The Camera's incredible! The Workflow documentation Sucks!

[Pierce Cook]

Take a breath, it's so easy that my kids are doing it on there Mac's and one of them is only 11.

Please write a book on good parenting. This is awesome (I'm not a parent, and don't plan to be for a while, but, cheers!).

Also, good to know for those of us who haven't yet had the opportunity and are or have been overly concerned about workflow, even with a good/great system.

p.s. I hope you can't get thrown off this forum for saying "piss easy" too many times, is that regarded as swearing in the US?

While a lot of people here in the States can't deal, I think REDuser is one of the last places you'll find those people! We're pretty open minded here, haha...

 

[RivaiC]

But Chris, when HD first released, other 3rd developers are slowly catching up too. There's no native solution like when F900 first released. I dont see how is it any different from RED. Yet, it's still true that there's camera that used existing format and have it a go. Both cases happened.

But since you tackling comment about Arri stuff, ok..i can accept it, just not generally. RED work really hard to open support big platforms right now and it's not something to be taken lightly. These are all due to open development of RED, just hope we dont take it on the wrong way.

 

[Johann Schulz]

Again, the need to wallow through the dirt gets ya muddy.

Brandon Kraemer made a great RED WORKFLOW document available here on reduser. I'm not posting the link because I hope you can use search and find it.

Search for me hasn't been up to par, adding to confusion and frustration, but it does seem to be finding it's feet. I tried searching for '7in lcd' and nothing came up. Maybe '7 inch lcd by RED'? It's a crap shoot, and we posters don't make it any easier by continually posting in the wrong forums.

Someone mentioned time. Yes, time takes. ..."And it won't wait for me"....

 

[Chris Pickle]

But Chris, when HD first released, other 3rd developers are slowly catching up too. There's no native solution like when F900 first released. I dont see how is it any different from RED. Yet, it's still true that there's camera that used existing format and have it a go. Both cases happened.

But since you tackling comment about Arri stuff, ok..i can accept it, just not generally. RED work really hard to open support big platforms right now and it's not something to be taken lightly. These are all due to open development of RED, just hope we dont take it on the wrong way.

Yes other vendors are catching up, and that's a good thing. I'm sure that Red doesn't want to build a stand alone post solution and instead have their format supported by others. Most agree that that would be the best.

I think Sony supported the 900 right out of the gate with their proprietary platform. It was clunky and expensive but it existed.

I don't really want to sound like I take exception with Red although I think they could find a way to compile the latest post info for easy referral. I know it changes fast, but I still think it would be doable.

Post with Red can be easy or it can be a bear, and as mentioned there is more than one way to do it, but none of us should be surprised that there are some very bright people that don't yet understand it.

Many teenagers know every word to every Green Day song, that doesn't make them smarter than the NASA scientist who doesn't know how to start a lawnmower.

The next new person that arrives on this board asking for post help because they find it terribly confusing might be Walter Murch.

Cheers,
Chris

 

[David Wyatt]

The fact is that the Red workflow really is as simple as Big T says if you want to keep it simple and maybe not have maximum quality (if you want that you just have to jump through a few more hoops). If you're going to standard def, as most people are day-to-day, then the "piss easy" workflow (as it will be known henceforth) is acceptable, if you're going to print on the other hand then it's definitely worth jumping through those extra hoops which is going to become easier and easier as more 3rd parties get involved with the release of the SDK. All in all it's not like we're building the Large Hadron Collider from scratch here (now that would take some research!! :wink

 

[Jon Farhat]

I don't usually post much, unless something really needs clarification. First of all, please stop being so harsh to people who ask serious questions. The people at RED are always respectful to their clients, and people on this forum should be too. Especially those 'defending' RED.

Next, Moggy here asks a question in a very thoughtful and respectful way, and frankly his point is true. Documentation is a 'huge' thing and this book won't be written for a while.

Until then, this is all I know...

1. Dropping the proxies in FCP and pressing Render, is fine. But until Apple supports the REDCODE as a sequence setting, "pressing RENDER" is a recurring thing as you edit.

2. REDrushes is the best thing I've seen so far. It's a front end which drives the REDAlert command-line codeset called REDline. Type REDline in your terminal window after you install the latest package of REDAlert.

3. Best quality Apple ProRES dailies I've seen, right out of REDRushes.

4. Are you working in Film? Have High volume, then you can work their DeBayer routines right into your workflow network. I like to set a color and create my 2k masters (dpx) and automatically create 3 other ProRES and H.264 versions for editing and logging. Here's a working tutorial. http://jfi.net/RED/REDline.php

Moggy, this workflow thing is evolving faster (and for the better) than we all can imagine. I share your excitement for the camera, it is simply stunning.

best.
j.

Hmmm, ok so this perfectly illustrates what I'm talking about. Two different opinions on whether exporting ProRes from an FCP proxy timeline gives you highest quality or not. A basic question that surely everyone who is editing with RED at any level needs to understand, and these are people have been using the camera for a while. I've only had mine for a week.
So what's the right answer? I bet someone at RED knows. Bueller? ....Bueller?

 

[Christopher Nagel]

my kids are doing it on their Mac's and one of them is only 11.

Will you adopt me?

Now I know how my teachers felt when I was developing color film and making Cibachrome prints when I was 11. Nice to have your dad "in the biz"!

Chris

 

[Mark Toia]

Sorry all,

I wasn't looking down at anyone, Was not trying to be condescending.
My Appogize if some of you took it the wrong way.
I was just trying to dispel the fair and loathing of RED and make it as easy for everyone as it is for us.

Ok..

TV AD's and FILM DI's are two completely different procedures. FACT

I'm sharing a TVC work flow only! that works perfectly, flawlessly that gives max information if done properly. FACT (cause we do it daily)

TV ad's that get finished in HD or in most cases SD don't need to go down the painful passage of RENDERING out miles and miles of rushes that can take hours upon hours, then go through another painful process of making DPX's and so an so on..

We have developed after months and months of R&D the fastest RED TVC (NOT FILM DI OUT) workflow known to man kind. (Actually, more known to RED)
We get the same range out of the R3D's as any other long winded process and I will prove it to any POST HOUSE GURU that begs to differ.

We do use RED ALERT when we need to suck more information out of a badly exposed image and we maybe once per ad output a single shot via RED ALERT RENDER (our settings) if the DOP has completely screwed the exposure up. (his eyes must have been closed).

In short, it is as easy as just dropping the QT's into the time line, edit it and when finished, RENDER out a PRO RES final from you time line.

The secret is make sure you are in REC 709 in your camera to get the best results. In other words shoot it like film not like video. Shoot for max information, Stop trying to trick the cameras ISO's . Just shoot it at either 250 or 320. If it gets dark. Use brighter lenses. At the end of the day, If you shoot crap, you get crap, same rule applies with film.

Now.. Knowing how to set up RED ALERT (our settings) so as to get and image to match the one that ends up in FCP and knowing how to save a shot that has been over exposed through RED ALERT. (our settings) is the key to making this all work.

I have not got time to share this very simple and super quick work flow, But I will team up with the FXPHD dudes and http://www.fxguide.com/redcentre and share the knowledge.
Cause I'm sick of hearing about all the pain and all the long winded processes out there.
Way to many DOP's and Directors are ringing me up saying, how are you doing it so easily.
I say, its not your camera, its your POST.

Like I said, My kids are doing it without a single problem.

I have just showed a very prominent POST HOUSE how do it, There jaws dropped, They saw our simple settings which intern solved a lot of issues for them. They are now cutting an AD up for one of our directors as I speak with ease and grace. No rendering, NO DPX's, flying!

I've uploaded an AD that never went near a single DPX, tiff nothing. All finished in FCP, A few shots went back to red alert with new QT's updated, Edited, graded in FCP.
The only thing rendered was the final 60 second edit on the time line to PRO RES 10 bit (HQ).(which took about 3 minutes). Ad was put out to HDSR ready for cinema and TV.
Ooops sorry I lied, we had a small bit of compositing done in INFERNO (removing a sign post) then put it out to tape.

3 hours of footage injested in 3.5 seconds, cutting in 2k (low video playback in FCP) 5 seconds after that. Finished cutting RENDERED the AD in 3 Minutes to PRO RES, put out to tape . 60 seconds

"PISS EASY" is a good name for this process.

Here's the ad.

http://www.zoomfilmtv.com.au/ftp/RED_EASY_2k_1K_WORKFLOW/

Talk to you all soon.

Keep your eyes open in RED FCP workflow forum for "PISS EASY"


cheers

 

[hanselblat]

Well done, Big T

 

[Paul Didier Mogg]

Well Big T,
Not to be too blazé about it, whistling: Oh what the Heck!!) ...but as a newcomer I'd already discovered what you call your "Piss Easy" workflow several days ago and posted it on the Apple Workflow forum under the title "Quick Event Footage Workflow?", asking if this was a valid way to do things quickly. Nobody answered, (and still haven't to this moment) So I posted here with a more dramatic title to get a response.

I guess people thought, just like me, that it was too simplistic and quick an approach to be viable. After reading the tons of very long workflow information that's out there on the forum it's easy to reach that conclusion. Now someone such as yourself who's been doing this for a while commercially says it's actually THE way to do it, (for certain jobs of course).

So why do you think it's so much quicker to render ProRes from an FCP timeline than from RedAlert?

I guess all we need now is to know the "Super Secret Sexy™" settings (SSS) you use in RedAlert to get comparable color output from FCP and we're all set!

So if anyone else is doing this with success, I'd be grateful if you'd share your RedAlert settings guidelines to achieve this. I'm sure it's partially a dark art .. but it would be very helpful to us newcomers.

Many thanks, (and I really didn't take your post as condescending!)

Moggy :greedy:



p.s. here's my post from the Apple Workflow forum from a few days ago for reference.

Hi Folks,
I'm new to the RED workflow possibilities and trying to sift through all this information to figure out what would be the right workflow for the actual jobs I have to do on a regular basis.
So... if I've shot a 2 hour event that requires little editing, requires SD DVD and 1080p master delivery, and has to be output as quickly as possible (like the same day), what's right and wrong with the following workflow?, and specifically, what quality am I getting on output by doing it this way?.

1) Shoot 4k 2:1 to get best initial captured quality camera allows.
2) Load footage into RedAlert, make color and look adjustments, then "Make Quicktimes".
4) Load _M proxies of footage into FCP (on MacBook Pro) and edit, accepting sequence settings FCP offers.
5) Export edited timeline using Quicktime Conversion (or Compressor) as 1080p ProRes HQ, letterboxed to preserve 2:1 aspect ratio. This is my master.
6) Use this master to produce my SD DVD.

I've tried this and it works, but what quality am I getting? As I'm editing Proxy files that access the 4k, when I output 1080p from the _M quality FCP timeline am I getting a high quality conversion derived directly from the 4k .R3D or not?

Also, rendering from the FCP timeline to Prores HQ seems to take only 5 times real time (approx), as opposed to 25 times real time if I convert to ProRes upfront from RedAlert in order to edit ProRes directly in FCP. This seems too good to be true and leads me to believe all is not as it appears. Please enlighten me as to what's actually going on.

Many thanks upfront for your input on this.

Paul

 

[Kenn Michael]

In my experience, rendering from the _H proxies (2K from 4K footage) uses the Standard Debayer (1/2 Res) which looks very good unless you've underexposed. The noise print is noisier in 1/2 Res Standard Debayer. This is the 5-6 times real time figure.

Rendering from Red Alert defaults to Full Debayer (although you can change this). Full Debayer is slightly softer than the 1/2 Debayer, but the noise print is smoother. This is giving you the absolute full best quality picture that Red can offer. I get 16 times real time on my 8 core Mac Pro.

Either choice, full debayer or standard debayer (half) works if it works within the context of your final master. _M quality is 1/4 debayer or draft (please correct me if I'm wrong here Graeme). Personally I would make my SD masters from the _H proxies if anything. _M for me is strictly offline - although it really does look very nice.

Sometimes I like the crispness of the standard debayer at 2K. Sometimes I absolutely need the full debayer and it's smoother noise print. It all depends on the project and the footage.

Many ways to skin a cat with this Red footage. If you're mastering at SD or 1080, if you don't have to push the image around that much using the FCP proxy/timeline render from 2K method can achieve beautiful results as Big T demonstrated with that great ad.

If you've underexposed and need every last bit of data from the RAW for your color correct outside of Red Alert, then DPX is the way to go.

Anyway, enough rambling. Hope some of this info helped.

Why am I still awake?

 

[Kenn Michael]

Also - all this is assuming you're strictly shooting 4K.

When you go down to 3K or even 2K, you have to pay attention to your proxy sizes. Using an _H proxy for 2K slo-mo is going to look pixelated because it is half the resolution - 1K. You'll have to use the _F proxies for that and for the 3K to match the quality of the _H 4K proxies.

If you're rendering Full Debayer from Red Alert, Red Rushes (which I LOVE) or Redline, the quality will scale properly for the smaller shooting modes. Doing the full transcode from these apps will also make all your shooting sizes common to your output format - ie 1080 or 2048x1024, etc...

 

[Charles Angus]

This is an incredibly unfair comparison! And actually untrue!

Arri and Aaton don't post workflows because their shooting format is not a brand new proprietary codec. It's an industry standard exclusive of their camera.

Sony certainly does publish workflows, and actually Sony built several systems to match their cameras. Panasonic and Canon introduced new cameras but with mostly existing workflows. i.e. DVC Pro on FCP and HDV which was quickly supported by Apple.

Red invented a brand new format out of thin air that couldn't be used on ANYTHING that currently existed. But you think they are not responsible to guide or be helpful in post?

Hey, here's a car that uses a special gas that has yet to be invented, but you're on your own as to who will or when invent it.

Yes I'm exaggerating a lot, and Red has certainly been very helpful in building tools and trying to get others to embrace their format, but RedRaw is not anything else. Yes that's good, but it can't live in it's own little vacuum with its fingers in it's ears.

I don't know why people that understand the current workflows are often so mean and condescending to people that don't yet understand. Is it really unfair to think that a set of white papers, even steadily evolving ones, are too much to ask.

I know lots of very bright people that are slow to grasp this new paradigm. I don't for a second look down at them.

Cheers,
Chris

HDCAM is a perfect example of how Sony invented a proprietary format "out of thin air", and then refused to offer the slightest support for it. HDCAM 24psf is only 144mbps, and should be editable on a modest desktop system (just like DVCPro HD 1080p at 88mbps), but Sony doesn't like that idea and so you can't do it. Your only options with HDCAM (or HDCAM SR, actually) is to ingest into your editor as uncompressed.

Is that a smart workflow? I've done it, and I gotta say, even the mild compression of SR starts to look good after uncompressed.

AFAIK Panasonic's AVC Intra is a similar story - you can go uncompressed or ProRes, but no native support.

RED has a basic offline/online workflow. Granted, there's a few hitches, but for those coming from film/high end HD it's not much of a stretch. For those used to editing HDV on their mac and onlining from square one, it's different. But, then, I'm sure using real lenses and needing real tripods and AC's and all that jazz is new to them too. Get your feet wet and get learning... (not directed at anyone in particular)

 

[Paul Didier Mogg]

Great information, thank you all!

So, is it possible to edit _M proxies if you have a slow machine, conform to _H proxies, then output ProRes, to get the higher quality. I guess I should just try it.

Moggy

 

[Lucas Wilson]

Hi "Moggy,"

It sounds like you want an all FCP workflow. If that's true - cool - and the people in this thread are giving you great advice.

But we have a lot of customers with very smooth and elegant workflows using SCRATCH in conjunction with FCP. If you're interested in those, please contact me offlist at lucas (at) assimilateinc (dot) com.

Best,

Lucas
------
ASSIMILATE, inc.
LA, CA, USA

 

[M Most]

HDCAM is a perfect example of how Sony invented a proprietary format "out of thin air", and then refused to offer the slightest support for it. HDCAM 24psf is only 144mbps, and should be editable on a modest desktop system (just like DVCPro HD 1080p at 88mbps), but Sony doesn't like that idea and so you can't do it. Your only options with HDCAM (or HDCAM SR, actually) is to ingest into your editor as uncompressed.

That's not quite the way it happened. When HDCam was released, the codec was licensed to both Discreet and, of course, Sony, as a software codec. There were, in fact, a few builds of Fire that had HDCam software codec support. There was also the Sony Xpri editing system, which had software and hardware support for native HDCam work. The primary reason it didn't take off is probably because hardware at the time (we're talking late 1990s here) was not very capable of doing a real time HD decode in software without some hardware assistance. Hell, it was barely capable of playing HD (uncompressed) in the first place. Since the alternative of simply using the HD-SDI feed was available, and tied in much better to online editing environments and the software driven devices at the time, the idea of a software codec for HDCam never really gained industry traction. There was never any real move to introduce it on things like PC's and Macs because they were not generally used in professional post facilities, which is where the HDCam machines were (and generally still are). The notion of doing HD work on a desktop computer really didn't take hold until Apple announced its support of DVCPro HD in 2004 - much more recently than most people here realize. And it took at least another year before uncompressed work was supported on small platforms.

 

[Mark Toia]

So why do you think it's so much quicker to render ProRes from an FCP timeline than from RedAlert?

I guess all we need now is to know the "Super Secret Sexy™" settings (SSS) you use in RedAlert to get comparable color output from FCP and we're all set!

So if anyone else is doing this with success, I'd be grateful if you'd share your RedAlert settings guidelines to achieve this. I'm sure it's partially a dark art .. but it would be very helpful to us newcomers.

Ok... I'm in the middle of a shoot so can't ramble on for to long...
You don't need to render anything out in RED ALERT once I give you the (SSS) settings to balance out your images, unless you have very bad bad image your trying to save. Other than that you will just make new quicktimes. They are instant. FCP will update your new media automatically.

It's not a dark art, its just lots of hours going through hundreds of different scenarios till I found the correct settings.
Its all there ready to go.
I will share the settings with you VIA FXPHD once I have time and once I figure out a easy way to explain it.

all will become clear very soon.

 

[Mark Toia]

Great information, thank you all!

So, is it possible to edit _M proxies if you have a slow machine, conform to _H proxies, then output ProRes, to get the higher quality. I guess I should just try it.

Moggy

yes... When you have finished your edit, you go to sequence settings, make your new HD size, reconnect media and select your H quick times, POOOOF!
Magic, all done, RENDER out your project to APPLE PRO RES.

 

[David Nardini]

RED Workflow section on RED One Wiki ?

RED Workflow section on RED One Wiki ?

There really does need to be one (simple) location where workflow options are noted, documented (and kept up to date) !

While RED are busy producing such great kit, why not put introduce a 'workflow' section at the following (in the spirit of collaborative knowledge sharing) :

http://en.wikipedia.org/wiki/RED_Digital_Camera_Company

Would save (time) having to mine all the info out of RedUser.net ... which is a great forum, but with the passing of time, it's getting harder to find consolidated info, options, best practices, etc ... :whistling:

EDIT : BigT ... I've started putting the section headings in the above Wiki site ... feel free to fill in (included your 'get going' instruction)

 

[Chris Pickle]

HDCAM is a perfect example of how Sony invented a proprietary format "out of thin air", and then refused to offer the slightest support for it. HDCAM 24psf is only 144mbps, and should be editable on a modest desktop system (just like DVCPro HD 1080p at 88mbps), but Sony doesn't like that idea and so you can't do it. Your only options with HDCAM (or HDCAM SR, actually) is to ingest into your editor as uncompressed.

Is that a smart workflow? I've done it, and I gotta say, even the mild compression of SR starts to look good after uncompressed.

AFAIK Panasonic's AVC Intra is a similar story - you can go uncompressed or ProRes, but no native support.

RED has a basic offline/online workflow. Granted, there's a few hitches, but for those coming from film/high end HD it's not much of a stretch. For those used to editing HDV on their mac and onlining from square one, it's different. But, then, I'm sure using real lenses and needing real tripods and AC's and all that jazz is new to them too. Get your feet wet and get learning... (not directed at anyone in particular)

I never said Sony's workflow options were pretty, or vey useful, or cheap, just that they existed--at least within the context of introducing a new format.

With Panasonic, I was mostly referring to their HVX release, in that it was somewhat timed with Apple's support of DVC Pro HD. Same sort of goes with Canon and JVC and HDV. Not saying those were better than RED RAW, just that the tools and workflow was figured out fairly quickly because the codecs didn't represent such a big paradigm shift.

Again, I wasn't trying to be negative against Red. Their work in this field has been game changing, and obviously they provided some initial tools to convert RR to something that could be used. But I still think that their release is so different that there should be no surprise that some folks are struggling with post.

Look, most of the people on this forum are used to being on the bleeding edge of technology, including myself, and have no problem "diggin in", but there are a lot of bright, experienced, talented people, that don't view it the same way.

There has always been an irony in Red's approach to the market. Get the bleeding edge, indie, scrappy filmmakers to "buy" in. These are many folks who come from "editing HDV on their mac and onlining from square one", but you can't in the same breath sell to those people, and then say well, you don't understand because our tools are more similar to Film and High-end HD.

Not bashing anyone, we're just all having a friendly discussion.

Cheers,
Chris