EPIC HDR...

[Tim Hole]

It would be nice. Though, if Scarlet did not do it, that would be a significant feature that would help differentiate Scarlet from Epic. So far, Epic is like 3x more expensive for overcranking being the only obvious big difference.

Also data rates being a big one. But it probably won't be long before one of the VDSLR company brings out a psuedo-HDR mode for the new...whatever

 

[Pawel Achtel]

I need to clarify one thing...

EasyHDR™ happens in camera and there is NO hit to the data rate. Files come out of the camera the same size as normal mode. You are, however, constrained to 5K at 48fps. The HDR is "baked in" to the file so there is no post option for combining frames.

HDRx™ is recorded as two streams, so data rate is twice normal mode. Combining them happens in post and there are a wide variety of options for doing this. Again, you are limited to shooting maximum 5K at 48fps.

Jim

How 'bout +12 stops (for a total of 25 stop DR) with four streams at maximum 5k 24fps? :devil:

 

[Ketch Rossi]

You can still use RED Drives up to a certain REDCODE... not exactly sure what that is right now.

Jim

Going with the SSD no doubt, but with 24fps and 48fps HDR for just a little slow motion effect on Hair, and clothing movement and such, will the added data rate be allowed to be utilized the same way in the Red Rocket for the Mobile Rocket, or will two Red Rockets be needed and only possible with the Tower Configuration?

Sorry if it was answered, didn't find it if it was

 

[Mark Andersen]

It would be nice. Though, if Scarlet did not do it, that would be a significant feature that would help differentiate Scarlet from Epic. So far, Epic is like 3x more expensive for overcranking being the only obvious big difference.

Yeah, but 5K 100 FPS is huge. It's the feature I look forward to the most.

 

[albert rudnicki]

I agree with you Shawn
If they share the same features then it's better & cheeper to get 2 Scarlet in place of one Epic (one for s35 and another 2/3 for slow motion)

 

[MikeCurtis]

We have successfully tested EPIC HDRx™™ to 18 stops of dynamic range in motion. You can use the mode in two ways... Easy HDR (in camera) and HDRx™™ (manipulate in post). It can be used several ways, including the "magic" way... which is very cool. Or there is a "normal" way to deal with motion blur. Since it is very rare to need 18 stops, you can set different HDR settings... like +3 stops, +4 stops, etc.

You are really going to love this. We are working with the Foundry for the HDRx™™ mode. Graeme and the Foundry boys make a powerful combination.

Jim

Very cool! Coming into this late and only read the first 15-20 pages of posts, after reading this and based on double data rate and mention of the Foundry for motion blur issues, I'm guessing:

-you guys are reading out twice without resetting the sensor back to "zero" state

- one shorter exposure reading out first, generating shorter motion blur on the shadows but preserving highlight detail,

-and a second "pass" that is longer but includes the first, with a brighter image, so the motion blur is longer on the highlights

-thus would need to integrate the two using the motion vector temporal math that the Foundry offers, to make the motion blur consistent

-the SDK method glues the two exposures together blending using HDR methods to get the highlight (edit, thinkin' backwards) detail from the first shorter readout but has potential motion blur inconsistencies between highlights and shadows

So expose from zero to a fraction of the 1/48th or whatever you're shooting, read it out while still exposing (preferably), then read it out again for the full 1/48th exposure

-therefore you're biasing the exposure a little on the hot/overexposed side, rather than balancing entirely, since highlight (edit, was thinkin' backwards) detail is coming from that first brief exposure period

-otherwise you have temporal mismatch - gotta expose with as much overlap as possible rather than sequentially.

To get the +3, then expose up to 3 stops overexposed?

How am I doing so far?

-mike

 

[Rob Ruffo]

While there will always be "more"... EPIC is a game changer.

You will likely not ever need more. You will get it because we will make it. But EPIC is the ONE that changed everything. I have seen a lot. I own over 1000 cameras. EPIC is the one that will change history. It is the one that covers everything. Size, price,performance. If I died and went (somewhere)... I will be content. EPIC will get to market.

Jim

You know I do see that, although I'd say we were already almost there with our Red MX - we are still getting used to thinking in terms of the latitude it gives us, and cutting down on light sizes when we rent - the other day I caught myself trying to figure out how we would cheat candlelight and realizing that would not at all be necessary. It is already more camera than we ever thought would exist, let alone that we would own.

I see that with our 5D (in terms of STILLS). I really don't any longer feel the need for anything else in terms of full-frame 35 stills cam. I am happy. I don't read new stills cam announcements anymore. I don't care, I have what I need for that, I'm done.

5k 18 stops in raw in a light package!? I'll buy an Epic and just keep it forever.

 

[MichaelHalsell]

Any comments on the 2K performance of the Epic? Same, better than MX?

 

[Larry McKee]

Mike, a shorter exposure would give you more detail in the highlights, not the shadows. As far as how RED is doing their magic, that, I can't answer.

 

[MikeCurtis]

Larry - thanks, duh, fixed

 

[Lee Fingersh]

OK, Mr. Troll, I'll bite. Why?

Very high-speed subjects - focus tracking and viewfinder lag would make keeping them in focus (tracking) and keeping them in the viewfinder (lag) extremely difficult. The subjects were covering the entire DOF approximately 10-15 times per second (close to 30, in one case) and the angular slew rate was approaching 160 degrees per second. Very, very hard to follow even with a real-time viewfinder just because I couldn't turn my body fast enough at times. With a viewfinder lagging a few tens of milliseconds, it would be far harder.

I suspect Jim doesn't shoot a lot of truly high-speed subjects and I'm sure EPIC will make a very solid video (motion/cinema, whatever) camera, plus it has engineering applications too. But I doubt that it'll do everything because it lacks a TTL OVF and PD AF.

 

[Steve Wake]

Is there possibility that at least the EasyHDR will trickle down to the Scarlet line-up?

Unclear to me which is more likely to trickle down to Scarlet: HDRx or EasyHDR, or possibly neither, or possibly either or both with restrictions. Scarlet Fixed can write 120fps of data, so it seems it could easily write a doubled rate of 24 fps or so, for HDRx. On the other hand, it seems EasyHDR could require a lot more realtime in-camera processing that would more likely be beyond Scarlet Fixed capabilities (or buffer capacity-whatever that may be). Hard to say what, if anything will trickle down since we won't know exactly how it works until the trade secret is no longer a secret. Personally, I'm guessing this is a very processing-intensive process, with a lot of important info carried in metadata, and for EasyHDR and the motion blur Magic, a lot of pixels at the high and low end are more-or-less ignored (with information used to modify some values in their HDR-paired frame). So HDRx would be more likely to trickle down to Scarlet, with the actual processing done in post and "Magic" a possibility.

In any case, 3 (or even 6!) extra stops limited to 24fps and requiring an extra step in post would be a nice option if that's all Scarlet Fixed can handle. And if I'm totally on the wrong track and it turns out the HDR has to be baked in and it cannot include "Magic," the extra stops would still be nice at a normal frame rate.

 

[Michael Hastings]

Very high-speed subjects - focus tracking and viewfinder lag would make keeping them in focus (tracking) and keeping them in the viewfinder (lag) extremely difficult. The subjects were covering the entire DOF approximately 10-15 times per second (close to 30, in one case) and the angular slew rate was approaching 160 degrees per second. Very, very hard to follow even with a real-time viewfinder just because I couldn't turn my body fast enough at times. With a viewfinder lagging a few tens of milliseconds, it would be far harder.

I suspect Jim doesn't shoot a lot of truly high-speed subjects and I'm sure EPIC will make a very solid video (motion/cinema, whatever) camera, plus it has engineering applications too. But I doubt that it'll do everything because it lacks a TTL OVF and PD AF.

It's good information/critique and needn't be taken as trolling. As an early adopter I'm champing at the bit for epic but don't much care about the stills function - particularly when $800-$2000 buys an extremely good dslr with ttl ovf and pd af and will fit in the tamrac expedition bag just fine with the epic -AND is a good backup/2nd camera. And when I want to be a soccer grandpa (hoochie gpapa? ) I won't be quite as nervous as I would be with $30k of epic.

 

[Gabe Conroy]

I suspect Jim doesn't shoot a lot of truly high-speed subjects and I'm sure EPIC will make a very solid video (motion/cinema, whatever) camera, plus it has engineering applications too. But I doubt that it'll do everything because it lacks a TTL OVF and PD AF.

I thought Jim announced an OVF module a while back?

 

[Brian Petrie]

Very high-speed subjects - focus tracking and viewfinder lag would make keeping them in focus (tracking) and keeping them in the viewfinder (lag) extremely difficult. The subjects were covering the entire DOF approximately 10-15 times per second (close to 30, in one case) and the angular slew rate was approaching 160 degrees per second. Very, very hard to follow even with a real-time viewfinder just because I couldn't turn my body fast enough at times. With a viewfinder lagging a few tens of milliseconds, it would be far harder.

I suspect Jim doesn't shoot a lot of truly high-speed subjects and I'm sure EPIC will make a very solid video (motion/cinema, whatever) camera, plus it has engineering applications too. But I doubt that it'll do everything because it lacks a TTL OVF and PD AF.

How well it will compare in still mode overall to a new DSLR I am not quite sure you may be right. However, for highspeed subjects like you are talking about couldn't the Epic be very useful by shooting motion then just extracting stills(which in Epic I believe is gong to be simplified). It is like shooting 120fps raw burst mode haha. Obviously, other things are taken in to affect but you get what I am talking about just my thoughts anyways.

 

[Leo Sardello]

Very high-speed subjects - focus tracking and viewfinder lag would make keeping them in focus (tracking) and keeping them in the viewfinder (lag) extremely difficult. The subjects were covering the entire DOF approximately 10-15 times per second (close to 30, in one case) and the angular slew rate was approaching 160 degrees per second. Very, very hard to follow even with a real-time viewfinder just because I couldn't turn my body fast enough at times. With a viewfinder lagging a few tens of milliseconds, it would be far harder.

I suspect Jim doesn't shoot a lot of truly high-speed subjects and I'm sure EPIC will make a very solid video (motion/cinema, whatever) camera, plus it has engineering applications too. But I doubt that it'll do everything because it lacks a TTL OVF and PD AF.

Couldn’t you adjust the settings on the EPIC to aid in extraction of a single image of a high speed object later? I thought that was part of the point of the motion stills thing, you can shoot at a high frame rate and extract a single high quality still later.

*EDIT* Brian you beat me to it

 

[David Quinn Carder]

Scarlet doing HDR at 60fps would be awesome.

 

[Sergio Perez]

I'm sure Red will include Easy HDR on the Scarlets. Its too much of a "killer feature" to be missing. Its the key to completely put it above the competition in its price range. Its a "viewable immediate WOW feature". Much more important than resolution for the target market in the price range.

 

[Shane Betts]

Scarlet doing HDR at 60fps would be awesome.

Scarlet doing HDR at 24fps would be awesome :emote_happyhappy: Anything else has got to be cream.

 

[Michael Hastings]

Couldn’t you adjust the settings on the EPIC to aid in extraction of a single image of a high speed object later? I thought that was part of the point of the motion stills thing, you can shoot at a high frame rate and extract a single high quality still later.
[/SIZE][/FONT]

You are missing the point. In a video camera or dslr in video "live view" mode the autofocus is inherently less accurate particularly on moving subjects where the mirror allows the use of phase detection and better prediction of moving subjects and adjustment/correction during the mirror down/closed shutter period. So besides a better technical means of autofocus using the phase detection that is possible with the mirror system, it also means that in stills mode it can focus and if it overshoots or undershoots there is time to correct it between frames. It is also easier to track these subjects through the OVF of an slr.

Contrast detection tends to " hunt " so if you just shoot 120fps in either autofocus or manual focus you are likely to be slightly out of focus on most shots (so unusable) whereas modern high end slrs tend to get most of the shots in focus even with fast moving subjects. Not to mention the problem of motion blur with video frames.

So shooting football with 8 fps of well focused frames at 1/2000 shutter speed 18mp from a $1500 dslr might make more sense than 120 not so well focused 13mp frames at 1/240 shutter on a $30,000 epic.