Who can produce accurate visual representations

Accurate visual representation competence is not a single qualification. It is a split between three people: a chartered landscape architect who decides what to show and judges the effect, a land surveyor who fixes the camera positions and the scheme geometry to known coordinates, and a visualiser who builds and renders the model. The Landscape Institute's technical guidance on visual representation sets out levels of accuracy, and the higher levels require surveyed input that no amount of rendering skill can substitute for.

Which is why a beautifully rendered image can be inadmissible while a plain wireline overlay carries real weight at inquiry.

What the levels of accuracy require

The guidance describes a graded set of representation types, from simple annotated photographs through to fully verifiable photomontages. The distinction that matters to a decision maker is whether the image can be checked by someone else.

At the lower levels, a photograph with an indicative sketch overlay is useful for design discussion and for public consultation. It is not evidence of effect. At the higher levels, the image must come with a methodology that records the camera position to a surveyed coordinate, the camera height above ground, the date and time, the lens focal length, the sensor dimensions, the horizontal field of view, and the intended printed size and viewing distance.

Stated optics are what make the image falsifiable. Give an inspector the coordinates and the focal length and they can send someone to stand in the same place. Omit them and the image is a picture.

A common failing in submitted work is a montage at the right standard of rendering with no methodology appendix at all. The visual material then gets treated as illustrative, and the written assessment in the Landscape and Visual Impact Assessment has to carry the whole argument on its own.

Survey of camera positions as a separate skill

Fixing a viewpoint properly is survey work. The camera position needs a coordinate and a height, the scheme needs to sit in the same coordinate system, and the terrain model under both needs to be consistent.

In practice that means:

• The viewpoint is located by GNSS to a stated accuracy, not by dropping a pin on an aerial photograph.
• Camera height is measured and recorded, usually at eye level around 1.5m to 1.65m, and stated rather than assumed.
• Scheme coordinates come from the engineering model or the planning layout, converted into the same grid and vertical datum.
• Control features visible in the photograph, such as building corners, pylons or field gates, are surveyed so the model can be matched against them.

That last step is the verification. Without surveyed control in the frame there is no way to demonstrate that the rendered model is sitting in the right place, and an opposing expert will say so.

Landscape architects do not usually hold this competence themselves and are not expected to. What they are expected to do is commission it and check it.

Focal length, sensor and printed size

Field of view is a function of focal length and sensor size together, and getting the pairing wrong distorts apparent scale. The convention most practitioners follow uses a full-frame sensor with a 50mm lens for single-frame images, which gives roughly a 39.6-degree horizontal field of view and approximates comfortable human perception of scale at a stated viewing distance.

Three errors recur:

1. A crop-sensor camera used with a 50mm lens, which narrows the field of view and enlarges the apparent size of the development.
2. A wide-angle lens used to get everything in, which shrinks the development and flattens the landform.
3. A panorama presented on A3 with no stated viewing distance, so the reader holds it at arm's length and sees a scene that does not match the site.

Printed size and viewing distance are part of the specification, not an afterthought. A cylindrical or planar projection panorama intended to be viewed at 300mm from the eye is a different document from the same file printed small in a committee report, and the methodology should say which is intended.

Rendering judgement and material reflectance

Once the geometry is verified, the remaining judgement is about materials, light and vegetation, and this is where honest practice separates from advocacy.

Render settings can make a building disappear. A matt grey cladding under flat overcast light reads very differently from the same panel under low winter sun, and a glazed elevation can be shown either as a dark void or as a sheet of reflected sky. Neither is wrong, but choosing only the flattering condition is misleading.

Vegetation is the other lever. Showing proposed planting at maturity in the year one image, or retaining a hedgerow in the model that the scheme removes, are both common and both get found out. Good practice shows the scheme as built on completion, with a separate image at a stated future year, and states the assumed growth rates.

Season matters for the same reason. Photographs taken in full leaf for a scheme in a deciduous landscape understate winter visibility, which is why the guidance favours winter photography for baseline work where screening is vegetative.

Who checks the output before submission

Three checks, by three different people, in this order.

The surveyor confirms that the camera positions, control points and scheme coordinates reconcile, and signs off the match. The landscape architect confirms that the viewpoints represent the agreed set, that the conditions shown are representative rather than selective, and that the written effects assessment matches what the images show. The project lead confirms that every image carries its methodology data and that the printed deliverable matches the specification.

Where an AI-assisted draft is used to structure the written assessment, a chartered landscape architect still has to own the judgements in it. The professional directory is the route for arranging that review, and the competence expectations for landscape work are set out in more detail in our note on landscape professionals and visual assessment.

Challenging images produced by others

When you are on the other side of an application, the methodology appendix is where you start. Ask for the camera coordinates, the height, the lens and sensor, the field of view, the control points used and the terrain model source. If any of those are missing, say so in your representation and ask the authority to treat the images as indicative.

Then test what you have:

• Stand at the stated coordinate with the stated focal length and compare.
• Check whether the viewpoint set includes the closest and most sensitive receptors, or stops short of them.
• Check whether vegetation in the baseline photograph still exists and whether the scheme removes it.
• Check the lighting condition against the orientation and the stated date and time.
• Check that the stated printed size matches the document you were given.

Selective viewpoint choice is the most common substantive problem and the hardest to prove. Mapping the submitted viewpoints against the zone of theoretical visibility usually shows it quickly.

Specifying the deliverable in the brief

Write the brief so that the output is checkable. A workable specification names the level of accuracy sought for each viewpoint, the number of viewpoints and who agrees them, the survey accuracy required for camera positions and control, the camera and lens combination, the projection type and printed size for panoramas, the lighting and seasonal conditions, the vegetation assumptions and future years to be shown, and the format of the methodology appendix.

Agree the viewpoint list with the local planning authority and, on larger schemes, with the National Landscape or National Park body before the photography is commissioned. Repeating a winter photography programme because the viewpoints were not agreed is a year's delay, not a week's.

Where a Townscape and Visual Impact Assessment is the relevant document rather than a full landscape assessment, the same competence split applies. The report index shows how the two sit alongside each other, and which other documents a visually sensitive scheme usually needs.

One honest caveat. Levels of accuracy are guidance, not a standard with a certification scheme behind it, so practice varies between consultancies and between authorities. Ask what level you are being quoted for, in writing, and ask who is doing the survey.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required. This is general guidance rather than legal advice. Check the position with your local planning authority or a qualified professional before relying on it.