Review the result before choosing the file
A completed ASCAND processing job gives you a result to evaluate. It does not automatically give you a model that is accurate, complete or ready for every downstream task.
Review, export and download are three different stages. First, inspect the reconstructed geometry and decide whether it contains the evidence your intended use requires. Second, choose the kind of representation you need. Third, select a compatible file format, generate or select the export, download every required file and validate the result in the application where it will actually be used.
Keep that order:
- confirm the result;
- inspect the geometry;
- choose the representation;
- choose the format;
- generate or select the export;
- download the complete output; and
- validate it in the destination workflow.
A preview can support the first review. A processing status can indicate that a stage finished. A download control can make a file available. None of these is a certificate of model quality or fitness for purpose.
Understand the review and download stage
Confirm what result you are reviewing
Begin by confirming that you have opened the intended project, object and processing result. A project can contain different captures, methods, processing variants or later derivatives. A merged result is not the same artifact as either contributing scan, and a mesh is not the same representation as the point cloud from which it may have been generated.
Identify:
- the project and object;
- the source capture or source set;
- the Vision, Laser, Combo or Multi Merge path;
- the relevant physical configuration;
- the processing or merge variant;
- the completion state; and
- the relationship between the result and its original sources.
Do not rely on a thumbnail, filename or extension alone. A file ending in .ply, for example, can represent different point-based or surface-based content depending on how it was produced. A familiar extension does not tell you whether a result is raw, cleaned, merged, smoothed or otherwise transformed.
Preserve the original recordings and project context. If several results are available, review them separately and use the current platform description to identify what each contains. Do not assume that the newest, densest or most processed-looking option is automatically the most suitable.
Inspect the geometry from more than one view
Open the available result in the current viewer and rotate it through multiple viewpoints. Inspect directions that were easy to observe during capture and directions where occlusion, support contact or limited visibility may have affected the evidence.
Check:
- whether the overall form and proportions are plausible;
- whether every region required for the intended use is represented;
- whether holes or unexpected closures are visible;
- whether floating points, spikes or isolated fragments remain;
- whether surfaces are duplicated, offset or incorrectly bridged;
- whether turntable, support or background remnants are present;
- whether smoothing has removed features that matter;
- whether orientation and scale appear plausible; and
- whether color or texture aligns with geometry where supported.
The viewer is an inspection aid. Lighting, shading, point size and camera perspective can hide or exaggerate problems. A lightweight preview may also differ from the final export. Use the preview to decide what needs closer inspection—not to certify dimensions, units, topology, hidden geometry or compatibility.
If the required geometry is visibly absent or distorted, stop before converting it repeatedly. Another container format cannot recover a surface that was not captured or correctly reconstructed. Return to the earliest stage that can address the limitation.
Choose the representation before the format
A representation describes what kind of geometric information the result contains. A format describes how that information is stored and exchanged. Choose the representation first.
Point cloud
A point cloud stores sampled locations in three-dimensional space and may include supported point attributes. It is often appropriate when the downstream task needs direct inspection of coverage, density, outliers, alignment or reconstructed evidence. Point clouds can also serve as input for further filtering, registration, merging or surface generation.
Mesh
A mesh connects geometry into polygonal faces. It is appropriate when the next workflow requires a surface, defined faces or an interior and exterior. Meshing interprets point-based or volumetric information. It may smooth, bridge, close or otherwise transform the geometry, so inspect the mesh independently rather than assuming that it preserves every property of the source.
Processed variants
Labels such as raw, cleaned, dense, merged, smoothed or watertight describe processing intentions or states. They are not a universal quality ranking.
- Raw evidence may retain noise that is useful for diagnosis.
- Cleaning may remove outliers without restoring missing regions.
- Greater density does not establish greater accuracy.
- A merged result depends on correct registration and compatible overlap.
- Smoothing can reduce noise while also reducing meaningful detail.
- Watertightness means that a surface is closed; it does not mean every surface was observed or reconstructed correctly.
Choose the representation that preserves the information the intended task requires. Then choose a format capable of carrying it.
Compare point clouds and meshes
Choose a format for the destination workflow
The correct format is the one that can carry the selected representation and required attributes into the intended application with acceptable interpretation.
Before selecting it, check the destination software’s current documentation:
- Does it accept point clouds, meshes or both?
- Does it preserve points, faces and normals as required?
- Does it support vertex color, textures or materials?
- Does it expect separate companion files?
- How does it interpret units and scale?
- Which axis orientation does it use?
- Does it retain the level of detail needed?
- Will another conversion step be required?
Formats such as PLY, STL, OBJ, GLB/GLTF, 3MF and other supported outputs have different purposes and attribute capabilities. Their presence in platform or engineering history does not establish that every format is available for every current result. Consult the current output list and verify the exact variant before publishing or selecting it.
An extension is not a quality level. STL does not make a model printable. OBJ does not guarantee complete materials or textures. PLY does not tell you whether the stored geometry is a point cloud or a mesh. GLB or GLTF availability does not establish that every source attribute is included. Always evaluate the actual output and destination workflow.
Understand output formats and post-processing
Generate or select the verified export
Use the current 3D-Scan.Online project controls to select the required result, representation and available format. Some outputs may already exist after processing. Others may require a documented conversion or mesh-generation step.
Before requesting an export, verify:
- the correct project and result are selected;
- the representation matches the intended use;
- the format is supported by the destination application;
- required attributes are expected to survive the handoff;
- any documented generation parameters are understood; and
- current access, retention and download conditions are known.
Submit the export request once through the verified control. Keep the original result and source material unchanged. If the platform performs additional processing, distinguish the request from its completion. A request can be accepted without the derivative being ready, and an export job can complete without making unsuitable source geometry suitable.
Do not infer current button labels, defaults, parameters, queue states, processing time, availability, limits or access conditions from an older screenshot. Follow the current platform documentation and confirm that the expected output becomes available for the intended result.
Consult current platform documentation
Download the complete output and validate it locally
When the verified output is available, download it using the current project controls. Some representations may consist of one file. Others may depend on companion material, texture or metadata files. Download the complete documented set and keep related files together.
Then:
- confirm that the download completed;
- preserve an unchanged copy;
- check the filename and extension against the selected output;
- confirm that required companion files are present;
- open the model in the intended destination application;
- verify dimensions, scale and units;
- verify axes and orientation;
- inspect points, faces, normals and topology as applicable;
- confirm color, material or texture behavior where required; and
- record any reorientation, repair, conversion, decimation or smoothing applied later.
A completed download proves delivery of the file. A successful import proves that the destination application can read it. Neither proves that scale, topology, appearance or geometry is correct. Validation is complete only when the output has been checked against the requirements of its real downstream use.
For 3D printing, continue to the dedicated preparation tutorial. Print readiness requires separate checks; downloading a mesh or an STL file is not enough.
Prepare the scan for 3D printing
Continue, correct, convert or diagnose
After local validation, choose the next action by the earliest unresolved issue.
Continue downstream when the downloaded output preserves the required geometry, representation, attributes and scale for the intended task.
Choose another verified output when the geometry is suitable but the representation or format does not meet the destination application’s needs.
Correct the scan or reconstruction when coverage, distortion, outliers, incorrect surfaces or merge errors already exist before export. Another file format will usually preserve the same underlying problem.
Diagnose export, transfer or import when the platform result is suitable but generation fails, the download is incomplete, companion files are missing or the destination application interprets the file unexpectedly.
Keep source evidence, processed results and downstream derivatives distinguishable. This makes it possible to return to the correct stage without losing the known-good artifacts.
Merge complementary scans · Return to TutorialsAn ASCAND Laser Scan uses a projected line and a controlled camera–laser relationship to derive sampled surface points through triangulation. A Combo Scan coordinates laser-derived observations with Vision-derived evidence and evaluates how the independent results can contribute to a combined representation.