RED Post Tool - EDL/Cutlist

[Mark L. Pederson]

You mean like Apple opened up ProRes?

Post of day.

 

[M Most]

I'm interested... why do you think this is too much to ask?

Well, in part because Jim has already stated that it's not going to happen, at least in the first release. That's why. And additionally, because with the release of the SDK, Red's approach has clearly been to empower third parties to develop the post production software support to take the burden off Red's internal development team and to accelerate the format's acceptance. They are adding hardware support for this effort with the Rocket in part to encourage even further development, because they clearly understand that this kind of thing is better done by companies dedicated to the task. I happen to agree with that.

Besides, Red is not selling cheap toys. They are selling professional level tools that require professional level support to operate optimally. If you look at consumer level video cameras, many of them ship with tools like Premiere LE - essentially a very cut down version of a much better tool that is included to give you the basics, but is primarily intended to entice you to upgrade to the "real" tool. Or, in the case of formats like Sony's XDCam EX, a "browser" application is included that aids you in getting the shots of your choice into the editing system of your choice. That is basically what Red is doing with their free tools, but with the added value that they bring to the table - the ability to translate to various formats, to do basic RAW conversion and color settings, to scale, and with the new RedcineX, to preview quick cuts on a timeline. And the ability to play shots directly for QC purposes at the time of shooting. I think that's pretty darned good.

 

[roryhinds]

I don't know Mike, I do thinking having EDL support in REDCine would have changed the RED world for sure. Being able to smoothly go from Offline to Online would certainly have saved a lot of headache and time on many productions.

As for this being new technology, come on EDL has been around since the first NLE and REDCODE is over 2 years old now isn't it?
Yes a lot of productions don't do a second render because its a pain in the arse and not a simply dedicated process, if it was most would take the finished edit into REDCine and tweak to get the best results for sure.

REDCine was dead in the water because of Scratch and now ScratchCine
Spending $18k on a onelight R3D software package isn't viable for most shooters
Baselight now have there cutdown tool for working with R3D files but again its out of reach price wise for most shooters.
Just like the "Cinema" cameras were out of reach until RED came along with there REDONE.

All we are asking for is a RED dedicated tool that works with 3RD files and can take you from Off to Online with out a hop skip and a jump.

If RED need to charge for that tool then so be it, as I'm sure it will be moderately priced just as the REDONE is.

Now that RED are developing REDCINE-X in house themselves I don't see why they can't bring out a useful tool for there hardware, which I think will change things in the RED world in a big way, it is after all the missing link.

 

[Cüneyt Kaya]

If you look at consumer level video cameras, many of them ship with tools like Premiere LE - essentially a very cut down version of a much better tool that is included to give you the basics, but is primarily intended to entice you to upgrade to the "real" tool.

which would be in this case...yes

no seriously i like all red psot tools, but redcine was always a real good viewing tool for me, the instability (i know on lucas magic macbook pro it never crashes...hehe) was a deal breaker.

but the rest?

well we are all finishing red projects today

 

[Kevin Wild]

I'm with you 100% Rory. I'm hoping there is some type of conform tool that happens...mostly for an XML file from Final Cut.

Of course, if Apple did a make-over of Color and both FCP & Color supported realtime 4k files with Red Rocket, I guess I would prefer even more that I don't need to conform at all and can just edit, cc & output all within the same studio app. Dreaming? Perhaps...

 

[Lucas Wilson]

Adding EDL support is not big deal, its been around for how many years?

Well - yes, actually, EDL support is a big deal. Here are some of the basic issues confronting an application that wants to support EDLs.

Let's start with the assumption that a conformed timeline needs to be accurate. If it isn't, then a conform isn't much good.

First:

1) Just CMX support, or GVG and Sony as well?
2) If just CMX 3600 - which is the most common standard, then...

Second:

All EDLs have to come from somewhere. The most common offline programs are Final Cut Pro and Avid.

Final Cut Pro:

1) Almost *never* exports a proper CMX list. Look at a standard FCP EDL. There is usually a line at the end of the event saying "FINAL CUT PRO REEL:..." It does not have an asterisk at the front of the line. The CMX spec says that any comment line must have an asterisk at the front to distinguish it as a comment.
2) Dissolves can be pretty weird. Sometimes they are per CMX spec, and sometimes they look like zerocut events. Non-standard, and unpredictable.
3) There has been a bug in the EDL exporter for a long time that will sometimes generate a "C" or "S" in the Source Column instead of the standard "V," "A," or "V/A1," etc.
4) OSX frequently inserts non-standard characters into what is supposed to be a plaintext file, making parsing it difficult.
5) Multi-layer timelines will have large record timecode discontinuities, which are hard to track, as it means assuming the missing time is contiguous, when sometimes it is not.
6) Edits are inclusive (Avid is exclusive.)

Avid: GREAT EDL support, but lots of different options, some of which are non-standard CMX

1) Reel names longer than 8 characters
2) EDLs longer than 999 events (the max in the CMX spec)
3) I'm pretty sure (michael P - correct me please) that Avid EDLs can be either inclusive or exclusive.

Third:

Are you going to support video transitions? If so,

1) There are a couple of different ways to notate dissolves. There is only one correct way to do it per CMX spec, but FCP especially will sometimes throw in what look like zerocut events at the beginning of a dissolve - although this is frequently due to how the timeline is being used.
2) M2 lines are notoriously hard to conform, as every editorial program uses a slightly different methodology for calculating percentages
3) What about the (support, but rarely used) key event? How will that be handled?
4) Wipes? If so, how?

Fourth:

Will you support some of the common use, but non-standard terminology, such as "BL" for Black, "AUX" for B-side, and "LDR" for SMPTE leader?

Fifth:

Will audio support be included? If so,

1) How will L-cuts be handled?
2) How many audio channels, and how will routing be done from digital files?

-------------

That is what I can think of off the top of my head. There is more. And that takes us *just* to the point of properly parsing an EDL.

This does not take into account the following decisions that must be made once an EDL is parsed:

1) How will an event be interpreted? There are 3 common metadata parameters available in an EDL event to match up source material: Source Timecode, ReelIDs, and a "FROM CLIP NAME" line which frequently contains the name of a source data clip. What if a match comes up with one hit (sourceTC) but the incorrect ReelID? Will that be placed in the timeline? How about the correct "CLIP NAME" but the wrong ReelID?
2) How are incomplete clips handled? What if a source event has 100 frames, and the clip you've loaded only has 50?
3) How are record timecode discontinuities handled?
4) What about drop frame vs. non-drop for NTSC land?
5) What about 30-frame timelines for 24-frame material? (happens all the time in NTSC)
6) How are repeated or missed event numbers handled?
7) Are M2 lines supported as interpolations or straight dropped-frame events?

----------------------

All this has to be done if there is a *perfect* EDL. I was an online editor for many years, working with some of the best, most methodical offline editors in the world. I can count on one hand the number of "perfect" EDLs I received from offline.

Also, online frequently has no clue which program generated an EDL, so there is no way to consistently and accurately track all of the little errors I list above. A ton of intelligent programming and parsing has to go into interpreting lists and trying to make sense of them.

And all this is just to get from the step of loading an EDL to putting media on a timeline.

Then the real work of a conform begins - lining up the offline, comparing, and doing editorial problem solving to figure out what's wrong and how to fix it.

So... long rant... but the point is - EDL support is a big deal, and it's hard to do it consistently and do it well. And EDL is just the bare minimum needed to do a conform.

Crippling the software to project other companies is crappy especially when we're wanting to work with RED files from the RED camera.

If you're talking about SCRATCH here - REDCINE was never crippled to protect SCRATCH. It was designed for a specific purpose. The fact that tools like Crimson have popped up to squeeze additional functionality out of REDCINE doesn't alter the fact that REDCINE was never designed to be a conform tool.

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

 

[M Most]

do you believe the software only grading tools will hold their price over the next 5 years or will get cheaper ?

i am not talking of the hardware/software combos

There are three factors: supply, demand, and the unpredictable "wild card" factor. Supply and demand will generally determine the price. However, in the color correction area, you have the wild card factor in the form of Apple. Because Apple's business is primarily hardware, they have the unique ability to undercut every software competitor on price. Their combination of hardware and software is usually in the "good enough" range for many users, so in the low-to-mid price range, what they do with their applications will largely determine what the competition - if any - will have to do. On the professional end, Apple is much less of a factor. In that market, it's primarily supply, demand, competition, and general economic conditions that will determine future price points. One comment I would make is that "hardware/software combinations" probably have a limited shelf life, as the use of GPU's and the pure speed of general purpose computer hardware is getting to the point - if it's not already there - that a lot of the hardware assistance is no longer required, even in the most demanding situations. While it's true that certain combinations of components will be considered optimal for specific applications, they will be comprised completely of off the shelf hardware. You can see this already in such devices as Assimilate Scratch, Avid DS 10 (the first DS with no proprietary hardware at all), RAVE HD (a DDR made completely from off the shelf components, as well as an open source operating system and even open source code), and even the latest versions of Flame and Smoke (which use completely off the shelf hardware - as does Lustre, with the exception of its control panels). And if you're interpreting this as a prediction that the days of proprietary hardware assisted systems like Pablo, Baselight (at least the Baselight 4 and 8 models), and DaVinci Resolve are ultimately numbered, I'm not going to discourage you. But I would also say that for the next few years, they are all very useful and valuable tools, and in the right business situation can be very appropriate.

 

[Lucas Wilson]

I'm with you 100% Rory. I'm hoping there is some type of conform tool that happens...mostly for an XML file from Final Cut.

Hey Kevin...

Ok - see the list I just posted about a basic CMX EDL? Multiply it by about 20x for XML.

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

:beatdeadhorse5:

 

[Mark L. Pederson]

So... long rant... but the point is - EDL support is a big deal, and it's hard to do it consistently and do it well. And EDL is just the bare minimum needed to do a conform.


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

That was a long rant Luki.


Let me take a shot -

EDL is a LAME ASS spec that need to go the way of Kodachrome. It is the PAST.

XML is the future and now.

Why EVERYONE keeps trying to support a HIGHLY PRIMITIVE, OVER-SIMPLISTIC "standard" that sucks in practice is beyond me ...

I can see mmost reaching for the keyboard now ...

 

[M Most]

Yes a lot of productions don't do a second render because its a pain in the arse and not a simply dedicated process, if it was most would take the finished edit into REDCine and tweak to get the best results for sure.

Believe what you will, but I can say categorically that at least in the area of high end television work, that is just not the case. That has a lot to do with the complexity of modern editing styles, but it is also a concession to time.

 

[Lucas Wilson]

EDL is a LAME ASS spec that need to go the way of Kodachrome. It is the PAST.

XML is the future and now..

Sure, but what kind of XML?

XML is not a standard for communicating timelines for a visual project. It is a standard for formatting text.

SCRATCH has XML. Final Cut has XML. Both are valid XML. And they are completely incompatible with each other.

The best analogy I have is...

XML is sort of like the Roman Alphabet. A = A. B = B.

But whether your write English or Hungarian is up to you.

Lucas

 

[M Most]

XML is the future and now.

Why EVERYONE keeps trying to support a HIGHLY PRIMITIVE, OVER-SIMPLISTIC "standard" that sucks in practice is beyond me ...

I can see mmost reaching for the keyboard now ...

Well, when you get all industry vendors to agree to offer identical feature sets (including effects and their associated parameters, layer interpretations, and nomenclature for things like clip speed), standard tags, and then get a standards setting body to write a spec using those tags, and get all of those vendors to support it, let me know.

Until even simple things like video layers and how they're implemented are commonly agreed to, anything beyond basic EDL information is almost impossible, regardless of the file format. So for the moment, the next best thing is for all post applications to understand and parse both FCP and Avid XML formats in their entirety. Good luck with that.

 

[Mark L. Pederson]

anything beyond basic EDL information is almost impossible, regardless of the file format.

This is a gross over-statement Mike and you know it.

 

[roryhinds]

Lucas, I'm not saying that isn't not a work load to bring EDL support and even XML support to a RED tool. A work load that can be bought in from an existing product or a dedicate coder/coding team.

Again all not rocket science as this have already been solved, its not a question that can't be answered or problems that don't have a solution. We're not asking to reinvent the wheel here.

I would have thought in today's climate there would be many coders available suited for this if not an existing solution RED could purchase and adapt to there own needs.

Until RED Rocket comes along and we see how it works with r3D's my thoughts are I'm not fussed with editing directly with the r3d files. Much like I don't want to work directly with the film negative so I'll do a rough and ready onelight telecine for the edit and then do a 2k scan for the online - oh isn't that what REDCine was supposed to be... hmmm

 

[roryhinds]

Okay so have R-X support Avid's XML and FCP's XML
Forget about transitions, speed changes or even more than 2 layers of video for now as you can sort out your timeline in prep for the online and recompile these once the cutlist has been rendered.

Anyone from RED reading this thread and want to chime in?

 

[Mark L. Pederson]

Sure, but what kind of XML?

XML is not a standard for communicating timelines for a visual project. It is a standard for formatting text.

SCRATCH has XML. Final Cut has XML. Both are valid XML. And they are completely incompatible with each other.

The best analogy I have is...

XML is sort of like the Roman Alphabet. A = A. B = B.

But whether your write English or Hungarian is up to you.

Lucas

Pretty simplistic way of looking at things.

FCP and Scratch XMLs are not "incompatible" - they just need translation.

There is a HELL of lot more information in both XMLs specs than any EDL.

Ever see CLIPSTER conform a reel of a feature film from an FCP XML?

It will make your head spin.

It's so impressive IMO - that I am committed to putting my own resources to scripting translation from CLIPSTER XML to SCRATCH XML.

None of this is rocket science. After all ... A=A and B=B ... and I know a coder in the Slovak Republic that can circles around most ...

 

[conrad gaunt]

My main problems with Redcine are still the following:

- Won't export anything on any of my windows systems, except garbage

- Will only export images from OSX on imac (ATI), not QTs (except with garbage images).

- Still can't export audio as an AIFF etc, or choose an audio codec for that matter

Thinks I'd put on the wishlist would also include

- A bigger curve editor (unless I'm missing something)
- Graded/Ungraded split screen preview

I guess I'm greedy..


Not sure why I got the Mac now ..oh yeah, I can at least get TIFFs ..

Apple only make about 7 computer models, surely we can get it running on the known "out-the-box" macs, with their known specifications?

I think Red need to ask Apple to lend them a Mac Book, Mac Book Air?, Mac Book Pro, Imac, and , whats that little one without a monitor called?..maybe not that one.. and with all flavours of GFX card (ati and nvidia), for testing purposes.. the soccormoms won't want to buy Mac Pros so they can swap out gfx cards till they find one Redcine likes..

Thankfully, Redrushes and Red Alert are reliable for me.. CPU power seems to rule here, my GPUs don't seem to want to play..

I create online edit lists using "paper and pencil", that way its application non-specific

 

[roryhinds]

don't forget REDCINE as it stands can't relink the media

 

[Miltos Pilalitos]

I am using the crimson workflow for almost a year now so all this effort i see of presenting XML support for REDCINE as an impossibility is laughable at least.

I don't want audio or dissolves or any kind of FX in Redcine. I just need a simple way to take an edit into first light and export a DPX or EXR sequences to my compositor of choice.

Today i had to drive for 2 hours to an editor to help him do the first light CC in REDCINE. They shot a tv commercial with my RED and after editing in FCP they couldn't figure out how to export the color corrected cut to Fusion for FX. When i showed up with crimson they couldn't understand why i needed this step to use RedCine.

It just doesn't make sense...

 

[roryhinds]

yes why doesn't RED buy Crimson and give iBoom a nice cheque for his hard work and he can work with them to bring a RED tool that works without having to jump through hoops.

iboom if they do purchase, please buy me a beer