Epic, dragon, efp/eng.. please adress the need for a more sensitive 5k sensor !

[Steve Gibby]

Could be that the Scarlet Dragon 5k sensor could be re-calibrated for a dedicated mobile EFP version of the camera. The thing is that in many mobile EFP genres we need high frame rate capability, so beyond 23.98, and 29.97, HFR capability of 59.94, 50p would need to be supported, and hopefully even 120 fps capability would be included. I don't know if the planned Scarlet 5k sensor based on Dragon 6k could do that...hopefully it could.

If they're engineered correctly internal ND don't have to take up a lot of space. What would be needed would be clear, 1/4ND, 1/16ND, and 1/64ND. Perhaps some of the ND technology Red has already developed for the motion mount could be incorporated, or at the least a Motion Mount included in the Brain package sale of each mobile EFP version camera, thus solving rolling shutter and ND that way.

An available service module with dual full sized XLR, SDI, and HDMI out would also be a must.

It should be a straightforward engineering feat to enable deletion of clips within camera.

Interesting times ahead!

 

[Denis Buhot]

Wondering too about not being able to delete clips in-camera. Is that something people out in the field could live with or would it be a deal-breaker regardless of other changes?

No, I would not say it is a deal breaker. As Steve pointed it, or so I read him, high frame rates + low-light capabiity is what counts most. ND flilters + deletable files are just very useful bonuses. Given the absolute need for higher sensitivity + framerates up to 120 fps, I guess one should not be asking for to much at at time...

 

[Blair S. Paulsen]

How about...

How about...

Dragon sensor with in camera scaling from 5K or 6K to UHD rez with adjustable in camera NR for "ready to air" applications.

"Back Half" module/host/AIO that includes many of the features noted in this and similar threads, perhaps even including a UHD RGB recording option like ProRes.

I still think making an ergonomic shell as a host for the brain that offers needed I/O in a ENG/EFP friendly form factor would ease integration with other parts of the production ecosystem in a wide range of genres. IMO, its worth the extra cost and weight vs building out custom module stacks that will only further the "science project" objection. From the perspective of a producer who may not grasp all the specifics of various modules they can just rent the AIO (all in one) version and know it will tick all the usual functionality boxes. Integrated adjustable shoulder pad, battery plate back and low, "click in" dumb side mount for touch LCD, integrated EVF mounting rail/bracket, etc for "grab and go" shooting scenarios. BTW, it should really be able to power the Cabrios right from the PL mount without requiring a separate power run.

I'm suggesting that a 7 lb (est) AIO rig, plus 5 lb brain (22 lbs total with Cabrio & Red Brick) that is well balanced would be great for 90% of shooting situations with the ability to quickly click the brain out of the AIO host when you need to go small/light for jib, drone, more still-style hand held, 3D, etc. On bigger features the camera department can still cobble together speciality rigs when required, but IMO; on a day in, day out basis, the "standard" AIO config and the ability to go brain only when desired is sufficient flexibility. The upside is the workforce can quickly orient and operate the rig with no more knowledge/training than most professional cameras require.

Furthermore, the multi module approach requires each module to have its own casing, contacts, internal power distro, etc - all of which adds weight, cost and potential failure points. By putting everything in one shell you only have the one set of contacts to the back of the brain, internal power distro serves all boards, case weight per function is lower with consistent weight distribution properties vs re-rigging module stacks.

Cheers - #19

 

[Les Hillis]

Just some food for thought, but if you look at the old AATON/EclairACL form factor, we're still dealing with basically the same thing now when it comes to the main body/head/brain of the camera - https://www.google.com.au/search?q=...tatela.edu%2Ftvf%2Fequip%2Findex.html;500;319

What's missing obviously is the massive amount of real-estate taken up by the film mags. Without going overboard, surely that's the place for all the modules or for the all-in-one add-on module to go if you want to take that kind of ergonomically proven shoulder-mounted approach. There's quite a bit of space there to be able to place things, and even get them tucked away and covered up if you want. If RED were to take a multi-modular approach to it there could even be spacers or blank modules somewhere in there to help spread and balance it all out.

But...does that mean all newly developed modules would have to fit that cat-on-the-shoulder form-factor (or whatever overall form-factor RED come up with)? Or would it mean you'd have modules serving the same purpose but taking different forms depending on whether they're intended for mobile ENG/EFP or studio use?

You could go for an all-in-one design intended for use with a specially tuned ENG/EFP sensor (if such a thing is even possible), but would you still be able to attach a regular Dragon (or even Scarlet/Epic) brain to it if you wanted to, and would the all-in-one module have some features that non ENG/EFP users would want in separate modules?

So many questions when you start to look at it in more detail. Too many for me to get my head around, especially when it's not clear if some of this stuff is even technically possible, or how far RED would be willing to go to make it happen even if it is.

There are no doubt things RED could do to make their camera's more suited to certain purposes. But it seems to me like the sort of high-stakes stuff that requires a lot of consultation and planning and designing and re-designing and refining to get right, and which could turn out to be a huge waste of time and resources otherwise.

Definitely the kind of thing I need to keep my nose out of and leave to the experts (and those with more at stake than me) to sort out.

 

[TheBied]

In camera there are options for improving performance in low light situations.

The frame-averaging feature will substantially reduce noise at the expense of overall frame-rate. Choosing to average 4 frames yields over a stop gained at the low end and allows for usable frame-rates starting at 5K 2:1.

That is true.

And I suspect there are other ways than bigger photosites to calibrate a camera for lower light. Problem is that a sensor has a fixed DR, and thus you would have to give up highlight stops to get lowlight stops.
In the "cinema" world I know, that would be a hard sell...

But I can see the use for it in other contexts.

Making "the same sensor with bigger pixles", is as far as I can understand things I know little about, much more complex than any other suggestions, as it is essentially a new sensor-program. Much harder (I would think) than modifying the body.

I would guess a re-calibrated "same" sensor would be more doable.

But gee... I am speculating now.

 

[Gunleik Groven]

In camera there are options for improving performance in low light situations.

The frame-averaging feature will substantially reduce noise at the expense of overall frame-rate. Choosing to average 4 frames yields over a stop gained at the low end and allows for usable frame-rates starting at 5K 2:1.

Thanks for chiming in!
THAT sounds really interesting.

 

[Ryan Purcell]

More info on that please.

 

[Lauri Kettunen]

I'm in the same boat with Denis and Steve and would welcome more sensitivity. But, of course, highly appreciate what the set-of-the-art Dragon is! It's simply an amazing package, especially considering its size and weight especially with the CF body.

Need to test averaging the frame rates. Thanks for the tip!

 

[Steve Gibby]

The gist of the posts on this thread is that most mobile EFP genres (action sports, adventure travel, wildlife, etc.) require a combination of good low light sensitivity plus high frame rate capability. Many documentaries don't require HFR, but they definitely require good low light sensitivity. Mobile EFP work is almost always in natural light, and must be shot quickly. There's no standing around as a crew debating "Well maybe we should light to 5.6, put a Joker over there, etc." Mobile EFP is a whole different world from narrative cine, and EFP studio work, in which lighting is closely setup and controlled.

Currently when shooting those mobile EFP genres with Dragon and MX sensors the HFR is definitely there, and good low light sensitivity at normal frame rates is there - but a combination of good low light sensitivity at higher frame rates is the challenge. Obviously with a sensor its a balancing act between those two parameters - to a certain extent when you gain something you need to give something up. But good low light sensitivity at 60fps is critical in mobile EFP genres, and moderately good low light sensitivity is still needed at 120fps.

Red has done a remarkable job engineering the Dragon sensor and the MX sensor for that matter. That said, both those sensors were designed and optimized for cine style narrative production. For mobile EFP applications, in view of what I summarized above, a potential tweaking of the Dragon sensor for 5k resolution (or 4k) and increased low light sensitivity, without sacrificing high frame rates would do the trick - in tandem with a module or shell enabling UHD RGB, containing full sized XLR, Dual HD-SDI, and HDMI 2.0 outputs. Internal ND is also critical for mobile EFP genres. That could be achieved by an internal modification, or externally via a package deal offering the Motion Mount on a mobile EFP model Epic.

It was obvious at CES, and will be even more obvious at NAB, that most EFP genre production entities are now seeking (or should be) to edit and master in UHD, then if needed simply resample down for a 1080 delivery. That way as UHD TV proliferates then UHD mastered programs are available.

Everybody in the motion media industry isn't seeking huge acquisition resolutions like 6k and upward for their intended delivery mediums. Cine and IMAX are definitely into that, but EFP really isn't, and doesn't need to be. In most EFP productions a good UHD master is the target - and 6k or higher resolutions at acquisition are overkill for that. Low light sensitivity and HFR capability are the mobile EFP "sweet zone".

The mobile EFP industry is the single largest portion of the motion media industry in terms of yearly number of productions, worker numbers, and amount of equipment bought. Arri services both cine and mobile EFP, Sony services both cine and mobile EFP. I'm thinking its in Red's best interest to continue to do likewise.

On this thread, and some others in the Off Topic forum we simply have a collection of experienced working professionals sharing the needs of their particular niches of the motion media industry - and that's exactly how Red User should function, and valuable real world feedback for Red.

 

[Blair S. Paulsen]

I've played around with some of the frame averaging settings but found the artifacts too prominent for the projects/shots I was engaged with at the time.

IAC, I'm not sure its in RED's DNA to put heavy noise reduction right in the early stages of the imaging chain, which, IMO, is the key to low light quick turnaround nirvana.

Does RED have the IP portfolio and personnel to make a nice ENG/EFP camera? Probably. Is that a priority for them? I doubt it.

My concept of a AIO "back half" module might provide enough space for boards capable of processing the RAW image data into "ready to air" RGB signal - probably with a decent amount of default NR. FWIW I still think an optical cable with plugs that utilizes the RedMag contacts to feed an off camera processing device is a more elegant solution for hardlined EFP.

As a reference, when reviewing F55 footage on a 4K display it has a slight "air brushed" look to it that I attribute to fairly aggressive NR. While the 4K maniac in me cries foul, I think Sony has made a savvy choice for their market. What do you think is a realistic timetable for how soon at least 10% of viewers in the US have robust enough signal, good enough displays and appropriate viewing conditions (seating distance, ambient lighting control, etc) to actually perceive the difference between UHD/4K programs with slightly less high frequency detail than others?

Cheers - #19

 

[Lauri Kettunen]

Steve, you put it nicely into words!

Let me add not so seriously, that remember well the time when shot with some well known camera brands which had plenty of tiny buttons and switches. When it got really cold down to -35C and below, we joked that it would be so nice if the camera designers came to see what it is to use those buttons when fingers feel as flexible as fire woods and all the cables resemble thin iron bars. Red cameras have changed many things from those days. ;-)

 

[Steve Gibby]

Very good points in that post Blair, and in your previous post. Good food for thought...

Everything in this industry isn't narrative, cinema, or IMAX. The right resolution, frame rate, camera, lens, and accessories for any particular production depends on its intended audience and delivery medium. In the overall motion media industry there really is no "one size fits all" scenario - just a "horses for courses" scenario. I'm hoping some tweaking and possibly some new fabrication will enable the "Red horse" to run a bit better in the courses those of us in mobile EFP need it to.

 

[Steve Gibby]

I totally agree Lauri. Me and my crews are constantly shooting in every environment imaginable - blazing hot deserts, Arctic chill mountains, steaming hot tropics, and everywhere in between. Recently I had Epic way high in the frozen Alps, then deep in Central American jungles, then way out in arid Southwest deserts. The lightness and modularity of Epic are definitely a plus factor for that...

 

[Hrvoje Simic]

While I possess zero interest in seeing 35 mm capture for ENG purposes, the thought of EFP optimized RED camera based on DSMC concept seems very appealing.
Gibby and Blair echoed much of my thoughts regarding feature requests.

Ergonomics, practicality and essential feature set being the priorities.

 

[Blair S. Paulsen]

KISS

KISS

I hear you Hrvoje, big sensors for ENG can make focusing a bitch.

Would love to try 3K windowed on Dragon at sync frame rates (24, 25, 30) for 1080/2K output. That should be enough pixels to overcome the Bayer pattern penalty, similar DOF to 2/3" vid or S16, and if you aren't over-cranking you should be able to keep the compression ratio low enough to avoid noise issues. It also seems more possible that an internal processing board could turn 3K RAW into 1080/2K RGB on the fly with decent precision vs 5K or 6K scaled and flipped to a UHD/4K/2K/1080 RGB signal.

At the end of the day my goal is to advocate a hybrid option that adds capabilities to the existing Epic brain essentially as is. Any scenario that requires development of a different camera line and/or major mods to the Epic brain feels like a steep uphill push to me. My concept is all about providing a "host" for the Epic brain that makes it dead simple to go from "matchbox" to a more traditional form factor. This same philosophy leads me away from any multi-module solution that requires skills and AKS that might be in the wheelhouse of hard core RedUsers and jock ACs but just generates blow back in the wider production community.

If the landscape offered...

1) Current Epic brain (w/ side SSD)
2) Epic brain in ergonomic AIO host shell
3) Epic brain custom rigged with small modules

... it is my contention that #1 & #2 would cover 98% of use cases with minimal compromise. Moreover, if taking the Epic brain on and off the host AIO shell utilizes the rear bracket/contacts the way they appear to be designed for, you can head out on set able to switch between #1 and #2 configs in the time it took the R1 to boot up.

Cheers - #19

 

[Steve Gibby]

I very much like your AIO host shell concept Blair. Whether that would be cost effective to produce is the question for Red.

But if it has to be modules to satisfy the needed functionality, ergonomics, and the higher sensitivity/HFR of mobile EFP work, then IMO that would be better than the current situation.

A 1080 delivery is fine for a lot of present work, but right around the corner is UHD mastering and deliveries, so future proofing masters and footage libraries is wisdom. Your 3k raw to 1080 RGB in camera idea still wouldn't future proof the show or footage for UHD. To me either 5k raw or UHD RGB acquisition seems to be both the present (1080 downscaling) and future (UHD deliveries) for mobile EFP.

By ENG I'm assuming you aren't talking about what that's an acronym for - Electronic News Gathering. Nobody is doing ENG with Epic or Red One, but lots of us have been doing non-hardlined EFP (Electronic Field Production) with Red cameras. Just jabbin' ya )

I've never had any problems hand racking sight focus while shooting 4k on Red One or 5k on Epic during mobile EFP style work - and I know you haven't either because obviously I've been DP for music productions where you shot for us, and your hand racked sight focus was always spot on. Of course we lit the Journey and Heart productions for 5.6, so DOF wasn't terribly shallow. All ten Red One cameras for Journey, and all seven Red One cameras for Heart were shot mobile EFP style, with each operator hand racking their own focus - and though we were shooting in 4k there were really no buzzed focus problems to speak of. Even though each camera had a small hardline into the IMAG big screen, it was still a non-hardlined EFP style production because of a number of operational factors.

As you know, almost all mobile EFP genres (action sports, adventure travel, most docs, wildlife, reality, etc.) are shot at apertures of 5.6 or smaller, and shallow DOF isn't really sought except for specialty mood montage shots - and those are usually the type of shot you can take time on to rack focus accurately.

In my recent round of shooting with the Cabrios - football, volleyball, sand Jeeping - I again found no problem at all sight racking focus on the fly in 4k. But as always, almost all shot sequences were shot in apertures of 5.6 or smaller, except a few specialty montage shots which were shot at f4.

 

[Blair S. Paulsen]

I very much like your AIO host shell concept Blair. (snip) A 1080 delivery is fine for a lot of present work, but right around the corner is UHD mastering and deliveries, so future proofing masters and footage libraries is wisdom. Your 3k raw to 1080 RGB in camera idea still wouldn't future proof the show or footage for UHD. To me either 5k raw or UHD RGB acquisition seems to be both the present (1080 downscaling) and future (UHD deliveries) for mobile EFP. (snip)
By ENG I'm assuming you aren't talking about what that's an acronym for - Electronic News Gathering. Nobody is doing ENG with Epic or Red One, but lots of us have been doing non-hardlined EFP (Electronic Field Production) with Red cameras. (snip)

That's a good point Gibby, perhaps the better compromise is to record REDCODE RAW 4KHD at low compression to one Mag and 4 to 1 (quarter res, integer scaling for speed) scaled 1080 RGB to another mag simultaneously. This presumably could be accomplished with a side SSD module per current design plus a second mag slot in the AIO shell. That would leave you with a mildly compressed RAW data set for archive and a oversampled RGB HD recordable if desired AND live feed.

BTW I totally agree that EFP is a more accurate term, but sometimes I choose ENG due to its greater frequency of usage in the wider industry. IAC, as you note its really about mobility/ergonomics/practicality/flexibility/etc.

Cheers - #19

 

[Blair S. Paulsen]

UHD/4K beachhead

UHD/4K beachhead

BTW, I expect hard lined EFP sports/event producers at NAB will be far more serious about UHD solutions for multi-cam/truck shows, etc this year. If the AIO shell included a tactical fiber port that could run signal to a device on the truck, like say a RedRocket-X in the right host/config for processing out to the switcher - Epic just got a lot more interesting to the legions of folks building out UHD solutions across a huge swath of the industry.

I'm not saying there's going to be a UHD stampede in 2014, or even 2015 (though that's my gut feeling), etc but it will come in time - at least for sports/live event/premium VOD (likely over IP). I also think the tech to make UHD (eventually even the 4320P variant NHK uses) trivial to distribute before the end of the decade. Let's just pray they don't cheap out, splinter the bandwidth and slap a higher price tag on minimal improvements in delivered signal to the home. I threw up a little writing that :-/

Cheers - #19

 

[Steve Gibby]

As usual, very good points Blair...

Somewhere in all this thread lies the nucleus elements for a cost effective mobile EFP solution. What Red ends up doing will naturally also depend on the behind the scenes factors we aren't privy to. But we've all taken our time to express our concerns, ideas, needs, and hopes.

The trades I've been reading indicate that for sports TV the infrastructure changeover to UHD is in deep planning, with some networks already well into field testing UHD components in their bases, trucks, and UHD capable cameras on parts of their crews. For mobile EFP there are a lot of new cameras appearing in various form factors and cost levels.

As an early Red adopter like you I obviously love shooting with Epic, and will continue to - but for each production I've ever done I've always analyzed what camera/lens/resolution/shooting style combo is best for that production and then used the appropriate combo. I'd love to keep using Epic a lot in my mobile EFP productions but as multiple UHD capable cameras with good run and gun ergonomics emerge I hope Red will pay attention to the needs of mobile EFP production and evolve or spinout some killer setups for it.

Some interesting times ahead my friend...

 

[Denis Buhot]

weight...

weight...

let's hope what hopefully comes out will keep the WEIGHT in mind ! I have seen lbs adding up some place in this thread... EFP/ENG situations often mean operating alone, or a maximum crew of two people, right in the middle of nowhere...