An image-quality test for real estate photography, however it was produced: photographed, human-edited, AI-processed or AI-generated. Measurable parameters across exposure, white balance, colour, geometry, artifacts, and semantic plausibility.
An open standard for real estate image quality. VQS, the Visual Quality Score, is an objective, versioned standard published by Standard Vision Lab: every image is measured against a fixed parameter set and scored out of 100, and a listing is scored as a set by its weakest photo.
Pass · Flag · Fail. Measured, not opinionated.
Real estate imagery is shot, edited, AI-processed and AI-generated at portal scale. The methods used to evaluate it have not kept pace.
Subjective taste cannot be a quality gate.
Listing imagery arrives at portal scale. Review by eye means review by sample, or not at all.
Without numbers, there is no standard to enforce.
Reproducibility over opinion. Every parameter has a numeric threshold. The same runner scoring the same image must return an identical result, and two conforming runners must agree within a published tolerance.
Public methodology. The plain English guide is public, and the full specification is available on request.
Versioned, not revised. A VQS score today means the same in five years. When the standard changes, the version changes.
Adversarial by design. We test the failure modes vendors hope you won't check.
Disclosed, not implied. Standard Vision Lab is incubated within a group that also operates vendors eligible for certification. Identical thresholds and protocols apply to every applicant, related or not, and any material relationship is disclosed in the certification record.
An image-quality test for real estate photography, however it was produced: photographed, human-edited, AI-processed or AI-generated. Measurable parameters across exposure, white balance, colour, geometry, artifacts, and semantic plausibility.
Each parameter is scored against a published threshold. Status returns Pass, Flag, or Fail. The composite is reported as a score out of 100.
To replace subjective review with reproducible measurement, so buyers, vendors, and platforms speak the same language about image quality.
Parameter falls inside the accept band. No reviewer attention required.
Parameter falls in the caution band. Human review recommended; the image may be acceptable in context.
Parameter falls in the reject band. Image fails the standard and should not ship without remediation.
Composite score is the weighted aggregate of all parameter results. A single Fail does not collapse the composite, but is reported in the certification record.
Every parameter is versioned with the standard. If the parameter set changes, the version changes. Old scores remain valid against old versions.
A real estate listing is not a single image. It is a set. Buyers compare frames side by side. Inconsistency between frames is itself a defect.
σ across the set for brightness, white balance, colour cast, contrast, saturation, and window WB: globally, by room class, and by sub-set.
Free to request. The plain English guide is public now and the articles go deeper; the full normative text is in preparation for publication and we will send the current draft on request.
Only Standard Vision Lab issues the VQS Certified mark, and issuance is automatic. Run a shoot through the canonical runner and a composite in a certifiable band is granted its exact tier badge with the result: an embeddable, verifiable mark, live for 12 months.
An objective, versioned image-quality standard for real estate. Certify a pipeline, adopt the standard as a quality gate, or request the v1.0 specification.