I had a lightning tour of the new version of Lyric software for the Chyron character generator last week, from Kristy Weir and Dave Wade. It’s an impressive box with a lot of functionality that most users might never see. We’re looking at its use in a News environment, with many different types of graphics to handle. It needs to be comprehensive, but still easy to automate and operate. The Lyric software tries to cover both 2D and 3D environments and copes pretty well, but there will be times when the combination of 2D and 3D will become complicated.
On the whole, the Lyric software shows its heritage of being a character generator. Chyron have a long history and despite having been in the wilderness for a few years (at least in the UK) they show that their experience shows through in the richness of the interface. The operator can do most anything they like to a graphic and an expert will be able to make speedy last minute changes. The interface gives the competent operator a real speed of operation that is hard to match by others. Images and fonts can just be dragged into the window without complicated imports. Macros may be recorded to do complex operations and automated further with a few edits. Transition modes and simple keyframing should be straightforward. There are glows and pixel shaders which show that they are trying to keep up with new developments, but there are many facilites that a sophisticated 3D user would be missing.
On the 3D side, the company has found the same problems as others in trying to bolt on 3D to a 2D interface. There are 3D models that can be imported, but not with all the common formats, neither of the popular fbx or Collada formats yet; there are some problems with scaling with items from other modelling systems and the animation values can be difficult to synchronize but this may just need to have a workflow set up to mange the import process. The objects that are brought in seem to work well, with textures UV mapped quite accurately. I found animations and lights difficult to control, with limitations in the keyframing but maybe I need a bit more tuition there; it’s early days.
Basic 2D text is not part of the 3D scene by default, which can be a benefit in some places, but it shows that the 3D is a separate thought process. Having the ability to create simple 2D pages does allow the software to be more comprehensive and many different styles of graphic may be dealt with inside one piece of software. Although VizRT has a full 3D engine, it has a separate character generator interface, Trio, which is able to restrict the operator to what they can adjust. Brainstorm recently took over the Aston character generator software. It too has a good 3D engine, but Aston also found that it is not a simple task to control everything in a 3D scene. There become too many options for a live operator interface and there has to be a compromise somewhere.
Animation is done using the normal keyframing system. Objects may be grouped together and flown around in front of the camera. It is possible to move the camera but its path cannot be seen in the 3D views. Splines can be used, with adjustment to the ease in and out but the interface is not the easiest to use. It has none of the acceleration graphs or easy adjustment methods and in some situations splines can be difficult to handle. The functionality seems to be there underneath though, so there could be plugins added later to adjust these parameters.
Automation, where I am mostly involved comes in several ways. The IntelligentInterface protocol has been around almost as long as I have (remember Chyron SuperScribes?). It is a series of simple messages that can set template fields, record pages, playback pages etc. There are events, either just page read effect that happens when a page comes on or flies off, keyboard events that happen under operator control or transition events that can be fired from other pages. Because the Chyron allows several pages to be on screen together, for example a lower third strap and a bug, when the bug comes on it could fire an event to the strap to change its view to match the bug content or positioning. It’s a clever enough system to allow for further pages to be added and receive the same effect. The original graphic may not need to be changed, there just needs to be some forethought on the groupings of different graphics types and to how their naming should be handled.
There is an iNews interface for the journalists; an XML plugin can be used to link to rss feeds or similar blocks of data; the database and Excel connections work well, if a bit verbose. There is a free plugin from Patrick systems to help with web based input; I haven’t seen the pro version but that adds some scripting as well. I was surprised not to see a basic rss text and images plugin, but it should be straightforward to bring this in with scripting. I suppose that I expected some open source script to be out there. Chyron are working on ways to bring data in more easily for designers, but a programmer could use the methods quite well for most automated feeds. There is vbScript for the programmers where they have access to a full ActiveX interface which provides ways to control the graphics in more sophisticated ways and for extreme cases there is a plugin sdk which allows access to the whole object model underlying the graphics.
The DVE functionality is really good. Several movie avi files can be played into a scene, but the Chyron can also have its own clip player based around Leitch hardware and facilities to bring in the clips with a DVE are quite comprehensive. One or two channels of clips may be used, they can be animated on the timeline and the layering of the clips and masks in relation to other text and object layers is extremely flexible.
As with all the other systems, Chyron are looking at touch systems now and maybe their event model might suit this better than others. I’m not sure that the full multitouch would be easily handled by a page based system; Brainstorm and VizRT are more likely contenders there. It needs a bit of programming power behind the scenes, but single touch and gestures should be straightforward.