Historical 3D Reconstruction: Make the Evidence Visible

A historical reconstruction should show both an interpretation of the past and the evidence behind it. A photorealistic roof does not become a documented roof because the rendering is convincing.

Begin by stating what the model represents: the surviving condition, a proposed restoration or a reconstruction of a particular earlier phase. The London Charter’s documentation principle calls for the intended representation and the nature of uncertainty to be made clear. That is especially useful when a finished image makes every surface look equally certain.

Choose a question before choosing a viewpoint

Imagine a vanished building known from a plan and one street photograph. A useful question might be how its entrance related to the adjoining street. Rebuilding every unseen room would add assumptions without necessarily helping answer that question.

State the time period and the purpose. A model combining a façade from one decade with an extension from another may depict a condition that never existed. If several phases are compared, identify them separately.

Keep an element-by-element evidence record

For each important part, record the source, what it shows, its date and the inference used in modelling. A measured plan may establish a footprint while a photograph constrains part of the façade. Neither automatically supplies the unseen roof structure.

In the hypothetical example, the entrance position might be well supported, the roof pitch approximate and the rear elevation unknown. Keep those distinctions in the working record and in an explanation accessible to the viewer. Do not turn “not visible” into “did not exist”.

Show alternatives where the evidence permits them

If two roof forms remain plausible, compare them as alternatives instead of selecting the more attractive one and presenting it as settled history. Use labels, separate views or a clear legend. Colour alone may be difficult to interpret and should not carry the whole explanation.

A reconstruction can still be visually coherent. The aim is to give its reader a way to distinguish a documented feature from an interpretive choice. An accompanying evidence view can serve that purpose better than filling the main image with dense annotations.

What the old Double Prism page documented

The former studio’s historical-visualisation page named examples including the coastal-defence ship Admiral Ushakov, a hotel in Skoczów described as “White Horse” in the English version, and the town’s former synagogue. It also embedded an external fortification model credited to szance.jablonkow.

Those references identify what the page displayed. This independent publication does not claim to have created the works, recovered their editable models or verified every historical detail in them. The external model’s visible credit should not be reassigned to the publisher.

A useful archive includes more than the final image

Preserve the source list, model version, assumptions, key camera views and a readable explanation. Where an interactive model is published, provide a static route to the important information too. A future reader should be able to understand the main argument even if a viewer or embedded service stops working.

Record the rights and access conditions for source images and models before redistribution. A source being visible in a web archive does not itself establish permission to repackage every asset.

The final test is straightforward: can another reader identify the question, examine the evidence and see where interpretation begins? If the only support is that the scene looks believable, the reconstruction is not yet a transparent historical argument.

Sources and historical context

This new independent guide revisits the topic of the original archived page. It is not the former studio’s service offer.