Grid Pro Manual - Optimizing Grid Pro

From VidvoxWiki

Grid Pro Manual
Getting Started
The Basics
  1. Managing Media
  2. Playing Movies
  3. The Mixer
  4. The Sequencer
  5. Using FX
  6. The Preview Screen
Advanced Features
  1. The Gesture Slider
  2. Controllers
  3. The Fontsynth
  4. The Audio Player
  5. The DVR
  6. Shortcuts
  7. The Preset Manager
Optimizing Grid Pro

Contents

Playback Settings

Frame Drive Mode (aka no audio)

When using media files that have no audio track, or for situations where the audio track is not needed for a performance, GRID Pro can switch to a more efficient playback engine known as Frame Drive Mode. On some computers, Frame Drive Mode provides smoother playback, especially under heavy processing loads. Frame Drive can be activated by clicking on the Frame Drive button in the General Settings section of the preferences pane.

Flash and Quartz Composer Support

For performance reasons, by default playback of Flash movies (.swf) is disabled in GRID Pro. To use Flash movies in GRID Pro, open the preferences pane and in the General Settings tab, click the Flash Enable button.

Note - as of GRID Pro 1.0.3b3 this preferences option is no longer needed, Flash & QC documents are automatically detected when triggered

Fast File Loading

In order to lower latency when triggering movies, GRID Pro will load the first few frames of each movie into RAM when the movie is initially loaded into the Media GRID. This process can be very time consuming. Turning Fast File Loading on will speed up the loading process at the cost of increased latency when triggering movies. While it is not usually recommended for use during performance, this feature can be accessed in the General Settings in the preferences pane.

Resolution Settings

The Resolution Settings in the preferences pane are used to set the resolution at which GRID Pro will process each video frame. All video files will automatically be scaled to this resolution when played back in GRID Pro. For this reason, to prevent scaling, when [preparing media for GRID Pro], it is recommended that all movies are rendered at the resolution you will be processing at.

Resolution is set in terms of the number of horizontal and vertical pixels. A higher resolution setting will result in a higher quality video output. When processing at a high resolution, each stage of processing (mixing, FX, etc.) will be more processor intensive.

Aspect Ratio

Regardless of the processing resolution, video output from GRID Pro is automatically scaled to fit the resolution of the designated hardware display. If the processing resolution has a different aspect ratio as the output display, you may wish to set a custom aspect ratio for the video output. This will place either horizontal or vertical black bars around the processed video to make it fit the custom aspect ratio.

Display Settings

Settings for the interface video previews and main output display can be configured in the Displays Settings section of the Preferences. The optimal display settings will vary from computer to computer, depending on the processor speed, free RAM, and graphics card of the computer. Finding the right settings can greatly improve the performance of GRID Pro on your computer.

Interface preview options

The interface preview screen can operate in one of three graphics modes: Quicktime Acceleration, OpenGL Accerlation, or no hardware acceleration. The preview window can also be put into a low priority mode, or turned off completely to save system resources for the main output screen or DVR module. ...

Main video out setting

The main video out is used to show the final video output from GRID Pro on a television, projector, or a second monitor. The main video out can also be displayed in a floating window on your main monitor. Like the preview window, there are three possible graphics modes that can be used by the main video out: Quicktime Acceleration, OpenGL Accerlation, and no hardware acceleration.

Preparing Media for GRID Pro

Recommended codecs

To get the most out of GRID Pro, it is important to prepare your media files by converting them to a format that can be easily decompressed and processed by a computer. Here is a list of some video codecs that are well suited for real-time processing.

PhotoJPEG

PhotoJPEG is the most common video codec used by VJs. Due to it's low data rates and fast decompression, PhotoJPEG is ideal for video scratching and playback of multiple video files, particularly on laptop hard drives which tend to be slower than hard drives in desktop computers. However, PhotoJPEG is a lossy codec and even when using a high quality setting (90% or above) the image quality is clearly worse than DV or a well-made DVD.

Component video

When video quality is the highest priority, a lossless video codec is needed. Component video stores video in an uncompressed format where the information for each pixel is preserved exactly. Because there is no compression, component video files contain a great deal redundent infromation. As a result, component files are very large with a very high data rate, making playback fairly hard drive intensive.

DV

DV is a both a video codec and a tape format. DV is considered to be a convenient high quality option for use with consumer and prosumer video editing suites such as Apple's iMovie and Avid Express. It uses a fixed bit-rate, high quality spatial compression which amounts to roughly 3.6 Megabytes per second or 4 minutes per Gigabyte. Although VJ softwares support playback of DV resolution files, due to it's slow decompression rate DV is not usually recommended when performing CPU intensive operations such as video mixing and effects processing.

What codecs not to use

Video compression schemes that utilitize temporal compression, such as Sorenson, MPEG-2, and MPEG-4 are usually not well suited for use in GRID Pro. In addition to being computationally intensive, most temporal compressions are lossy, meaning that they discard information, which can result in a significant loss in image quality. Additionally, it is not possible to randomly access any frame in a movie file - only keyframes. This makes temporal compressions particularly bad for video scratching and high speed playback.

Note on YUV vs. RGB video processing

You are probably accustomed to thinking of digital images being made of of dots [pixels] that contain varying amounts of red, green and blue. However, most video that we see is a varient of another colorspace known as YUV. In YUV each pixel is still made up of three peices of information. Y is for the brightness of the pixel and the U and V values are combined to determine the color information. This may sound like a rather contrived way to represent an image but all DVDs, television broadcasts and DV tapes are already in a YUV colorspace.

There are several reasons for the YUV color scheme. Early specifications for televisions only accounted for black and white video signals. When color television was introduced years later, the new specifications had to be backwards compatible, meaning that color television broadcasts would need to work on both black and white and color TV receivers. YUV solves this problem because the color information is sent on a seperate signal from the brightness information, so black and white televisions simply ignored the color information. Because the human eye more sensitive to differences in brightness than differences in color, YUV makes it possible to reduce the quality of the color information without affecting the brightness information. Most digital video formats store either 2 or 4 times more brightness information than color information, without a noticable change in image quality. As a result, video codecs that use a YUV colorspace will use far less disk space than their RGB counterparts. Because processing of brightness and color infomation can be done independently, working in a YUV colorspace is generally faster for operations such as motion analysis and color correction. A task as simple as fading to black would be roughly 3 times more computationally intensive in RGB than it would in YUV because the same operation would need to be done to each of the red, green and blue channels seperately, whereas in YUV only the brightness value needs to be adjusted.

Other optimization techniques

In order to get the best performance out of GRID Pro, it is sometimes useful to turn off unnecessary system activity. To get an idea of all the little programs that are running in the background on your Macintosh, open the system Activity Monitor located in the Applications/Utilities/ folder.

In this section we will provide a few examples of system processes that you can deactivate to save processor cycles and improve performance of GRID Pro.

Disabling network connections

If a network connection is not needed during performance, it is a good idea to turn deactivate these system completely. A few examples of network activities that can be turned off are:

  • Airport
  • iCal
  • Mail alerts
  • Date & Time auto-updating
  • Filesharing (such as Personal FileSharng and Apple Remote Desktop)

Shadow Killer

The Mac OS a is very slick, modern looking operating system. One of the things that makes OS X look the way it does are the distinctive drop shadows on each window. Drawing these drop shadows can take away a significant amount of resources from the computers graphics card. By removing these drop shadows using a program such as ShadowKiller by Unsanity, the graphics card can spend more time displaying video output and less time drawing purely cosmetic shadows.