A PHOTOGRAPH of a painting, even a photograph of a sculpture, can give a quite adequate sense of its character. But a photograph of a piece of architecture never can. Even a filmed documentary generally falls far short.I'm not fully certain why this is so, except that the light is often artificially set up for the camera - while walking into a building you have to adjust to dimmer light, to shadows, reflections, and nuances. Moreover, the film image, however kinetic, is still two-dimensional, strictly limited to the confines of the frame - while the human eye takes in, with the slightest movement, a wide and peripheral scope. And it is also able to take in simultaneously a large overview and the minutest details. A cam era has to move from one to the other, and often loses the context of the detail in the process. For these reasons alone, books, slide shows, and films about architecture are rarely substitutes for visiting buildings. And in that lies some justification for the onerous and expensive business of traveling. Take the work of Antoni Gaudi (1852-1926), for instance. I had long owned a book of photographs of Barcelona's Park Guell, one of the incomplete masterpieces of the Catalan architect. But the image I had of this park, with its fantastic variety of buildings, walls, closed and open spaces, was fragmentary. From the photographs, I knew there was an extraordinary serpentine arrangement of park benches decorated with colorful broken ceramics. I knew there was an impressive flight of steps and an astonishing sculpture of a lizard. The photos showed that there was a tower on a building capped with a red toadstool. And almost incongruous with all the fantastic freedom of these details, there was an array of classical, fluted Doric columns supporting a strange ceiling. But it was only in the fierce midmorning heat, after climbing up out of Barcelona through long steep streets and arriving finally at the meandering stone wall that surrounds this hillside park (built originally as a garden suburb for exclusive middle-class homes), that I could see how each photograph in my book pictured elements that led into and were dependent on each other. Everything was much closer to everything else than I imagined. The steps went up each side of the lizard and led directly to the classical columns. The "ceiling" the columns supported was, in fact, the underside of a large sandy area with a continuous snake of seating all around. Thus, these remarkable benches were actually the edge of this open, high-up space (meant originally for theatrical performances) and acted as a barrier that prevented people from falling off the edge. Gaudi, much in tune with his times, believed in an architecture that was a complete work of art. His work incorporated color, craftsmanship, the old and the new, down-to-earth practicality, and elatedly inspired imagination, planning, and spontaneity. It was vernacular and popular but at the same time aimed at the highest solemnity. It could suggest music, painting, and sculpture, or include these literally. The bell towers of his best known work, the church called the Sagrada Familia (still not remotel y finished more than 100 years after he started it), have hundreds of lipped openings designed to deflect the polyphony of the bells to the crowds below. One writer on Gaudi has described them as "huge resonance chambers." Visually these unusual towers stand as monuments to a unique architectural vision. Gauds architecture is entirely unlike anyone else's, even though Barcelona at the turn of the century had quite a few "modernista" architects of startling originality. He, however, seems to have had an isolated spirit that questioned virtually every established architectural norm. Above all, he was determined to get back to the "origins" of architectural structure - aware that "originality" means just that. Amazingly, Gaudi found wealthy clients keen to support him and willing to allow for the unconventional procedures he needed in order to make the sort of architecture he was inspired to make. Eusebio Guell, a textile industrialist, was his most important patron. Gaudi designed an extension to Guell's already large house on a Barcelona street; the hillside "park" already mentioned; and about 4 1/2 miles out of Barcelona, a church for a colony built by Guell for his workers. Like the park, this church was never finished. In fact all that exists is the crypt - the lowest and darkest part of any church, and even that took a long time to conceive and build. When we saw it, a wedding party had just emerged and was being photographed. A superb work of art it may be, but the bride and her retinue served to remind me that a building worth anything is useful. GAUDI took to extremes the basically medieval notion of an architect working alongside his builders and craftsmen, experimenting, altering, responding to the demands of the building as it evolved. Plans seem to have remained very largely in his head. But if such spontaneous and unpredictable methods suggest an architecture that must be piecemeal or improvisatory, the facts are otherwise. It is clear that his aim was to develop structures based not on other buildings, but on natural principles. In particu lar, he continually referred to the tree as his "master roots, trunk, branches, twigs, and leaves. The Guell crypt exploits this idea. Its supporting stone and brick columns seem to grow out of the ground in emulation of the leaning pines that surround the building. There isn't a vertical or a right angle to be seen. Gauds most basic structural element is the parabolic arch, which might be roughly pictured as the rise and fall of a jet of water fountained into the air at a leaning angle. A parabolic arch does not meet the ground at right angles. The towers of Sagrada Familia are based on this structure, as are the roof-supports, arcades, arches, and entrances in his other buildings. Here is an architecture that escapes from the rigid tyranny of square, rectangle, vertical, and - though not quite so easily - horizonta l. The Batllo House (currently up for sale at the mind-boggling price of about $93 million) tells the same story. Here, pre-Gauds intervention, was a conventional house wedged between others on an elegant street in Barcelona. Gaudi transformed it. He set the lower windows in projecting, egg-shaped frames. Slender columns like bones divide them and support overhanging balconies. The wall-face undulates like rippling waves, and its upper reaches are surfaced with blue and green ceramics. The roof has been lik ened to a dragon, with a contorted backbone and scaly tiles. The interior (and here I can only go by photographs) is no less astounding: banisters like vertebrae, ceilings like currents eddying in deep water.Everywhere Gauds buildings seem fluid, twisting, growing. How could such architecture be strictly foreseen and rigidly outlined before it was built? The answer is that it couldn't and wasn't. And even in his day, and in spite of his reputation, Gaudi sometimes came into conflict with the city's planning authorities. How on earth would today's brand of fastidious urban bureaucrats and picky conservationists have dealt with such a rebel?