RAW workflow - a discussion

[Frank Cueto]

If it's a project that I care for, this does not happen to often. Most clients does not have the money time or interest. So most of times i do not bother.

Normally we get 10 bit DPX files form Resolve and finnish in Flame. This to me is not really the best way to get things done. Grading and online should be a combined process. And I find it quite amazing that people think that they can create a better grade in resolve then in flame. That is nonsense, I find flame superior to any grading toolset. There is simply so much more thing that you can do in a flame and the colors are all the same. This is something clients will get their head around soon but now still people thing the grading tools are sharper for grading than the compositing tools. Thats wrong. I think it comes from the time when we still used film and the clients DP's etc got amazed over how the colorist could tweak the colors in the scanner, now they are still quite amazed and they think it's the balls in the grading suite that's the trick... But if People do not leaner then I will not teach..

So I let them do their grading with the colorist, here in sweden it's depending on budget, if the budget is good they take the plane to london to get a "good grade" if they do not have that money then they do it at The Chimney Pot or Stockholm Post Production, if they do not have that money then they do it some where else... I actually think it's quite nice to get a disk with dpx files to do the online and efx from. the grading discussion can be a rather long one and most times i find it better to stay out of it, and do my composites etc and take the money and leave.

But then there are the projects I do care about, then I do this.

I bring in the DPX files from their resolve session with a XML. Then I use the same XML and reformat it in flame so the XML points at the r3d and conform from them as well.

Later in flame I make containers for each clip holding both the r3d and the HD 10bit DPX. Then I have all the options in the world, I can tell the flame to chew on the 5k r3d or look at the pre graded dpx files for any given clip. If I do green screen and such I can use the native source and make a key-inline pass that is from a different iso /whitbalance etc. I can also make sure I know that the highlights info in the graded shot is all thats there etc.

I can also use the flame color match functions to perfectly match and save the colorist grade. This I find quite amusing... it's like... thanks super expansive london colorist.... I just snagged your grade ...

And while working as above Im quite supriced how sloppy they are in the grading sessions and how little they know about colors. Some try to grade bluscreen material and then to help me out they key back the blue in the screen so that I should be able to key it... thats nuts. They also zoom and crop and such because they think that it's better done in the grading toolset than in compositing... ( this is something left from the time when there was a film strip in the scanner, and people naturally did pan in scan to get as much out of the SD signal as possible) today it's a different story... zooming into and panning in a alexa frame before online only does one thing... it ruins it for 3D tracking...


So to me I think the future will be the following.
Grading will be advanced lut's and metadata that is forwarded to the online sessions together with the raw files.
Flame can do this allready, I only have flame 2012 so I did not get Gamma3 and such just yet. But I can actually choose the redlinexpro seting for each clip or camera seting or flame seting or make a mix of the 3 etc. And thats how it should be.

I would like to have resolve just creating r3d meta data. If there is advanced layering and such that should Im sorry be done in composting I think. But I guess some splines etc could go through in future metadata swell.


For me I get happy every time when I see a r3d node in one end of my compositing tree and an "final output node" in the other...

WE master all in 10bit uncompressed Quicktime files. So in the end we are not raw anyway. I think that is good enough, 16bit is only for things that later should go through extreme grading or keying processes.

that how we role.

Sure, you can "snag" the grade with the color match features on flame. What about the multiple power windows intersecting with keys/mattes/shapes/wipes? That won't work with your approach. All you can hope to snag is a palette, not the full intent.

 

[Gunleik Groven]

Don't be so angry

It is only different POV, and I cannot see from any of Björns posted words or work that he doesn't see that point.

The great thing abpout THIS discussion, is that it is open for differnt POVs

 

[Frank Cueto]

Sorry if it came across as angry. Not my intention by far! Just wanted to point out that it's not possible to snag a grade a 100% with that technique, at least not all the time.
No harm intended

 

[Brice Ansel]

It's not supposed to be dead just yet either, I have just been a bit busy with other stuff. More to follow!
G

Good to know!
I thought that thread was still missing a conclusion from your part.
Cheers
B.

 

[Gunleik Groven]

LOL

I am not sure I will pull any definite conclusive rabbits out of any hats.
But I'd like to further the discussion.

I think it is good to be open for different views and approaches and watch where they lead in practical terms...

Good to know!
I thought that thread was still missing a conclusion from your part.
Cheers
B.

 

[Brice Ansel]

LOL
I am not sure I will pull any definite conclusive rabbits out of any hats.
...

Hé hé hé...
Secret sauce indeed,
I wish I could be in your kitchen to see what and how you cook your digital ingredients.
Cheers
B.

 

[Gunleik Groven]

Bring back the zombies! )

But I want to work a bit more on thisone...

I have promised myself that I have only one more putely geeky thing to do, and that is a discussion on linear light and its implications from set to delivery...

Thoughts on this is appreachiated.

After that article I think I'll only be a shooter... )

 

[PatrickFaith]

RED RAW Linear Workflow for VFX

RED RAW Linear Workflow for VFX

Here's a compilation of my notes for RED RAW LINEAR workflow in AE and Blender.
General what is linear:
http://greyscalegorilla.com/blog/20...and-how-can-it-help-your-renders-look-better/
http://filmicgames.com/archives/299
pdf explaining why linear is superior for vfx workflow:
http://www.pixsim.co.uk/downloads/T...n_of_Gamma_Correction_and_Linear_Workflow.pdf
Showing AE workflow with linear:
http://prolost.com/blog/2009/9/30/passing-the-linear-torch.html
Linear Workflow site:
http://www.lysator.liu.se/~zap/lwf/
openexr ILM explaination(basically openexr is a linear format):
http://www.openexr.com/TechnicalIntroduction.pdf
sony open color linear workflow:
http://opencolorio.org/configurations/spi_pipeline.html
explaining linear flow in blender:
http://www.loramel.net/blender_minutes/2009/01/linear-workflow-in-blender/
blender page on linear color management:
http://www.blender.org/development/release-logs/blender-256-beta/color-management/
(both nuke and blender do linear color management as thier internal method)
Comparing camera's and linear color:
http://www.fxguide.com/featured/the-art-of-digital-color/
Motion Picture Academy's push to linear workflow(ACES):
http://www.oscars.org/science-technology/council/projects/pdf/iif.pdf
Reduser discussions related to Linear workflow:
http://www.reduser.net/forum/showthread.php?84809-is-REDLOGFLM-a-look-or-Camera-RAW
http://www.reduser.net/forum/showthread.php?78453-RAW-workflow-a-discussion

--------------
AE and Blender using Linear Settings
Adobe r3d to vfx linear workflow
project settings menu set:
-32 bits per channel (float)
-working space to sRGB IE61966-2.1
-Select Linearize Working Space
-select compensate for scene-referred profiles

on r3d file ... on interpret footage main:
-on color management tab:
-select working space sRGB IE61966-2.1 (linear)
-select interpret as linear light On for 32 bpc
-click more options to go to r3d source settings
-select redcolor3 and linear


(when rendering in AE vfx ... click on color management tab ... make sure they are set the same as above).
When outputing AE to vfx(for textures), make sure output is linear and select openexr sequence.
When outputing from AE to ColorGrading (i.e. Davinci) select DPX/Cineon on the output module settings (after you clicked render composite), select "format options", select "show all available profiles", click on "output profile" select "hdtv REC709"

When doing VFX in blender i use the "cycles render" set at 200 iterations, i create panels using the textures that came out of AE, set the panels to emmiter shading, and the object to not be seen by the camera.

 

[Nathaniel Opgenorth]

Wow.....this thread has been EXTREMELY educating...I wanna thank Gunleik for that allot! I just read this entire thread post for post (took a while!) and learned allot. Your posts in this thread and else where are always nice to read. Been sort of a lurker for a few months since I don't personally own a RED camera but have been looking at renting a Scarlet to use with my assortment of Canon lenses for a serious short film. I love the workflow described in this. When I shoot on my Canon DSLR I shoot RAW (when I do stills) and adjusting white balance in post is nice as is being able to adjust other settings (although NOT an excuse to slack on a shoot, I despise "Oh well we'll do it in post then!")...its almost as if RED has taken the RAW workflow associated with DSLR's and applied it for their digital cinema cameras (and I think I remember seeing a post on that in here)! I've not shot on RED but I've been given samples and the workflow is simple and sensible even for someone like me who is far from a hardened DP or experienced colorist. I look forward to giving my business to you guys in the future as allot of your ideas have lead to things that the industry needed but know one stepped up to the plate and even things that have totally changed cinematography as we know it.

 

[Adrian Jebef]

Sweet Thread.

G, I think it would be wise and very informative to take the trip from analogue to digital. Let's discuss how a digital sensor actually works. There's a lot of technicality and confusion involved here but I'll put forth a brief summary (I'm sure Graeme will volunteer to correct me):

We light our scene. Peter O' Toole stands before us in the searing desert. Light breaks out from behind the clouds and cast shadows across his eyes and the enormous dunes miles off in the distance. We steady our hand and hit "Record".

Light from the scene (analogue wave energy) passes thru our camera lens and hits the digital sensor. Each photosite is filled up with a FINITE amount of light at 1/48th of a second. The sensor reads this energy and converts it into a digital representation of luminance intensity (this is where we go from analogue to digital, and logarithmic to linear). This digital representation is what is known as "coded values". The number of coded values available to represent the Dynamic Range (DR) of our scene is dependent on a camera's bit depth. The higher the bit depth the more precise the camera's ability to faithfully represent our scene becomes. SD TV was an 8-bit system comprising of 256 discrete coded values. HDTV is frequently using a 10-bit coded value scale representing 1024 discreet values. Most high end digital cameras and image formats support higher bit depths on up to 16-bits but for practical purposes the 10-bit scale is more than enough for viewing. Once the analogue to digital conversion has been processed into coded values we have what we can technically call a RAW image.

Unknown to most, a RAW image is actually not an image. It is just a bunch of discretely sampled, black and white, linear light values. The RAW file MUST BE processed before it can be viewed. This processing can take place in the camera or in post production or in both. ISO, WB, tint, gamma curve, etc. are all different adjustments that MUST BE made in order to create an image. When we open a RAW file in RedCine or Resolve we are still "working from the RAW file" but what we see is an INTERPRETATION of it and, as such, can be less than perfect. The thing to remember here is that the RAW file in itself is NOT perfect. It was created by a limiting device. Our scene with Peter in the desert actually comprised of 17 stops of DR (lots of cloud details...) but our sensor could only record 12. Herein lies much confusion. Yes our sensor is 16-bits of awesome and can account for 65,356 discrete coded values (times 3!) but it is also limited by it's own inherent Dynamic Range. The sensor, like our scene, has a discrete DR. If our sensor DR is bigger than our scene DR we are OK. But if our scene DR is bigger than our sensor DR we are in trouble. And this is where we decide how to set our exposure.

Setting exposure is difficult. Do we expose for the RAW or ETTR? Or use a light meter or just assume that because it looks good or "good enough" on our on-set monitor the RAW file will get all of that AND MORE! Some things to remember beforehand:

1 - Our camera is not perfect and requires help in order to produce a beautiful image.

2 - Each digital sensor has a discrete Dynamic Range it is capable of recording.

3 - Each sensor's processing from Analogue to Digital is different and comes at a cost.

4 - Each sensor's processing from RAW to a video image is different and comes at a greater cost.

5 - Each camera's ability to record a RAW or video file comes at a cost (unless it is uncompressed).

6 - Each post application that processes a RAW or video image does so destructively but this is a GOOD thing as we want to make our image look better (or at least prepare it to look better) at this stage of the game (going back to RAW just gives us thousands of 1's and 0's and puts us back at the start of this race).

7 - A vote for Obama is a vote for forward progress. Chipping in an extra $20 even if you live outside of the US is not frowned upon.:sifone:

8 - Sometimes the best rule of thumb is to simply expose for what is most important in your scene (in this case the eyes).


What can we learn from all of this? That every camera (and this does indeed include the EPIC) has a DISCRETE native sensor sensitivity to luminance. So changing the ISO on our EPIC as we shoot or later when we process may not technically change the amount of information contained in the R3D file but changing the T-Stop sure the hell does. Don't fool yourself into thinking that your camera will record all of the detail you can see. Far from it.

There can be a disconnect from what can be achieved technically and artistically on set and what can be (t&a) achieved in post. The complexity and cost-basis of the RED camera systems has allowed a lot of us to cross over from one world to the other. This has been hugely helpful in understanding the artistic and technical challenges before us. But it has also completely confused a whole population of new recruits.


G, this thread is GOLD. Now back on point. Is there any reason to maintain a complete RAW (or maybe even linear - except S. Shaw) workflow if we can't actually see the benefit but just trust that it is there?

 

[Nathaniel Opgenorth]

Great post. I think its important to understand the tedious stuff like that...well at least for a DIT, DP, Colorist, etc. or anyone that is going to be manipulating the files directly. I believe they need to understand what they are doing. I hate talking to professionals and when I ask them how something works and they say "It just does it and then it works" or something stupid like that to cover up the fact that they don't really know how it works...thats like flying blind to me! I wouldn't dream of that!

Now I'm not saying everyone needs to know every tiny ounce of scientific data about RAW but knowing that say your T/stop changes ALLOT but your ASA/WB is not set in stone is important...everytime I pick up a new camera I play with it and test it...its an extremely important part for me, some people think its weird but if you throw me a camera and say "I need footage of this this and this and it needs to look like this" I'm making a project a proving ground when it shouldn't be that at all!

About staying in RAW though, I can see the benefits but lets say we stay in RAW tell we adjust our ISO and white balance for the scenes WE KNOW we are going to use and then convert to an uncompressed TIFF sequence or whatever robust RGB format you prefer/need, correct me if I am wrong but you can adjust white balance in a RGB format or say Apple ProRes 4444 indirectly right? I know I've screwed up white balance before (on cameras that shoot H.264, MPEG2, aka NOT RAW) and fixed it up a bit by adjusting color highlights or whatever....I only mention ProRes as thats what I use in FCP on my laptop just so it doesn't barf...Also I know with RAW formats like Canon's CR2 a JPEG is stored as the "core" so it viewable, REDCODE has a representation in H.264 or some other format correct? I'm learning and confirming things as I go everyday!

 

[Adrian Jebef]

Great post. I think its important to understand the tedious stuff like that...well at least for a DIT, DP, Colorist, etc. or anyone that is going to be manipulating the files directly. I believe they need to understand what they are doing. I hate talking to professionals and when I ask them how something works and they say "It just does it and then it works" or something stupid like that to cover up the fact that they don't really know how it works...thats like flying blind to me! I wouldn't dream of that!

Now I'm not saying everyone needs to know every tiny ounce of scientific data about RAW but knowing that say your T/stop changes ALLOT but your ASA/WB is not set in stone is important...everytime I pick up a new camera I play with it and test it...its an extremely important part for me, some people think its weird but if you throw me a camera and say "I need footage of this this and this and it needs to look like this" I'm making a project a proving ground when it shouldn't be that at all!

About staying in RAW though, I can see the benefits but lets say we stay in RAW tell we adjust our ISO and white balance for the scenes WE KNOW we are going to use and then convert to an uncompressed TIFF sequence or whatever robust RGB format you prefer/need, correct me if I am wrong but you can adjust white balance in a RGB format or say Apple ProRes 4444 indirectly right? I know I've screwed up white balance before (on cameras that shoot H.264, MPEG2, aka NOT RAW) and fixed it up a bit by adjusting color highlights or whatever....I only mention ProRes as thats what I use in FCP on my laptop just so it doesn't barf...Also I know with RAW formats like Canon's CR2 a JPEG is stored as the "core" so it viewable, REDCODE has a representation in H.264 or some other format correct? I'm learning and confirming things as I go everyday!

Going thru the basics of how things work is absolutely necessary. When we shot film the process was transparent. You could actually see the mechanics of a film camera, how it functioned, and why. You also had the opportunity to touch the film stock, visit a lab and watch it being processed and then printed. There were real, physical limitations present in the system that were easily understood over time. With digital capture we started off in a similar fashion with SD and HD video camera systems. The DITs and video engineers of 2005 had a background in DSP (digital signal processing) and really understood the physical dynamics at work in creating a digital image. Here in 2012, we have seemed to have bought the argument that the cameras of today are so much more advanced than the cameras of yesteryear that understanding how they actually work is A) pointless because they are just F-ing AWESOME and B) that understanding something this AWESOME is simply too complex. The truth is that today's high-end digital cinema cameras are really not that different from yesteryear's, they are simply much more refined. Understanding how it all works allows us to anticipate the inherent shortcomings of CCD and CMOS sensor technology. And the more you know about the shortcomings the better off you will be in compensating for them when you light your scene.

As a quick take on my own personal experience, you really need to spend the time trying to understand the basics of the digital process over and over again. The first time you try to make sense of all of this it can be a bit overwhelming. The 4th time you'll feel you have a grasp. The 8th time you'll realize you've had it all wrong for the past 3 years. But at a certain point all of it will come into focus and instead of a complete technical mastery of the process you'll have arrived at a relaxed understanding of what it takes to be a better artist and not just a more perfect technician.

I have a collection of old links on the subject that you may find helpful:

http://www.adrianjebef.com/index2.php#/vendors/

Onward !

 

[PatrickFaith]

...
As a quick take on my own personal experience, you really need to spend the time trying to understand the basics of the digital process over and over again. The first time you try to make sense of all of this it can be a bit overwhelming. The 4th time you'll feel you have a grasp. The 8th time you'll realize you've had it all wrong for the past 3 years. But at a certain point all of it will come into focus and instead of a complete technical mastery of the process you'll have arrived at a relaxed understanding of what it takes to be a better artist and not just a more perfect technician.

I have a collection of old links on the subject that you may find helpful:
http://www.adrianjebef.com/index2.php#/vendors/
Onward !

Still some cool stuff on your link, ... found the link http://www.quantel.com/repository/files/library_DigitalFactBook_20th.pdf still perty good.

I think I'm somewhere between my 4th and 8th time cycling through my workflow. My current tool set is Redcine, Adobe Master suite , Nuke, Modo, blender, davinci. I'm finding everything i don't know about them ... i make a fundamental mistake in my workflow. I am working a linear r3d flow between all the packages, each package in itself is an entire world of complexity. I think this is the big challenge of raw work flow and high quality production, keeping current with these package suites and using the right tool for the right job. My thumb is black and blue from workflow misses, as I use sledge hammers when trying to hang a small picture on the wall.

 

[Nathaniel Opgenorth]

Going thru the basics of how things work is absolutely necessary. When we shot film the process was transparent. You could actually see the mechanics of a film camera, how it functioned, and why. You also had the opportunity to touch the film stock, visit a lab and watch it being processed and then printed. There were real, physical limitations present in the system that were easily understood over time. With digital capture we started off in a similar fashion with SD and HD video camera systems. The DITs and video engineers of 2005 had a background in DSP (digital signal processing) and really understood the physical dynamics at work in creating a digital image. Here in 2012, we have seemed to have bought the argument that the cameras of today are so much more advanced than the cameras of yesteryear that understanding how they actually work is A) pointless because they are just F-ing AWESOME and B) that understanding something this AWESOME is simply too complex. The truth is that today's high-end digital cinema cameras are really not that different from yesteryear's, they are simply much more refined. Understanding how it all works allows us to anticipate the inherent shortcomings of CCD and CMOS sensor technology. And the more you know about the shortcomings the better off you will be in compensating for them when you light your scene.

As a quick take on my own personal experience, you really need to spend the time trying to understand the basics of the digital process over and over again. The first time you try to make sense of all of this it can be a bit overwhelming. The 4th time you'll feel you have a grasp. The 8th time you'll realize you've had it all wrong for the past 3 years. But at a certain point all of it will come into focus and instead of a complete technical mastery of the process you'll have arrived at a relaxed understanding of what it takes to be a better artist and not just a more perfect technician.

I have a collection of old links on the subject that you may find helpful:

http://www.adrianjebef.com/index2.php#/vendors/

Onward !

Wow...Great post. The part about people assuming its "just so fricken awesome but too complex to understand" is true in allot of fields. I find DSP to be an interesting field myself so reading a white paper on a codec is fun (most of the time at least!) however I'm sure for 80% of people they just want to soak the jist of it into their head and start shooting. Trying to understand the basics of digital like you described is EXACTLY what I've experienced...its amazing just HOW many times I realize I've assumed something would do something automatically or completely missed an important setting after the fact! I think "Okay I'm manually controlling my f/stop, 2.8 is just where I want it my ASA is where I want it, shutter angle is as close to 180º as I can get and lighting is not as bad as I thought" then I get in post and I realize my white balance was set for the previous day in the sun and I'm in a church with lights that are near orange! :O Fortunetly now its like not cleaning a filter thats a bit dirty or massive mistakes like grabbing a memory card thats full of important stuff I can't delete or just not bringing one at all! :banghead::yikes: Now something like white balance is about as vital as other settings since I don't get RAW video with the cameras I work with.

Most of the people I do video and photos with would probably ban me from their house if I started "lecturing" on CMOS technology or DSP! Yet friends who are electrical engineering students completely get what I'm talking about but don't do photography....its like I need to start finding DIT's and DP's and start forcefully breeding them into perfectionist DP's from birth...of course I'm kidding but you see my point hopefully. Anyways! Even if people don't want to take the time to read they should take the time to test their cameras....my family looks at my like a 6 year old with some new legos when I have my camera "Aw look at him play with his camera he must want to be a photojournalist or a photographer!" when I'm really just trying to see what happens when I use a camera in a room with X lighting or whatever, I guess thats "playing" though and if by photojournalist they mean cinematographer then yes. If a picture is a thousand words than a film is 24,000 words a second (well technically 23,976 with NTSC...) but I digress...When I hit a wall (figuratively!!) I go online and read about lens, CMOS or CCD tech, DSP and uh those 1s and 0s...

Also I wanted to mention with RAW I feel like its going back to the way film photography was. I mean lets say you take a JPEG image...you simply CANNOT manipulate that say 4000x3000 image like you can with a 2K film scan and thats because no one in their right mind would do a film scan in JPEG image sequences or a codec like H.264! I mean sure if your doing a film scan with catalog films for a DVD/Blu-ray re-release the film doesn't need too much adjustment but if your shooting a movie on RED say and you want to do a sequence in Super 35 your going to want to be able to manipulate it like the REDCODE, an obvious statement but yeah.

Anyways looking at some RED reels recently I feel as though RED has made a true "digital film" unlike other company's...I see some other good cameras and I say "thats a great video!" But I instinctively say "film" more often when I see something like The Girl with the Dragon Tattoo or the numerous other films shot on RED. If I said that on another forum I might get labeled a fan boy but thats just how I feel...not saying something like the Arri Alexa (which I think is incredibly cool and smart for TV with its ProRes recording) or the Panavision Genesis isn't an amazing piece of equipment because they are (don't hang me out to dry guys!). After reading Mr. Galt's (believe I got that right) (VP of Advanced Digital Imaging at Panavision) article on 4K, 2K, etc. on CreativeCow I really felt a stronger understanding of imaging technology, I can't find the article but its a great read, he really explains things like the Bayer sensor amazingingly well and while it might have some jabs at RED or Dalsa its really not "propaganda". In fact although the article explains that 4K (at least for end consumers) might be overkill it further solidified my thoughts on things you guys at RED are cooking up like the 6K dragon sensor....if the argument that after de-bayering you get half or whatever resolvable lines then thats fine because with RED its still an AMAZING picture and half of 4K or 6K Bayer sensor is still allot of lines.

The Link is much appreciated though....I take that kind of thing from people who have been doing this for a while with much appreciation.

 

[Gunleik Groven]

Adrian, thanks!

Not finnishing this thread, is one of my bad conciousnesses...

Great description.
Thanks!

And i happen to think that this thread is of some value...

I won't finnish it today, and my "linearlight for dummies" article is still not written, maybe because i see that as my last purely geeky act... Hahahaha.
Degeeking myself heavily these days.

I usually describe a sensor as a photon-weight.
Everytime you double the intensity/weight, those levels are mapped to a new bit in the AD
That new bit allways have twice the theoretical precission of the previous one.

What is important to get, both for dp's and post, is that what is true of luma values, will also be true per chroma channel, and not at least how the whiteballancing works, as that affects the image beyond technical over and underexposing.

If you in-camera ballance from let's say 5600 to 3200 kelvin (just picking a random example here), that involves not only gaining the blue channel, but also clamping the red, as the ballancing happens with the green channel as the neutral "zero"-value.

First. There is nothing wrong with this approach. It is just how it's done.

Most people seem hung up in what happens with the amplifying of noise in the blue channel through this process.
Well that is just as expected, and as the blue channel will have more than 1 stop of lower exposure under tungsten than daylight, the precission in that channel SHOULD be lower, even when digitally amplified to reach the same whitepoint. No surprises there.

I have become much more concerned with the clamping of the red in high exposures.

As the clamp hits in, the top of the red channels first flattens, and them gets a hard limit, causing the oh-so-infamous flat yellow-metallic "video" skintones under tungsten, even though the photon-weight/sensor doesn't technically overload/clip yet.

There are ways to handle this on-set and in post, but this issue is the main reason for my "expose for the sensor, even when using tungsten lights" approach.

A nice test, is to take a live-image on any camera, a scene which is lit with tungsten and exposed to just bellow clippinglevel, at 5600/ daylight, and change the wb only and look at how the skin plasticifies as you change to a lower wb.

Happens on all digitalcams i have tested it on to a varying extent (as there are better and worse ways of implying the gain and the clamp) but it happens even on the good ones, 'cause it has to. NOTHING wrong with it. It is just math.

But managing that last stop of exposure under tungsten to get an elegant result, is a bit underfocused subject by most camera manufactureres, even though that is exactly where most digital cameras have a hard time coping.

The epic does better than most, though, as it has overhead for the calculating process (having an internal processing with a higher bit-depth than actual stop/dr capacity). While the r1 mx suffers from the exact opposite issue, methinks...

Anyhow.

Thanks for keeping this thread alive

It is one of my pet threads over here!

 

[Björn Benckert]

Sure, you can "snag" the grade with the color match features on flame. What about the multiple power windows intersecting with keys/mattes/shapes/wipes? That won't work with your approach. All you can hope to snag is a palette, not the full intent.

Well, offcouse if you use power windows, mattes etc it takes more than a few clicks to snag the grade... But honestly I never seen a grade that I can not match in less then 30 minutes. I'm not saying copying the grades with color matching tools is the way to go. On the conterary I think grading should be a skin to the raw file thats all. Grading before composting and burring the grade into DPX files.... is simply not the future. I think the colorist tool that will let the artists export their setups into some sort of metadata or such that is applicable / adjustable downstream in/ on other platforms will do better than whats there now...I mean only being able to export affected quicktime files or image sequence is not a perfect solution since it's far more accurate to add the CG / do the composting onto log raw files while being able to watch the comp with the applied grade ontop of the CG / layering.

So with the above in mind I think one way of doing things kind of right would be. Do a first round of color timing, kind of what they do in the lab process for positives. i.e only use a lut editor and export a 3DLut for each scene. Then do the composting in LOG while viewing trough these 3D Luts provided buy the colorist. Then finnish the composting without the luts aplied in log space and ship it back for final grade. This way the composting can be done in the right direction for the grade. But thats more of a workaround for not having anything more advanced in terms of settings than 3D LUT's that transfers between the platforms of grade and composting.

So I think a proper grading tool should be able to export source picture data in a good format and skinned setups that is readable for other platforms.

just as 3D packages exports models /cameras etc to each other it would not be so hard to make a Resolve to Autodesk importer converting the resolve setup to a node group in the flame batch or nuke tree... as there is no grade that you can create in resolve that these programs can not mimic.

I repeat that last sentence and then you prove me wrong.... There is no grade done in resolve, or film tool or any other grading tool for that matter that Flame or Nuke can not mimic.

There is still some people that think that grading software is unique in their way of handling colors or such, I guess that is coming from the times of the scanners... So yes, you can not scan a negative in Nuke. That is correct, but you can do all the mathematical transformations that a grading software can do.... and since you can write your own functions /plugins etc for those kind of softwares there is really no end to it, the grading softwares on the other hand are rather limited... They are all more about turning nobs and adding windows...

 

[Gavin Greenwalt]

Still some cool stuff on your link, ... found the link http://www.quantel.com/repository/files/library_DigitalFactBook_20th.pdf still perty good.

I think I'm somewhere between my 4th and 8th time cycling through my workflow. My current tool set is Redcine, Adobe Master suite , Nuke, Modo, blender, davinci. I'm finding everything i don't know about them ... i make a fundamental mistake in my workflow. I am working a linear r3d flow between all the packages, each package in itself is an entire world of complexity. I think this is the big challenge of raw work flow and high quality production, keeping current with these package suites and using the right tool for the right job. My thumb is black and blue from workflow misses, as I use sledge hammers when trying to hang a small picture on the wall.

Ughhhhh they're calling Video "Linear". I would expect better out of something that proposes to a definitive guide.

 

[Björn Benckert]

Ughhhhh they're calling Video "Linear". I would expect better out of something that proposes to a definitive guide.

Do not listen to much to what quantel have to say these days... time have proven them wrong, and wrong again... It was different back in the days when they invented and had patent for the cool stuff like blur and dynamic rounding

 

[PatrickFaith]

Ughhhhh they're calling Video "Linear". I would expect better out of something that proposes to a definitive guide.

Well ... was 2008, they are also re-using the word linear for editing and other things. I've been looking for a good book that goes over color science math/physics/biology in a more or less exact way, but all I find is books that do a lot of hand waving when it gets down to the details. The jpeg2000 books for dcp are expensive, and that really just touches on a small part of color science. I have noticed though, that the ILM donated code (and some other heavy hitters) for openexr 1.x and 2.x is really well documented, has a lot of great content that seems very accurate ... problem with that is it's only readable by C++ and fanatical color science experts(lot of big names on the production side involved with that code). Not having a definitive color science book hurts the industries understanding of raw. The sony opencolorio code is also real good to read, but again, isn't really a definitive "book".

 

[Adrian Jebef]

Ughhhhh they're calling Video "Linear". I would expect better out of something that proposes to a definitive guide.

Don't take it so bad. If there's anything I've learned over time in this business is that definitions change as much as the technology does. Heck, just think about "Gamma". That word confused the hell out of most people in 2005 (and before . Now everyone feels like they totally get it. Yeah, right... :banghead: