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 :)

Whats Next After Dragon From Red????

R

Rohan B

Well-known member
Joined
Feb 18, 2011
Messages
371
Reaction score
0
Points
0
I know maybe it is too early to discuss right now, but still, i was wondering,and i am sure many more like me too are, about what the next sensor or camera from red will be like and when will we be seeing it 2017/2018. will it finally kill film in terms of overall image quality??? I thought since alexa's dr is so close to film(the closest to film), the next sensor will beat the film dr but it seems the new 65mm sensor is just the oversized version of the Alev iii sensor, increasing only the resolution without shrinking the pixels,so just maintaining the previous dynamic range but not enhancing or increasing it like the dragon has over mx. So maybe the next sensor from red finally kills film truly in terms of overall image quality and technical superiority. Film will die anyway in a few years even if the improvements in digital technology comes to a halt, simply because of many other factors, apart from image quality, which have already been discussed many times before. But i would surely want to see a digital cinema camera or sensor that can truly be called better than film before film completely becomes obsolete. And i would want film to be completely retired by a successor that is worthy enough or worthier to take its place, and not just because the "old one" is getting old and needs to be replaced. I would want acclaimed film loving Dops and directors like Christopher Nolan and Wally Pfister among many others to embrace digital happily and for it's exceeding Film's image quality and overall technical superiority and not just unhappily using digital cameras because of films unavailability and shutting down of all labs and still boasting about film's image quality and indulging in the old memories about how good and unbeatable and irreplaceable it was in terms of image quality........... because that would be truly unfortunate.

By the way i love my dragon and Alexa and the f65 too, and i am not saying that i am not happy with theese greeeaaaat digital cinema cameras and their sensors, all i am saying is that there still isn't a digital sensor that i can truly call superior than film, they are all verry close, verry verry verry close, but still a little behind,and i would soon love to see that "little" be eliminated.............................:thumbsup:
 
Rohan B said:
I know maybe it is too early to discuss right now, but still, i was wondering,and i am sure many more like me too are, about what the next sensor or camera from red will be like and when will we be seeing it 2017/2018. will it finally kill film in terms of overall image quality??? I thought since alexa's dr is so close to film(the closest to film), the next sensor will beat the film dr but it seems the new 65mm sensor is just the oversized version of the Alev iii sensor, increasing only the resolution without shrinking the pixels,so just maintaining the previous dynamic range but not enhancing or increasing it like the dragon has over mx. So maybe the next sensor from red finally kills film truly in terms of overall image quality and technical superiority. Film will die anyway in a few years even if the improvements in digital technology comes to a halt, simply because of many other factors, apart from image quality, which have already been discussed many times before. But i would surely want to see a digital cinema camera or sensor that can truly be called better than film before film completely becomes obsolete. And i would want film to be completely retired by a successor that is worthy enough or worthier to take its place, and not just because the "old one" is getting old and needs to be replaced. I would want acclaimed film loving Dops and directors like Christopher Nolan and Wally Pfister among many others to embrace digital happily and for it's exceeding Film's image quality and overall technical superiority and not just unhappily using digital cameras because of films unavailability and shutting down of all labs and still boasting about film's image quality and indulging in the old memories about how good and unbeatable and irreplaceable it was in terms of image quality........... because that would be truly unfortunate.

By the way i love my dragon and Alexa and the f65 too, and i am not saying that i am not happy with theese greeeaaaat digital cinema cameras and their sensors, all i am saying is that there still isn't a digital sensor that i can truly call superior than film, they are all verry close, verry verry verry close, but still a little behind,and i would soon love to see that "little" be eliminated.............................:thumbsup:
Click to expand...


I strongly feel that Dragon is superior to motion picture film stock in nearly every technical merit. Aesthetic-wise, that's a different conversation.

Resolution in relationship to film format size? Advantage Dragon.

Dynamic Range, depending really on how you shoot it, Dragon can perform better. Sort of the same thing really with film stock and exposing for the appropriate and recommended ASA and lighting conditions.

Texture? Film has grain, Dragon has texture. Some hate texture and grain, some don't. Dragon's feel is different than film. That's sort of a toss up to me really.

Color? Tough one as there's also a subjective answer here. Dragon with the Skin Tone - Highlight OLPF is what I've done the most extensive film related testing with and it's a very controlled and consistent color, which is important in these times.

Highlight Roll-Off? Well Dragon doesn't exactly have a digital look to highlights through clipping. It's actually pretty nice.

Cost, efficiency, and general production speed related thoughts? Well, you likely aren't investing in film labs these days.


As for what Red has baking in the R&D oven? Well a couple things. It's clear that Dragon is a developing story. Especially this week! Dragon is now and an ongoing digital cinema camera receiving various advancements as we move forward through time and space. I think it would be foolish for Red to discuss what's next as it's A. too early to announce a thing, B. giving competitors way too much information, and C. in the midst of a Dragon storm.


Thoughts on DPs who prefer film? They prefer film. There's reasons for it outside of the technical merit that land more in aesthetics and a slight intangible magic. I have demoed Dragon for very film biased DPs and nearly every one of them had good things to say or were just blown away from the the potential of the whole thing. In fact it seems to me that the "film guys" like Dragon more so than the "video guys" and I get that. However, for some of those folks all of this is a bag of tangled wires. Dragon though is a camera that you can shoot like film and "works" like film. It's really the first time we've had that when you add everything into the mix. That's the entire reason I own one and have been testing harder than ever because it's hit a certain magic spot and more improvements/advancements continue to develop.


I personally have very strong thoughts about what Red "should" do next, but I'll chew their ear off about it and Red has been very good at listening to rather demanding shooters in the past.
 
well i am not gonna disagree with you on most of the points you made Phill, but still didn't alexa did most of the things that dragon does today three years back in terms of color and dr, the only thing it did not do was resolution and frame rates and of course then ther is size weight and cost but thats a different story all together.................
 
Bob Gundu said:
Considering RED is going to more photography conventions, I expect the next round to be more photo happy.
Click to expand...

hmmmm thats a possibility......:001_huh:

high end professional red dslrs sounds good
 
Rohan B said:
well i am not gonna disagree with you on most of the points you made Phill, but still didn't alexa did most of the things that dragon does today three years back in terms of color and dr, the only thing it did not do was resolution and frame rates and of course then ther is size weight and cost but thats a different story all together.................
Click to expand...

Not exactly and resolution does play a rather large part in relevance and also looking towards the future/now here.

If Alexa "is" that camera for you though, that's fine. I've been down that alley before.

However, you referenced film "something that has a little bit more" and Dragon is that in so many ways.
 
Phil Holland said:
Not exactly and resolution does play a rather large part in relevance and also looking towards the future/now here.

If Alexa "is" that camera for you though, that's fine. I've been down that alley before.

However, you referenced film "something that has a little bit more" and Dragon is that in so many ways.
Click to expand...

well i feel film still is the gold standard interms of dr and latitude and the closest to that in digital cameras is alexa. though the difference is little alexa is a little behind film in terms of dr and overexposure latitude and dragon as i have seen in many tests is still a little behind alexa in terms of dr and latitude ,verry little though, but still dragon clips a little before alexa and is comparatively noisier in the shadows and sees a little less than alexa, the alexa has got higher luma values...i think i would rate film at 14.5 stops and alexa at 14 stops and dragon at 13.5 to 13.75 stops. i would love to see a digital cinema camera that can have 15 or maybe the "precious" and highly advertised 16 stops that dragon was supposed to have but didn't.......
 
Rohan B said:
well i feel film still is the gold standard interms of dr and latitude and the closest to that in digital cameras is alexa. though the difference is little alexa is a little behind film in terms of dr and overexposure latitude and dragon as i have seen in many tests is still a little behind alexa in terms of dr and latitude ,verry little though, but still dragon clips a little before alexa and is comparatively noisier in the shadows and sees a little less than alexa, the alexa has got higher luma values...i think i would rate film at 14.5 stops and alexa at 14 stops and dragon at 13.5 to 13.75 stops. i would love to see a digital cinema camera that can have 15 or maybe the "precious" and highly advertised 16 stops that dragon was supposed to have but didn't.......
Click to expand...

Kodak will tell you film is 13 stops and if shot "perfectly" you can squeeze 15 out of it, and I mean absolutely perfectly. I don't have to think about it, it was a large part of my job for a long while. I've measured 19 visible stops on Dragon via REDlogFilm, though you can argue what's usable. Shooting/rating around ISO 800-1280 will give you what you are looking for.

These are measured full stop patches:



Most of what people are experiencing with Alexa is a variation in LOG, general exposure across response, and "that other thing that I dare not mention" that involves a small piece of glass.

It's cool to disagree on this one. I've been whacking the poor dead horse for far too long.

if you are curious I can add something into the mix that makes Dragon see about 20 stops of measurable DR and it's an optical component, not HDRx. It will give you the luma you are looking for and the highlight "effect" that you find pleasing about Alexa.

Sadly I'm at a point where this has more or less become "meh". I'm 100% doing things with Dragon that I could never do before with film. I'm very sure of that.
 
Rohan B said:
.......By the way i love my dragon and Alexa and the f65 too, and i am not saying that i am not happy with theese greeeaaaat digital cinema cameras and their sensors, all i am saying is that there still isn't a digital sensor that i can truly call superior than film, they are all verry close, verry verry verry close, but still a little behind,and i would soon love to see that "little" be eliminated...........:thumbsup:
Click to expand...

IMHO, close and behind are relative paths , equivalency is there. One mustn't forget that its a new era, a new media for the same purpose that opens roads whereas we had come to an end (at least relative to whats is feasible on todays limits and practical barriers).
 
Phil Holland said:
Kodak will tell you film is 13 stops and if shot "perfectly" you can squeeze 15 out of it, and I mean absolutely perfectly. I don't have to think about it, it was a large part of my job for a long while. I've measured 19 visible stops on Dragon via REDlogFilm, though you can argue what's usable. Shooting/rating around ISO 800-1280 will give you what you are looking for.

These are measured full stop patches:



Most of what people are experiencing with Alexa is a variation in LOG, general exposure across response, and "that other thing that I dare not mention" that involves a small piece of glass.

It's cool to disagree on this one. I've been whacking the poor dead horse for far too long.

if you are curious I can add something into the mix that makes Dragon see about 20 stops of measurable DR and it's an optical component, not HDRx. It will give you the luma you are looking for and the highlight "effect" that you find pleasing about Alexa.

Sadly I'm at a point where this has more or less become "meh". I'm 100% doing things with Dragon that I could never do before with film. I'm very sure of that.
Click to expand...

Personnally, I am pretty curious. It would be nice to see some highlights from dragon with a low-contrast filter ;-).
 
I would be very interested to understand what are the real bottlenecks are producing a high speed 4k image to record.

With a few Altera 14nm FGPA 'simulated' ARM cores - and the rest of your 'code' also flashed on the FGPA, no encryption, no immediate conversion to .R3D. ... leave all processing only when reading the data ... I guess RED would be read out rate limited?

But if this is the only limitation ... why stop at 1000fps (which I guess is about 2 x current sensor limit).
At what point people would people not feel the need to get a Phantom .... 5000 fps?

And another random thought ... has anyone experimented with using hexagonal pixels with 3 x parallelograms (R+G+B) rather than squares (Bayer with 2 G + R + B). Wouldn't this give more saturated colours & less pixels to process & a larger proportion of sensor vs gap?

AJ

Marcos Montenegro said:
Next? a 1,000fps slow motion 6K camera
Click to expand...
 
Jeffrey Loewe said:
I could get behind the High speed heh heh heh... Especially if it came in the form of an upgrade to the current body :)

A dream come true that would be!
Click to expand...

+1 would love to see higher speeds at 4k / 5k and maybe even 6k without any new hardware.
 
1) Lightfield
2) Broader ISO range
3) Color
4) Size

Those are the 4 big metrics I expect RED to tackle next.

Lightfield for one-camera stereo, VFX tracking assist and roto-assist.
Broader ISO range to tackle lowlight A7S style work all the way to daylight without ND.
Color because it can always look better. Filmstock emulation and other additions would be nice.
Size, because it can always be smaller. Also presumably better power management so that a 30wh redvolt could power the camera for over an hour.
 
Nick Pujic said:
+1 would love to see higher speeds at 4k / 5k and maybe even 6k without any new hardware.
Click to expand...
I imagine there would have to be a decent hardware upgrade. I'm just saying more than anything I hope RED keeps the DSMC system that it has developed. I would guess that you'd have to have quite a few internals that would have to change to achieve Phantom like frame rates, even at half of what the Phantom can do, i'd guess you'd have quite a bit of internals that would have to be changed out.
 
I guess there are disadvantages to lowering camera inertia to below that of the Dragon ... and so the 'size' is unlikely to get much smaller until this benefits professional workflow.

I guess until now the bandwidth limitation meant that using an off the shelf tethered connector would have introduced a bottleneck for high speed export in real-time:

Dragon Single image = 6144 x 3160 x 16bit = 37MB a frame of RAW data
TB2 (20Gb/s = 2.5GB/s) = 67.5 frames a second.
Even TB3 (40Gb/s = 5GB/s) = 135 frames a second ... not exactly high speed.

And writing high speed in real time to a physical media (in a small package) isn't ready for prime time yet.
Lets say 10 secs of 5000fps = Storing 50,000 Frames a 37MB / frame = 1.8TB storage ... which is probably too much RAM (too much energy to keep in RAM running, not to mention heat, space & expense). Also TB size RAIDED SSD would still be a bottleneck on writing speed.

I am guessing that a new sweet spot will emerge for tethered devices that uses 1.6Tbit/s bandwidths pipes.


MXC (intel) = 1.6 Tbit /sec = 200GB/s = at most 5405 frames a second (probably 5k after making allowances for comms protocol)
ARM Corelink = 1.6 Tbit a second (allowing easy hook up of FPGAs off the bus)

And as for the sensor design. We need high speed - with LOW noise .... I guess we'll could ultimately end up with sensor sensors with tuned optical 'antenna' whose length is tuned to the frequencies of light that we wish to collect (so no OLPF). I believe that refraction (prism) could be used to spread the colours at each pixel into - IR,R,G,B, UV - focusing the light onto an antenna (or antenii) for each frequency band.

All the energy for that frequency of light can be read (with very little noise, and no crosstalk) .. and so No Bayer coloured filters.

I make this 1db (OLPF) + 6db (increase in sensitivity in Red and Blue channels from not having a Bayer pattern.) = 5 x better S/N.
Using a 65mm sensor over the Dragon would given 2.5 times the surface area = 2.5 better S/N

This would give a combination of ~12.5 times increase in Signal to Noise.

If the current High speed frame rate is 400 fps -> this could translate into a sensor that is > 4000 fps.

There are some other materials that better CMOS in their electron / hole mobility out there (InGA). I think this will be more pragmatic than waiting for the the process of fabricating two-dimensional materials to mature.

AJ


===
http://www.intel.com/content/www/us/en/research/intel-labs-silicon-photonics-mxc-connector.html
http://www.arm.com/products/system-ip/interconnect/corelink-ccn-508.php
 
Let's be realistic for a moment... I want a cheap knob that works with Offhollywoods Hotlink.
 
You must log in or register to reply here.
Back
Top