Welcome to our community

Be a part of something great, join today!

Register 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 :)

Raptor X

R

Remington Chase

Well-known member
Joined
Feb 11, 2016
Messages
1,077
Reaction score
16
Points
38
Now that it's out, Looking forward to getting one tomorrow. Also will the upgrades start immediately and can we upgrade the S 35 raptors?

It looks like they cost $5k more than the current non X but the upgrade price is $12k to bring current Raptors up to X.
 
Last edited:
How does the extended highlights work ?
Is it based on a dual AD sensor? Or On a merge of an X track (with a different capture time?)
Who is Global vision ? ;-)

Patrick
 
Patrick Tresch said:
How does the extended highlights work ?
Is it based on a dual AD sensor? Or On a merge of an X track (with a different capture time?)
Who is Global vision ? ;-)

Patrick
Click to expand...
Yes, I'm wondering if it is some kind of temporal trickery like on the old Epic or something completely new?
 
Patrick Tresch said:
How does the extended highlights work ?
Is it based on a dual AD sensor? Or On a merge of an X track (with a different capture time?)
Who is Global vision ? ;-)
Click to expand...
As I understand its done in a similar way as Hdrx, seams logic as the camera als ocan do doal captures.... Something that the original raptor also should be very much capapble of from looking at its bit and frame rates. So yes, seams very much like a software price differentiation thing not to put it in og raptor firmware.
 
Yeah, from what I can tell Global Vision looks like HDRX, but without the visual artifacts incurred by using a different shutter speed for the x-track (I presume global vision uses the global shutter to keep the shutter angle the same on the "x-track")... To be fair, It's easier to record two streams than having to process/combine/gamma map them in camera (Ala Canon's DGO or Arri), so its a bit of a hack under the "but you get to control it!" guise that's been RED's MO since the beginning.

I'm more interested in Phantom Track (and it's compatibility/use/function) to get a higher quality "raw" source capture as well as a daily vfx capture (which I presume isn't raw). If it's easy to do, I can see that being value-add for virtual productions.
 
Mike P. said:
Yeah, from what I can tell Global Vision looks like HDRX, but without the visual artifacts incurred by using a different shutter speed for the x-track (I presume global vision uses the global shutter to keep the shutter angle the same on the "x-track")... To be fair, It's easier to record two streams than having to process/combine/gamma map them in camera (Ala Canon's DGO or Arri), so its a bit of a hack under the "but you get to control it!" guise that's been RED's MO since the beginning.

I'm more interested in Phantom Track (and it's compatibility/use/function) to get a higher quality "raw" source capture as well as a daily vfx capture (which I presume isn't raw). If it's easy to do, I can see that being value-add for virtual productions.
Click to expand...
If its HDRX then Its not the same as what Arri does. Arri do two different digital gains to the signal that hits their sensor and then comp them together in camera taking the lowlights from one and highlights from the other and by doing so they get a wider DR than if they just read single gain signal.

HDRx is something else. It does part readout of each frame. For example if your exposure time is 1/48 of a second. then HDRx can dump to memory the first half of that i.e 1/96 of a sec exposure. But while dumping that it continues to fill the photo sites for the remaining other half of the frame. That gives you two frames one that has the first half of the frame captured and one that had the whole duration of the frame captured. So the first one has half the exposure and half the motion blur second track has full motion blur and exposure. Then you can mix those two in post... But when you do, naturally if there is action / movement in the shot the motion blur of the two captures will not line up. As the second track has twice as much motion blur as the darker first capture... red cine x had something I might be wrong but that was called "magic motion" to combine the two tracks, it had some little composting trix going to try to do its best of hiding the difference in motion blur of the two tracks but it was not very advanced. Red could take that further, in their new global capture program, even use AI etc to create a much better blend between the two tracks, using motion vectors and introduce post motion blur to the first shorter read out. But however you slice it... its a hack solution, it comes with artefacts and errors that gets more visible the more motion there in the shot and to do it well it gets quite GPU taxing. So sure you can get out more than twice the DR with the HDRx way but it has drawbacks and is as I see it not fair to compare with what for example arri is doing. Both have its benefits the dual digital gain way is more fail safe, does not come with artefacts, but sucks more power and is more costly to put together than HDRx, but yes HDRx can be pushed much further.
 
Using HDRX technique or dual AD either way you will have to combine or map two streams. Both will have some quind of artefact.

But from the sample I see it's really good. Global shutter in VV with those FPS speed and DR is by the numbers impressive.

RED did a killer camera.

Patrick
 
Patrick Tresch said:
Using HDRX technique or dual AD either way you will have to combine or map two streams. Both will have some quind of artefact.

But from the sample I see it's really good. Global shutter in VV with those FPS speed and DR is by the numbers impressive.

RED did a killer camera.

Patrick
Click to expand...
What Im saying its very different the dual AD way you combine one bright and one dark image with each other. Thats easily done there is perfect composting math for doing so. When comping HDRx tracks together you try to blend between two different captures / two different looking images, its hack solution prone to make producer, directors, post people etc reason to say -"oops what's this."
 
I believe HDRx doesn't dump halfway through the frame, if it did there would be a moment where the photosites are resetting and you'd have a blip in the motion blur. Instead it's just sneaking a second frame in while the shutter is 'closed'. Basically shooting 48fps with alternating shutter angles, and having a behind the scenes way of compositing them together to appear as 24fps. Iphone does this to take photos as well.

Same issue when red said you could do frame averaging to get more dynamic range, well it only really works for still images. Otherwise there are blips in the motion blur when the shutter is 'closed' or the photosites are being read.

My understanding is that arri reads the photosites with two different ADC paths and then adds them. Downside is the noise of one path will show up in the midtones of another path, and so fixed pattern noise, cmos smear etc has to be eliminated before the adding, and the overall image is unavoidably noisier, although this has a nice effect of making the noise more even through the image. Blackmagic's early sensors also did dual gain but didn't manage these issues and it gives them strong unpredictable fixed pattern noise.

Canon instead basically has 2 4k sensors on one chip, one nestled in the corners of the other, each with a different photosite size and so naturally different sensitivity (or is it just different sensitivies in the electronics? i'm not clear here). In other words, instead of reading a photosite twice, they have twice as many photosites. And the problem with this is the same as arri's style as well as that readout time is much longer, giving it worse rolling shutter and worse high ISO performance from the smaller overall photosite size.
 
Last edited:
Chris Webb said:
Answered in the video below around the 8 minute mark.
Click to expand...
Supper bummer.... That would be like taking a Red Tesla in for a update or upgrade and they return it with a black hood saying sorry thats all we got..Hopefully they will see the light and at least offer a Storm Trooper version for maybe a cost to have the matching "hood" Can figure out why they didn't think there would be at the least 10 20 50 that would want to keep the Strom Trooper look throughout the camera..
I wonder if Jared's ST was given a black front when he had it changed out..
Im sure they will have a new camera in the future and offer only a few in different colors like the Komodo's
Ive talked to my rep's before about how if RED offered to let you personalize your camera color if only a few colors for a premium, that they would probably move a few more. Or at least a 3rd party to powder coat their cameras....Seems modular only applies to hardware not color for your particular production needs.
 
Last edited:
You must log in or register to reply here.
Back
Top