What makes buildmedia’s visual simulations so accurate?
At buildmedia our visual simulations are developed to portray the development at the correct scale and position, while at the same time being scientifically produced using measurable methods. Each ‘vis sim’ is produced in a way that minimises possible human error through each step.
What makes our visual simulations so accurate? Over the many years of development our visual simulation methodology has been tuned, tweaked and optimised. Every year we improve on the methodology to achieve the highest level of accuracy through a scientifically measurable process. Below I have listed some of the steps that enable buildmedia to attain the highest levels of accuracy in producing visual simulations.
Bunnings New Lynn - 90 degree Visual Simulation - Existing
Bunnings New Lynn - 90 degree Visual Simulation - Proposed
We use a robotic camera head
We use a robotic tripod that rotates a full-frame camera around the optical centre of the lens, which in turn eliminates any lens shift within the panorama, removing distortion and parallax error. It is very important to remove parallax to get the absolute best results and ensure that the multiple photographs overlap and line up accurately within the final high-resolution panorama.
The robotic camera head also accurately obtains shots at exact degree increments, removing the possibility of human error. Around 70-100 individual photos make up the final panorama composite that is many hundreds of megapixels in size. These extremely high resolution panoramas are simply impossible to obtain using conventional photography.
We use relative tie points to generate panoramas
Instead of using the Photoshop built-in panorama creator, we use similar methods to an aerial mapping company and how they would generate aerial maps. We use relative control points to stitch images together eliminating any distortion of the lens and providing us with a measurable result. The final panorama accurately replicates a rectilinear image of a known field of view. Relative control points are features that coincide between source imagery. Using this method, we can then calculate the accuracy and check for errors.
We use high-accuracy RTK GPS equipment
We do not use handheld GPS devices – we use high-accuracy surveying equipment to capture the camera position and tie points that are used to align the 3d model. Using RTK GPS we can tie the survey into known geodetic marks located near the site and match coordinate systems used by other consultants. This allows us to align the camera position accurately to GIS information and validate the accuracy. We produce a result that we know is as precise as it can get!
Comparing an accurately-mapped GIS model to a visual simulation capture.
We use accurate lighting and materials
Over the years the team has spent considerable time developing render techniques and workflows to increase the realism of images within a budget. If we have plenty of time we can create some highly-realistic-looking images!
We have a robust and tested workflow
Over the years we’ve spent numerous hours developing in-house scripts and tools to accurately align and test the alignment of the 3d model in the panorama. These eliminate the possibility of human error and provide us with an alignment that is as accurate as the survey.
What are the visual simulation limitations?
Visual simulations are photographs which lack a number of depth cues, are not in stereo and only capture a specific moment in time. They are great tools for the landscape professional to help make an informed decision. As per the New Zealand Institute of Landscape Architects Best Practice Guide:
“The primary purpose of a visual simulation is to accurately portray, in a realistic manner and context as possible, a proposed activity, modification or change to the viewed landscape.
“Visual simulations are not ‘real life’ views. They are however, very useful tools to assist in the assessment and decision-making process whereby better informed and more transparent judgments on appearance and effects can be made.
“Visual simulations illustrate a two-dimensional view of a proposed activity from a particular viewpoint as depicted in a photograph – not as it would appear as a three-dimensional image as seen in the field with the human eye”.