Soon you will be able to visit virtual reality versions of the world's great historical sites. But will the heritage industry turn them into another Disneyland? Janet Robson reports
W ant to walk around inside Stonehenge, or wander through the caves of Lascaux? Too bad. In the name of preservation, access to many of the world's greatest historical artistic achievements is barred to ordinary mortals. But we should soon be able to enjoy artificial versions as good - or better - than the real thing, as the heritage industry embraces virtual reality.

The National Trust is carrying out feasibility studies on introducing VR versions of the 12th-century Fountains Abbey in Yorkshire and of the Avebury stone circle, and English Heritage has just commissioned a VR version of Stonehenge. Professor Robert Stone, director of the UK Centre for Virtual Environments, which is creating VR Stonehenge, explains that it will allow the viewer to fly in to the site and see it in great detail. The model will primarily be a research tool, allowing archaeologists to see how the monument was constructed, but it will also allow "tele-tourism". Intel is a key partner in the project and it wants to put Stonehenge on to CD-Rom and the Internet.

At the world's first Virtual Heritage convention, held recently in Bath, VR projects on show included models of Pompeii, an ancient Egyptian fortress, Roman baths, Cluny Abbey and the prehistoric cave paintings of Lascaux. One reason for all this activity is the rapid development of VR technology. "Many VR tools such as software, modelling tools and advances in networking have been developed only in the past couple of years, or even months," says Eben Gay, a VR consultant. "We've now reached the stage of being able to do real work."

The cost of high-level platforms is falling, which will allow widespread use of VR in schools and museums. Another spur in the UK is the Millennium Fund, which many heritage bodies see as a one-off bonanza to provide funding for VR projects.

The choice of subject matter demonstrates one obvious advantage of VR in the heritage industry; it is a great way of enabling people to see buildings and sites that no longer exist. Cluny Abbey was destroyed in the French Revolution, the Egyptian fortress at Buhen was submerged by the Aswan Dam project in 1970, and Lascaux has been closed to the public since 1963. And, despite the popular image of VR, people will not necessarily have to wear helmets to visit heritage sites. Most heritage VR projects display 3D models on a flat computer screen.

At present, despite the technological advances, there is still a trade- off in VR between interactivity and the quality of the graphics. VR reconstructions involve building up 3D shapes on the computer out of polygons, and then rendering surface textures and patterns on to these shapes. The more detailed the rendering, the more realistic the model. This is a laborious job - a typical model runs at 30 to 60 frames per second and it can take up to an hour to render a single frame. The problem is that the model may end up being too big to run in real time, so truly interactive VR models that allow the viewer to move around at will have to sacrifice some of the detail.

IBM's solution has been to produce VR models using computer animation techniques. The result is a short video "fly-through" of a building. Its first video, made four years ago, was a two-minute fly-through of Cluny Abbey, made to be shown in the visitor centre on site. Since then, IBM has worked on VR versions of the Roman baths at Lutece, outside Paris, and of the Frauenkirche at Dresden, which was destroyed by Allied bombing in the Second World War. The result is glossy but controversial. VR geeks reckon it is not really VR at all, since it is neither interactive nor "immersive". Heritage professionals and academics balked at the soundtrack on the Lutece video, as a husky male voice purred about diving into warm baths and being rubbed down by handmaidens.

Authenticity is the biggest concern of the purists. When reconstructing a historic building, should unknown elements be left incomplete or be borrowed from elsewhere? In the reconstruction of the Roman baths, IBM filled in the gaps with fresco material from Pompeii and some additional material from the Baths of Diocletian in Rome, both of which were historically justifiable. Somewhat incongruously, the seating looks like Seventies MFI. But if all this gives visitors to the site a better idea of what it would have been like to be a Roman going to the baths, does it matter that the reconstruction is not entirely historically accurate?

This is not a simple equation. "All history is in some sense fantasy," says Bryan Mills, "either the historian's or the reader's." Mr Mills is part of the consultancy team developing an ambitious project in Northumbria, which will include a VR reconstruction of Hadrian's Wall. One problem is that the requirements of academics who want to study a site in depth and those of tourists making a quick visit are completely different. A typical museum or visitor centre exhibit is looked at for only three minutes.

Professor Benjamin Britton, who has spent four years creating a VR version of the prehistoric cave paintings of Lascaux in southern France, agrees that reconstruction is subjective. The Lascaux exhibit was the hit of the Virtual Heritage convention, combining historical accuracy, tremendous detail and artistic imagination. It is also extremely beautiful and compelling. Professor Britton's exhibit is not accompanied by any text or educational soundtrack, but by ambient cave sounds and flute playing. He wants to allow people to experience the site in their own way.

Other projects are attempting to fulfil more obviously educational ends. Virtual Pompeii is a reconstruction of the Theatre Complex of the ancient city, being developed in collaboration by the Archaeological Institute of America, Silicon Graphics and the Carnegie Mellon Simlab. Professor Lynn Holden, of Simlab, explained that the project will attempt to combine an authentic reconstruction with written and spoken knowledge. Virtual Pompeii's interface allows viewers to "pick up" objects, which will reveal written information. Each object is linked to an "agent", a historical character who will speak to the viewer. This idea of a virtual guide is also being used in the fortress of Buhen. Viewers will be able to take a VR tour of Buhen, accompanied an Ancient Egyptian scribe, who will answer their questions.

The purists fear that VR will turn the world's greatest historical monuments into Disney Heritageland. Their fears were fuelled by a VR game demonstrated by IBM at Virtual Heritage, in which players gained access to an Egyptian pyramid by blasting the artefacts to smithereens. But VR proponents point to the academic benefits. For instance, the technology can be used to present alternative views of reconstruction, enabling archaeologists to test their theories about a site. Eben Gay pointed out that only when archaeologists started constructing the VR model of Buhen did they realise that some of their ideas about how the fortress had been constructed did not work in practice.

It will also be possible to disseminate knowledge by passing VR around the Internet. In the networked future, children will be able to experience heritage sites at first hand from their classrooms, examine artefacts, call up archive material and greatly increase their understanding and experience of history. "VR will become the de facto educational standard," enthuses Professor Stone.

But even some VR pioneers believe the ability to disseminate VR heritage should be handled with care. Professor Britton, for instance, has resisted pressure to put Lascaux on to CD-Rom or the Internet. He believes these works are creations of a communal human experience and he wants people to experience them communally, by going to a museum. CD-Rom and the Internet encourage solitary and casual browsing which may devalue what is being experienced. "It is vitally important to retain the specialness of cultural places that people care deeply about," he says.