Red Cine X debayer options

[fredjeang]

Hi,

This is a question regarding the export from Red Cine X bin for editing in NLE

In the export window, I shall create a preset according to the output, wich I did.

But what I don't get very clear is the debayer settings options. (linked a screen capture)

Can somebody explains in easy words how work those debayer options?

Thanks.

[fredjeang]

An article in creative cow on that matter brought me more confusion than light: http://forums.creativecow.net/thread/242/1343

[bcooter]

Hi,

This is a question regarding the export from Red Cine X bin for editing in NLE

In the export window, I shall create a preset according to the output, wich I did.

But what I don't get very clear is the debayer settings options. (linked a screen capture)

Can somebody explains in easy words how work those debayer options?

Thanks.

Fred,

I think you have to realize from that link that the respondents are editors.  Good editors, but non the less editors first.

Most indie, business, commercial editors don't really care about the format they receive as they consider themselves editors first, colorists, conformists, effects specialists way down the line if they provide those services at all.

In fact most editors do not want to be concerned about anything but editing the story.

It really depends on what you shoot, what you believe (at the moment) your client needs today and will need in the future.  Is 4:4:4 necessary.  Not really, unless your laying in a lot of green screen or effect work and depending on who you outsource this to, then it can make a difference, though it could easily be that you do a half debayer for the main cut, a full debayer to send out for effects that will be converted down to fit the edit.

Really this highlights a lot more than just 4k, 5k, 4::4:4, full debayer editing.

Most people cannot see the difference in high end compression, some can depending on what you shoot.

This is all a moving target in the world of people that sell equipment, software and new services, but in the real world, it probably doesn't  matter that much as 99.999999% of all motion imagery either goes to the web, or business projection with some kind of 2k projector.

I get about 20 e-mails a day offering new software, equipment, etc. and I know that the majority of editorial houses don't run out and buy the newest annouced every time it's offered.

Not everyone is shooting cinemax quality, including the major motion pictures.

A lot of this talk comes from the fact that the RED is one of the few semi affordable raw motion cameras.   If there was a smaller, faster, lighter 2k raw shooting camera, we probably would not be having this discussion about the RED as well as dslrs, a motion jpegs, or hdv 2k prosumer cameras.

This make you wonder where the industry is going.  It seems to be consumer, prosumer, the high end prosumer/eng, then the RED/Arri.   

There is a huge hole in the market for easier to use raw imagery capture cameras regardless of resolution.

IMO

BC

[BFoto]

Fred,


There is a huge hole in the market for easier to use raw imagery capture cameras regardless of resolution.

Which is why everyone was waiting for the RED scarlet.

[fredjeang]

What really catch my attention is that even people in the cine-video spheres are also confused to some extend. I thought that because I came from still image and starting to learn video from zero I was logically confused and that the video gurus have it all under control but in fact the all picture in much more messy than I thought and managed to put almost everyone into different degrees of confusion.

In the end the job is done but I have a sensation of being immerged into a new lenguage who's tools and technical rules are not yet settled and probably the today's workflow will be changing in a matter of months wich makes the all process a non stop self-actualization with an overall sensation of that even the industry is making endless prototypes without really knowing where they go.

It reminds me a lot the early days of web html and flash and even today we have to deal with lack of standarts and unsecure procedures. For example I just saw yesterday in the I.pad that it does not read properly certain javascript that are basics for years when I wanted to put my user-password in a web I'm afiliated to, and it wasn't flash.

I think that James hit the point in another thread when saying that in the end it is probably more important to find a workflow from prod to that works for one, stick with it and forget about the pp because it seems that there is nothing stable to relly on.
I really think this is the best approach in the sense that it brings the stability that is lacking everywhere else in the process.

In short, if there is a mess of standards, make yours that work for you and your clients. Indeed.

[canmiya]

What really catch my attention is that even people in the cine-video spheres are also confused to some extend. I thought that because I came from still image and starting to learn video from zero I was logically confused and that the video gurus have it all under control but in fact the all picture in much more messy than I thought and managed to put almost everyone into different degrees of confusion.

In the end the job is done but I have a sensation of being immerged into a new lenguage who's tools and technical rules are not yet settled and probably the today's workflow will be changing in a matter of months wich makes the all process a non stop self-actualization with an overall sensation of that even the industry is making endless prototypes without really knowing where they go.

It reminds me a lot the early days of web html and flash and even today we have to deal with lack of standarts and unsecure procedures. For example I just saw yesterday in the I.pad that it does not read properly certain javascript that are basics for years when I wanted to put my user-password in a web I'm afiliated to, and it wasn't flash.

I think that James hit the point in another thread when saying that in the end it is probably more important to find a workflow from prod to that works for one, stick with it and forget about the pp because it seems that there is nothing stable to relly on.
I really think this is the best approach in the sense that it brings the stability that is lacking everywhere else in the process.

In short, if there is no standarts, make yours that work for you and your clients. Indeed.

Can you give some examples as to the kind of "standards" you are looking for/talking about?
Regards,

[fredjeang]

We have a 1.89:1 as a standard (so for 4K it would be 4096x2160, for 2K 2048x1080).

Then we have other standards: 2048x858, 1998x1080, 4096x1716, 3996x2160

RED files shoot: 4520x2540 (1.77), 4480x1920 (2.3), 4096x2034 (2), 4096x2048 (2).
None match with the DCI specs. Normally 30fps at full res for Red and progressive.

I think that Avid is the only NLE, the MC 5 that can actually support the RedRocket card, if I'm wrong please correct me.

I'm not sure about the FCP with native Red files wich is somewhere a non-sense (if it can't but again I'm not sure, I do not use FCP) because FCP is a "standard" in many prods but if you're lucky enough to get Autodesk Smoke then you're safe with latest version.
Does Smoke runs with other NLE ? it does not to date that I know. It's a modern powerfull complement married with an outdated NLE and we don't know what Apple will do.

Why do we still have interlaced ? Does that make any sense today ? But we do have to deal with it. One standard interlaced, the other one progressive.

ARRI is FF 35mm at, (arriraw), 30fps wich is to me more into the tendency we will need because I beleive like BC that we'll have, or we'll need squared standards very soon because of the evolution of medias displays and distribution and because of the need to get one material for both stills and motion.
The good think about the PL mount is that you can mount via an (expensive, about 300-400euros) adapter to Canon 35mm or even Pana GH (but then remember the sensor is smaller)

In all that simplicity and field of clarity you can create a funny frame composer in Arri website for ex according to the jungle of formats that can be required. That's very practical but it also mean that you actually loose part of the sensor recorded in the area virtually cropped.

Red algorith is not lossless (although I think it can be considered lossless) while Arri is, all the other shooting codecs that aren't RAW aren't lossless. It means that the degradation has to be stopped at the stage of editing.

The obliged debayer path for raw video is tunable according to where the footage has to go on the chain, so again, convertions and convertions and more convertions at any stage and that is true whatever the system is.
One for shooting, convertion(s) for editing and zillions possibilities for the delivery.

Europe broadcast is 25fps, US is 30 cine is 24 or 48  and they want to go to 60. Why "or"? and why this silly 24 that belongs to the stone age?

That means that you have to have a master edited and then dispaches to this and that. More rendering on the chain.

Broadcast compress to standards, cine is flexible and adaptable. Broadcast cine-orientaded is full hd while sports-orientated is 720.

I do not have 2 technical sheet for clients that are the same in terms of requirements except the size and generally the sound bit but not always. That is why I'm going crazy and I'm not the only one.

In the case of an inter-frame compression wich it's been regarded as the new holy grail because it has a higher compression, if for a reason or another, the single frame is lost due to a tech issue (and it happens), the all group of frames that belong to it is lost. I'm not sure every expensive prod is willing to take risks on such a fancy stuff and actually they do not.

Not 2 NLE handle the same way the AVCHD editing (wich is absurd editing in this format). Avid is truly workable and Premiere much slower (at a same computer performance).
Converting to a lossless editing is different if you are on Mac or PC, and there are too many choices. So yes, it's stick with one like prores but not always you can export with prores. It's like you won't be able to export a Tiff on a PC.
At first the 5D2 only had 30fps, (europe is sometimes difficult to find on a map fron Japan) then they finally released a 25fps and joined the SMPTE time code standards not without vacillations.

Etc...etc...etc...

It seems that every brand has its own vision, it's own solution, while standards are been established in a great confusion that probably will not match soon anyway the needs of publishers with the new medias and will be again in another mess.

http://www.dcimovies.com/press/

Please, if I made tech mistakes in those lines I'm writting while I'm actually render with a fancy codec that will probably be obsolete tomorrow morning, correct me.

I'd just like a little bit more simplicity.

[fotogurl]

We have a 1.89:1 as a standard (so for 4K it would be 4096x2160, for 2K 2048x1080).

Then we have other standards: 2048x858, 1998x1080, 4096x1716, 3996x2160

RED files shoot: 4520x2540 (1.77), 4480x1920 (2.3), 4096x2034 (2), 4096x2048 (2).
None match with the DCI specs. Normally 30fps at full res for Red and progressive.

I think that Avid is the only NLE, the MC 5 that can actually support the RedRocket card, if I'm wrong please correct me.

I'm not sure about the FCP with native Red files wich is somewhere a non-sense (if it can't but again I'm not sure, I do not use FCP) because FCP is a "standard" in many prods but if you're lucky enough to get Autodesk Smoke then you're safe with latest version.
Does Smoke runs with other NLE ? it does not to date that I know. It's a modern powerfull complement married with an outdated NLE and we don't know what Apple will do.

Why do we still have interlaced ? Does that make any sense today ? But we do have to deal with it. One standard interlaced, the other one progressive.

ARRI is FF 35mm at, (arriraw), 30fps wich is to me more into the tendency we will need because I beleive like BC that we'll have, or we'll need squared standards very soon because of the evolution of medias displays and distribution and because of the need to get one material for both stills and motion.
The good think about the PL mount is that you can mount via an (expensive, about 300-400euros) adapter to Canon 35mm or even Pana GH (but then remember the sensor is smaller)

In all that simplicity and field of clarity you can create a funny frame composer in Arri website for ex according to the jungle of formats that can be required. That's very practical but it also mean that you actually loose part of the sensor recorded in the area virtually cropped.

Red algorith is not lossless (although I think it can be considered lossless) while Arri is, all the other shooting codecs that aren't RAW aren't lossless. It means that the degradation has to be stopped at the stage of editing.

The obliged debayer path for raw video is tunable according to where the footage has to go on the chain, so again, convertions and convertions and more convertions at any stage and that is true whatever the system is.
One for shooting, convertion(s) for editing and zillions possibilities for the delivery.

Europe broadcast is 25fps, US is 30 cine is 24 or 48  and they want to go to 60. Why "or"? and why this silly 24 that belongs to the stone age?

That means that you have to have a master edited and then dispaches to this and that. More rendering on the chain.

Broadcast compress to standards, cine is flexible and adaptable. Broadcast cine-orientaded is full hd while sports-orientated is 720.

I do not have 2 technical sheet for clients that are the same in terms of requirements except the size and generally the sound bit but not always. That is why I'm going crazy and I'm not the only one.

In the case of an inter-frame compression wich it's been regarded as the new holy grail because it has a higher compression, if for a reason or another, the single frame is lost due to a tech issue (and it happens), the all group of frames that belong to it is lost. I'm not sure every expensive prod is willing to take risks on such a fancy stuff and actually they do not.

Not 2 NLE handle the same way the AVCHD editing (wich is absurd editing in this format). Avid is truly workable and Premiere much slower (at a same computer performance).
Converting to a lossless editing is different if you are on Mac or PC, and there are too many choices. So yes, it's stick with one like prores but not always you can export with prores. It's like you won't be able to export a Tiff on a PC.
At first the 5D2 only had 30fps, (europe is sometimes difficult to find on a map fron Japan) then they finally released a 25fps and joined the SMPTE time code standards not without vacillations.

Etc...etc...etc...

It seems that every brand has its own vision, it's own solution, while standards are been established in a great confusion that probably will not match soon anyway the needs of publishers with the new medias and will be again in another mess.

http://www.dcimovies.com/press/

Please, if I made tech mistakes in those lines I'm writting while I'm actually render with a fancy codec that will probably be obsolete tomorrow morning, correct me.

I'd just like a little bit more simplicity.

I am new here, but from where I m sitting, a significant part of the problem is too many people picking up video and trying to do everything themselves, just as they are used to doing with stills. We are going to end up with a bunch of "jacks of all trades, and masters of none." 

Just as it is with stills, there are different video cameras for different purposes.  Some of the resolution issues are easily figured out by looking  end use.  From a business perspective, some of the standards you are referencing would require open architecture:  How would this be beneficial to companies like Red, Arri, Panavision, et al?

[fredjeang]

I am new here, but from where I m sitting, a significant part of the problem is too many people picking up video and trying to do everything themselves, just as they are used to doing with stills. We are going to end up with a bunch of "jacks of all trades, and masters of none."  

I do agree. But the cake is served and a part of it already eaten.

Yesterday I was going to a good friend's opening in this gallery http://www.galeriaesquina.com/html.php/exposicion_actual and on my way there was a demostration covered by many medias. What caught my attention where 3 things: 1 is that there where much more medias as usual, 2: crews where reduced even for big chanels, 3: there was woods of 5D2 filming that we could reforest the amazon and the old still guard of always with their 1D as usual, the overweighted video pros with their fossilized sonythomsonandyouaddtherest looking at the 5D2 mess with perplexity and envy. Then on both sides of the avenue, thousands of passers-by with their colored compacts immortalizing the event and a few dodgers infiltrated in the pros with their multifunction rangefinders digital replicas... the demostration was convoked through I.phones and Blackberry gadgetery and did an unexpected snowball. On morning, I could watch bad but authentic pictures and video coverage of the event on the I.pad and on recycled paper free newspapers that are now the most read after of course the latest of Real Madrid football team.
http://www.20minutos.tv/video/cmpqFWY5-batalla-campal-en-el-centro-de-madrid/0/  for example.

[Graeme Nattress]

The DCI spec is for a container - depending on the aspect ratio you want to use, you fit (with padding) your image into that container. There is no technical necessity to shoot the exact resolution of an output container, and indeed there are many advantages to not do so. For instance, in Social Network, Fincher shot wider than necessary so that he could post-stabilize every shot (think the antithesis of shaky-cam) and he shot RED so he still had more than enough resolution for a really nice image on the cinema screen.

Graeme