What we CAN do...NEW

[Jeff Kilgroe]

I think that's an interesting flip Jeff. Are you saying you will charge more for 2K or 4K than for transcode? In which case, what if they really don't want it. Then don't you eat that labor cost?

Actually, I keep changing my mind on that... Not sure. I guess I won't know until the camera is in my hands and I do a couple real jobs with it. I think most jobs will have to be approached and quoted on an individual basis anyway, so this discussion isn't very likely to come up in this way. If the client wants a DVCPROHD transcode of the footage, that's fine, I'll just incorporate that into the bid for the project and not make any specific issue over it.

But on small projects or where clients are nit-picking every aspect of the job, I'm thinking that I will probably charge more for the 2K / 4K master. I'm looking at it as if that's the "negative". 99% of the time I will have to copy the data over to a hard drive and most likely convert to a format they want as part of the job anyway. So may as well figure my rates based on that and charge them a few extra bucks to have the 2K/4K footage as well.

Someday when everyone wants the 2K/4K footage, I'll adjust my rates accordingly. Or maybe it makes more sense to just give the 2K/4K footage and include one transcoded copy in the price as well -- price is the same whether they take advantage of that service or not.

 

[Blair S. Paulsen]

1080 crop flag

1080 crop flag

I like the scaled to 2k with the 1080 crop flag. Why?

1) It leaves the door open for some really nice "smooth cam" or motion tracking in post when needed.

2) The crop should avoid any of the potential softening of 1152 to 1080 and cropping should be much faster than re-scaling.

3) The simple math and increased light sensitivity of the "4 to 1" super pixel makes sense to me but as has been stated several times that's a Graeme question.

4) The crop flag hides any "technical complexities" from the editor who is finishing for HD.

 

[Gavin Greenwalt]

I only wish to have the best camera for my dollar, can you blame me for that?

I agree. Every customer for themselves! It's red's responsibility to deliver a camera that appeases as many customers as possible. It's each of our responsibilities to ask for what *we* need not what we *think* someone else *might* need.

Only problem is EFP and High-End Cine seem to be under represented. So.... ummm... Gibby gets his opinion multiplied by x10 and each of the beta tester's crews get x20. And me.... divide mine by 5 because I'm only a renter.

2c.

 

[I Bloom]

4K REDCODE RAW @ 1-60 fps to RAW Port
4K REDCODE RAW @ 1-24 fps to Compact Flash or RED Drive

2K REDCODE RAW scaled @ 1-60 fps to Compact Flash or RED Drive (Under development)
2K REDCODE RAW cropped @ 1-100 fps to Compact Flash or RED Drive

Jim

Jim,

I really appreciate your candor. I have given a great deal of thought to this issue since you first introduced it. I was initially in favor of keeping 1080pRGB and 720pRGB because I felt these formats would be important for non-cinema RED applications. I've been persuaded otherwise.

Once I had realized that RGB would not be helpful as an in camera format I was worried because I saw no in camera format that sufficiently reduced the amount of data to something near to current HD formats while still giving us a 35mm FOV.

The introduction of 2K Scaled REDCODE RAW has changed this. If this new format can deliver 24fps at around 7MB/sec then it fills the gap.

I want to clearly state that I now support RED dropping all RGB formats (including 1080p RGB and 720p RGB) in favor of only recording RAW. I believe this will lead to a more streamlined system that will best serve users on both the cinema and EFP sides of the debate.

By dropping RGB and focusing on RAW we also open up the possibility of greater performance gains as RED continues to update their processors and FPGAs and possibly faster turnaround time on camera developement with fewer formats to deal with. If we dump RGB now we won't be stuck with it in years to come.

2K Scaled RC RAW will also most likely intercut well with 2K Windowed RC RAW making a 2K only pipeline a very viable and flexible option for many productions, in particular documentaries.

Thanks again for including us in this process,

Ian Bloom

 

[Michael Brennan]

Please pick an argument. For theatrical projection, you shoot 4K. For 1080p home delivery, it makes more sense to have a singular 1080p option than creating the process of having to shoot 2K and resizing it again.

Here Here.
If possible enable 1080p and 2k.
Otherwise 1080p as that often requires fast turnaround.
4K for large screen presentation where fast turn around is usually not an issue.



Mike Brennan

 

[Alexander Nikishin]

H
If possible enable 1080p and 2k.

It is possible to enable both 2K and 1080P, but not without losing 2K windowed at 96fps, which is a major selling point to the cinema oriented shooter.

60fps just doesn't cut it for many shots.

And if you can have 2K scaled (meaning the full 35mm frame of the 4K sensor) then why would you complain when you can just convert that to 1080 (possibly) in camera or in Redcine, after all, you will need to go through Redcine at some point anyway.

 

[Stephen Webb]

2) The crop should avoid any of the potential softening of 1152 to 1080 and cropping should be much faster than re-scaling.

You're downressing from 4096x2304 to 2048x1152 already, and you're worried that going from 2048x1152 to 1920x1080 will soften the image???

Still, the idea of a metadata tag to crop to 1080 (complete with guides in the viewfinder) is a good one.

 

[Gavin Greenwalt]

This is a 1:1 crop of my experiment in photoshop. There's a candybar in it for the first person who can identify which got rescaled twice and which got rescaled once (with proof!). :gun:

Good luck.

 

[Häakon]

I'll play, since I feel like I'm the one who got us into this mess and I think people are really tired of going back and forth on this.

The one on the left looks much better to me... assuming you did the test in the same way we've discussed and there's no other variables going on, do I win a candybar or a nice glass of "you may now shut the f*ck up?"

 

[Peter McCully]

I'm a leftist too.

 

[Gavin Greenwalt]

Left is double sampled. But I'll send you a hershey's chocolate bar for trying. PM me your address.

Partly because I'm only 90% certain. I had a little jingle to help me remember but there is that nagging doubt the jingle wrong.

One way to know for certain... one sec...

Edit: Whoops. I lose. Left is 1 rescale. Choose your poison Haakon what'll it be? Crunch? Butterfinger?

 

[Häakon]

Hey, I was all ready to offer my humility if I stood corrected!

The one on the left is clearly retaining more resolution than the one on the right, which is why I chose it. The one on the right exhibits the exact kind of softness I was worried about in being sampled twice.

I would gladly concede ditching 1080 if my instincts were wrong, but this demonstrates exactly the fears I was having. And 2K is just not enough of a difference in resolution to justify wanting it simply because it's bigger on paper. I know this may not be what people want to hear (and granted, this was just one informal test), but at least I know that my educated gut isn't steering me wrong. Besides, Pete is a leftist too.

Feel free to bring my Twix to the LART.

 

[Gavin Greenwalt]

<errata>In 2k's defense I did use the "wrong" bi-linear for downscaling and there are algorithms that are tuned to down-rezing but they'll probably begin to introduce the almight sin all re-samplers bring: sharpening.

Speaking of which. Greame how can you guys claim to "not speak of sharpening" but then use any sort of rescaling algorithm? They all sharpen to some extent just usually very very very subtley.
</errata>

-- Twix it is.

 

[tj williams]

Jim
If we have built in crop to 1080 or 720 then we can also have SD all you have to do is give us another set of viewfinder scribe lines?

 

[Rob Lohman]

I'm sorry if this is a dumb question . . . . but isn't 4K RAW and 2K RGB the same thing? Since a square consisting of four RAW pixels actually consists of 1 red, 2 green, and 1 blue pixels, if you treat the four as one RGB pixel (perhaps taking the average value of the two green pixels), you end up with 2K RGB, right?

That would produce artifacts. You need to do a proper down sample or different debayer form.

 

[Rob Lohman]

Graeme, with the death of on-board RGB recording on the horizon, will you still pursue the quicktime RGB codec?

What do you mean by this? QuickTime codec will still decompress your footage to RGB/YCbCr. For the moment we are focussed on the decompressor side of the codec.

 

[krd]

Only problem is EFP and High-End Cine seem to be under represented. So.... ummm... Gibby gets his opinion multiplied by x10 and each of the beta tester's crews get x20.

There's another multiplier (or is it divisor?) here: independent film (or independent D-cinema) in the U.S. does not make money.

What keeps the "business" going is successive waves of aspiring filmmakers and financiers, hopeful or deluded, who are likely to buy or rent their Reds from previous waves of Red owners who've already failed. The 2000 or so(?) Reds already on order may be enough to supply the 800-1000 American independent features which get submitted to Sundance every year, for years to come.

Maybe Red should listen to Gibby, for the sake of continued cashflow and expanding market base.

 

[planet e]

i really don't get where you are coming from, alexander. not a single movie that you've worked on, from your imdb database listing, illustrates a need for 2k at 96fps, and yet you are the loudest proponent of dropping the bread and butter format of *my* business (which consists primarily of DVD and TV delivery), for some pie-in-the-sky that may (or may not, for all we know) be in your future. for you to make your dream martial arts film, you are willing to cheerlead for dropping the ax on dozens, perhaps hundreds, of thriving, existing businesses.

i'm not entirely averse to dropping 1080p RGB, if 2K scaled RAW is the alternative, but this characterization of 1080p as a "loser" format or as backward-thinking is the height of ludicrous. the vast majority of delivery systems *still* don't support 1080p output. until everyone in the country is tuning into their blu-ray player every night, until Apple starts supporting blu-ray output in DVD Studio Pro (duh, this should have been a no-brainer by now....), the 1080p is still largely the future. maybe when jim revolutionizes the nation's broadcast system (not being facetious here....) and, then that will change. that could take some time.

but shouldn't there be some testing of conversions or cropping or something, before such a radical move is made? this choice has the potential to break a long-standing promise to reservation holders to "make obsolescence obsolete" --a promise to make high-end HD accessible to lots of EFP and corporate shooters. i'm not saying this is the outcome, i'm saying this is *the possibility*. i think the options should be explored (thank you, haakon, for keeping this conversation civil and on-task....), before a crucial move is made.

if jim wants to make the camera that jim wants to make, that's his prerogative, and i have nothing but gratitude for the grand experiment and the amazing ride that he has taken all of us on. whether i complete this purchase or not, i have the utmost respect for the transparency he has brought to this process--it's truly remarkable and at least as revolutionary as the camera itself. but my responsibility, before i drop $55-60K on camera gear, is to make sure the final incarnation of this camera serves the purposes of my business. i guess LART will tell. i really want to complete this purchase. but i have to be sure that it is the best use of these funds, period. i don't have that kind of money to throw around, unless i can feel confident that it will return itself.

i have had nothing but confidence in RED, until this conversation started. now i feel quite uncertain about the direction this camera is headed and its efficacy for the type of shooting my studio does.

 

[Steve Gibby]

Firstly, I’d like to again congratulate RED Team on the unbelievable amount of planning, R &D, and development work they’ve been able to pull off on RED One – in a time frame that many simply thought impossible.

Secondly, I’d like to again say that I know each of the RED Team members and they have my utmost respect – and nothing has changed or will change that.

The outer layer of this discussion, to use an onion analogy, is the technology in question. But the inner layer, and the heart of the matter, is business. RED is a business, movie production is a business, television production is a business, all motion media genres are businesses, and each of us has a business to run.

In terms of pure volume of productions each year worldwide cine-style production is the tip of the iceberg, but the rest of the iceberg beneath the water is EFP (electronic field production). ENG is just a tiny sub-genre within EFP. EFP isn’t just B4 2/3” production. Vast amounts of EFP productions use cine lenses, accessories, and techniques – a natural market for RED. As a business, RED has generated a lot of promotional attention in the indie world (RED User and beyond) and features world (Jackson, beta testing, etc.), and yes RED One has been positioned from the beginning as a D-cinema camera that can also be used for EFP work, but I feel RED as a business would be best served by paying close attention to the cine market (the marquee market) and EFP market (the cash cow) as they finalize their camera system specifications. Its best for RED as a business.

What’s best for us as RED adopters? Thinking like a businessperson is what’s best. If we’re so intent on just working in a narrow niche of production, functioning as a “starving artist”, and then go bankrupt, how does that get us to our final goals? If we’ve already “made it” in the motion media industry, we know we have to fight every day to stay on top. This is a dynamic industry – every day we’re either climbing or dropping.

As RED One was initially described, and has been ever since, it was intended to be a scalable, utilitarian, flexible, and affordable system that could be quickly re-lensed and re-accessorized for use across multiple cine style and EFP style genres and sub-genres. Cutting edge technology enables RED One, and the potential result of that tech package is the enabling of massive revenue generating potential for RED One adopters (if they maximize the system’s potential) and RED One (if they keep feature sets that satisfy the needs of both the cine and EFP industries). That’s good business my friends – a win/win situation for everyone concerned.

RED User is a great board and an excellent central info source for “all things RED”. I love this board! That said, the professional demographic of the membership of RED User is heavily weighted with indie cine types, and not representative of the overall market for RED’s products (see iceberg analogy above). If you indie guys will open your minds, spread your skill sets to multiple genres of production, and thus generate exponentially higher revenue with your RED camera systems, you just might then enable your achieving your main goal of doing only high-level features. If RED will pay close attention to their largest market for the RED One system (EFP, not ENG), they will sell tons more products overall.

The days of just working in a narrow niche of production genres are rapidly coming to a close, enabled by two factors: technology and market. We all need to keep our ear to the rails – there’s a freight train approaching. It’s “jump on and ride” or “or keep walking and get mowed down”.

Now that we’ve established that it is business we’re talking about here, let’s specifically talk about the technology in question. I think what RED has proposed, with the scaled 2k RC RAW @60fps is awesome. I produce/direct/shoot across multiple cine and EFP genres, and a big portion of what I’ll shoot will be 4k RC RAW in camera, and the new 2k RC RAW scaled (if it makes the final spec). I shoot DSLR projects extensively and I’m used to a RAW workflow – I love it and I’m excited to use it with RED! That said, I also work in multiple genres of EFP production for national television and high-end business media. That’s the largest single industry market for me and for RED, and the rest of you guys if you’ll bring yourself to work in those industries. Now that RED has verified that scaled 2k RC RAW is possible and potentially in the spec, and we know that transcode and/or processing speed and ease of delivery to clients is the crux of the issue, I think that RED needs to pay close attention to a workflow that will satisfy the delivery speed to clients issue and we’ll all be fine. If the final solution to that conundrum entails a spec of scaled 2k RC RAW @60fps and scaled 1080p RGB @60fps, that will be a business decision that RED needs to make. Like wise, if a 1080p crop from a scaled 2k RC RAW frame is the final solution, then it’s workable. Quickest delivery to clients is the X factor that must be addressed. If it is – voila! I see some preliminary speed tests are already being conducted by RED User members – very cool, but more extensive speed tests need to be done and analyzed before hard conclusions can be reached.

Let’s heal the division that this “drop 1080p/720p” subject has raised. We’re all part of the overall motion media industry family. Outwardly this debate has been a tech discussion – but the core issue is that it is really a business discussion – and rightly so, because we’re all in businesses of various sizes. My position is that I simply want what’s best in a business sense for all of us, including RED, because a fiscally strong RED filters down to all of us adopters.

The entire RED Team has my deepest respect – these guys are actually doing what many thought impossible.

 

[redhead]

That would produce artifacts. You need to do a proper down sample or different debayer form.

Thanks.