Navigation

 ·   Wiki Home
 ·   Categories
 ·   Title List
 ·   Uncategorized Pages
 ·   Random Page
 ·   Recent Changes
 ·   RSS
 ·   Atom
 ·   What Links Here

Search:

 

Create or Find Page:

 

View Linear Workflow

Linear Workflow relates to the image and render pipeline, and how input imagery (typically textures and backgrounds) are processed by the rendering engine, and how the rendering engine applies adjustments to the final resulting images and movie files.  This can get real complicated real fast, so lets keep to the basics for now.

Problem: If you’re not paying attention, you can easily create assets with different color management values.  This will cause the renderer to give less than adequate results in return, and may cause much work to be wasted.

Solution: Insure that all color management values across the pipeline are consistent, and conform to the requirements of the rendering engine.

Most people don’t really think about color management until it bites them in the a@#.  The reason being is that this issue is not something that most people typically deal with in their daily lives.  In fact, color management is effectively limited to printing and motion imagery.  Compared to the rest of the world, that is really a small number of people overall.  That said, color management effects the entire world.  We are a visual species, and in addition to the various rules of design that have evolved over the years, they are all influenced by good or bad color management choices.

In the preceding problem/solution paragraphs, the takeaway should be that a little upfront setup can save you from a world of hurt.  The reasons why we have to deal with color management come down to some simple realities when dealing with contribiting and resulting assets.  Let’s examine the “why” in a bit more detail: Rendering engines use linear math to process incoming images, render the scene, and output the resulting imagery.  Linear math is fast, produces reliable results, and is arguably the best overall approach to take.  But in order to linear workflows to perform optimally, the “color space” for the input imagery needs to be known and understood. If you use inconsistent color models, you will confuse the renderer, to the point of getting garbage in return in some cases.  You will also limit your color gamut which will limit the ability of a compositor to really push the limits of the rendered imagery after the fact. Part of getting that right is to understand how your output imagery is to be used or processed after it comes out of the rendering engine.  Let’s start with gamma.

Historically, gamma has been one of the biggest bugaboos, especially when working with multi-machine environments, such as Mac/PC/Flame/Infermo/etc. Each of these physical machines have different gamma values, and will require the rendering engine to understand this, and handle any contributing imagery from these platforms accordingly.  In addition, NTSC video has a gamma for video is 2.2; gamma for the typical film stock is .625 or thereabouts.  Nad Macs and PCs fall somewhere in the middle. That is a rather large range to deal with.  Add to the fact that the gamma that display cards are set to is 1.0.  The gamma settings affect the color perception and contrast, etc.  Fortunately, some file formats carry this information with them in a tag called metadata.  When present, the gamma of the source imagery can be correctly handled at render time.

Color space is another issue, though typically less of one than gamma, as most display systems use a variation of RGB.  But realize that you may see the CMYK, LAB, and other color spaces come through, especially if your are dealing with assets from clients that you are required to handle.  Fortunately, applications like Adobe Photoshop or After Effects can convert these formats into something the Cinema 4D can directly manage. 

Color format is next.  This article is on Linear workflows, in part because Cinema 4D works in linear space when rendering.  However, the Cineon format and the Flame/Inferno systems frequently operate in the Cineon format, though it is falling into disfavor.  The most recent, and perhaps most comprehensive format is Open EXR, developed by ILM.  If you need deep color, you should be using Open EXR.  Most content applications handle it, and the format itself can handle multiple pass layers in addition to variable deep color options. In still images, RAW is similar to Open EXR, and the benfits of using RAW over other formats are the same.

Delivery formats will likely ease this issue somewhat; television is now digital, and digital projection is starting to replace film projectors in many metroplexes.  Some projects still shoot live action on film, but, save for a few holdouts, digital is becoming harder to argue against. Its important to discover what the best options for your delivery format may be. 

Gotchas

All graphics cards come with color space drivers, and the typical driver is sRGB, which was defined by Adobe, but there are plenty of other formats, which you will likely see if you browse the color management settings for your graphics card.  Note that only very expensive cards/monitors can show more than 8 bit color.  This is an issue with deep color formats, as the monitors cannot show you all of the color gamut actually contained within the image.  In fact, as of this writing, the best that can be done with color display is 10 bit.  Even so, the difference between an 8 bit color component display and a 10 bit display is striking.  More than likely higher bit depths will eventually filter down to less expensive cards, but that will take a few years to happen.

Procedure

So, the goal is to mitigate these issues in order to deliver Before you start ANYTHING, the first step is to make sure that you work with the driver you need, or the driver defined by your facility.  Make sure that all of your input imagery is created or edited with that setting, and make sure that setting is maintained as you work throughout the process.  Changing that setting mid-stream will break the process.

The next step is to make sure that your various software programs are all in sync as well, or that they pass through the data unmodifed.  In the cases where that is unknown, or worse, cannot happen, then you will need to make some asset conversions manually, maintaining the highest quality possible.  That means that the JPG format and others should be avoided at all costs!

Third is to make sure your target format supports deep color, such as Open EXR.  You can choose different work formats as long as they do not contaminate the image data, or reduce its color space or color depth.

Following the above recommendations should help you deliver the highest quality and most versatile output possible.  This is not a comprehensive discussion (this is a glossary entry after all.)  Best to continue the research on your own, especially if you are a professional in your field (and if you are, you should already know this stuff, right?)

 

Category:Glossary

Category:Rendering

Category:Compositing

Categories: