AEC magazine looks at two visualisation specialists that create designs from simple sketches and block models.
This aerial image by visualisation specialist Atelier York was designed with minimal architectural input in a single week, covering an existing site to which the client proposed changes.
It had to show not only where the new development south of the town would be, but also the connection between it, the hill and the railway line running north connecting the two.
Working from provided survey data, most of the architectural information came in the form of hand-drawn sketches of the proposed buildings, pathways, roads and other changes.
So far, so vague — even the aerial photographs were outdated — so the sketches were traced over in CAD to create splines to connect cleanly with existing spline data from the CAD drawings, and images were used for rough referencing. “The existing buildings were created by extruding the spline data, converting to Editable Poly, insetting the roof elements and extruding back down to create the illusion of parapets,” says Alex York, creative director and founder of Atelier York.
“Along with some subtle variation in tone this enabled us to create a relatively detailed and random distribution of structures. Proposed buildings were modelled from scratch and rotated and moved into place manually to exactly match the architect’s plan.”
Paths, roads and the railway line were created using a mix of CAD data and manual modelling. “Quite a bit of artistic license was afford to us for areas not covered by either the existing survey data or the proposed masterplan,” says Mr York.
A combination of Autodesk 3ds Max, Vray, Mental Ray, Forest Pack Pro and Photoshop was used to create the finished image. Many of the matte channels and individual passes for each element — buildings, roads, paths, trees, hedges — were rendered and composited in Photoshop, or Forest Pack Pro. “We manually painted much of the grass and general landscape to make the various zones within the image appear random,” says Mr York. “We then used some loose artistic techniques in Photoshop to give the image a more illustrative feel, as requested by the architect, including exaggerated but aesthetically useful depth-of-field, subtle noise and a second version of the image that fades to white at the edges.”
This modern residential planning visualisation is by Wonder Vision, and is typical of the work it does to assist architectural firms win pitches and competitions.
“The main [goal] is grabbing people’s attention, looking to make an impression with the visuals,” says director James Elderton. “Planning visuals tend to be more scrutinised.”
Projects can arrive at the specialist visualisation agency in all shapes and forms, from basic sketches on paper to fully built Revit models. “It’s a real mix — some AutoCAD and some 3D.
“Some clients have a sketch and want a design. Some clients, mainly architects, use Google SketchUp, which we can work with.
“Sometimes, if the client is looking for high-end visualisation, the model isn’t quite up to standard, so we remodel it or amend their model,” says Mr Elderton.
“We work in Maxwell Render software. We take their model as a base and put in extra details that bring it to life and populate the scene with furniture and light fixtures and fittings.
“It is the best way to get the high-quality image they are looking for.
“When we model it ourselves it allows us to control what is split out into different bodies. We will try and model it as separate parts; for example a wall is a different part so we can apply the correct material and texture to that independently.
“It speeds up the process for us to be able to do that.”
Wonder Vision uses a wide range of techniques within Maxwell to illuminate the right colours and tones, including interiors.
HDRI lighting generates accurate tones and reflections, while the Physical Sky System within Maxwell has real life geographically accurate data.
The company also uses retouching techniques in Photoshop, adding contrast layers, highlighting individual textures, and using the z-depth channel, rendered from Maxwell, to use as a layer in the retouch to add depth and atmosphere to really grab attention in the final image.