How to Identify Antique Paintings

Old Masters (1300–1800)

The term "Old Masters" covers an enormous span — from late medieval panel paintings through Baroque and Rococo:

• Renaissance (1400–1600): Panel paintings with egg tempera or early oil techniques. Leonardo, Raphael, Titian, Dürer, van Eyck. Genuine works are almost exclusively in museums. Any appearing on the open market warrants extreme scrutiny.
• Baroque (1600–1700): Large-scale oil on canvas. Rembrandt, Rubens, Caravaggio. Workshop production was standard — distinguishing "by Rubens" from "studio of Rubens" from "follower of Rubens" is a matter of degree.
• Rococo (1700–1770): Lighter palette, decorative subjects. Watteau, Boucher, Fragonard. Smaller-format works are more common on the market.

19th-Century Movements

• Romanticism (1800–1850): Dramatic landscapes, historical subjects. Turner, Delacroix, Friedrich. British and German Romantic landscapes are among the most frequently reproduced.
• Impressionism (1860–1890): Broken brushwork, plein-air color, light effects. Monet, Renoir, Degas. Extremely valuable and correspondingly heavily forged.
• Post-Impressionism (1880–1910): Van Gogh, Cézanne, Gauguin, Seurat. Van Gogh forgeries are among the most common in the art market.

Early Modern (1900–1960)

• Expressionism: Kirchner, Nolde, Schiele. The Beltracchi forgery scandal demonstrated that even museum-quality fakes could deceive experts for decades.
• Cubism and Abstract: Picasso, Braque, Kandinsky. Stylistic variety within a single artist's career makes authentication complex.

Chinese and East Asian Paintings

Chinese scroll paintings, Japanese screens, and Korean ink paintings follow entirely different traditions from Western art:

• Chinese scroll paintings: Ink and color on silk or paper. Mounting, seals, colophons, and provenance chains are key markers. The tradition of copying masterworks as homage complicates attribution.
• Japanese screens and scrolls: Kanō school, Rimpa school, ukiyo-e. School attribution rather than individual artist is often the relevant question.
• Korean ink painting: Joseon dynasty literati paintings. Authentication relies on calligraphy quality, seal analysis, and historical documentation.

Brushwork and Paint Application

Brushwork is the painter's handwriting — deeply personal and extremely difficult to replicate convincingly. Experts analyze:

• Brush rhythm: Confident painters produce fluid, consistent strokes. Copies often show hesitation — the copyist pauses where the original artist was continuous.
• Impasto: Thick paint application has a three-dimensional quality difficult to reproduce. Under raking light, ridges and valleys reveal direction, speed, and loading of the original brush.
• Underlayers: Many painters built up images in layers. Areas of wear that expose earlier layers provide authentication evidence. A copy applied in a single layer will not show this depth.
• Paint quality: Pre-industrial oil paint was hand-ground, producing uneven pigment particles and a texture that differs from modern machine-ground paint.

Canvas, Panel, and Paper

The support provides strong dating evidence:

• Canvas weave: Hand-woven canvas (before mid-19th century) has characteristic irregularities. Machine-woven canvas on a painting claimed to be earlier is disqualifying.
• Wood panels: Poplar in Italy, oak in the Netherlands, lime in Central Europe. Panel thickness, sawing marks, and aging evidence are dating indicators.
• Paper and silk (Asian paintings): Hand-made paper has different texture and aging than machine-made paper. Paper composition changes over time.
• Stretcher bars: Hand-cut mortise-and-tenon joints predate machine-cut. Expansion keys became standard in the late 18th century. The stretcher often tells a more reliable story than the front.

Pigments and Color

Pigment analysis is one of the most powerful authentication tools because certain pigments have precise dates of invention:

• Prussian blue: First synthesized 1704. Its presence before that date is an anachronism.
• Chrome yellow: Available from 1809. Famously used by Turner and Van Gogh.
• Cadmium yellow: Commercially available from the 1840s.
• Cobalt blue: Available from 1802 (distinct from natural cobalt minerals used earlier).
• Zinc white: Available from 1834. Replaced toxic lead white in many applications.
• Titanium white: Not available until 1920. Its presence in any painting claimed to predate 1920 is definitive evidence of later production.

Signatures and Inscriptions

Signatures are the most commonly faked element. A genuine signature was typically applied as part of the painting process. Warning signs include:

• Paint mismatch: A signature painted over aged, cracked paint surface was added after the painting was completed.
• Stylistic inconsistency: Artists' signatures evolved over their careers. A style that does not match the claimed period is suspicious.
• Too perfect: A signature that is mechanically precise, as if traced, lacks natural variation.
• UV fluorescence: Under ultraviolet light, newer paint fluoresces differently. A signature applied over original varnish appears as a distinct patch.

The Attribution Game

Most painting fraud is not outright forgery. More commonly, it involves:

• False attribution: A genuine old painting by an unknown artist is given a more famous name. The painting is authentic — the fraud is in the attribution.
• Enhancement: A genuine period painting is "improved" by adding a signature or retouching damaged areas.
• Studio work passed as autograph: Workshop paintings presented as entirely by the master's hand.
• Period copies: Copies made in the same era are genuine antiques but not by the attributed artist. The hardest cases because materials and techniques are period-correct.

Modern Forgery Techniques

Sophisticated forgers take extensive measures to defeat examination:

• Aged canvases: Purchasing genuinely old paintings and scraping off the original paint, then painting a new composition. The canvas passes dating tests.
• Historical pigments: Using period-correct pigments sourced from old paint tubes or specialized suppliers.
• Artificial aging: Baking paintings to accelerate varnish darkening, applying tinted varnish, using ink to darken craquelure.
• Craquelure simulation: Induced by temperature cycling or chemical treatment. Tends to look more uniform and less organic under magnification.

Photography-Based Fakes

High-resolution photographs printed on canvas, sometimes with additional brushwork applied on top. Surprisingly convincing casually but immediately identifiable under magnification — the regular dot pattern of printing is unmistakable. AI for antiques can detect these patterns efficiently from high-resolution photographs.

What Good Provenance Looks Like

A strong provenance includes:

• Exhibition history: Documented inclusion in exhibitions, especially at major museums, with catalogue references.
• Publication history: Reproduction in scholarly catalogues or catalogue raisonné entries.
• Sales records: Auction records that document the painting's passage through the market.
• Collection stamps or labels: Many collectors applied stamps or inventory numbers to the backs of paintings.

Provenance Red Flags

• Gaps in ownership: No documented history before a certain date invites the question: where was it?
• Lost collection claims: Forgers often fabricate provenance involving collections destroyed in wartime. "From a European private collection" with no further detail should prompt scrutiny.
• Over-elaborate provenance: Excessively detailed documentation can be suspicious — genuine provenance is often incomplete.

Stylistic Consistency Analysis

AI excels at comparing visual characteristics across documented works:

• Brushwork pattern matching: AI analyzes brushstroke direction, density, and variation and compares against documented works by the claimed artist.
• Compositional analysis: AI compares arrangement of figures, spatial relationships, and color distribution against the artist's known body of work.
• Period style verification: Cross-references stylistic elements against conventions of the claimed period and school.

Surface and Condition Assessment

From high-resolution photographs, AI for antiques can analyze:

• Craquelure patterns: Compared against databases of known aging patterns. Artificial craquelure has detectable characteristics.
• Restoration mapping: Areas of overpainting or retouching that differ in surface texture and reflectivity.
• Print-on-canvas detection: AI identifies the regular dot patterns of printed reproductions.

What AI Cannot Do

Photograph-based AI analysis cannot:

• Perform pigment chemical analysis (XRF, Raman spectroscopy)
• Assess paint layer structure (requires cross-section or X-ray)
• Evaluate canvas thread count and fiber composition with laboratory precision
• Authenticate provenance documentation
• Determine physical condition (flaking, delamination, stretcher damage)

Essential Shots

• Full painting, front: Photograph straight-on with even lighting. Natural daylight is best. Ensure the camera is parallel to avoid perspective distortion.
• Full painting, back: The reverse carries critical information — stretcher bars, canvas type, labels, stamps, gallery stickers, inscriptions.
• Signature close-up: In sharp focus with good lighting. Include enough surrounding area to show whether it sits over or under the varnish layer.
• Surface detail (raking light): Position a light at a steep angle to reveal brushwork texture, impasto, craquelure, and restoration. This single technique reveals more than almost any other approach.
• Detail of key passages: Close-ups of faces, hands, fabric, or sky — areas where the artist's technique is strongest.
• Labels and inscriptions on reverse: Individual close-ups of any labels, stamps, or handwritten notes.

Common Mistakes to Avoid

• Flash and direct lighting: Flash creates specular reflections on varnished surfaces. Use diffused natural light.
• Glass or glazing reflections: Photograph at a slight angle to avoid your own reflection, or remove the glass if safe.
• Skipping the back: The reverse is often more informative for authentication than the front.
• Low resolution: Brushwork analysis requires fine detail. Use the highest resolution available.
• Cropping the frame: Include the frame in at least one photo. Frame style and construction are additional dating evidence.