Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

1080p out - a slippery slope

Andy Jarisz said:
At some point you have to decide what you are doing. Are you making a camera, or changing the industry?

If you're making a camera, then just give a 1080p out and be done with it. Satisfy everyone.
Click to expand...

Both.
Epic and Scarlet (scarlet via module) have 1080P HDSDI out so you can record via a Ki Pro, etc to Prores and other codecs.
 
Deanan said:
Both.
Epic and Scarlet (scarlet via module) have 1080P HDSDI out so you can record via a Ki Pro, etc to Prores and other codecs.
Click to expand...

Deanan,

THANK YOU!!!!

Somebody please tell Jay Kelly this so he can pop the champagne early this year.

David
 
Post Production.

Post Production.

RED CINE and RED ALERT were the only solutions in the early going.....
Click to expand...

That is (almost) true, but it also illustrates my point.

We have seen other camcorders languish for a year or more after launch with no NLE support other than via baseband HD-SDI ingest.

With REDCINE and RED Alert! and QuickTime proxies, then the .R3D file format SDK, RED has been able to bridge this "post gap".

I expect native .R3D support will become the norm across all aspects of post, but it's likely to take a while to complete that work.



The reason to provide a clean feed 1080p 4:2:2 signal as an available video output has nothing to do with the above.

It's been a common request that we could respond to. What you choose to do with it is up to you.

Which obviously includes using that output to record to an HD video codec if you so wish.
 
Stuart English said:
That is (almost) true, but it also illustrates my point.

We have seen other camcorders languish for a year or more after launch with no NLE support other than via baseband HD-SDI ingest.

With REDCINE and RED Alert! and QuickTime proxies, then the .R3D file format SDK, RED has been able to bridge this "post gap".

I expect native .R3D support will become the norm across all aspects of post, but it's likely to take a while to complete that work.



The reason to provide a clean feed 1080p 4:2:2 signal as an available video output has nothing to do with the above.

It's been a common request that we could respond to. What you choose to do with it is up to you.

Which obviously includes using that output to record to an HD video codec if you so wish.
Click to expand...


Stuart.

You are mistaking me for a guy who wants this 1080p feature really badly. I am the guy who wants to see the envelope pushed in .r3d and improving the RAW workflow. That is what I bought into and believe in.

IN my very first response to Jay A. Kelly, I said the following:

"I think RED is created around a Tapeless workflow with a RAW image capture at the center.

...

I'm not sure a move toward focusing camera resources on making the HDSDI a 4444 signal would be of great benefit, but it would fit the Swiss army knife camera we were expecting.

I respect the fact that you'd love to make more money with your camera but I am not sure the red is best suited to live 1080p recording. If I wanted that flexibility I'd buy an f900 but I do not really demand that of my red.

I want the resources poured into improving RAW. If we get a decent video tap that's a bonus. But thus is a digital film camera and I don't want things to stray to far from that core idea. "

David
 
Stuart English said:
That is (almost) true, but it also illustrates my point.

We have seen other camcorders languish for a year or more after launch with no NLE support other than via baseband HD-SDI ingest.

With REDCINE and RED Alert! and QuickTime proxies, then the .R3D file format SDK, RED has been able to bridge this "post gap".

I expect native .R3D support will become the norm across all aspects of post, but it's likely to take a while to complete that work.



The reason to provide a clean feed 1080p 4:2:2 signal as an available video output has nothing to do with the above.

It's been a common request that we could respond to. What you choose to do with it is up to you.

Which obviously includes using that output to record to an HD video codec if you so wish.
Click to expand...

Very elegantly put Stuart.

Yes - REDCINE and REDALERT - as well as the consecutive release of REDRUSHES - shows that you guys are more than just a camera manufacturer.

Hopefully people in post will further recognize the advantage of "look metadata" and the power of RAW - and not fall victim to the convenience of 1080p output. It definitely makes life "simpler" for some workflows, but ignores the other quality advantages of RED that are possible in a data workflow.

Advantages that are important to cinematographers.
 
Yes I am happy to hear this.. And while I am passionate about it, I sure hope people don't run around saying "Kelley hates RAW, all he want's is to record off the 1080p".

This would be a terrible thing to think. I shoot RAW on RED. I have a post workflow I love (It's not popular in some circles, but I love it).

The 1080p out simply allows me to say "yes" more often.. That's all.

Jay
 
Jay A. Kelley said:
Yes I am happy to hear this.. And while I am passionate about it, I sure hope people don't run around saying "Kelley hates RAW, all he want's is to record off the 1080p".

This would be a terrible thing to think. I shoot RAW on RED. I have a post workflow I love (It's not popular in some circles, but I love it).

The 1080p out simply allows me to say "yes" more often.. That's all.

Jay
Click to expand...

OK, how about if we say, SOME of Kelly's CLIENTS hate raw ;) JK!!

Congrats on getting your desired feature!
 
Dylan Macleod said:
Hopefully people in post will further recognize the advantage of "look metadata" and the power of RAW - and not fall victim to the convenience of 1080p output. It definitely makes life "simpler" for some workflows, but ignores the other quality advantages of RED that are possible in a data workflow.
Click to expand...

Well put, Dylan. I couldn't agree with you more.

I do have to point out though, that 4K, when scaled to 1080P, with color correction to follow, can look amazing. Before we aquired our Scratch, the most efficient, cost effective method to maintain an HD workflow was accomplished by playing out of the camera to an HD D5. It was via this process that we were able to successfully introduce RED to many of our commercial clients. Today, many of these clients are well versed in the data workflow process. Going back to a tape based workflow would seem unlikely.


Do I feel that a 1080 record option will cause others to ignore the quality advantages of RED and it's data workflow? Not necessarily.
As more & more producers are becoming savvy to RED, it's benefits and it's workflow options, offering a 1080p recording is merely one more opportunity to utilize this camera. For many potential users, it will now be a chance to raise the bar in terms of production value, as well as removing the potential "uncertainty" regarding the transfer of files.

As was wisely stated, it will be one more reason to say "yes".
 
Even the the good old (bad old?) days of 2007 there was a pretty nice proxy based option for post. If the editors were too lazy (to quote an earlier post) to 1) keep the r3ds in the same folder/drive as the proxies and 2) retain the original file names - is two rules too many, really? - then it worked pretty well two years ago.

IMHO the real problem was not laziness, lack of recognition of post issues by RED or glaring workflow gaps - the problem was that post production practices in most facilities were based on a different foundation. Film would be scanned to tape - DigiBeta, HDCAM (hopefully the SR variant), etc and then join video camera generated material in a tape based ecosystem. The steps in the dance were from tape to NLE to tape to color to tape (or some version of that) that fit the standard division of labor and use of dedicated hardware/software for specific tasks.

RED presented an unfamiliar format that resisted conventional treatment. I would argue that if I took a random sample of editors with creative skills and no technical knowledge, separated them into groups and trained each group in only one ecosystem - tape or data - that the data indoctrination would be easier. In the real world the population of tape savvy post personnel and the engineered infrastructure they interact with at their facility represents a huge investment in systems engineering and developed skill sets. RedCode RAW was the proverbial square peg / round hole scenario.

IMHO as more post personnel become familiar with data centric pipelines and more facilities are designed to efficiently handle them, Red generated footage will be widely welcomed. The only real question is how soon will that happen in your particular market. On a side note, when the economy picks up and the credit markets return to normal I expect a lot of post houses will make the upgrade and happily embrace the data centric world. Tape decks are high cost, high maintenance items and once producers get used to NOT paying for digitization/ingest hours they aren't likely to want to go back.

Cheers - #19
 
I've never had a problem with editors, but I've often had clients who won't even let me talk to the editors to figure things out! All the "just send us the tape" or "just burn it to DVD" horror stories you can imagine.

The one thing I would love is some sort of consistent metadata system. My professional workflow revolves around Clipfinder, RED Alert! and REDrushes. My one-light corrections create .RSX files, and they carry through my complete workflow.

The trouble is, other software ignores those .RSX files, and most use the old Color Science, so whether I've done one-light corrections, or just left the footage as-shot on Build 20, I can't pass those R3Ds to someone else and expect them to see things the same way if they are using a different workflow.

It's become clear that although the SDK was adopted into plenty of apps, they haven't done a good job of keeping up to date. I wonder if there is some way of creating a frameserver type program, something that RED would release for Windows, Mac, and Linux. RED could update it with the things like new Colour Science and 4.5K support on a regular basis, and make the SDK a simple little thing that would not need updating by the other software companies, but simply request frames from RED's program that you'd leave running. Something like AVIsynth, perhaps.

In a world like that, I would be much more comfortable passing R3Ds to other companies, and they would be much more comfortable working with them. That would eliminate all the cases now where I am delivering 1080 ProRes Quicktimes so I can be sure people can work with them, and have them looking the way I intend.
 
Post Production.

Post Production.

Mike, Blair and Stephen, those are three very insightful posts...
 
Dylan Macleod said:
Case in point is the Phantom. I did a Phantom shoot in Toronto where the director (bless him!) insisted we record the RAW data to hard drives - and not go directly to tape (which has become the norm in Toronto). The rental company had very little in the way of support for recording data only (i.e. Cinemags) and we were very close to being forced in to shooting to tape.

All of this because once editors found out that the camera could shoot to tape, they would insist DP's use this option. As a result the rental house had no motivation to invest in the gear required to support data recording. We would have had to ship Cinemags from New York - prohibitively expensive.

Like I have said - tail wagging the dog.
Click to expand...

Hey Dylan:

The real reason the phantom is being used as it is in Toronto is that producers got used to this method. Remember we had one of the first phantoms (actually bought from Vision Research, before Abel started to rep them.) There was no other option for the phantom at that point....the download process was a time killer on set. Producers got used to the cost of just dumping to tape....and editors got used to having tapes dropped off already to start with.

I dont think its been editors demanding the tape format, rather, production is not willing to spend the extra dollars on the rental of the cinemags. Those "prohibitively expensive" cinemags from NYC would just come from Toronto....and production has not shown that they are willing to pay for them. We have done over 400 commercials, of which 6 (that I know of....may have forgotten one or two) have gone to hard drive. Its really economics. (although I do like to blame the editors for every other problem in the world.....just not this one!)

Think outside the box for the 1080 output.....say a multicam shot on RED....Music concerts, even reality shows. Live switched...it does raise interesting ideas!

John DeBoer
Director of HD Sales
SIM VIDEO International
 
At the risk of hijacking the thread I want a module that gives me 4K live output over quad SDI for IMag. Imagine if you had the equivalent of a Red Rocket in a module on camera that had 4 SDI spigots that could be run to a 4K projection screen live.

It may not be necessary, or even desirable for a CEO speaking to a ballroom but it might be darn compelling shooting a Shakira concert :001_tt1:

Cheers - #19
 
Stephen Strangways said:
The one thing I would love is some sort of consistent metadata system...The trouble is, other software ignores those .RSX files, and most use the old Color Science, so whether I've done one-light corrections, or just left the footage as-shot on Build 20, I can't pass those R3Ds to someone else and expect them to see things the same way if they are using a different workflow.

It's become clear that although the SDK was adopted into plenty of apps, they haven't done a good job of keeping up to date. I wonder if there is some way of creating a frameserver type program, something that RED would release for Windows, Mac, and Linux. RED could update it with the things like new Colour Science and 4.5K support on a regular basis, and make the SDK a simple little thing that would not need updating by the other software companies, but simply request frames from RED's program that you'd leave running. Something like AVIsynth, perhaps.
Click to expand...

Stuart English said:
Mike, Blair and Stephen, those are three very insightful posts...
Click to expand...

Alright, you brought me out of my silent reading cave with the very intimation that RED might ever consider implementing what Stephen posted. I'd probably cry in public for it [tears of joy].
 
Stephen Strangways said:
The trouble is, other software ignores those .RSX files, and most use the old Color Science, so whether I've done one-light corrections, or just left the footage as-shot on Build 20, I can't pass those R3Ds to someone else and expect them to see things the same way if they are using a different workflow.

It's become clear that although the SDK was adopted into plenty of apps, they haven't done a good job of keeping up to date. I wonder if there is some way of creating a frameserver type program, something that RED would release for Windows, Mac, and Linux. RED could update it with the things like new Colour Science and 4.5K support on a regular basis, and make the SDK a simple little thing that would not need updating by the other software companies, but simply request frames from RED's program that you'd leave running. Something like AVIsynth, perhaps.

In a world like that, I would be much more comfortable passing R3Ds to other companies, and they would be much more comfortable working with them. That would eliminate all the cases now where I am delivering 1080 ProRes Quicktimes so I can be sure people can work with them, and have them looking the way I intend.
Click to expand...


ditto....well said
 
Hi guys,

please forgive my lack of knowledge. Does the RED1 can shoot scaled HD? I mean, without cropping the sensor?

Or do I have no choice than shooting quad HD, then downscale?

Basically we'll need to finish HD, and the editor is concerned about dealing with heavy file (no monster computer, tiny budget). But I am concerned about keeping RAW in the process!

So I want to shoot HD AND RAW. Am I dreaming? Cause I can see this specs on the Epic and Scarlet, but I'm not sure it's available on a RED1...

The perfect workflow should be : shot scaled HD, then edit & post .r3d files into CS4, then export HD master (prorez for exemple), then build the bluray.

Thank you for helping me to understand if I'm asking to much:wink5:
 
The RED ONE does not shoot scaled HD. That will only be on the EPIC and Scarlet.

Definitely shoot quad HD (the RED ONE calls it 4K HD), as it requires far less processing power when you downscale to 1080 or 720.

When you're editing it, you can always switch to 1/2, 1/4, or even 1/8 res debayer within CS4 to improve playback speed of the .r3d files, or you can make proxy files and edit with those.
 
Julien Deka said:
(snip)I am concerned about keeping RAW in the process! So I want to shoot HD AND RAW. Am I dreaming?(snip)
Click to expand...

IF the RedOne is your axe of choice then live output scaled HD is not on the menu. Best advice I have, is to get a RED Rocket™ in the mix - rent one or hire a DIT that has one. The Rocket does the heavy lifting so even a modest computer set up can manage your project. That said, be realistic - in our business it pays to have fast modern computers. Just sayin'

Cheers - #19
 
You must log in or register to reply here.
Back
Top