Visual Evidence for Land Use Attorneys

Quick answer: A visual becomes evidence when somebody else can reproduce it. For a land use matter that means four things: the inputs are stated, the model provenance is documented, the output is reproducible from the archived file, and the person who made it can explain every choice without improvising. A beautiful image with none of that is advocacy, which is fine in a brochure and a liability in a record.

The difference between a persuasive image and an exhibit

Firms commission visuals for two different jobs and the jobs have different rules.

The first job is persuasion: showing a board, a council or a neighbourhood meeting what a proposal will look and feel like. Here the standard is clarity and honesty, and a good image carries the room.

The second job is the record. An exhibit has to survive somebody hostile reading it closely, months later, looking for the choice that was quietly made to flatter the applicant. That reader does not care whether the image is attractive. They care whether its assumptions are written down and whether running the same inputs again produces the same output.

Most disputes about visuals in land use matters are not about geometry. They are about an unstated choice: a camera lens that compressed a street, a vegetation state that happened to be in full leaf, a context model that stopped at the property the applicant would rather not discuss. Every one of those is defensible if it is declared and indefensible if it is discovered.

The stated assumptions record

The most useful single document a firm can insist on is a short assumptions schedule issued with every visual exhibit. It takes a page and it changes the character of the whole package.

• Source geometry. Which drawing set, which revision, which date. A visual built on an earlier revision than the one submitted is the most common and most damaging inconsistency in an application.
• Survey basis. Whether ground levels and the context came from survey, from public data, or from assumption, and dated.
• Orientation. How true north was established, and for any sun dependent output the latitude, longitude, dates, times, and whether the clock is standard or daylight saving.
• Camera data. For any view from a specific position: the position as a coordinate, the height of the lens, the direction, and the focal length or field of view.
• Context extent. How far out existing buildings were modelled, from what source, and whether approved but unbuilt schemes were included.
• Vegetation. Modelled, omitted, in leaf or out of leaf, and on what basis.
• Materials and finishes. Whether colours and reflectivity represent specified products or are indicative.
• Exclusions. What the exhibit does not purport to show.

Every one of those is a question somebody will eventually ask. Answering them in advance converts eight possible weaknesses into eight pieces of evidence of care.

Model provenance and the archive

Provenance is the chain from the submitted drawings to the issued image. It needs to be documented at the time, because reconstructing it later from memory is exactly the testimony that falls apart.

What a firm should require: the native model files archived and retained, the source drawing set and survey archived with them, a record of which model version produced which issued exhibit, and a revision history where later versions carry new numbers rather than silently replacing earlier files.

Silent replacement is the trap. A visual that was updated after an objection, under the same filename, with no revision note, hands the other side a question that cannot be answered well. A visual that was updated as revision B, with a note saying what changed and why, is simply a document that was corrected.

Ask for the archive at the point of commission rather than at the point of dispute. Suppliers who work in this area keep it as a matter of course. Suppliers who do not will tell you so, which is itself useful information.

Reproducibility, which is the whole test

The practical question in any challenge is whether an independent party given the same inputs would produce the same output. If yes, the exhibit holds. If the answer depends on judgement the record does not describe, it does not.

That is why study work and presentation imagery are different animals in a record. A shadow or sight line study is deterministic: stated position, stated date, stated level, one answer. The conventions that make those checkable are set out in shadow study deliverables and what reviewers check and sight line and story pole studies explained.

A photorealistic rendering involves aesthetic choices that are not deterministic: time of day, sky condition, how busy the street looks, how warm the light is. None of those are dishonest, and all of them are arguable. Where a record needs certainty, lead with the deterministic products and let the rendering support them rather than carry them. See massing study, what it shows and what it proves.

Building the exhibit set

A visual exhibit set that works in a land use matter usually has three layers.

The base layer is the key plan: every viewpoint, every observer position and every tested receptor located on one plan with north and a scale bar. Nothing else in the set makes sense without it, and a set without one invites the question of what was left out.

The middle layer is the deterministic studies: massing, shadow and solar, sight lines, view analysis. These carry the burden in a record because they are checkable. The approach to view work is set out in view analysis and view corridor studies, and the solar basis in sun path and solar access studies explained.

The top layer is the presentation imagery, clearly labelled as illustrative where it is illustrative. Labelling costs nothing and removes an entire line of attack. An image captioned as indicative of materials cannot later be characterised as a false representation of them.

Keep the layers visibly distinct. A set that mixes a surveyed sight line section and an artist's impression on the same sheet, with the same weight, drags the credibility of the first down to the level of the second.

Consistency across the application

The fastest way to lose a land use argument with your own documents is to submit documents that disagree.

It happens routinely. The massing diagram is built from revision C, the shadow study from revision D, and the rendering from a design that was never submitted at all, because it was produced first and nobody went back. A board that notices one discrepancy treats every number in the package as unreliable.

The fix is procedural rather than technical: one context model, one source drawing revision per issue, a single issue sheet listing every visual document with its revision and date, and a rule that any drawing change triggers a check of every visual built on it. It is tedious. It is also the thing that distinguishes a package that holds up from one that does not.

Preparing the person, not just the document

Where visual work is going to be spoken to, the supplier who produced it should be able to answer, without hesitation, where the camera was, how high, what lens, which drawing revision, where the ground levels came from, what was excluded, and why.

Three questions come up almost every time. Why this viewpoint and not another one. Why this time of day or time of year. What is not shown here. A presenter who has answers to those three, prepared and consistent with the assumptions schedule, is treated as a technical witness. One who improvises on any of them converts the whole exhibit set into a question mark.

The useful preparation is adversarial: have somebody who wants the application refused read the package and list every choice they would attack. Most of what they find is answerable with a sentence and a line in the schedule, provided it is found before the hearing rather than during it. See visuals for hearings versus community meetings.

What to ask a visualization supplier for

A short list separates suppliers who have worked in a contested context from suppliers who have not.

Ask whether they issue an assumptions schedule with exhibits as standard. Ask whether native models and source drawings are archived and for how long. Ask how revisions are numbered and noted. Ask whether somebody who worked on the model is available to explain it, and in what form. Ask how the context model is built and from what source. Ask what they will not do, because a supplier who has never declined a request to make a massing look smaller is not a supplier you want on a record.

Also ask the plain commercial questions early, because they shape the brief: which products are deterministic studies and which are illustrative, and what each requires as input. Published figures on the 3D rendering cost page cover the visualization deliverables: standard exterior stills run $750 to $1,250 and upwards, premium exteriors $1,250 to $2,500 and upwards, interiors $1,250 to $3,500 and upwards per space, and animation work $3,500 to $15,000 and upwards. Study packages are scoped after the code requirement, the receptors and the survey information are reviewed, because the cost sits in the context model rather than in the output.

Where this fits in the matter timeline

Visual work commissioned late is visual work commissioned twice. The deterministic studies should be run while the envelope can still move, because their purpose is to find the problem before a board does, and a setback decided in response to a sight line test is cheap while the same setback decided after a refusal is not.

Presentation imagery comes last, once the design it illustrates is the design being submitted. The sequencing and the lead times are covered in planning a rendering around a hearing date, and the wider process in the land entitlement process guide.

The honest summary

Treat every visual that will touch a record as a document somebody hostile will read closely. State the inputs, archive the model, number the revisions, keep the deterministic studies visibly separate from the illustrative imagery, make sure no two documents in the application disagree, and prepare whoever will speak to it for the three questions that always come.

None of that makes an image less persuasive. It makes it survivable, which in a contested matter is the same thing as useful.

Rendimension has produced this work since 2004, architect led, across more than 1,000 projects. See also winning neighbour support before a hearing.

Frequently asked questions

What makes a visual usable as evidence rather than advocacy?

Reproducibility. The inputs are stated, the model provenance is documented, the archived file produces the same output again, and whoever made it can explain every choice. A beautiful image with none of that is advocacy.

What should an assumptions schedule contain?

Source drawing revision and date, survey basis, how true north was established, camera position height direction and focal length, context extent and source, vegetation treatment, whether materials are specified or indicative, and what the exhibit does not purport to show.

Why does model provenance matter?

Because the chain from submitted drawings to issued image has to be documented at the time. Reconstructing it later from memory is the testimony that falls apart. Native models, source drawings and survey should be archived with a record of which version produced which exhibit.

What is wrong with updating a visual under the same filename?

It hands the other side a question that cannot be answered well. A visual updated as a numbered revision, with a note saying what changed, is a corrected document. A silently replaced one looks like something being hidden.

Should renderings or studies lead in a record?

Studies. Shadow, solar, sight line and massing work is deterministic: stated inputs, one answer. A photorealistic rendering involves aesthetic choices that are arguable. Lead with the checkable products and let the imagery support them.

How should an exhibit set be structured?

Three visible layers: a key plan locating every viewpoint and receptor, the deterministic studies that carry the burden, and the presentation imagery clearly labelled as illustrative where it is. Mixing a surveyed section and an artist impression at equal weight weakens both.

What questions come up most often about a visual exhibit?

Why this viewpoint and not another, why this time of day or year, and what is not shown. Prepared answers consistent with the assumptions schedule make the presenter a technical witness. Improvised answers turn the whole set into a question mark.

What should a firm ask a visualization supplier before commissioning?

Whether an assumptions schedule is standard, whether native models and source drawings are archived and for how long, how revisions are numbered, whether somebody who built the model can explain it, how the context model is sourced, and what the supplier will decline to do.