HDRx motion interpolation

[Jannard]

But essentially we're frame interleaving here, right? There's no surplus bandwidth to write the frames at "the same time". And the X track exposure correction is a function of time between read-reset cycles. This would account for the unique motion blur that be achieved. Also the efficiency of making this part of a single R3D stream so the media can handle it. But in theory, the X track could be the "first" frame in the sequence giving the motion streak of the A track as a trailing, comet-tail effect, right?

The interesting thing about HDRx is the the two tracks are conjoined. There is no time gap between the two frames. The sharp/blur are always connected.

Jim

[Shane Betts]

There's no surplus bandwidth to write the frames at "the same time". And the X track exposure correction is a function of time between read-reset cycles.

I don't think that's right Jeff. I believe that there is the bandwidth to write both frames at once. That's why there are no motion artifacts and it's also why HDRx halves your available frame rates and it's why the Red One will never do HDRx - because the R1 doesn't have the surplus bandwidth but the Epic does.

The comet-tail effect is a result of two exposure lengths, one short and sharp, one longer and more blurred.

[Jeff Coatney]

I don't think that's right Jeff. I believe that there is the bandwidth to write both frames at once. That's why there are no motion artifacts and it's also why HDRx™ halves your available frame rates and it's why the Red One will never do HDRx™ - because the R1 doesn't have the surplus bandwidth but the Epic does.

The comet-tail effect is a result of two exposure lengths, one short and sharp, one longer and more blurred.

Your theory being that the two exposures are happening simultaneously? As in: within the same read/reset cycle? The X track committing to the media at the beginning of this cycle and the A track following milliseconds later? Explain, please.

[Jeff Coatney]

The interesting thing about HDRx™ is the the two tracks are conjoined. There is no time gap between the two frames. The sharp/blur are always connected.

Jim

Yes, this effect I noticed in some images in the barn shot. I will be shooting at-sea, in the waves, where the bright foam will be backlit against the dark trough of the wave. I'm interested in HDRx's ability to handle the motion of breaking surf. I will be happy with whatever I get in such a situation, just trying to predict what that will do from a motion perspective.

[Jannard]

Yes, this effect I noticed in some images in the barn shot. I will be shooting at-sea, in the waves, where the bright foam will be backlit against the dark trough of the wave. I'm interested in HDRx™™'s ability to handle the motion of breaking surf. I will be happy with whatever I get in such a situation, just trying to predict what that will do from a motion perspective.

Please read my 1st post carefully...

Every motion camera ever made has "motion artifacts". Nothing reproduces exactly like we see. Some "artifacts" look better than others. Some we have gotten used to. "Magic Motion" and "MNMB" just create different types of motion artifacts... or motion "looks". There are limits to motion with every camera. HDRx™ and these two flavors of motion have their own different limits. Learn what they do and use them when they give you what you want. Turn them off when you don't get what you want.

Jim

[Jeff Coatney]

Please read my 1st post carefully...

Every motion camera ever made has "motion artifacts". Nothing reproduces exactly like we see. Some "artifacts" look better than others. Some we have gotten used to. "Magic Motion" and "MNMB" just create different types of motion artifacts... or motion "looks". There are limits to motion with every camera. HDRx™™ and these two flavors of motion have their own different limits. Learn what they do and use them when they give you what you want. Turn them off when you don't get what you want.

Jim

Trust me, I read all Your posts carefully :).

[Blair S. Paulsen]

Assuming I am interpreting Jim's words correctly the "quick" X frame is read out during the time the longer base A frame is being read. Perhaps they start reading at the same time but the X "quick" frame finishes it's read sooner. Perhaps the X "quick" frame starts reading later then ends at the same time as the A frame. Perhaps it happens in the middle of the exposure time of the A frame.

Based on some characteristics of the Vegas sample footy the best current guess is that the X frame starts reading later in time than when the A frame starts reading but they both conclude their reads at the same time. In any case, it will be fun to watch this tech mature. While certain instances may "break" the illusion, I expect Graeme and the Foundry and others in the future to come up with very advanced ways to recombine the two frames to cover virtually any situation.

The more interesting aspect of this to me is something Jim has mentioned several times, what is the real world human visual experience of seeing fast moving things (or spinning your head quickly). When I wave my hand quickly what do I see? You can test this yourself without an EPIC assuming you have reasonably good normal (that's a thread in itself) vision. Obviously there is a whole debate about what we are conditioned to when watching movies or TV vs real world.

Here's to 48fps base frame rate, HDR, 4K measured resolution after decode imagery and the future - Cheers!

- #19

[chrislancaster]

Please read my 1st post carefully...

Every motion camera ever made has "motion artifacts". Nothing reproduces exactly like we see. Some "artifacts" look better than others. Some we have gotten used to. "Magic Motion" and "MNMB" just create different types of motion artifacts... or motion "looks". There are limits to motion with every camera. HDRx™™ and these two flavors of motion have their own different limits. Learn what they do and use them when they give you what you want. Turn them off when you don't get what you want.

Jim

Hey Jim wouldn't it be awesome if the camera could have have a peripheral view instead of everything being tunnel vision or box vision .. i guess screens would have to be somewhat concave.. i love the whole HDRx idea

i called it .. i knew secret #21 was hdr look it up the post is still there muahaha evil genius strikes again

but i digress

[Jeff Coatney]

Based on some characteristics of the Vegas sample footy the best current guess is that the X frame starts reading later in time than when the A frame starts reading but they both conclude their reads at the same time.
- #19

This is what I observed as well. Since shots are usually edited into a larger context, and the head and tails of them are cast aside, perhaps by shifting the X frame upstream or downstream in time a frame, one can affect the appearance of any motion artifacts within shots? Perhaps if this is currently elusive, at some point in the future, this may be possible? Seems simple enough in theory-- although I am very satisfied with the current capabilities of HDRx as she stands now. A huge Bravo! And I can't wait to play with it.

[David Quinn Carder]

Every motion camera ever made has "motion artifacts".

At a sufficient frame rate these artifacts would be virtually non-existent, correct? I know most people seem to be interested in sticking with the legacy 24fps format…