What are survey accurate visual simulations?

Visual simulations are a great tool to accurately understand the visual impact of a development or change in the landscape. An image really is worth a thousand words. However, it is important to understand the limitations of visual simulations and the purpose for which they are created.

Over the past few years, considerable time and effort has gone into developing the tools and techniques around buildmedia’s visual simulation methodology. I’ve been lucky enough to be involved with many great projects here and in Australia, which have allowed me the opportunity to test, scrutinise and refine the workflow using previous experience in programming, surveying, aerial mapping as well as CGI.

Very recently, questions regarding the field of view, lens, panorama projections and print size have been popping up, and what I have found is that there is a lot of misinformation and misunderstanding regarding these subjects in the landscape architecture community. Below I hope to demystify some of the visual simulation process and clear a few things up.

Mangatangi Mine - 90 degree Visual Simulation - Existing View
Mangatangi Mine - 90 degree Visual Simulation - In Operation
Mangatangi Mine - 90 degree Visual Simulation - Rehabilitated

What is a visual simulation?

Visual simulations in New Zealand are defined by the New Zealand Institute of Landscape Architects – Best Practice Guide for visual simulations, which states that:

“The primary purpose for a visual simulation is to accurately portray, in a realistic manner and context as possible, a modification or change in the viewed landscape.

Visual simulations illustrate two dimensional views of a proposed activity from a particular viewpoint as depicted in a photograph.

Visual simulations are not ‘real life’ views. They are useful tools.”

I would like to stress that a visual simulation is a photograph of a view that is taken from a specific viewpoint, at a known field of view. As it is simply a two-dimensional image it does not include any of the depth queues that would be apparent when visiting the site in real life. Therefore, one should really think of it as a photograph of a proposed design rather than a recreation of human vision.

Human vision is far more complex than what is captured in a photo. For example:

we see in stereo

our detailed vision is approximately 3-5 degrees off centre of our focus

our peripheral vision is great at capturing movement but poor at detecting colour and detail

our perception of an environment is also influenced by our experience of the world.

It is a very complex process to capture all within a single image. If we were to represent exactly what a single eye would ‘see’ focusing into the centre of a visual simulation it would look something like the example below.

New-zealand-institute-of-landscape-architects-Best-practice-guide

Visual simulations are not “real life” views.

Meridian Mt Munro Wind Farm - 90 degree Visual Simulation

Your eyes naturally focuses only on the target - this is called Foveal vision

Your eyes only see the target point in colour.

As this where the highest amount of cone photoreceptors are concentrated on the retina.

Plus we also have a point on our retina where there are no photoreceptors called the blind spot.

When you look at a scene your eyes naturally dart around and build up a mental 3d "map"

This phenomena is called "saccade" eye movement.

Saccade is quick, simultaneous movement of both eyes building up a mental picture of a scene – https://en.wikipedia.org/wiki/Saccade

Field of view

Before I explain the lens we must first understand what ‘field of view’ (FOV) means for human vision. The field of view is the extent of the observable world that is seen at any given moment. In the past, visual simulations have been created using a FOV from anywhere from 40 degrees up to 124 degrees depending on the requirement. In earlier years a FOV of 40 degrees was common because it matched a 50mm camera lens.

environconhumanfact_fig2_3Now upwards of 124 degrees are accepted in the industry as it is determined approximately by the same FOV as human two-eyed stereo vision.

Note, that if we include monocular peripheral vision then it is a greater FOV than 124. Depending on each individual viewer’s facial structures, it can sometimes extend upwards to approximately 180 degrees.

Bunnings New Lynn - 90 degree Visual Simulation - Existing

Bunnings New Lynn - 90 degree Visual Simulation - Proposed

90 degree visual simulation showing the proposed Bunnings Warehouse New Lynn, Auckland.

The camera lens

From the NZILA Best Practice Guide:

“Camera lenses of different focal lengths create images with different fields of view. None of these fields of view are the same as the human field of view. A camera lens does not encompass the same horizontal and vertical ‘degrees of arc’ that are captured by human binocular vision. This is why a picture taken with a ‘non-human’ receptor such as a camera does not represent what we actually see.”

Which lens to use when capturing the photography for the visual simulation can now vary widely. In the past a 50mm lens was considered the standard for visual simulations as it was the most readily available lens and has minimal distortion, giving a natural unforced perspective and a similar field of view to the cones in our eye. It was also easier to align a single frame to the digital model using older software packages.

Today, the industry now accepts many different field of views from 40 degrees (a 50mm lens) to 124 degrees (a 9.58mm lens – though there are not too many of these available). The lens has less significance because a panorama is stitched together using many photographs. If you were to stitch a panorama using a 50mm lens and a panorama generated using a zoom lens of 80mm the only varying difference between each panorama would be the resolution of the final image and the number of photos required to generate the same panorama.

robotic-camera-panorama-building-vis-si

Various comparative focal lengths and field’s of views within the same scene.

Panorama and mapping projections

This is a technique of photography using specialised equipment or software that captures images with elongated fields of view.

A panorama (in terms of visual simulations) is created by stitching together many photos to generate an image with a wider FOV than that of a single photograph. There are many software packages to stitch together panoramas and some are superior to others. At buildmedia we utilise photogrammetric methods as this is the most accurate and can provide a measurable result.

A panorama can be presented in many forms using different mapping projections. A map projection is a systematic transformation of the latitudes and longitudes of locations on the surface of a sphere or an ellipsoid into locations on a plane (2d page). There are many projections available depending on the software packages used, however, in the visual simulation industry there are two that are widely used. These are rectilinear and cylindrical.

buildmedia-geometry-of-panoramas-1080x284pxRectilinear and Cylindrical

Each projection will stretch and manipulate the images in different ways, and therefore must be viewed differently.

NZILA Best Practice Guide:

“The ideal method of viewing a cylindrical panorama is with the image presented in a curved format, viewed at the correct radius from the centre of the curve (distance D in Figure 14A). If mounted on a flat surface, it should ideally be viewed by one repositioning along it’s length, maintaining distance D as one moves.

Where a planar or flat panorama is viewed, one must look directly at the centre of the image without moving one’s head, and rely on peripheral vision to see the extremities of the image. Movement of the head to view the extremities of the panorama will result in a viewing distance that is larger than the optimum distance of D – shown as distance E in Figure 14B”.

A word of warning: If you are after a single projection that will map a spherical (even partial) panorama on a flat surface without bending lines…you won’t find one!!

Mighty River Power - Puketoi Wind farm 90 degree Visual Simulation - Existing

Mighty River Power - Puketoi Wind farm 90 degree Visual Simulation - Alignment

Mighty River Power - Puketoi Wind farm 90 degree Visual Simulation - Proposed

Print size

And lastly, print size. This is quickly and easily explained in the NZILA Best Practice Guide for visual simulations:

“A 50mm focal length lens using 35mm film would produce a 36x24mm image that would need to be viewed approximately 50mm from the eye. A simple scaling up of the image dimensions by a factor of 10 would result in an image 360x240mm and with a correct reading distance of 500mm. In other words, if a photograph is taken with a 50mm lens on a 35mm camera and the image is printed at a size of 360mmx240mm, standing at the point from which the photograph was taken, it will be possible to hold up the image at a distance of 500mm from the eye and see the photographic image line up with the real scene.”

Tim’s time is spent managing a team of talented 3d artists, while also researching and developing new workflows and exploring the latest in 3d technology.

Testimonial Tim Johnson
Tim Johnson
Studio Production Manager