Pages: 1 [2] 3 4   Go Down

Author Topic: RED carpet bombs NAB  (Read 27330 times)

ChristopherBarrett

  • Guest
Re: RED carpet bombs NAB
« Reply #20 on: April 18, 2012, 11:28:06 pm »

Early reports are that the R1 shoots cleaner than the Scarlet.

My Epic shoots cleaner than my R1.  Less noise in the shadows, better color fidelity.  More DR.  The R1 just came back from RED with a new OLPF and a clean bill of health.  It's still a friggin awesome camera.  The SSD module made it way more convenient to shoot and I can't wait to get it down to Louisiana for our next short.

When I'm not shooting I spend a lot of time thinking about gear... so that when I AM shooting I never have to think about gear.  Get shit that works and go work.

Yada yada etc
Logged

bcooter

  • Guest
Re: RED carpet bombs NAB
« Reply #21 on: April 19, 2012, 03:29:28 am »

Early reports are that the R1 shoots cleaner than the Scarlet.

My Epic shoots cleaner than my R1.  Less noise in the shadows, better color fidelity.  More DR.  The R1 just came back from RED with a new OLPF and a clean bill of health.  It's still a friggin awesome camera.  The SSD module made it way more convenient to shoot and I can't wait to get it down to Louisiana for our next short.

When I'm not shooting I spend a lot of time thinking about gear... so that when I AM shooting I never have to think about gear.  Get shit that works and go work.

Yada yada etc


I dig your enthusiasm for photography and moving forward.

I also think a lot of good things about RED, but sometimes get kind of put off on the secret club/fan boy thing that goes on.

I'm sure you've done this  . . .  read a RED manual on a camera.   The scarlet is the first manual I've read front to back and regardless of what I read the camera does different things than what's buried in the PDF.

I also am a little turned off by the pixel race.  Got very bored with it for still photography, even more so with motion, given the fact that 90% of everyone will edit in 2k, given the fact that probably 99.999999999999% of all video will end up on the web for viewing.

Yesterday, we had an issue with one of our R-1's.  No matter what we set the camera at, the monitor would not  reflect changes of  wb, or iso, or tint.

What looked pretty good at iso 1000, in reality was a stop over, so it required us puling the SSD, looking at the computer, then adjust, then back to the computer.

Last night I tried to find the answer on RED user net, which sent my head spinning, so once again, back to the manual and running tests.  It's still kind of goofy and makes me worry about the camera as I have one more country to finish up on before I can get the camera into service.

We can switch to the Scarlet which I find shoots smoother than the R-1 buy not as pretty and we can't find a way to play through a quick scene on camera and hear the sound.  It's there so once again, back to the computer to be sure.

So during a day like yesterday, we do think about the camera.

Now these cameras have been to 4 countries on this gig and carried in and out of trucks a gazilion times in the last few weeks, so they've taken a beating, though the Sony fs100 we've treated it like an old shoe and today no issues.

This could all be an anomaly, but yes today as we head to the airport I am thinking about the cameras.

IMO

BC

« Last Edit: April 19, 2012, 03:38:37 am by bcooter »
Logged

fredjeang

  • Guest
Re: RED carpet bombs NAB
« Reply #22 on: April 19, 2012, 04:56:22 am »

The problem I see is, as it's been pointed several times, the coincidence with our still digital short history. It seems that the pixel race has been launched, it seems that DR button has been pressed etc...
In short, it will be more or less the same BS. Point of no return. We're there in the deep shit again.

What happened within the MF still industry is that they focussed on those easily sailable arguments (pixel-DR), putting artificialy price to the stratosphere (= speculation) and almost nothing has been done on
usability and design. They are paying a high price for that actually.

To give an example, one of those suposdly "high-end" backs cost the same amount of money as a Sony F35, wich is a hyper advanced 14 stops DR pro video camera and much more advanced than any Red. It gives an idea of the grade of nonsense and robery MF manufacturers are doing commerce with.

I'm afraid this is what will also happen in motion. Same drawing, same story-board, same nonsense.

Then, it's gona be hilarious (or not !...) when clients are going to ask 4K, without even knowing why but just because they've heard that everybody's in 4K on-the-cheap. It's going to be 4K just for the buzz. No rationality. 4K for 4K. Then 5, then 6...

What about broadcast standart then? Are we already in the situation that the HD standart is now to relegate to the dump and they are re-thinking it ? or can we hope this is going to be "stable" for some more years?

Coot talks about the web as a lowres output. But I'm afraid this 4K mantra is also going on the web inevitably. With the progress of connections, Youtube already has 4k capabilities. So yeah, we're about to experience another round of fences and garden shots, my cat my dog and my bird in 4K with all the pixel peeping megalopsychopathology involved. Times are closed indeed when we'll have the equivalent of this DxO stuff here and all the pseudo-engineers crowd of the forums will post graphics and comparatives of all kinds as a scientific approach of the motion imagery.

It's going to be the wild west even much more than now. Every manufacturer will go out with its format that nothing will read. I doubt a lot on the Raw DNG for motion. It's going to be a complete mess and we'll have to ugrade computers, cameras and softwares every 2 years to keep the race. It's an enormous lucrative business, maybe more than stills and will be under pressure and forced to spend way much more in equipment and derivated.

Not only the current pros are pushing from the top towards us but all the wanabee facebook crowd with 5D4s and Boris color is going to push from the bottom and we'll be sratched.
The New stars aren't going to be the Avedons but the bloomedons. Guys testing gear for the wanabees, doing networking and using their own image as a brand with a myriad of gear sponsors that flash like a Las Vegas avenue at night.

There was a time I thought that motion imagery was going the healphy way. Now I think it is the same circus as the still world became.

Beleive me guys, the more we stay away from all that BS, the best for the business and peace of mind it will be.
« Last Edit: April 19, 2012, 09:17:08 am by fredjeang »
Logged

stewarthemley

  • Guest
Re: RED carpet bombs NAB
« Reply #23 on: April 19, 2012, 05:42:57 am »

Fred, that's pretty much how I see it. Went down the sheep track, following the MFDB hype, which is actually BS (I said so on this site and got panned for it) and came out the other side knowing that for me, all I care about is content. Luckily, so far, that's all my clients care about.

Now I'm back doing video, I'm resisting the BS and getting the minimum that gets the job done and gives the clients what they want. It's vital that my gear works when I turn up at a gig. Once something lets me down, it's out the door. No second chance. And that's one of the attractions of gear at the lower end, the Sonys and Canons. They just work. You don't have to wade through manuals and forums looking for reasons why the thing failed and workarounds to try to get the thing to do what you paid for it to do. You just push the record button and go. Ironically, that's almost priceless.

I guess one of the (few) good things about the latest economic mess is that this endless chasing of "better" gear has been slowed right down.
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: RED carpet bombs NAB
« Reply #24 on: April 19, 2012, 06:22:30 am »

What about broadcast standart then? Are we already in the situation that the HD standart is now to relegate to the dump and they are re-thinking it ? or can we hope this is going to be "stable" for some more years?
I am sure that broadcast will evolve (in ways that necessitate the purchase of new gear and more expensive subscriptions), but I am not so sure that >1080p is happening in the close future. I think that 3D, framerate, color depth, DR etc is more likely to eat the increasing complexity/bandwidth budget.
Quote
...With the progress of connections, Youtube already has 4k capabilities. So yeah, we're about to experience another round of fences and garden shots, my cat my dog and my bird in 4K with all the pixel peeping megalopsychopathology involved. Times are closed indeed when we'll have the equivalent of this DxO stuff and all the pseudo-engineers crowd of the forums will post graphics and comparatives of all kinds as a scientific approach of the motion imagery.
...
I don't see the reason for this hostility towards technical angles or amateurs making content.

The general public have shot images and videos of their cats, kids, trains (or whatever interested them) for as long as it has been technically/economically possible, and they will continue to do so as long as it is possible. Seeing a video of a cute cat may be boring in 4k, but it is equally boring in 480i.

People will try to understand the stuff that they are interested in. Some do it using formulas that they may or may not apply with skill. Some do it with practical tests that may or may not be relevant for real-world usage. Some of the most interesting results probably come from those who master both theory and practice to a sufficient level.

-h
Logged

fredjeang

  • Guest
Re: RED carpet bombs NAB
« Reply #25 on: April 19, 2012, 08:51:55 am »

I don't feel hostility on the technical-engineering sides of the imagery, on the contrary.

I only feel hostility -or been better expressed- tired of, the pseudo technical noise that abunds in internet. Imagery is technical and artistical. The parameters involved are complex in terms of skills, techniques and gear is really important.
But that's the point. I rarely (very rarelly) read really really serious high-end technical discussions made by experienced people. All the noise turns arround one thing: a camera+sensor. point. All the other parameters like lightning techniques, optic, post-production techniques, narrative techniques, the MUAs, the pre-prod etc... are very rarely debated. It's reso reso and reso at the dr sauce. This, I have an indigestion to be honest.
All they do is exhibit curves and graphics taken here or there in some public websites that target the general public as "information", "proofs" of all kind (to proove what, is still a mystery to me), and always those pseudo debates based on pseudo infos end in wars, accusations and Michael having to do the policeman in the end or close threads.

This forum was safe of this plague so far. Not for long.

Now that we have 4K and the race is lauched, you'll see the wars between what isn't 4K and what is. Then graphics posts by the wanabees will fall from the sky to proove that we don't need 4K etc etc...
Then the ones who won't be able to access high-end equipment will try desperatly to demostrate their usefullness bombing threads with sparkling tittles like: "Canon kills RED with the latest 6K"

I'm not against scientists. I'm not against tech.

I'm against pseudo-scientists and tech arguments that come from people with in general little information and little technical skills. That's what I criticize.
« Last Edit: April 19, 2012, 09:00:44 am by fredjeang »
Logged

Graeme Nattress

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
    • http://www.nattress.com

This actually sounds like a good fit to the coming generation of 4K cinema projectors: 6K with RED's Bayer CFA is about what is needed to make the most of those 4K projectors, which deliver all three primary colors at all 4096x2160 pixels.

Likewise, for resolution this is about on par with the Sony F65, which has 20 million total pixels vs about 18 million for the RED Dragon), with 10 million green pixels laid out on a rectangular grid to roughly match the 4096x2160 output, and then half as many of each of red and blue, arranged diagonally (like in the old Fujifilm SuperCCD.) Details at Sony's site here

But that Sony F65 costs about $65,000!

The F65's 20mp is not all used for image, so the comparison number is 17.6mp I think, so there's no resolution advantage there.

Graeme


Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: RED carpet bombs NAB
« Reply #27 on: April 19, 2012, 10:32:43 am »

I only feel hostility -or been better expressed- tired of, the pseudo technical noise that abunds in internet.
The internet is noisy. With experience, one learns how to extract the information that one needs.
Quote
Imagery is technical and artistical. The parameters involved are complex in terms of skills, techniques and gear is really important.
Agreed
Quote
But that's the point. I rarely (very rarelly) read really really serious high-end technical discussions made by experienced people. All the noise turns arround one thing: a camera+sensor. point. All the other parameters like lightning techniques, optic, post-production techniques, narrative techniques, the MUAs, the pre-prod etc... are very rarely debated.
"really, really serious" stuff can be found in research journals. People (with a few exceptions) rarely have the time or inclination to spend months or years researching a Science or Nature-worthy piece, having it peer-reviewed to publish it in a discussion forum. Most of us are unable or unwilling to comprehend those papers anyways.

I have found some authors on this forum able and willing to correct my own misunderstandings and gaps of knowledge using arguments, references and an attitude that ease the change of mind. I appreciate that a lot.

You probably are right that many people are more interested in the latest electronics than in lighting technique. Optics, though?

I do some exposure bracketing. I have learned from technically minded posters that Nikon right now does DR somewhat better than my Canon system. I don't think that I would have known or accepted this without some "curves, numbers and references", or highly controlled side-by-side shots that seems very hard to produce. The equally evil twin of "pseudo-scientists" may be the "touchy-feely subjective position" that may be even more common on the net.
Quote
always those pseudo debates based on pseudo infos end in wars, accusations and Michael having to do the policeman in the end or close threads.
This is not how I perceive things.
Quote
I'm not against scientists. I'm not against tech.

I'm against pseudo-scientists and tech arguments that come from people with in general little information and little technical skills. That's what I criticize.
I guess that we are all still learning, and given any one group of people and narrow topic of interest, typically only one person of that group will be most knowledgeable.

For the rest of us, participating in the discussion, suggesting explanations and receiving criticism is a way of learning, just like posting images can be a (harsh) way of learning the art. As long as people contribute with a positive tone and don't try to appear more knowledgeable than they really are, I see no problem.

-h
« Last Edit: April 19, 2012, 10:36:26 am by hjulenissen »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
4K vs 6K vs 8K: spec sheet wars and interpretation
« Reply #28 on: April 19, 2012, 10:39:56 am »

Graeme,
Agreed: I mentioned the total photosite counts only to indicate rough equivalence in resolution, not to claim a 10% advantage for the F65.

To put it another way, I was just pointing out that the resolution difference is probably not what some people might think by reading just the headline "Red Dragon 6K > Sony F65 4K" specs.

On the other hand, Sony sometimes refers to the F65 sensor as "8K" (because photosites fall on 8000 different vertical lines, due to their diagonal layout), but a comparison based only  on "Sony F65 8K > Red Dragon 6K" would be equally misleading!


Whatever happened to describing resolution in good old-fashioned "lines per picture height"?! Not a serious question: I know that such things are not simply determined from the sensor hardware description alone.
« Last Edit: April 19, 2012, 03:46:52 pm by BJL »
Logged

smthopr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 612
    • Bruce Alan Greene Cinematography
Re: RED carpet bombs NAB
« Reply #29 on: April 19, 2012, 01:19:00 pm »

4k observations...

I was just at the NAB convention and saw the 4k demo projections from the major 4k cameras in 4k.

The 4k looked beautifully detailed...until the camera panned.  Shot at 24fps the motion blur / judder created when panning dropped the detail very significantly.  The Canon demo movie carefully avoided panning, I think for this reason.

So, perhaps more important than 4k or 8k, we should be thinking about 48fps or higher for high resolution motion pictures.  The challenge is that today, 4k post is not cost justified for almost all productions, and doubling the frame rate will double the already taxing data load that makes 4k so expensive.

Just my 2 cents :)
-bruce
Logged
Bruce Alan Greene
www.brucealangreene.com

fredjeang

  • Guest
Re: RED carpet bombs NAB
« Reply #30 on: April 19, 2012, 02:10:20 pm »


So, perhaps more important than 4k or 8k, we should be thinking about 48fps or higher for high resolution motion pictures.  The challenge is that today, 4k post is not cost justified for almost all productions, and doubling the frame rate will double the already taxing data load that makes 4k so expensive.


My thoughts too.

ps: If I had to invest right now from scratch (in serious prod) I think a camera like the Sony F35 would be my choice over Ks.

Here is a contrasted point of view: http://www.definitionmagazine.com/journal/2009/12/1/why-i-bought-a-sony-f35-cinematography-camera-by-dan-mulliga.html

It's also interesting to note that the british company Take 2 Films, is working with the F35.
« Last Edit: April 19, 2012, 03:00:38 pm by fredjeang »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Does 4K need 48fps? Jackson and Cameron seem to think so
« Reply #31 on: April 19, 2012, 04:40:19 pm »

The 4k looked beautifully detailed...until the camera panned.  Shot at 24fps the motion blur / judder created when panning dropped the detail very significantly.  The Canon demo movie carefully avoided panning, I think for this reason.

So, perhaps more important than 4k or 8k, we should be thinking about 48fps or higher for high resolution motion pictures.
Indeed, Hollywood's two most prominent big budget high tech. film-makers, Peter Jackson and James Cameron, are adopting 48fps: Jackson is already using it for the Hobbit (with RED Epic 5K Bayer CFA cameras too) and Cameron says he will for an Avatar sequel to two. So 4K and up and 48fps and 3D ... this should keep the mass storage device industry happy anyway!

http://the-hobbitmovie.com/peter-jackson-discusses-new-filming-standard/
http://latimesblogs.latimes.com/movies/2011/03/james-cameron-champions-faster-film-projection-rates.html
Logged

Graeme Nattress

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
    • http://www.nattress.com
Re: RED carpet bombs NAB
« Reply #32 on: April 19, 2012, 08:00:38 pm »

BJL: I've been unable to measure a F65 yet, but some Sony documentation presented it's sensor as 6k x 3k which fits in much better with it's pixel count and what I figure it's measured resolution would be. My only thing is that Sony keep saying 8k, which it's not and 20mp which it isn't either because over 2 of those mp never get seen or recorded.

Bruce: I can't wait to see what Peter Jackson with Hobbit is doing with his 48fps

fredjeang: a contrasting view indeed, but most of what he says is wrong, and it's very out-of-date, (and double wrong from a "today" pov) and looking at the F65 I hope Dan Mulligan made his money back on the F35.

Graeme
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: RED carpet bombs NAB
« Reply #33 on: April 19, 2012, 08:57:59 pm »

BJL: I've been unable to measure a F65 yet, but some Sony documentation presented it's sensor as 6k x 3k which fits in much better with it's pixel count ...
It seems fairly clear from the documents that I link above that the F65 sensor uses diagonally aligned photosites (like the black squares on a chequer-board, for example) with the green ones on a 4096x2160 rectangular grid (so, "true full 4K" for green) and then an equal number of red plus blue photosites located diagonally between them.  (That is, a chequer-board with red and blue squares, and then a green dot at every corner between the squares.) So the total count is about 4096x2160x2, about 1.77MP. Rounded up to one significant digit to get 20MP?!
Logged

Graeme Nattress

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
    • http://www.nattress.com
Re: RED carpet bombs NAB
« Reply #34 on: April 20, 2012, 12:24:15 am »

Yes, 17.7 with a diagonal array, hence the confusion over horizontal pixel count which is 4k or 8k depending on how you count, but the math and the early Sony documentation give a figure of 5.6k and 6k accordingly. Nope, not rounding to get 20mp, but just counting the edge black and unused pixels on the sensor (active pixel count compared to full sensor count).

Graeme
Logged

fredjeang

  • Guest
Re: RED carpet bombs NAB
« Reply #35 on: April 20, 2012, 03:41:16 am »

The Sony pixel count allows false marketing claims, it reminds me to use an image the Foveon system and interestingly how Sigma exploits it.

Graeme, yes this is an old post on purpose, refering to (or compeat to) material that have evolved since in both brands. It's because I had in mind the second-hand market according to a previous post in this thread.

A different point of view is healphy, specially after realizing the grade of fanboyism the Red forum has, wich I think it's never good for us consumers as any critic is automatically rejected by the users. (Red isn't the only brand in this case, fair enough). Dan's was very aware of it in the article. I got the guys points, so as yours. So far, every cat that have critized Red, experienced or not, the answer is always "that's not true". Maybe it isn't, maybe nobody on earth has brought on the table to date truth reliable contrasted opinions or critics on the system because it's never thruth and there's always a clever tech answer from the magic hat.

This reminds too much MF early days. Hope you guys over there won't fall on the same nonsense, familly business mentality.  
« Last Edit: April 20, 2012, 03:53:55 am by fredjeang »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: RED carpet bombs NAB
« Reply #36 on: April 20, 2012, 10:28:52 am »

The Sony pixel count allows false marketing claims ...
But likewise, I would call the RED "5K" and "6K" claims a bit misleading or at least open to misinterpretation, though not outright false. Because in video, and particular with video projection and also with many professional video cameras, resolution spec's like 1920x1080 or 4K have traditionally referred to a count of pixels each with full three primary color information. Many professional video cameras use three or more photosites to get each of these full color pixels, through methods like:
a) striping, using three or six photosites (as in the Sony-made sensor of the Panavision Genesis?)
RGB
RGB
b) Canon's creation of one pixel from four photosites in the C300:
RG
GB
c) using three sensors and beam splitters (only up to 2/3" format though AFAIK.)

But on the other hand, with still cameras and now with recent large sensor video cameras, stated pixel counts are sometimes counts of single color photosites. RED seems to be leading the charge, due to its adoption of Bayer CFA sensors, and Sony (or its marketing department) is now following suit.

When you dig into Sony's documents on the F65, it is acknowledged that what it actual offered is full 4K resolution in green, and from 4K down to 2K resolution in red and blue, depending on the direction in which resolution is measured. (4K horizontal and vertical, 2K at 45ยบ angle, due to the "checker-board" pattern of the photosites for each of these colors.)  Does RED describe such details of actual sensor resolution limits anywhere?
Logged

Graeme Nattress

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
    • http://www.nattress.com
Re: RED carpet bombs NAB
« Reply #37 on: April 20, 2012, 10:59:51 am »

4k, 5k, 6k etc are always in reference to the sensor and not regarded as measured resolution. With the original R1 in 4k, we found we could get a good measured resolution of around 3.2k, and now with current cameras the "factor" from sensor k to measured resolution is around 80%. That is not 100% obviously, but neither would you want it to be unless you really like aliasing artifacts, which we don't. Although the stills guys can somewhat "get away" with more aliasing because they can always go and fix up that one frame, for motion it can be atrocious. (yes, we've talking about this extensively in public forums)

a) yes, Sony F35 and Panavision Genesis share the same heritage and use a RGB stripe pattern that uses more pixels to produce less resolution and more aliasing and chroma moire than the REDs, so a "theoretical" gain is turned into a practical dis-advantage.

b) Canon's idea is reasonable enough, but if you were to take their sensor data (as you'll soon be able to do with the C500) and properly demosaic then properly downsample to 1080p you'll get a better looking result. I've done that exact experiment with RED data to confirm how that demosiac arrangement works. In other words, there are better ways to decode that raw sensor data, but they're more expensive to implement (in terms of processing power, heat, weight etc.)

c) works well enough at 2/3", but not really practical beyond that, although NHK did work with a 4 way prism for their 8k camera so it can be done. Still, you're trading again and you're getting prism issues and lack of good lens compatibility for some theoretical advantage that doesn't typically pan out to a practical one. Interestingly many 3 chippers used a deliberate offset in the green sensor to achieve better results for 4:2:2 chroma sampling images. And you can't forget you're still fighting sampling theory, still need decent optical low pass filters.

Of course, people often come back with the three layers of film argument, but then when you look at a high quality film scan you see a very different amount of actual detail in the three channels, so this concept of unique RGB values is very much a computer graphics idea, not a traditional video concept, especially as practical video has always recorded and/or broadcast chroma sub-sampled signals.

Oh, no, we're not leading the charge! Very rarely if ever has measured resolution equalled pixel count. For "4k" I believe Dalsa were the ones to start with that description.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: RED carpet bombs NAB
« Reply #38 on: April 20, 2012, 11:15:48 am »

Oh, no, we're not leading the charge!. ... For "4k" I believe Dalsa were the ones to start with that description.
OK, Dalsa was first. (Whatever happened to Dalsa in digital cinema cameras anyway?) But ...
Very rarely if ever has measured resolution equalled pixel count.
I somewhat disagree: true in digital photography and in video overall, but in the high end of video, like digital cinema, a lot of cameras have used specs like 1920x1080 to mean full three color pixel counts. In particular, in the era when higher end gear was mostly 3CCD --- But that is water under the bridge, and I am not arguing for going back to 3 sensor cameras or striping for the sake of "honest" pixel counts! I agree with you for the rest; as far as I can tell anyway, a Bayer CFA plus good demosaicing algorithms is the current best strategy, despite some occasional forum dogmata about "use all the light!" or "measure resolution by the lowest value amongst the color channels!".

Again my point was just that there is some ambiguity in attempts to indicate resolution by a single sensor spec. number, due largely to the persistent difference between pixel counting in capture [usually one color] vs pixel counting in display [usually all primary colors].
« Last Edit: April 20, 2012, 11:20:05 am by BJL »
Logged

Graeme Nattress

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
    • http://www.nattress.com
Re: RED carpet bombs NAB
« Reply #39 on: April 20, 2012, 11:28:41 am »

I think the number of HD cameras that actually managed a measured HD resolution (never mind equal in RGB) with negligible aliasing is somewhat slim. It is by far and a way not the majority, hence my "Very rarely if ever" statement. Even the best of the beast, the F35 had significant aliasing and chroma moire and didn't really measure 1920 across. To get fully sampled recording, I think all the solutions were off-board recorders too.

" as far as I can tell anyway, a Bayer CFA plus good demosaicing algorithms is the current best strategy," - yes, but only at a decent enough resolution. If you try that for sub-3k resolution on the sensor, I think you're getting into the realm where the other approaches begin to make more sense. Turning it around, the bayer CFA sensor only began to make sense when it went large and high resolution (like a DSLR or in a digital cinema camera). I was terrified of bayer CFAs for motion cameras before RED, but practical experience of DSLRs and then of what we could do at RED taught me otherwise.

"Again my point was just that there is some ambiguity in attempts to indicate resolution by a single sensor spec." - absolutely, which is why we've often posted MTF plots to tell a better picture of what is going on.

Graeme
Logged
Pages: 1 [2] 3 4   Go Up