Scanning impressions instead of pouring is a straight saving when the impression is a clean addition silicone, the tray is rigid, and the scanner can capture a deep sulcus without a powder cloud. It is a false saving when the impression is alginate that has been sitting for an hour, or when the margin sits under a fold of soft tissue that no optical path can reach. The decision is not about whether the technology works. It is about whether your impression is good enough to be the last physical artefact in the chain.
When scanning impressions instead of pouring models is the better route
Direct impression scanning removes three error sources at once: the stone's setting expansion, the pour technique, and the transport of a poured model between rooms. It also removes a day from the schedule, because the impression can go from the operatory to the scanner before the material has fully recovered.
The reason to skip pouring models is not speed but the removal of stone expansion, and the preconditions are specific. The impression must be a dimensionally stable elastomer, most often a polyvinyl siloxane, in a rigid tray with an adhesive that was allowed to cure. The margin must be visible in the impression, which usually means the clinician used retraction cord or a retraction paste and took the wash before the tissue rebounded. And the scanner must be able to handle the depth, because a dual-arch impression has vertical walls that a short-depth system cannot reach.
| Situation | Scan the impression | Pour the model |
|---|---|---|
| Clean PVS impression, margins visible | Yes | Unnecessary |
| Alginate more than 20 minutes old | No | Yes, but pour immediately |
| Deep subgingival margin with tissue overlap | Risky | Yes, with a die trim |
| Border-moulded edentulous impression | Possible with a high-depth scanner | Yes for a definitive cast |
| Bite registration in a separate record | Yes, scan both and align | Yes, mount on an articulator |
| Impression with voids at the margin | No | No, remake the impression |
Where scanning impressions instead of pouring loses accuracy
Four mechanisms dominate. First, impression distortion, which begins as the material recovers and continues as it sits. A PVS is stable for days in a rigid tray at room temperature, but an alginate is not, and a flexible tray distorts under its own weight. Second, powder or spray coating, which is still used on reflective impression surfaces; a thin brushed coating is around 1 μm while a traditional aerosol can be an order of magnitude thicker, and the difference shows up at the margin. Third, scanner alignment error when two impressions are stitched together. Fourth, stone expansion, which the direct route avoids entirely and which is one reason a scanned impression can be more accurate than a poured model of the same impression.
Preparing the impression for the scanner
Most scan failures on impressions happen before the impression reaches the scanner. A short preparation routine removes them.
- Rinse and dry the impression. Water beads scatter light and produce holes; a few seconds of air is usually enough.
- Trim flash and tray material that overhangs the occlusal surface, because overhangs create undercuts the scanner cannot see past.
- Check the margin in the impression under magnification before scanning. If it is not there, remake the impression now.
- If the surface is highly reflective or translucent, apply a thin brushed powder layer rather than an aerosol cloud.
- Mount the impression on a stable holder at the working distance the scanner expects, not held in the hand.
- Scan the arch in a single continuous pass where possible, and close any hole manually rather than accepting an automatic fill over the margin.
- Scan the opposing impression, then align, then verify the occlusion against the clinical record.
Step-by-step: a lab workflow for scanned impressions
Set the workflow so that the impression is scanned on arrival, before any other task. The value of the digital route is that the impression stops being a perishable item the moment it is captured. In an impression scanning dental lab, that means the impression can go back on the shelf while the case is already in design.
During scanning, use the scanner's depth capability deliberately. The AI-S3 and AI-S9 both support direct impression scanning, and the S3 handles a single impression in 15 to 19 seconds while the S9 takes around 29 seconds with its higher-resolution capture. The S3 adds automatic dual-impression alignment and AI screw channel closure, which matters for implant cases where the scan body sits deep in the impression.
Powder, reflections and the surfaces that fight back
Impressions are usually matte enough to scan without any coating, which is one reason the direct route is popular. The exceptions are the areas where the impression captures a polished surface, such as a scan body, a metal post or a highly glazed provisional. Those reflect light back into the scanner and produce a hole exactly where the geometry matters.
A thin brushed coating solves it without the geometry penalty of an aerosol. A brush-applied powder at roughly 1 μm deforms the surface far less than a 12 μm aerosol layer, and a 10 g bottle covers roughly 1,000 to 3,000 units, which makes the cost per case small enough to ignore. The powder is water-soluble, so it rinses off the impression without residue if the case has to be poured afterwards.
Criteria for when pouring still wins
Direct scanning is not a universal replacement. Pouring remains the better route in four situations, and a lab that pretends otherwise will produce avoidable remakes.
- Alginate impressions, which distort as they lose water and should be poured within minutes rather than scanned later.
- Impression surfaces with folds or drag marks at the margin, where a physical die trim gives the technician a clean edge to work from.
- Edentulous cases with functional border moulding, where the definitive cast is inspected physically before processing.
- Cases where the clinician and the technician need to discuss a fit issue on a physical model, which is still the fastest way to resolve an argument.
Comparing the two routes on cost and time
The economics are easy to get wrong because the visible cost, the stone and the pour, is small. The real cost sits in labour, storage and the delay between impression and design.
| Factor | Scan the impression | Pour the model |
|---|---|---|
| Time to a design-ready file | Minutes after arrival | Hours, including set and trim |
| Labour per case | One scan pass, one alignment check | Mix, pour, set, separate, trim, inspect |
| Material cost per case | Low, plus occasional powder | Stone, model base, consumables |
| Dimensional error | Impression distortion plus scanner error | Same, plus stone expansion and pour technique |
| Storage | Digital file, negligible footprint | Physical models, shelving, retrieval |
| Rework path | Rescan the impression if retained | Repour from the same impression if retained |
Making the switch without breaking the schedule
A digital impression workflow changes who does what, so change one product line first. Pick the crown and bridge work with clean PVS impressions and run it in parallel with pouring for a month, comparing fit results on the same clinicians. Do not switch implant cases in the same month, because the scan body geometry and the alignment step add variables you will not be able to separate.
To review which scanner configuration suits your impression volume, see the impression scanning configuration and the wider dental lab scanner range. The scan powder page covers the coating options for reflective surfaces, and if you want the workflow costed against your own case mix, talk to our lab team.
Frequently asked questions
Can you scan alginate impressions?
You can, but the result depends on timing. Alginate loses water and distorts within twenty to thirty minutes, so a scan taken an hour after the appointment captures a shape that has already changed. If alginate is the only option, scan on arrival and treat the result as a short-lived record.
Is a scanned impression as accurate as a poured model?
It can be more accurate, because it skips stone expansion and the pour technique. It can also be less accurate if the impression distorts, if the surface needs a thick powder coat, or if the two arches are misaligned. The material and the handling decide the outcome more than the scanner does.
Do I still need to keep the physical impression?
Keep it for the period your quality system or your customer requires, then dispose of it. The scan becomes the working record, and a shelf of old impressions is not a substitute for a verified digital file. Some labs keep the impression until the case ships, which is a reasonable compromise.
Which scanner specification matters most for impressions?
Depth of field and the ability to capture vertical walls without losing tracking. Accuracy figures quoted on a flat reference are less useful than a demonstration on a deep dual-arch impression, so ask for that specific test before you commit.
The switch is worth making where the impressions are stable and the margins are clean, and it is not worth forcing where they are not. Export the scan as STL and any CAD package in the lab can open it without a converter. If you are weighing the two routes for a specific product line, review the model scanner options or ask for a workflow review with your monthly case volume.
