The next generation features of cameras

[Hrvoje Simic]

Why would one use software and develop additional hardware for software image stabilization if Steadicam, gimbals, dollies, jibs etc exist to enable smooth camera motion and capturing imagery which doesn't need post fixing with motion and geometry artefacts ?

 

[Arturo Sánchez]

What about internal nd’s? It’s pretty obvious but necessary for some small setup situations.
The RF adapter with variable nd’s has polarizers in it and it doesn’t allow to have rf lenses with nd’s

 

[Phil Holland]

What about internal nd’s? It’s pretty obvious but necessary for some small setup situations.
The RF adapter with variable nd’s has polarizers in it and it doesn’t allow to have rf lenses with nd’s

Revolva for Komodo, RF>PL and likely other options. Physical glass filters via rotary turret/cartridge. 4 of whatever you want in there.

DSMC3 I imagine will reevaluate internal NDs on the body side of things. But it will come at a cost. And electro-ND tech is just getting started. Way better solutions out there already than what we've already seen, just not available yet.

 

[Blair S. Paulsen]

Smart phone smarts

Smart phone smarts

I can't help thinking about the potential for integrating a smartphone into a camera system. Modern smartphones have so many built in capabilities that can be exploited - while being small, light and having their own internal battery. Durable, high resolution, high brightness touch screen - check. Gyros, accelerometers, Wi-Fi, BlueTooth and soon 5G TX/RX - check. Ability to internally record audio, perhaps proxy video and lots of metadata - check.

To make such a pairing really rock would take some serious development work. Custom apps, perhaps custom cradles to support positional data accuracy, monitoring latency low enough to be useful, etc. IMO, the juice would be worth the squeeze and the Komodo program is likely addressing several camera/phone interface issues already - so dev wouldn't be starting at ground zero.

The form factor of a typical premium smartphone seems perfect for an on board monitor - few shooters use anything significantly larger and it weighs less than typical LCDs.

Perhaps even more alluring is an opportunity to get even closer to the Obsolescence Obsolete mantra of the Madman. Instead of having to iterate the camera itself to access new tech, you could swap in a newer phone with better gyros, screens, etc. Moreover, the cell phone vendors would be doing most of the heavy lifting at budget levels commensurate with the massive mobile phone marketplace. The Hydrogen project may not have panned out as hoped, but if it fosters a deep dive into phone/camera integration potential... it could end up gestating a bountiful tech marriage. Some companies would avoid such a program to protect their display sales - but, for a disruptor like RED, I think they would see how such a combo would potentially increase sales of camera brains enough to offset whatever they make on displays.

Cheers - #19

 

[Hrvoje Simic]

I can't help thinking about the potential for integrating a smartphone into a camera system.

Glass, internal battery, multiapps, battery/processing hogging OS, multiple software makers, bloatware, data mining, UI fiddling, multicorp dependency, third party device access...kind of no-no for professional production camera system.

 

[DJ Meyer]

Why would one use software and develop additional hardware for software image stabilization if Steadicam, gimbals, dollies, jibs etc exist to enable smooth camera motion and capturing imagery which doesn't need post fixing with motion and geometry artefacts ?

Why would one use an auto body repair shop? Can't they just learn to avoid accidents?

 

[Christoffer Glans]

Why would one use software and develop additional hardware for software image stabilization if Steadicam, gimbals, dollies, jibs etc exist to enable smooth camera motion and capturing imagery which doesn't need post fixing with motion and geometry artefacts ?

There are hundreds of applications for further stabilization. There's a Steadicam operator who uses Stead XP together with his Steadicam in order to make movements super smooth. You might be happy with just the regular stabilization you get from gimbals and steadis, but someone like Fincher would benefit a lot from having a stabilization slider in post instead of going through Syntheyes every time. It also enables you to stabilize when you don't have stabilization. If the shot is sudden and there's not much time to get it, you won't have time to set up a gimbal. Or for times when you can only fit the camera in a barebone config.

With hardware-based stabilization, you don't get the same post-artifacts as with other methods.

But these arguments are in my view the reason we never get these tools. For some reason, many argue against these kinds of tools. I find that both counterproductive and keeps holding development back. There's no reason not to develop these tools as an internal part of a camera, there are only benefits to all of those who see the benefit of them. The rest can ignore the options, but I guarantee there will be times when such tools would save a shot that would otherwise have been doomed in post.

I have a feeling RED understands where the industry is going.

There's an interesting thing happening on the 3D side in post. While all of the major players like Maya and 3D studio max innovate slowly, the community driven and free Blender keeps pushing forward faster than everyone else. Much because it's open-source and everyone can help to add features. I have no doubts that Red will try and be ahead of the industry, but I also know that when there's a business model behind something, innovation can be very slow. There are very few who ask for the features I mentioned because there are very few who understand how to use them or who cares about them on the camera side.

Cinema camera manufacturers care for cinematographers and directors first, post-production second and VFX last. They will go where the priorities are and won't touch VFX and post before giving cinematographers and directors what they want, because that's where the money is.

I want to challenge that perspective a bit. I would say that Arri is in such a killer position right now that it will be hard to push Red based on just sensors and R3D. If Red would focus on giving a perfect VFX production camera that implements all the features I mentioned in my first post, it would be the go-to camera for any production using even a small number of VFX shots. If a camera could cut down post to a fraction because so much is in the metadata, it would save productions millions. If such a number would say anything it would be that this would be a very big incentive for any production to choose a Red. I know I wouldn't want to use anything else if I can get camera track data and RAW based stabilization directly in post.

Figure out high-resolution Z Depth data and the days of greenscreens are gone. All in-cam, everything ready in metadata for every shot taken with a camera, whatever the setup. That would be powerful and I think this is a point that camera manufacturers should start to pay attention to, not just sensor sizes, DRs and color science.

 

[Les Hillis]

I think RED should stick to their modular DSMC approach.

But maybe they could explore the possibility of spinning off a separate business entity to develop a SFX camera and SFX (or other non-standard application) modules for RED and other brand cameras.

It seems to me like too much of an ask for one company (of RED's size) to make a single camera or a range of cameras to cater for all possible uses.

Over time new features and capabilities could be integrated as standard in all their cameras, but at the moment the technology and its practical applications are way too underdeveloped for that, imo.

The amount of work (and investment) to get the new technologies developed to a professional standard that meets an as-yet-undefined and changing demand, would be too much.

I think the last thing RED should do is compromise what they've already established in order to try and integrate incomplete or 'of the moment' features into their camera's.

I'd vote for them to keep refining and developing their DSMC camera bodies, while keeping new or extra features as modules (until or unless they can be seamlessly integrated into the camera bodies without any compromise).

As others have said and pointed out, there's clearly an opportunity for new and emerging technologies to become standardized into cinema-quality cameras, but the 'who' and the 'how' of it is not so obvious or easy.

 

[Christoffer Glans]

But maybe they could explore the possibility of spinning off a separate business entity to develop a SFX camera and SFX (or other non-standard application) modules for RED and other brand cameras.

This is exactly what I think not should happen. The whole point is to integrate the tech into primary cameras.

It seems to me like too much of an ask for one company (of RED's size) to make a single camera or a range of cameras to cater for all possible uses.

Red is a major player in the camera manufacturing market and the technology isn't new or needs to be "invented", it needs to be implemented into current cameras. You can lift up your phone and record realtime tracked footage with 3D elements directly tracked into it.

I think the last thing RED should do is compromise what they've already established in order to try and integrate incomplete or 'of the moment' features into their camera's.

I'd vote for them to keep refining and developing their DSMC camera bodies, while keeping new or extra features as modules (until or unless they can be seamlessly integrated into the camera bodies without any compromise).

As others have said and pointed out, there's clearly an opportunity for new and emerging technologies to become standardized into cinema-quality cameras, but the 'who' and the 'how' of it is not so obvious or easy.

There's nothing new to the DSMC system. What should be refined? The Ranger already has the features that were missing in the DSMC lineup, the Komodo is much better for small camera setups. But when Arri now has the Mini LF there has to be more than just a stop more of DR, global shutter and more Ks.
If I'm gonna upgrade from a Weapon 6K, what is it that I actually need? What do productions need?

The features mentioned aren't anything new, your phone has them and the hardware cost is insignificant compared to other development costs. Compare that to the cost of post-stabilization and camera tracking in post. This is also what I mean by camera manufacturers focusing on cinematographers and directors primarily, they never ask for these things and therefor the implementation will never happen. But for anyone who does work where these things are done in post, it's such a huge postproduction cost that it's baffling that more people won't lobby for this implementation into high-end cinema cameras. At the moment I'm doing a tedious tracking and VFX job that has so far taken three days because of its complexity. I would have been done with it within an hour if I had the tracking data as metadata.

People keep saying these technologies would result in compromises, I would like to know what compromises that is because I can't see them. There doesn't seem to be many compromises to our phones when it comes to accurate tracking, so why would there be on a much larger system?

 

[Blair S. Paulsen]

I would note that part of the disconnect might be the way most projects are budgeted. Production is typically expected to capture a lot of material in a short time as cheaply as they can. A more wholistic approach that actually evaluates the total cost from planning to deliverables might make other choices.

Cheers - #19

 

[Les Hillis]

This is exactly what I think not should happen. The whole point is to integrate the tech into primary cameras...

I'm not against the development and incorporation of 'new' features into the RED ecosystem, I was just trying to be realistic about what it would take for it to not only happen, but also be successful (which includes factoring in resistance from established/entrenched players in the wider 'industry') .

I don't actually think RED would make any of the kinds of compromises I've suggested (eg. limiting functionality or making other sacrifices to get things working together), but I think that itself suggests why some of the things people are asking for might not happen (at all, or in the form they'd prefer).



Anyway, it's all just a bit too complicated for me to write about here. I get the appeal for incorporating new features into cine-camera's, but it also makes sense why it hasn't happened, yet.


Makes me wonder what Jarred and others in-the-know at RED wish they could do camera-feature-wise if they had unlimited resources.