Tech Briefing: Projection Systems

Be the first to comment on this article

Using multiple screens and sources in a projection installation requires the images to be electronically processed and managed. Peter Lloyd explains the basics of the technology.

Latest News

Entries to the AV Awards 2012 close this Friday

AV Europe editor in Madrid

Market Survey – Distribution: Heading for the cloud

Big screen presentation using a mix of multiple projection and subsidiary images in the Information and Communications Pavilion at Expo 2010

In its simplest term, a projection system is a projector, a screen, a stand, a PC (or other input device) and a screen (or a white-washed wall).

But there’s a whole world of projection and image management systems which both allow multiple images (and multiple projectors) to be used on a single screen and manipulate pictures so that they fit on irregularly-shaped surfaces. Used in boardrooms and live events, they let picture-in-picture ‘windows’ show remote camera or conferencing links – and they can even be used to project on and track objects as they move across the stage.

Using multiple projectors
People started using multiple projectors on a single screen to increase brightness. The idea was pioneered in the multi-image business, when the only way to get sufficient screen brightness on large screens from slide projectors was to segment the screen into small areas and use a projector on each. Producers soon discovered the creative possibilities of using overlapping images on multiple screen areas and the whole multi-screen/multi-image concept was born.

When video projection took over, most live event shows (and a lot of visitor centre and experiential installations) moved away from the widescreen formats used in multi-image to 4:3 aspect ration ‘tv’ screens that worked easily with PowerPoint and video footage.

‘Simple’ widescreen
In corporate communications, the trend towards very wide screens started in live event staging, where show designers for both stage events and exhibitions (such as motor shows) realised the potential impact of using multiple 4:3 or 16:9 aspect ratio screens to show products, convey emotional messaging, or simply combine a lot of information elements on screen.

‘Simple’ multi-screen is still widely used for events such as awards ceremonies and shows at visitor centres, where the producers want to devise something that has more audience impact than the screens people have at home. It is also used in applications such as simulation, gaming and the creation of immersive environments.

The simplest way to project a multi-screen image is either to use a device that ‘splits’ a high-resolution picture edited to the right overall aspect ratio across the number of screens required, or to use multiple synchronised sources (PCs or video players) that play ‘their bit’ of the picture. The images are then butted together where they join (the least complex way of doing it, but one which demands very precise mechanical alignment of the projectors), or slightly overlapped and ‘blended’.

Edge blending
This blending – known as soft edge masking (from multi-image days), soft edge blending or simply image blending – uses computer manipulation of the image to gradually fade one picture into the other.

In the days of slides, producers used a ‘soft edge mask’ – an additional film chip placed in the slide mount that had a matrix of very thin black lines covering the picture area that was to be masked out. The ‘partner’ slide had a mirror image mask inserted over its picture. Then, as now, slight variations in light areas of the picture – such as the sky – could give the game away.

Doing the same thing electronically, applying a mask by removing pixels from the image being projected, uses a lot of computational capacity in order to modify a series of images being projected at 50 or 60 fields per second, but is relatively easy to set up.

Several manufacturers market projectors with built-in edge-blending capacity, and there are a variety of ‘outboard’ systems available to do the job of taking in an image feed, splitting it and applying edge blends – including, where required, blends at both the sides and the top or bottom. The projectors used have to be colour matched and their light output has to be adjusted accordingly.

Information overlay
The next projected image manipulation to take root was ‘windowing’ or picture-in-picture (PIP), which uses more complex picture handling to create areas on the overall screen in which different images (data, graphics, moving pictures or stills) can be shown.

The initial impetus came from live event staging and rental, but the use of on-screen windows has spread into corporate meeting room installations (in which, for example, spreadsheet data and multiple video conferencing inputs could be combined on to a single screen) and simulation. Although the technology is apparently simple – after all, we are all used to windows on our PC or Mac screens – the projection requirements are complex.

First of all, the projection system has to set up the background (sometimes known as a canvas, as in Photoshop) complete with masks, colour correction so the pictures match, and a consistent brightness level.

Next, the system has to cut out the right sized and shaped ‘hole’ in the background for the source material going into a window – obviously, it can’t be projected over the background. It then has to configure the source material going into the masked out window so that it’s the right size, shape and pixel count and tell the projector where to place it.

It has to do all that on the fly and at high resolution. But projection systems also have to cope with the fact that producers like to move their windows around the screen and resize them in real time, and to use (and move) several windows at a time.

Add soft edging to the windows to create overlapped, apparently multi-layered, content and things are beginning to get really complex. Then add the ability to move and modify windows along pre-arranged paths (such as having a window make its appearance small top right, then moving along an arc to bottom left, growing in size as it moves) and there’s a lot of processing going on.

Windowing systems
In practice, there’s too much going on for most windowing to be handled in-projector, so it tends to be handled by external devices.

A relatively simple windowing processor is fine where set patterns of windows have to be called up and an input assigned to them. For example, many meeting room installations use a series of screen layout templates (with different options for different applications) that are pre-programmed with window sizes, positions and input selections. Such installations are pre-programmed and managed from the room control system.

Where the application requires dynamic windowing (with image areas whizzing around the screen and being re-sized as they go) complex external systems using frame stores, image processing and vision mixing controls are required. Most can be pre-programmed, with a routine set up before the show runs ‘live’, but they are usually operated by a specialist team with manual over-ride (to cope with show changes) and cues set up for each show segment.

Image warping
So far, the projection system has edge-blended stills or video, created dynamic windows and filled them with content – all without missing a beat. The next piece in the jigsaw – almost literally – is image warping.

At its simplest, image warping involves ‘bending’ the picture so that it can be projected on a surface that’s curved in one dimension, such as a curved screen surface. Projecting a ‘flat’ picture on to a horizontally curved surface would mean that either the centre or the edges would be slightly out of focus (or both, if a compromise was aimed at). Electronically warping the image can cure that.

Things get more complicated when images (and particularly multiple, edge-blended images) are to be projected on to a convex or concave surface that’s curved in three dimensions (such as a planetarium dome, or a sphere). The pixels in the image then have to be ‘mapped’ on to the projection area by warping (ie. compressing and/or expanding) the image dimensions.

Using film technology, specialist lens were used to create distorted images that could appear correctly-proportioned when projected, and some sphere projection systems still use lens-based arrangements. Electronic systems, however, increase flexibility. So, for example, a simulation programme or specially-made presentation can be transferred from a dome or immersive environment to another.

Buying image management
Finally, some words of advice. If you need to edge blend projected systems, manage windows (and their content)  warp images on to an irregularly-shaped surface then have a hard look at what’s available and decide what you need to be able to do, now and in the immediate future.

Key factors to consider include programmability and future expandability. As a guide, most in-projector systems are capable of only limited upgrades, but ‘outboard’ systems can be changed and new features or capabilities added at will.

KEY SUPPLIERS

Manufacturers of projectors with built-in edge blending and image warping capabilities include Barco, Christie, Digital Projection, Panasonic, projectiondesign and Sanyo. Some projectors – including Barco’s IQ series – also have windowing capabilities and many offer picture-in-picture.

Suppliers of ‘outboard’ image processing systems fall into roughly two camps: Manufacturers whose products are designed for fixed or pre-programmed installations and vendors aimed at the live events market. The former includes suppliers such as Analog Way, Calibre, DWI, FSR, Kramer, nVidia, RGB Spectrum and TVOne, as well as Barco, Christie, Digital Projection and Sanyo. The biggest names in live event processing: Barco’s Folsom Encore systems and Christie’s Spyder range.

Your Comment

*