RED FOCUS..

[Jeff Brown]

And to add to the recap....

In 2007 the Red One entered a high end market that was dominated by tape 1080p cameras or film. It was buggy had some issues and post was not easy. But it had no competition so it flourished.

in 2014 Dragon Epic is entering a high end market place crowded with viable alternatives for high end production. Without useful modules (as discussed here) and the need for a new Rocket X to debayer full quality in realtime whether it flourishes this time is very much anybody's guess.

I'm really, really hoping that before Dragon is fully "out there" we have:

1. Full GPU SDK for post FINISHING. Forget FCP X R3D timelines etc - footage will be transcoded by DITs as ever but it HAS TO BE GRADABLE with the hardware post houses have NOW. Trust me, all those tears ago when I went and conformed projects pre-rocket in a lot of the major post houses in Soho just to get the camera in use it was a quirk but impressive. This time it will be unforgivable.

2. Modules / Studio shell of whatever to fix all the ergonomic/usabiilty issues that makes crews and DP's bitch and moan. As mentioned earlier in the thread Dragon can have all the resolution, sensitivity and frame rates as you like but without these things DOPs will be reaching for the familiar, balanced , tactile feel of the grey box with the name beginning with "A".


The clock is ticking and the upgrades are shipping. I'm already getting calls as "The guy who knows Red" from shoots renting from owner operators (who managed to be online before me because I was at Red's stand!) suddenly realising that this isn't just a normal Epic and their post schedules and budgets are suddenly going off the rails.

You only get one chance to make an impression and I fear that unless things get straight before Dragon is out in numbers then the rental/ high end industry will have absolutely no reason to change its mind about "same-old" RED.

Hopefully the Off Topic threads still get read and I will be proved wrong......

 

[David Battistella]

Hey Jeff,

FYI and to clear things up. This wasn't moved to off topic, I put it here so that it wasn't mixed in with epic, dragon, workflow or anything else. I just wanted it to be a broad range of ideas so I selected off topic, it wasn't moved here.



Battistella

 

[David Battistella]

I appreciate you are adding post into the equation here Jeff.

It it can not be ignored.

When you you choose RED, it's a whole system and that means looking at it right through the signal chain.

Problem now is that as data rates increase with dragon, the bottlenecks in computing power start up again. While it's great to have more data and 6k and more dr, that kinda breaks a lot of existing workflows. I get that RED is on the leading edge, but a perfectly solid 5k workflow yesterday, is seriously strained with the introduction of Dragon.

Why?

Well, the best post tool for RED became obsolete overnight.

What it means?

Users need to invest in new hardware, either an RRX, dual GPUs, a new Mac, etc,. To shoot dragon, you have to be on the front end of computing power. There is a sense that we are kinda back in 2007 trying to push 4k down the pipe on hardware that "almost" does it.

Its different in many ways and much more mature than it was then, but let's face it, you dragon upgrade really needs to go hand in hand with a RRX in an old machine or a nMP loaded to the tits. Again here, an owner op might be nimble enough to eat this, but when that dragon footage starts pouring into post houses that have not converted to nMP's and all the very most up to date stuff, it also messes things up.

5k on an iMac, doable. 6K not so much (we'll see what shakes down with expansion chassis for RRX cards). So you see folks aren't going to want to put down 10k beside every imac in their facility to "handle" dragon. This is where the bottlenecks start and where the proxy module talk comes in because if I drop an Alexa Prores 4444 on a USB 3 Drive and plug it into an iMac, I'm cutting inside ten minutes.

Believe me, producers and editors notice this and it makes r3d a non viable real time, wide usage editing solution. This is where the split happens. A drive shows up and I can cut on an iMac, another drive shows up (where a DIT was not involved) and it could be a couple of days of transcoding. I know I might hear that premier and FcP x are real time, BUT FOR THE MAJORITY, it is not. That is where the unfriendly to post thing comes in, because that guy who was cutting an Alexa show yesterday does not want to be told, "well, go upgrade you whole setup" he made the same amount of money yesterday with less hassle.

I think post has a larger influence on all this than anyone wants to believe. So there are two solutions.
1. Edit ready files from the camera
2. SDKs that really open r3d up to being real time on 1 or two year old hardware.

Dont underestimate post.

Every movie goes through that pipeline and those professionals have a say. If they think R3D is a pain in the ass, it's as important to fix that as it is a build a camera module.

battistella

 

[Gunleik Groven]

Cool input everyone!

I have kinda refrained from commenting on post. Not because it is irrelevant, but because it doesn't really feel like the quantum leap to go from Epic to Dragon that David suggests.
But I have not tested. I am just sorta "happy".

I appreciate all the input here.
Would be fun to see more people participating in the same constructive spirit.

There is a few cues to this conversation. Meitzler making it almost out the door (as it seemed), and then not, is one of them.
GPU vs RR/RRX is another.

I am open to seeing this all the wrong way. Maybe the current Epic formfactor is right, and could be just adjusted to more versatility.
Small is good, and there are a lot of uses for small.

Is it possible to bring more voices to this discussion?

I would love to hear from dedicated ACs, rental-houses, and people super-happy with the current body.

As a one-man-band, I dig the current config...
So I am sitting solidly on both sides of this argument...

I actually have a pretty long list of what I think RED does "really good" now, which have been issues at earlier releases.
As stated a couple of times earlier: to me it looks like this is part of a change v3 for RED. And so far they've been pretty good at hitting a few nails squarely on their heads with the previous transitions.

 

[Jeff Brown]

Exactly. The single most important point in relation to Producers not looking epic is it's "difficulty in post".

For the last 3 years we had a situation whereby a DIT on a Prores shoot has basic kit and charges accordingly. On a Red shoot there is an additional Rocket hire to that kit. It costs more for the Producer and that's a real world, every day example. Added to that the post house tells them that 1080p/2K Prores is much quicker and easier for them to ingest and grade. Decision is made - shoot Alexa.

Epic is now niche in my market - Movi rigs/ Octocopter/ Plate shots where resolution counts or post heavy work. Here the trade of is balanced - the added cost for DIT post / post is outweighed by the camera's strengths.

Everything else, however, goes Alexa.

Things have changed slightly and evened the plaint field. The XT brings brings Arriraw and yet even when you try to explain to Producers that this means the Alexa is the same as Epic i.e. (Shoot Raw/ Transcode for editorial / conform and grade.) they still hold on to belief the Epic is harder. Maybe that's legacy and maybe it's DIT Rocket hire and maybe it's post houses not wanting to buy Rockets for finishing. Probably all of those.

So the SDK / GPU support is crucial if Red have no intention of going down a lower resolution Prores recording route. Opening up the old rocket to debayer Dragon even at less than realtime is also a massive deal for all those "set in stone / 5 year lease decision/ don't you dare put a square peg in my round hole" Post houses.

Let's not forget:

Big movies choose their post route based on their camera.

Owner Ops and small shops choose their post route based on their camera.

TV / Drama / Commericals/ Lower budgeted films (i.e. everything else and the largest percentage) choose their camera based on their post route.

And if you've ever spent days of your time in numerous meetings at large broadcasting establishments (BBC, SKY, CHANNEL 4 etc) as I have talking to them about Red post etc and seeing what their bureaucratic infrastructure means you will understand why they all shoot Alexa.......

As regards 6K I don't think a majority of shoots will shoot that. The S35 of the 5K on Dragon will be the favourite resolution for a lot os shoots (lenses will match/ no coverage issues/ less data). Dragon will be chosen not for resolution but for sensitivity / feel and frame rates for jobs finishing in 1080p for the small screen for the foreseeable future.

So it's not 6K it's the cost of buying/ replacing Rockets with Rocket X without a viable alternative in place currently. And by that I mean Redcine X is not a finishing tool of choice and neither do any post houses want to go via it to extract DPX sequences for grading. When they get a load of Dragon R3Ds and press import on their Baselight what do you think will happen? And who will get the blame? Correct - the camera.

And just to make this a real world discussion and not just hypothetical today I'm DIT on a very large worldwide commercial for a very well know sports brand featuring some of the highest paid sports stars on the planet.

We have 2 Epics on Movi rigs - Guess why they chose Epic for this? (See above for answer).

We have 2 x Alexa XTs - Guess what format they are shooting that's easy for them despite this going to cinema? (See above for answer).

We are transcoding everything on set for the editors to with LUT correction - Guess what platform they are editing on which necessitates this (PS. it's not Premiere or FCPX like 100% of the jobs I do)

It's 4 - 6 cameras shooting high speed and I have a fair bit of gear here including a couple of Rockets - Guess what the Producer said? (I'll tell you that answer for this one- "Is it that much because we are shooting Reds - they always cost me more I can't stand the things. On Alexa I never have to pay that much - it's so easy")

So real world examples for real world problems.

And at the moment I'm just glad they are on MX chips and not Dragons so I don't get THAT call when they online..........

 

[Gunleik Groven]

Jeff...

So I guess supplying something like a 1-minute Dragon 8:1 5kHD file to test postworkflows with, would be a good thing?

BTW, I see your points. RED post is not hard, when you have set up the workflow.
But you have to set up a RED workflow.

 

[Jeff Brown]

Jeff...

So I guess supplying something like a 1-minute Dragon 8:1 5kHD file to test postworkflows with, would be a good thing?

BTW, I see your points. RED post is not hard, when you have set up the workflow.
But you have to set up a RED workflow.

I've got plenty of Dragon stuff here and it makes my 8 core towers creek (new dustbins next week). I know my purchase/upgrade cycle is quicker than post so that's my concern.

What's the problem with Red Workflow? The bit that says "Red" i.e. specific and proprietary. (See my point about camera decisions above).

I drive an Audi A5 coupe which I really love. The design and the look are great. When I go to the garage though I don't want to buy Audi fuel, I want to buy Fuel.

(Long day I'm tired and running out of analogies...).

 

[David Battistella]

Cool input everyone!

I am open to seeing this all the wrong way. Maybe the current Epic formfactor is right, and could be just adjusted to more versatility.
Small is good, and there are a lot of uses for small.

Is it possible to bring more voices to this discussion?

I would love to hear from dedicated ACs, rental-houses, and people super-happy with the current body.

As a one-man-band, I dig the current config...
So I am sitting solidly on both sides of this argument...

I actually have a pretty long list of what I think RED does "really good" now, which have been issues at earlier releases.

I am all for more voices here, I genuinely want to read about what people think and their experiences. And. I have not plugged dragon into my workflow so it's a bit unfair for me to over categorize. I find it a shame that the rr could not at least do some of the lifting, but I'm happy I got through three generations of red footage with it.

My laundry list of amazing things about RED would have many people calling me a fanboy, but I'm less interested in restating what is obvious to me and more interested in rolling up the sleeves and making this whole thing even more perfect.

So please folks, chime in. I hope (and think) RED are open to listening. They always have been very accessible, that has been my experience because super busy guys like Jim and Jarred actually take the time to answer my emails and be here participating all these years. They care, we care.

Battistella

 

[David Battistella]

BTW, I see your points. RED post is not hard, when you have set up the workflow.
But you have to set up a RED workflow.

and....

someone has has to be willing to pay to implement that workflow. PAY.

Jeff's providing a pretty honest real world account of what I have seen and heard. It's not only about stellar images.

Battistella

 

[Jeff Brown]

I agree Davd and I'm hoping this discussion is being seen by those that count.

I don't post as much as I used to in the early days as I'm usually too busy.

I been following Reds developments since the first DVX user thread which became Reduser. Name down in 2006, Red one in 2008. Attended Reducation no. 1 and now I run a rental company with 8 Epics, 2 Scarlets and a Monochrome (most soon to be Dragons).

It's been a long and hard journey. Exhausting in fact.

I'm just hoping that some of the things that I have to deal with daily can get addressed. Everything I say here is from a supportive standpoint.

This is stuff that I don't just think, but know, would make a HUGE difference in Red's market penetration and use.

 

[Russ Campbell]

I appreciate you are adding post into the equation here Jeff.

It it can not be ignored.

When you you choose RED, it's a whole system and that means looking at it right through the signal chain.

Problem now is that as data rates increase with dragon, the bottlenecks in computing power start up again. While it's great to have more data and 6k and more dr, that kinda breaks a lot of existing workflows. I get that RED is on the leading edge, but a perfectly solid 5k workflow yesterday, is seriously strained with the introduction of Dragon.

Why?

Well, the best post tool for RED became obsolete overnight.

What it means?

Users need to invest in new hardware, either an RRX, dual GPUs, a new Mac, etc,. To shoot dragon, you have to be on the front end of computing power. There is a sense that we are kinda back in 2007 trying to push 4k down the pipe on hardware that "almost" does it.

Its different in many ways and much more mature than it was then, but let's face it, you dragon upgrade really needs to go hand in hand with a RRX in an old machine or a nMP loaded to the tits. Again here, an owner op might be nimble enough to eat this, but when that dragon footage starts pouring into post houses that have not converted to nMP's and all the very most up to date stuff, it also messes things up.

5k on an iMac, doable. 6K not so much (we'll see what shakes down with expansion chassis for RRX cards). So you see folks aren't going to want to put down 10k beside every imac in their facility to "handle" dragon. This is where the bottlenecks start and where the proxy module talk comes in because if I drop an Alexa Prores 4444 on a USB 3 Drive and plug it into an iMac, I'm cutting inside ten minutes.

Believe me, producers and editors notice this and it makes r3d a non viable real time, wide usage editing solution. This is where the split happens. A drive shows up and I can cut on an iMac, another drive shows up (where a DIT was not involved) and it could be a couple of days of transcoding. I know I might hear that premier and FcP x are real time, BUT FOR THE MAJORITY, it is not. That is where the unfriendly to post thing comes in, because that guy who was cutting an Alexa show yesterday does not want to be told, "well, go upgrade you whole setup" he made the same amount of money yesterday with less hassle.

I think post has a larger influence on all this than anyone wants to believe. So there are two solutions.
1. Edit ready files from the camera
2. SDKs that really open r3d up to being real time on 1 or two year old hardware.

Dont underestimate post.

Every movie goes through that pipeline and those professionals have a say. If they think R3D is a pain in the ass, it's as important to fix that as it is a build a camera module.

battistella

Well said. I really wish that RED would acknowledge that this is one of the primary reasons for Alexa dominance, not the form factor of the camera. Yes, AC's and grips moan, but it's the producers and directors at the end of the day who have the final say and their reasoning is that Alexa post is simpler, safer and more affordable due to Prores.

 

[Nick Morrison]

Well said. I really wish that RED would acknowledge that this is one of the primary reasons for Alexa dominance, not the form factor of the camera. Yes, AC's and grips moan, but it's the producers and directors at the end of the day who have the final say and their reasoning is that Alexa post is simpler, safer and more affordable due to Prores.

Do we need a prores module though? Wouldn't the new Convergent Design Odyssey fit the bill? It looks good enough to replace your SmallHD or TVLogic and provides focus assist & proxies, all at once.

 

[D Fuller]

Do we need a prores module though? Wouldn't the new Convergent Design Odyssey fit the bill? It looks good enough to replace your SmallHD or TVLogic and provides focus assist & proxies, all at once.

Do you want to shoot 120fps? If you do, then recording to any external recorder is kludge. It will slow you down.

 

[Burke Doeren]

This is purely speculation, but out of curiosity, do you think if RED released a rear connection module with dual linked HD-SDI outs, could we use the Odyssey 7Q to record 120fps in at 1080p in DNxHD or ProRes?

 

[Blair S. Paulsen]

Live UHD/4K output over SDI and HDMI for high resolution monitoring and live applications. I've had to sub-rent F55s just for this feature.

Cheers - #19

 

[Adrian Jebef]

The point here is not that RED Post can't be done, or that it will cost you real $$$ to set it up, it's that it's been a proprietary pipeline for way too long. GPU acceleration in RCX is a GREAT start but we need it in Resolve and all the 3rd party Apps as well. I get better than realtime performance with ARRIRAW and F65 raw files with just one Titan in an old Mac Pro. For the cost of a single RedRocket I can hit 85fps+ rendering out full rez Arri and Sony deliverables.

In the professional LA/NY market that encompasses high-end national commercials, TV shows, and Features we are required to deliver greater than HD deliverables maybe 10% of the time. And that's 9% for VFX on Features and the other 1% for everything else. On these productions it's all about scale and the final master. 90% of the Features are still only doing a 2K finish. Why? Because when it's done well it looks fantastic. And making a 4K master requires 3 times the amount of time to render VFX and 3 times the amount of data storage and 3 times the money. For a $100 million dollar movie that's an extra $30 mil. For TV shows and commercials it's a basic no-starter.

Proxies are for conforming back to the R3Ds for a 4K finish. If the camera could record a compressed log-encoded video file at HD resolution you would have a lot of content people out there who never wanted a 4K finish in the first place.

 

[Hrvoje Simic]

100% agree. It is worth noting that if RED built a "pick it up and shoot" camera for the EFP/ENG, higher-end YouTube kids, reality TV shows and docs - that could ONLY record UHD ProRes (3840x2160) - no RAW - Dragon sensor - priced less than the Amira - they would sell a ton IMO. Not suggesting I think they should do that - just saying it would be another way to split a product line and tap into a bigger marketshare without cannibalizing existing product line. Interesting that many folks on this thread (myself included) talk about getting a bigger share of the high-end features & TV world when obviously - sports, ENG/EFP, docs, reality, higher-end YouTube kids, etc. is a much, much, much, much bigger market. "Broadcast" is now "Broadcast+Mobile/Internet". Massive. And it's all going to UHD aka 4K.

RED has the 4K bull by the horns ... and the competition is awake. Dragon slays the other sensors out there and the competition knows that - so they will play to their strengths and RED's perceived weaknesses.

I agree on the potential benefit regarding no RAW for different segments, but I see a bunch of potential issues with Prores UHD. The reason for referring to this is not just your post but many similar notions, for the route which would likely bring benefit mostly to the same segment Red cameras already cater.

With that route the benefit would be avoiding a de-bayering step, but one important issue would still remain.
Which is recording data rates.

Which goes back to RAW data rates. Roughly fourfold increase in files sizes (slightly less with Prores compression) is a huge issue and roughly 600 GB per hour of footage for UHD 12 Bit Prores 4444 would take a huge bite out of practicality. Having only UHD Prores would significantly limit usage in that market segment.


This is why (numbers added) :

1. Big movies choose their post route based on their camera.

2. Owner Ops and small shops choose their post route based on their camera.

3. TV / Drama / Commericals/ Lower budgeted films (i.e. everything else and the largest percentage) choose their camera based on their post route.

For 3) I don't see an issue with TVC's with UHD RAW capture route due to less amount of material and fast and proven workflows, and I would add - based also on their shooting priorities, crew, budget, etc...but the main point of different market segment related priorities is still here.



One example from a good friend and a colleague.
He's currently finishing a 60 min. documentary he shot over the period of roughly 6 months.
55 hrs of footage.

If that was UHD 10 bit Prores 4:2:2 that would roughly take 20 TB of data.
30 TB in 12 bit Prores.

I asked him today if given the choice between two high-end cameras for that type of project, HD (downscaled from a higher resolution sensor, Alexa and up) and UHD, other picture quality factors and camera practicality and functionality being the same, having to use Prores in both cases, would he go with UHD acquisition.
I already knew the answer but I asked him anyway.

/dramatic tone/

"No way."


I gave the same question to another good friend, very experienced AC and DIT.
I also already knew the answer but I asked him anyway.

/dramatic tone/

"No way."


There is very good reason for that choice being the same and predictable, whether seen from the shooter's or production perspective.
Freedom and practicality of creating the content.

In many cases the rest is secondary once the picture quality level meets the general criteria.
With today's cameras, 12 stops and up, 12 bits and up, large sensors, realistic colors, good sensitivity, no ugly highlight clipping, no visible compression etc...for many picture quality is satisfactory.

Many will confirm that how "good for the job" the camera is depends not only on its picture quality, but also on "how good you can shoot your project with it", and that also includes keeping constant fiddling and hassle around recording media, media management, additional hardware, etc. on a minimum. For many types of projects this is a ball and chain wrapped around the ankle of creative potential. If any of those workflow factors lead to a shot missing to fully tell the story, or take away the time, focus and energy to tell it properly, details in eyelashes or clouds in the background fail to compensate.


This is one of the reasons Sony went with XAVC 10 bit h.264 for UHD in that segment.


This is also one of the reasons I proposed the consideration of broadcast quality HEVC route for UHD as a non-RAW recording option.
Hardware support in TV sets, mobile devices and computers is around the corner.

Something around 500 Mb/s 12 bit 4:4:4 might be a sweet spot.
300 Mb/s 10 bit 4:2:2 for space saving, basic color correction or light grading and direct to broadcast.
Both visually lossless.

 

[Gunleik Groven]

Hrvoje,

I think you have an excellent point...

For 1080/2k prores is a very good codec.

Possibly not for 4k

And I think (frankly) that the main useage for a proxy-module is towards 1080/2k delivery.

I am also aware that RED is all about 4k, so a module would probably have to have a useable solution for a 4k format. What you describe is that.

A 4k REDraw file is around the same size as a prores 444 1080 file.
The decoding needed for the prores is significantly less and it plugs into "all" workflows without extra hardware (transcoding/LUT or not)

AT 4k, not so much.

One thing which is a philosophy vs a practical reality for many, is that 4k, while having become a deliveryformat and definitely is the route of TV, for the most part, 4k is NOT a deliveryformat for "bread and butter" projects.
I have now finished a few projects in 4k. And it is doable. Screening has been in 1080/2k, though...

For some projects the futureproof argument holds. For others, not so much.

If finishing in 4k ONLY adds hazzles and has no gains, there is little point in finishing in 4k for a 1080 screening. (for TV it is actually often screened at 720 p50 from a 1080 p25 master...)

And if you need to finish in 4k at a later stage and the project is not VFX heavy, it is mostly a re-render or re-conform issue...

It does actually depend a bit on the project what we need to deliver.

For a Drama series which will be screened again and again, a 4k master MAY hold some relevance.

For a TVC for an event/product-release/whatever happens "NOW"... not so much.

And the proxy-thing, is of course about having access to a quick turnaround solutions for "here and now" productions.

Problem with KiPro/PIX etc, is that it is considered "extra", and not "part" of the camera.
But the time may have gone from a proxy module, or not. Not sure.

The Odyssey looks good...
From the productions I have followed, there have so far been issues with all the external recorders. Maybe this is "the one". It def has the right price...
I do think that the proxy-recording would have a place in a "complete body module" or a re-designed body for different markets , though.

Now... Problem is if RED thinks it s wrong of us to deliver 1080...

I love the flexibility of using 6k for stills and superwides, 5k for S35 1080 or 4k and most of all the image of the Dragon.
But I guess I will continue to finnish in whatever size/format that suits the production...

Working on a project now going to 4k finnish.
Then 4k is no problem. Just a huge gain.


I guess one of the things we are discussing here in a way is "modularity of workflow options, and not only the physical camera"

And also about letting RED plug-in with "standard" equipment (like XLRs and Lockit boxes) and "standard" workflows, when those are more suited for the situation.

The image is there all the way now, and has arguably been there all the way.

"Reality" has changed a lot since 2006, though. Both as to where computing bottlenecks are and what alternative good images and workflows are out there.

And I loved seeing 6k playback at realtime from GPU at IBC.
I think THAT is the future, but for a lot of people, the adaption time to even that (not RRX) will include a few hurdles.


Title of the tread, is RED Focus.

For the post-side, I guess it is a bit of if there should be a focus both on "here and now" and "the future".

From the sentiments I see.

Noone is claiming that RED workflow is undoable...

We do it all the time.

But market and competition has changed quite a bit since 2006. Not least due to REDs entering of the marketplace

 

[Adam Johnson]

As far as tech advancement goes I would LOVE for RED to continue developing the light field technology to implement in future cameras. The moment focusing is no longer "baked-in" and RAW in combination with a camera sensor like Dragon... Mind explosion

 

[Scott Brown]

Been shooting RED from the very start and we have NEVER had to deliver 4K, I know it'll come soon BUT it has been 1080p all the way so far and the workflow could have been made so much simpler and easier with the option to use 1080 ProRes from day one!

We have always transcoded to ProRes and have cut and graded from this, so for us RAW has not been a big sales point. In the early RED One days, the novelty of shooting shallow DOF with a cinematic look was fantastic. There really were no other affordable options and RED gave us a USP.

In 2014, things are very very different. Jim and RED have always chased Hollywood and it always seemed that the priority was to get the camera into the hands of the big studio productions, perhaps at the expense of the people who actually put down their money for the camera. Today, Arri's Alexa has become the preferred camera for most high-end productions, certainly here in the UK and this has been down to RED ignoring their hard core user base. I can't help thinking that if RED had adopted all the feedback/advice given through this forum over the years, we would now be in a very different position and RED would be in a much stronger position.

While the Dragon sensor is wonderful and a major step forward, RED needs to address lots of other issues if they are to survive and thrive moving forward.

If I was RED, I would completely abandon the current form factor. There are now just too many modules and options and costs just continue rising to get a camera that is workable for most users. Keep things simple and look at how most people are using the camera. Most users are not shooting Hollywood epic big budget productions, I suspect the majority of RED users work with small crews and often modest budgets.

Form factor: make the camera slightly larger. This will allow much better thermal properties meaning the camera will run cool and above all quietly. Add built in ND's, full size XLR's and work with Sound Devices (or similar) to add the best audio circuits in the world to the camera. A larger form factor should also make future brain upgrades easier!

Media - technology is changing rapidly and media costs are dropping. CFast 2.0 flash memory cards and many others are appearing so being tied to RED's proprietary and very expensive media is a BIG negative and yet another excuse for productions to use Arri.

Workflow - talked to death here already BUT nail this urgently and above all keep is simple and make it fast and STOP giving producers and excuse to choose other cameras. RED is getting a lot of negative press in the industry and we don't want to be left with cameras that nobody want to work with. A popular camera is a profitable one! The simple fact today is that our crews are choosing other cameras (C300's, F55's, Alexa's) and our expensive Epic is looking rather lonely sitting on the shelf.

In 2014 everyone will start moving from HD to UHD, and the competition is heating up rapidly. RED's dominance is on the decline and urgent action is needed or I for one will move on and I suspect I won't be alone.

I'm not being negative here, I really want to stick with RED but my fan boy days are over and I for one, seriously wish Jim was back in the driving seat!

Best wishes

Scott