and complain that I didn't include Red but still have the other "BIG WIG" companies represented, well I have not much to say about that.

I could write the entire article just around quotes pulled from Graeme from these message boards, but Graeme has also declined to speak in print. My only aim is to as objectively as possible, lay a ground-floor understanding on the development, design parameters and implementation considerations of digital imagers so that in the future any story discussing a digital cinema camera can have something to stand on. Who wants to have to read the same remedial explanations of what a CCD or a CMOS imager is, or why MTF is important (and what the fuck does it even really mean) in every future article talking about a camera's performance?

I KNOW people who are on this forum LOVE Red. And I too am a member of this community for a reason. I also KNEW I'd catch flack when my article came out, but I wrote what I felt and that was the assignment. It is our job to be critical of the manufacturers that supply the industry and encourage and reward progress. Likewise, it is also a right to demand reliable performance out of the products we use. I know of MANY cinematographers that were frustrated with the Red when it came out. Red has admitted openly here that they were learning as they went. And that's fine, but it doesn't make one happy on a multi-million dollar project when you're the DP and the camera system fails. Whether it's the DPs fault or not, all eyes are on him, so we DEMAND our camera systems perform under pressure.

Okay, I'll crawl back under my rock now.
Click to expand...


Christopher,

I understand where you are coming from, it’s a difficult situation.

There’s a terrific British expression that states “If you sit on the fence you get splinters up your arse!”


What I’m getting at is that it is possible to essentially present the same information, but in a more balanced tone where you manage to say what you want to say but in a fairly diplomatic manner that most parties can respect. I think the problem is that most writers no matter how good, generally do not seek the advice of a good editor. I’m sure as you appreciate these are different skills. The big problem is that especially with a publication it’s very difficult to get enough time and distance from the things that you write to have an objective opinion about your own work. No matter how many iterations and circles and agonizing hours tearing your hair out to get to final draft, it is impossible to get that higher level view on your own. So it’s easy to edit someone else’s work, it’s very hard to do the same for your own work. It’s taken me decades to understand that.

Similarly, good writing and a bit more thought might have helped the rather odd Q67 presentation from SONY where things had been over simplified almost to the point of lying. I’m sure eventually they will find a better way to explain things. The other thing that bothers me (slightly) is that Graeme will say something like and I paraphrase, this is not a direct quote [[[“ we’ll the images look nice..]]], the implication being that you don’t really need to know exactly what’s going on under the hood, we have that all taken care of, don’t worry about it. In this case Graeme is probably right!!!!, BUT I have to say this is almost like the British “nanny” state where the idea is that “we know what is good for you”. I don’t think that plays so well in the US, where this is the land of the individual, and individuals like to make informed choices. Graeme probably does know what is good for you, but again here I believe that good technical writing has the capability to express deeper and more useful concepts WITHOUT giving away trade secrets. However this really does require a massive amount of effort to get right; that is why I was little shocked at the SONY presentation of the Q67.


Armed with good technical editors and writers you can make a real difference, spouting technical facts and woes is not enough and occasionally things go rather “pear shaped” in print as a result.

Ta.

Eric

I’ll crawl back under my rock too! Actually I’m going to take my eight year old daughter down the trail on her fat little pony “meatball”.

Seeyah/good luck with de-Bayer
 
There is a misconception in the industry that manufacturers are required to tell you what they are doing. We don't subscribe to that concept. We get to tell you what we want and nothing more. Apparently we are leading the way and pushing the industry along in rapid advancing cinema technology (am I wrong?). Every competitor is keen to find out what we are doing, how we are doing it and what we will do next. We have learned the hard way that talking too much is bad for our business. As Soderbergh said... "I don't know how they do it... magic?"

Jim
 
Obviously, motion picture film's physicality made it a poor candidate for competing with the quality of large format stills; however, sensors are not limited in this way, so why anyone would curtail the development of a technology that might approach the superlative qualities of large format, is beyond me.

Here's to "pushing" technology. Let the artist's abilities, and not his petty preferences, be the only limiting factor.
 
Jannard said:
There is a misconception in the industry that manufacturers are required to tell you what they are doing. We don't subscribe to that concept. We get to tell you what we want and nothing more. Apparently we are leading the way and pushing the industry along in rapid advancing cinema technology (am I wrong?). Every competitor is keen to find out what we are doing, how we are doing it and what we will do next. We have learned the hard way that talking too much is bad for our business. As Soderbergh said... "I don't know how they do it... magic?"

Jim
Click to expand...

That’s pretty clear.

In industry as you know it’s a horrible situation, you either have to go the patent route or the trade secret route. With the patents (as you know), by law one has to print or quote the patent numbers on the products themselves. The specification, if granted is published in the public domain, and theoretically the description has to be sufficient so that “one of ordinary skill” in the relevant arts can perform the invention. So at least that grants one the right to enforce a monopoly for about 14 years until such point the patent expires and passes into the public domain, then everybody has it (that’s one of the main tradeoffs). There is also the grey area where academic institutions might or might not have the right to perform the invention. A trade secret HAS to be maintained as a secret, so no easy licensing options there, and if someone independently re-discovers what you have done, then there is very little recourse.


It’s a shame you can’t have a patent that is also a trade secret that is not publicly published that you can take to the grave! That actually might help smaller entities and independent inventors do battle with giant corporations, and even then you still need a considerable war chest for court proceedings and litigation to defend one’s IP.


As time moves on I genuinely think that the basic RED vision and Red team and approach is what will in the long term make the real difference. I firmly believe that you guys will always be the best at what you do. No one else will ever be able to OUT-RED RED. That's pretty clear, and something to be proud of. Other players large or small are simply not “geared up” to do what you do, and they never will be.

Perhaps “magic” makes a better adjective than “nice”.

Cheers,

Eric
 
Here's a question I've asked before Graeme and been ignored..

I'd like to deconstruct the debayered RGB image generated by a R1 back to pseudo raw for motion calculations, what is the photosite order for the different R1s resolutions, do they alternate Red Green or Blue green on the first line etc..It would be very useful information if you could enlighten.. and are the photosite orders the same for each resolution (I'm guessing yes). Or is this top secret info ? ;)
 
You must log in or register to reply here.
Back
Top