RED FOCUS..

[Jeff Kilgroe]

As a shooter, a 4K "proxy" module would make selling the camera to productions much easier. The reason most of my jobs go with other cameras is because they have the ability to record direct to an efficient codec and that leaves less possibility for error on the post end of things (what most producers like). I believe a 4K (with options of 1080p, 2k, 4k) proxy module is really necessary to overtake a larger segment from Arri. Like I said in the other threads, this module needs to record slow mo real time. As a shooter the camera has everything I need, but that doesn't speak for the rest of the production. Proxies, Proxies, Proxies.... We can always go back to the R3D's for color, etc...

The point about less possibility for error [with proxies] comes up every now and then. While it is often an appreciable concept, isn't it interesting how the process of creating additional files to work with, which then link back to the reliable source media, is often viewed as "more reliable"? When in most situations, adding extra steps, extra data, additional layers of abstraction is usually what causes things to break.

Proxies are a solution for the now, not the future. As RED is able to roll out broad support for GPU acceleration, combined with the ever-increasing abilities of computers, R3D native workflows will eventually become ubiquitous. Until that day, RED is stuck between that proverbial rock and hard place. Do they give in and offer a proxy module? Is it really worth the R&D when they're a rather small operation to begin with? Or is it better to focus their resources on what is to come? After all, the industry has embraced RED, in some circles more than others, and the company thrives. People have got on for over 6 years now without an official proxy solution from RED and the R3D native workflow is easier now than ever before.

That said, I think the biggest reason we don't see a Proxy module is because of what has happened with the Alexa. At least a healthy 90% or more of everything shot on Alexa is recorded straight to 1080p ProRes or DNX with a baked in look. People see this as "good enough" and go all the way through post and finishing with these files. Often never recording a single frame of ARRI RAW to begin with. RED does not want that. They don't want to give people using their cameras the misguided opportunity to settle for "good enough" and negate nearly every key advantage there is to shooting on a RED camera. One could argue that they could construct proxies in such a way that people would have to re-visit the source media, by way of lower resolutions or some other penalty in quality. But we all know that would be bad. It would be misinterpreted. People wouldn't understand why they couldn't give beautifully baked and detailed 1080p "proxies" in the way the Alexa records beautiful 1080p.

Just my opinion, but I don't see the Proxy module as a winning strategy for RED, no matter how it's implemented. IMO, the only way for them to win that game is to not play. Better to keep working toward broad support and ease of direct use for a native workflow.

 

[David Battistella]

Just my opinion, but I don't see the Proxy module as a winning strategy for RED, no matter how it's implemented. IMO, the only way for them to win that game is to not play. Better to keep working toward broad support and ease of direct use for a native workflow.

ive been a big defender of RAW period and took a lot of flak in thread about proxy modules for this about a year ago. The point is RAW and that is the cornerstone of the whole system. The increased opening of the debayering through software GPU, pushes RAW deeper into post.

Just as any serious photographer shoots raw so it should be with red motion, but dslrs do offer JPEG as a functionality/space thing, I'm not sure a the "proxy equivalent" for red motion footage would be as good an idea.

me, I stick with RAW.

battistella

 

[Gunleik Groven]

@ Jeff

I agree on the reasoning, but have oposite conclusion...

1st... The internal debayering in REDs cameras are better than in the Alexa. Giving a foolproof RLF image with only wb baked, would probably give many a lot better result than all the metadata-tweaks in the world. (sorry for saying that loud).

2nd. I see REDS not chosen a lot, because they do not have this option. Not for Alexa as much as C300/F55. For broadcast.
In thsese situations, if the RED recorded prores, too, it would have been an obvious/better choice

3. we can discuss offline/online workflows as much as we want. I think it will/should be around for certain kinds of productions for quite a while still, especially larger productions where you have 3+ editors working in sync with color/audio/VFX. Typical large-scale TV-Drama/feature film stuff. Even if the machines could work with the RAW files, the signalchain/image-control drawbacks are quite significant.

People are not that scared of RAW anymore. Good prores files still have a lot of uses, and speedier turnaround and a bit more foolproof "baked" files, would not hurt a few of the situations I see.

It is the final image that counts.
And that is not allways bettered with RAW, even though one COULD argue that the RAW image could have a better potential. But that potential is not realized without people figuring it out. And THAT sometimes leads to the "RED is hard" assumption.

I have seen this quite a bit. It does not go away with a quality argument...

Giving some the option to use proxies, does not take away Davids and my (or your) chance to work in RAW as much as we desire... (I use proxies a lot, though...)

It is also an extra security in case of copy fuckups...

there is a lot of "smaller" productions that would be better off with a RED/prores workflow than the practical alternatives.

And the external recorders are not as "one body" as a module... People are a bit tired of that option...

 

[Mark L. Pederson]

Just my opinion, but I don't see the Proxy module as a winning strategy for RED, no matter how it's implemented. IMO, the only way for them to win that game is to not play. Better to keep working toward broad support and ease of direct use for a native workflow.

Yes. 10000% agree.

Perfectly good solutions that are available TODAY - I love using ATMOS BLADE on Dragon. And I have the flexability to mount it OFF THE CAMERA if I want - and still get auto record/stop features wirelessly. Any money RED wants to spend developing a proxy recorder ... is MUCH better spent developing the modules that ACs and Operators really want and need.

 

[Tim Morten]

How about making a RedMote that actually works.

Stills mode that is power sipping and only shoots the frames you press the button for instead of tagging them in a video stream

Better power distro (dare I say DTAP options?)

WiFi module for camera control, slating, circle takes, etc.

Great thread!

 

[Mark L. Pederson]

How about making a RedMote that actually works.

Stills mode that is power sipping and only shoots the frames you press the button for instead of tagging them in a video stream

Better power distro (dare I say DTAP options?)

WiFi module for camera control, slating, circle takes, etc.

Great thread!

WIFI makes the REDMOTE obsolete - just use your iPhone or Android or iPad as the REDMOTE. Put the REDMOTE screen on the side of a new PRO IO WIFI module and add nice buttons like I described in my post - then you have what AC / Operators want.

DTAP needs to die a fast, hard, bloody, painful death.

 

[Tim Morten]

Auto focus that is competitive with Canon/Nikon for stills

Built-in image stabilization.

Keep becoming lighter/smaller/cheaper.

Provide pre-set looks for the DIT-challenged.

 

[David Battistella]

I totally see your points Gunleik.

But what if the onboard processing/scaling does not provide/produce a rc/Rlf file that is "worthy of the raw". What if creating that file (which people will use as final image an never go back to raw) involves having to do line skipping, throwing away information ,etc. which of course may be "good enough" for some people but not good enough for RED (kinda rhetorical). The camera and engineering design is to record RAW and provided LUT'd signal monitoring paths.

Record in a proxy format sorta blows up in the face of all that Jim and red have stood for, so I see why the proxy stuff creates a drug like catch 22. It's tempting, but if people become addicted to it, then you are dropping down to be on Alexa's level or Sony's instead of maintaining what you built the company for, to create a digital camera that captures a file format that replaces film.

If people wanna record that monitoring feed at their own risk with some hackin-rig, great, but should RED endorse any Prores workflow that would record a flat matrixed, 444, uncompressed, open DNG, full quality solution? That also flies in the face of RED's very best patented invention, the R3D image container, which already beats the rest of these formats in many ways hands down, in size, compression scheme, quality and flexibility (and next ease of use as computers have caught up)

So I see why development on proxy has been dropped, because it ain't RED, I'm sure it's tempting as hell, but it's sorta admitting the red way of doing things is not good enough, that whole red concept has flaws (which I don't believe because r3d is pretty brilliant) no?

i think it's been really good that they held firm on this one, sometimes you have to stick to ideals and what sets you apart and above the crowd.

battistella

 

[Tim Morten]

WIFI makes the REDMOTE obsolete - just use your iPhone or Android or iPad as the REDMOTE. Put the REDMOTE screen on the side of a new PRO IO WIFI module and add nice buttons like I described in my post - then you have what AC / Operators want.

DTAP needs to die a fast, hard, bloody, painful death.

RedMote still compliments WiFi... Nice to have a compact device with physical buttons to mount to a wireless 1st AC rig.

I understand the vitriol about DTAP, but despite its reversability, DTAP is the standard that most owner/operator (rather than big production accessories) use. We're not going to buy new cables... We'll just get one lemo to DTAP and keep using what we've got. Might as well just make us happy and hive us the connector we use.

 

[Mark L. Pederson]

I totally see your points Gunleik.

But what if the onboard processing/scaling does not provide/produce a rc/Rlf file that is "worthy of the raw". What if creating that file (which people will use as final image an never go back to raw) involves having to do line skipping, throwing away information ,etc. which of course may be "good enough" for some people but not good enough for RED (kinda rhetorical). The camera and engineering design is to record RAW and provided LUT'd signal monitoring paths.

Record in a proxy format sorta blows up in the face of all that Jim and red have stood for, so I see why the proxy stuff creates a drug like catch 22. It's tempting, but if people become addicted to it, then you are dropping down to be on Alexa's level or Sony's instead of maintaining what you built the company for, to create a digital camera that captures a file format that replaces film.

If people wanna record that monitoring feed at their own risk with some hackin-rig, great, but should RED endorse any Prores workflow that would record a flat matrixed, 444, uncompressed, open DNG, full quality solution? That also flies in the face of RED's very best patented invention, the R3D image container, which already beats the rest of these formats in many ways hands down, in size, compression scheme, quality and flexibility (and next ease of use as computers have caught up)

So I see why development on proxy has been dropped, because it ain't RED, I'm sure it's tempting as hell, but it's sorta admitting the red way of doing things is not good enough, that whole red concept has flaws (which I don't believe because r3d is pretty brilliant) no?

i think it's been really good that they held firm on this one, sometimes you have to stick to ideals and what sets you apart and above the crowd.

battistella

And the perception of ProRes workflows from Alexa being so much "easier" is nonscense. Every serious production I am aware of that shoots on Alexa - records LOG-C ProRes 444 or ARRI RAW. THIS GETS TRANSCODED with color look files applied (almost always CDLs). So ... RAW transcode to ProRes/DNxHD vs. LOG-C Prores 444 transcode to ProRes/DNxHD.

Last time I checked - you could not record ARRI RAW at the same time as ProRes/DNxHD proxy - and also - last time I checked you could not record LOG-C ProRes 444 at the same time as a ProRes/DNxHD proxy with color correction for editorial.

How come everyone on the Alexa forum isn't bitching for a proxy recorder??

 

[David Battistella]

Exactly,

the shoot edit marketing campaign was brilliant, in a madmen kinda way.

battistella

 

[Gunleik Groven]

How come everyone on the Alexa forum isn't bitching for a proxy recorder??

First:
Again see your point and agree but, I am talkingabout productions that decide to go MXF on C300 or F55. Documentaries and smaller TV Drama.

Secondly, a prores 444 + LUT has a much lower toll on any compute I know, than any RAW file I know of. Usually a transcode would not be needed.


we are looking at different scenarios here...

I am talking about where Prores is not "good enough" but "extremly better than the alternative"

 

[Marcos Montenegro]

* A game-changer, platform (Windows, Apple and Linux) and NLE independent delivery codec that blows Apple's PRO RES out of the water, and has 4K to boot.
* True power management; if a REDVOLT could be made to last 1 hour, that would make lots of people here very happy.
* True DSMC implementation (to finally dump our DSLRs for timelapse/photo shoots)
* Ability to zoom 1:1 while recording
* A true HD 1080p quality EVF
* An built in ND-only mount, not a motion mount, that also can bypass all ND filters (clear).
* Different pattern screens, like a grid screen.
* Ability to delete files from SSDs.

Audio- I think we can all agree that RED's audio approach to the DSMC body is the weakest link of the entire concept, so lots of ideas for that.

 

[Mark L. Pederson]

Secondly, a prores 444 + LUT has a much lower toll on any compute I know, than any RAW file I know of. Usually a transcode would not be needed.

This is true, except not many folks do it that way because they feel the need to have CDLs attached on a per shot basis for editorial and viewing dailies. It's silly - but it's how it's often done. Now, there is a show in NYC that shoots on ALEXA without a DIT and CDLs - but that show doesn't just put a 709 LUT in editorial for a transform ... it pays the lab to do a "best light color correct' and transcode to DNxHD and post the viewing dailies online.

 

[Hrvoje Simic]

How come everyone on the Alexa forum isn't bitching for a proxy recorder??

Because they don't need it with Prores. Prores holds pretty well while it runs nicely on many machines. 12 bit 4:4:4 version does well in grading too and I prefer it to 10 bit dpx. 12bit 4:4:4 Prores out of Epic ?

I would suggest considering 16 bit 4:4:4 h.265 as an onboard camera recording option.
Not to replace raw, to replace 12 bit Prores in the future.

 

[Mark L. Pederson]

Becaus they don't need it with Prores. Prores holds pretty well while it runs nicely on many machines. 12 bit 4:4:4 version does well in grading too and I prefer it to 10 bit dpx. 12bit 4:4:4 Prores out of Epic ?

I would suggest considering 16 bit 4:4:4 h.265 as an onboard camera recording option.
Not to replace raw, to replace 12 bit Prores in the future.

I guess the point I'm trying to make is that it's NOT shoot>edit>post with MOST Alexa TV shows and films.

It is shoot>transcode to proxy>edit>post

Which is the same as shooting RED right now. What comes off the camera - LOG-C most of the time - does NOT go to editorial. A proxy is rendered. Sound familiar?

 

[Isaac Marchionna]

Biggest changes I'd like from an Operator/AC point of view:

Cabling/Ports - These have to get out from under the camera. Terrible design decision as they're annoying when there's no dovetail attached, and impossible to access without a multitool/spelunking headlight when on a longer dovetail plate. Just terrible. Also, I'd say ditch HDMI, I'd rather have 2-3 SDI outputs than HDMI, which has been shown to break. Go with an ALEXA style breakout/in setup. Look at the F55 for how to do a shit load of ports correctly.

Dumbside Controls - No brainer. I shouldn't need to either use the DSMC side grip or have to bother the operator to use his RED TOUCH, which wouldn't be an issue if...

More than one LCD/EVF port - Honest to god, stop nickel and dime'ing us with modules that remove features to enable features...+1 Module I'm looking at you!

Proxies, please! - I don't need proxies that often, but I like having a 1080p backup just in case, and often times the clients don't need 5K still images, and would rather just hand off 4:4:4 Prores files to their editor that day. It also avoids situations where some houses aren't setup for .R3D files, which usually turns into a conversation of "why don't we just get an Alexa anyways for this gig."

I want RED to stop showing off in progress designs, and start showing off shipping products. Make modules that we want, in a time frame that makes them feasible. Maybe NAB will be a sea change, but at this point with RED's track record even if they show off a design it could be two years till it trickles out...or gets quietly cancelled.

I love my Epic, especially the initial promise of modularity. But that modularity hasn't paid off. I don't know if it's too late.

 

[Mark L. Pederson]

I love my Epic, especially the initial promise of modularity. But that modularity hasn't paid off. I don't know if it's too late.

It's not too late. If anything ... it's exactly the RIGHT time.

 

[Isaac Marchionna]

It's not too late. If anything ... it's exactly the RIGHT time.

2 years in, with things like the meizler, proxy, etc cancelled, dragon trickling out SLOWLY, it shows that the product design/release cycle is very slow and very hit or miss on release.

If RED can get modules out that add features we need THIS year it'll be nice, otherwise we're still just playing catchup to design we should have had already.

 

[Stacey Spears]

I want digital audio in for my dual system sound. Would probably fit nicely on the time code module.