Published October 27, 2026

How to Identify Antique Bronze Sculptures From a Photo: Marks, Patina and Fakes

Why Bronze Identification From a Photo Is Worth Doing Carefully

A dark metal figure on a mantel can be a lifetime cast by a named Paris foundry, a later reduction of a famous model, a spelter shop piece painted to look expensive, or a resin tourist copy sold last decade. Those four objects photograph almost the same in a dim living room. The difference in price is often tenfold or more.

Photo-based identification works on bronze because the useful evidence is visual: foundry stamps, sculptor signatures, the word after, seam lines from the mould, the colour and wear of the patina, and the way the base and underside were finished. What a photo cannot do is weigh the piece, feel the chill of the metal, or swear that a signature is lifetime rather than later. Treat every image-based result as research, not a certificate.

Tools like Antique ID speed the first pass. Upload a full view plus close-ups of the base, any stamp, and the underside, and the app suggests a likely model, period, and market range from comparable sales. That estimate is not a certified appraisal, insurance valuation, or authentication. For estate division, insurance, or a high-value purchase, take the piece to a sculpture specialist who can handle it.

This guide covers the shot list that actually helps, how to read foundry marks and signatures, lost-wax versus sand-cast clues, how to separate bronze from spelter and resin, the reproductions you will meet (including honest after casts), and when to stop guessing.

What to Photograph on a Bronze Sculpture

Most failed identifications fail because the seller sent one glamour shot from above. Bronze hides its identity on the edges of the base, under the rim, and inside hollow casts. Photograph as if you were listing the piece for a specialist auction house.

1. Full figure, straight on and three-quarter. Include the whole sculpture and the base. Step back so the proportions are true. A cropped torso hides whether you have a reduction, a fragment, or a later marriage of figure and plinth.

2. The base from every side. Walk around it. Signatures and foundry names usually sit on the upper edge of the bronze base, on a rockwork outcrop, or along the back. Take one photo per side, then a raking-light close-up of any lettering.

3. Underside and interior. Tip the piece if it is safe and not too heavy. Photograph the bottom of the base, any felt or marble plinth, screw holes, and the open interior of a hollow cast. Sand-cast interiors look different from lost-wax interiors; modern resin looks different again. This is the photo most listings omit.

4. Marks at true close range. Fill the frame with the stamp. Hold the camera parallel to the letters so they do not foreshorten. Use a desk lamp from the side rather than flash. If the mark is in a recess, take two angles.

5. Patina and wear. Photograph high points (knees, mane, hat brim) where handling has rubbed the surface, and sheltered areas that still hold colour. A uniform chocolate spray with no wear is a warning; honest rubbing on the places hands actually touch is not.

6. Seams, plugs, and repairs. Look for a faint line around the waist or along a limb (mould join), circular core plugs, solder-coloured patches, and drilled holes from later lamps or clocks. Those details date the manufacturing method and flag later conversion.

Daylight from a window, camera a few feet back, no filter, no heavy shadow. For lighting and focus technique that applies across categories, see how to photograph an antique for identification. Bronzes add one extra rule: never rely on a single pretty face shot.

Foundry Marks: The Name That Cast the Bronze

On nineteenth- and early-twentieth-century French bronzes, the foundry mark is often more useful than the sculptor's name. The sculptor modelled the clay or wax. The founder poured the metal, chased the surface, and applied the patina. Collectors pay for both names, and later recasts of a popular model often keep a sculptor signature while dropping, faking, or muddying the foundry stamp.

Common Paris marks you will meet in photos:

F. Barbedienne / F. Barbedienne Fondeur. The great reduction house. Often paired with an A. Collas reduction medallion (see below). Lettering is usually crisp, incised or stamped, and sits on the base or reverse.

Susse Frères / Susse Ftes. Another first-rank Paris founder. Marks can be small and sit on the rim or underside. Lifetime Susse casts of Barye and others are a different market from unsigned later editions.

Alexis Rudier / Georges Rudier Fondeur Paris. The Rudier family cast much of Rodin. Alexis marks appear on lifetime and early posthumous casts; Georges marks are later. A Rodin signature without a plausible foundry story needs extra work, not a shrug.

Thiébaut Frères, Siot-Decauville, Hébrard, Valsuani, C. Valsuani Cire Perdue. Each has a documented lettering style and period of use. Hébrard and Valsuani matter for Degas and other lost-wax editions. Photograph the exact spelling, including cire perdue if present.

Read a foundry mark the way you read a maker's mark on silver: transcribe every letter, note whether it is incised, stamped, or raised from the model, and check that the mark's working dates overlap the style of the sculpture. A crisp modern sans-serif stamp on a supposedly 1840 animalier is a stop sign. A faint, slightly irregular incised name that matches known examples is a start, not a finish.

Foundry marks can be added later. Compare the colour of the letters with the surrounding metal. Fresh-cut letters in an otherwise worn patina, or a stamp that sits on a ground-down patch, deserve scepticism. The mark is evidence. It is not a hall-pass.

Sculptor Signatures and Where They Hide

Signatures live where the modeller could reach the clay: the top of the bronze base, a rock, a tree stump, the hem of a robe, occasionally the reverse of a shield or the underside of a hoof. Photograph the whole base first so you can prove the signature belongs to this object, then move in.

A signature that is part of the model — raised letters that were in the wax or sand pattern — is different from letters cold-stamped after casting. Raised-from-the-model names travel with every later recast of that mould. That is why a clear BARYE on the base does not, by itself, make a lifetime Barye. It tells you the model claims Barye. The foundry, chasing quality, edition clues, and comparable casts do the rest.

Hand-incised signatures, especially on unique or small-edition works, can be stronger evidence if the hand matches documented examples. Compare letter shapes, not just the name. Forgers copy famous names; they are sloppier about the way a particular sculptor formed a capital R or a terminal E.

Check spelling and language. Barye, A. L. Barye, BARYE all appear on authentic casts. Garbled French, a missing accent on a name that always used one, or an English marketing line cast into a "Paris 1870" figure are later tells. Photograph the letters so they are readable without guessing.

After Casts, Reductions and the Collas Machine

The word after (French d'après) is not automatically a fake. It is a legal and commercial label: this bronze follows a model by someone else. A Barbedienne reduction of a David d'Angers bust, stamped after the sculptor and bearing the Collas reduction medallion, is an honest nineteenth-century decorative bronze. A modern resin figure sold as "original Rodin" is not.

Achille Collas patented a mechanical reduction machine that let Barbedienne issue the same model in several documented sizes. Many of those casts carry a circular stamp reading some version of Réduction Mécanique A. Collas Breveté, often with a profile head. If your close-up shows that medallion, you are looking at a reduction edition, not a unique modelled sculpture. That is useful: it places the object in a known commercial system and sets a realistic value band. It is also a favourite thing for later copies to fake badly, so compare the stamp's relief, lettering, and wear with museum examples.

Size matters. Measure height including the bronze base (exclude a later marble or wood plinth) and put the number in the photo caption or notes. A 22 cm "Thinker" on a desk is a different object from a museum-scale cast. Online listings that omit height are hiding the most important reduction clue.

Posthumous editions complicate famous names. Rodin, Degas, and others were recast after death by authorised foundries and by everyone else. An authorised posthumous cast can still be a serious object; an unauthorised recast of a recast is a decorative metal souvenir. Photos can raise the question. They rarely close it. If the name on the base would move the price into four or five figures, stop and get hands-on help.

Patina: Honest Age Versus a Weekend of Chemicals

Patina is the chemically coloured skin of the bronze — brown, black, green, reddish, sometimes two-tone. Nineteenth-century founders treated patina as part of the art. Later restorers and fakers treat it as makeup.

In photos, honest old patina is usually uneven in a way that matches handling and air. High points thin toward the warmer metal colour. Recesses stay darker. The colour sits in the surface, not on top of it like paint. Look at the mane of a lion, the folds of drapery, and the sharp edge of the base. Wear should make sense.

Warning signs in a photo:

Uniform spray. The same dead chocolate from nostril to tail, including the underside and the inside of a hollow cast, often means a recent chemical or paint job. Founders did patinate undersides, but a perfectly even modern coating that fills chasing lines is a different look.

Paint sitting in the grooves. Cold-painted bronzes (Vienna bronzes, some animal figures) are a real category — the colour is meant to be paint. The fake version is a spelter or resin figure sloshed with bronze-effect enamel. Zoom in: paint that bridges across fine hair or feather chasing, or that chips to a silver-grey zinc or a creamy plastic, is not a foundry patina.

Bright green on high points only. Outdoor bronzes develop verdigris. Indoor mantel bronzes that have lived in a parlour for a century usually do not look like a fountain in a park. Artificial green on the nose and nothing in the recesses is theatre.

Fresh file marks under a dark wash. Someone cleaned, then stained. The chasing looks scraped; the colour looks wiped on. That can still be an old bronze that was over-restored. Disclose it. It is not "original patina".

Do not polish a bronze to "help the photos". Bright raw metal destroys the surface collectors pay for and makes dating harder. Dust it. Photograph it. Leave the metal alone.

Lost-Wax Versus Sand Cast: What You Can See in Photos

Two historic methods dominate antique bronzes, and each leaves traces you can photograph.

Sand cast (fonte au sable). The model is pressed into sand moulds, often in pieces. You may see a faint mould join around the figure, slightly softer fine detail, and an interior that looks sandy or pebbled if you can shoot inside a hollow cast. Many nineteenth-century French decorative bronzes, including a large share of Barbedienne's output, were sand cast. Sand cast is not a quality insult. It is a method.

Lost-wax (cire perdue). A wax model is invested and melted out; metal takes its place. Fine tool marks in the wax can survive in the bronze. Interiors of hollow lost-wax casts often show the irregular, dripped look of the investment rather than a sandy grain. Marks such as cire perdue next to Valsuani or Hébrard are claiming this method. Later marketing stamps the phrase on objects that were nothing of the kind, so treat the words as a claim to check, not a proof.

In a photo, look for: sharpness of eyelids, claws, and lace; a visible seam that was chased down but not erased; circular core pins or plugs; and the texture of any visible interior. A figure with mushy detail, a heavy unchased seam, and a glassy-smooth interior is more likely a later commercial recast or a non-bronze copy than a carefully finished nineteenth-century sand cast.

You will not reliably call "lost-wax versus sand" from one phone picture of the face. You can, with underside and seam photos, put the piece in the right manufacturing neighbourhood and avoid paying lost-wax prices for a coarse recast.

Bronze Versus Spelter, Resin and Cold-Painted Fakes

This is the distinction that saves people the most money. Real bronze (copper plus tin, sometimes with zinc or lead in the mix) is dense, chip-resistant, and takes a chemical patina. The common stand-ins are not.

Spelter (zinc alloy). Lighter than bronze. Chips to a bright silvery-grey, not a warm yellow-brown. Edges of bases and fingertips are the usual wound sites. In a photo, a pale metallic glint under missing "bronze" paint is the giveaway. Spelter was widely used for clocks, lamp figures, and cheaper salon sculpture from the late nineteenth century on. It is a historical material, not always a cheat — but it is not bronze, and it should not be priced as bronze. Listings that say "bronze" while the chip is silver are mislabelled.

Resin, bonded bronze, cold-cast bronze. Modern polymer mixed with bronze powder, or painted resin. Warm to the touch in the hand (photos cannot show that). Broken areas show a creamy, brown, or grey plastic core, sometimes with air bubbles. Seams look moulded, not chased. Screw-in marble-effect bases and felt that is still fluffy and black are common. Weight is low for the size. A magnet will not settle this — neither bronze nor resin is magnetic — so the chip and the interior photo matter more.

Brass and gilt metal. Yellower, often used for ormolu mounts and some small figures. A magnet still should not stick. Colour in daylight, not indoor yellow bulbs, is the photo clue.

If the surface is flaking like paint, assume you are looking at a coating, then identify the metal underneath from a chip or the underside. Do not scrape a piece to "test" it for a blog photo. Use damage that is already there, or leave the test to a conservator.

For the wider habit of comparing construction and materials rather than trusting the sales title, see antique versus reproduction: how to spot the difference.

Weight, Magnet and Sound Tests (and What Photos Cannot Prove)

People ask for a magnet test because it is cheap. On bronze it is only a negative screen. Bronze is not magnetic. Spelter is not magnetic. Resin is not magnetic. Brass is not magnetic. A magnet that does stick means you have iron or steel in the mix — a later armature, a screw, a repaired iron pin, or an object that is not a copper-alloy bronze at all. A magnet that does not stick proves almost nothing.

Weight is more useful and almost invisible in photos. Bronze of a given size should feel surprisingly heavy. Spelter is lighter. Resin is lighter still. If you are sending pictures to someone, include a kitchen-scale reading in grams and the height in centimetres. "Heavy for its size" in a caption is not evidence. A number is.

A gentle tap with a fingernail: bronze rings; thin spelter is duller; resin is dead. Do not ping a fragile cast or a marble-mounted group. And do not pretend a phone video of a tap is science. It is a clue.

Photos also cannot show whether the interior is packed with plaster (common in some spelter and lamp figures), whether a marble plinth is original, or whether two figures on one base started life together. If value depends on those points, the photo job is finished and the in-person job starts.

Common Reproductions You Will Meet

The bronze market has been reproducing itself since the models were new. Knowing the usual patterns is more useful than hunting for a single "fake" tell.

Documented reductions of famous models. Barbedienne, Susse, and others issued authorised smaller versions of salon successes. These are period decorative bronzes. Price them as such, not as unique works. The Collas stamp, foundry name, and a standard height are your friends.

Later recasts of popular animaliers. Barye, Mène, Fratin, Cain, and P. J. Mêne models were recast for a century. Some recasts are competent twentieth-century foundry work. Many are coarse, with blurred whiskers, heavy bases, and a signature that is the only sharp thing on the piece. Compare the modelling of ears, claws, and ground texture with a museum cast of the same model.

"After Rodin" cottage industry. The Thinker, The Kiss, and Balzac exist in every gift-shop size. A Rodin name on a 20 cm resin or a bright-brass bookend is décor. Lifetime and early Rudier casts are a specialist field. Do not use a phone estimate as permission to pay a specialist price.

Vienna cold-painted animals and "Bergman" marks. Real Franz Bergman / Namgreb painted bronzes are collected. The market is flooded with later painted metal and resin copies, including nonsense stamps. Paint quality, bronze weight, and a mark that matches documented forms all have to agree.

Lamp and clock conversions. A decent figure drilled through the head or back for wiring, or bolted to a later clock case, is a marriage. Photograph the holes. Value follows the quality of the bronze minus the damage, not the fantasy of an untouched sculpture.

Asian ritual bronzes and later tourist casts. Archaic Chinese forms (ding, gu, animals) have a serious scholarly market and a warehouse of recent copies with fake burial patina. That is a specialist lane. Photograph inscriptions, underside, and encrustation, then find someone who handles archaic metal — not a generalist who "likes antiques".

Using Antique ID on a Bronze

Once you have the shot list, run the images through Antique ID. Start with the clearest full-figure photo so the model can be recognised, then add the mark close-ups and the underside. The app compares visible features with a large set of sold and documented objects and returns a suggested identification plus a value range.

Use that output as a hypothesis. If it names a sculptor and foundry, go and look at documented casts of that model: height, mark placement, chasing quality. If it returns a wide range, the photos may be too poor or the model too common for a tight number. If it suggests a famous name and a high figure, treat that as a reason to slow down, not a reason to list tonight.

The range reflects recent comparable sales for similar objects in similar condition. It is a market sketch. It is not a formal appraisal and it does not certify that the metal is bronze or that a signature is lifetime. For more on how photo estimates sit next to sold prices, see how to estimate antique value from a photo.

When a Photo Estimate Is Not Enough

Escalate when any of the following is true: the suggested value would change an estate split or an insurance schedule; the name on the base is a blue-chip sculptor; the piece might be an authorised posthumous edition rather than a decorative recast; you are about to pay four figures or more on the strength of a listing photo; or the metal itself is in doubt and a chip test would damage the object.

A sculpture specialist can check alloy, look at chasing under magnification, compare foundry stamps with known punches, and write a document you can actually give to an insurer or an executor. Photo tools, including Antique ID, do none of that. They get you to the right questions faster. That is the job.

If the bronze arrived in a house clearance, photograph it in place before anyone "cleans it up", and keep the marble plinth, screws, and any paper labels. The inherited-object workflow in how to identify inherited antiques from photos applies directly: do not polish, do not separate parts, do not throw the heavy base away because it looks modern.

A Practical Shot List You Can Reuse

Before you open the app or write a listing, you should have:

One full front, one three-quarter, one back. Four sides of the bronze base. One parallel close-up of every line of text. Underside of the bronze. Underside or edge of any marble or wood plinth. A raking-light shot of seams or repairs. Height in centimetres and weight in grams written in the notes. A colour-accurate shot in daylight, not a warm bulb that turns spelter into "gilt bronze".

That packet is enough for a useful first identification on most nineteenth-century French decorative bronzes, animaliers, and reduction busts. It is not enough to authenticate a Rodin, a Degas, or an archaic Chinese vessel. Knowing which packet you have is part of identifying the object.

Frequently Asked Questions

Can a photo tell me if my sculpture is real bronze?

Sometimes. A chip that shows warm yellow-brown metal, a heavy documented model, and a period foundry mark together make bronze likely. A silver-grey chip, a plastic core, or a feather-light figure make bronze unlikely. Many intact pieces need a hands-on check. Do not file a sound surface just to answer the question.

Does a sculptor signature mean the sculptor made this cast?

No. On edition bronzes the signature is usually in the model and appears on later recasts. The foundry mark, edition evidence, chasing quality, and documented sizes are what tie a cast to a workshop and a period.

Is an "after" bronze worthless?

No. Period after reductions by houses such as Barbedienne are collected as decorative sculpture. They are worth what that commercial edition is worth. They are not worth what an original modelled work by the named sculptor is worth. Modern unsigned "after" copies sit lower still.

Will a magnet stick to real bronze?

It should not. A magnet that sticks means iron or steel is present. A magnet that does not stick does not prove bronze — spelter, brass, and resin also fail the stick test.

Should I restore the patina before I photograph it?

No. Photograph the surface as found. Cleaning and repatinating are conservation decisions. They also erase the wear pattern that helps date the object and they can cut the value of a good original skin.

Is an Antique ID value a certified appraisal?

No. It is an automated estimate from photographs and comparable sales. It is useful for sorting, listing, and deciding whether specialist help is worth paying for. It is not a certified appraisal, an insurance valuation, or a guarantee of authenticity.

From a Photo Packet to a Decision

Identifying an antique bronze sculpture from a photo is a packet problem, not a single-image problem. The model lives in the full figure. The workshop lives in the foundry stamp. The commercial story lives in after, Collas, and height. The material lives in chips, underside texture, and weight. The age lives in the patina you have not yet polished off.

Take the photos. Read the marks without wishful translation. Run the best images through Antique ID for a first name and a range. Then check that name against documented casts. If the numbers get serious, or the metal looks wrong, hand the object to someone who can pick it up. That sequence — evidence, hypothesis, check, escalate — is how you keep a mantel bronze from becoming an expensive story.