Historical Character Art Commissions: Getting the Period Right

The first historical commission I ever fumbled was a Norman knight. The client, a polite history-PhD named Tomasz, had sent me a four-page brief with footnotes. I painted a beautiful kite-shielded knight in chainmail. I shipped it on a Friday. On Saturday morning, Tomasz wrote back, very kindly, to point out that my knight's mail coif sat under a great helm that wasn't invented for another ninety years. I had jumped a generation. The painting was period-accurate the way a Hollywood costume is period-accurate, which is to say: not really.

I repainted his helm that weekend. I also started taking historical briefs seriously in a way I hadn't before. This guide is everything I've learned since, written for the kind of client who notices when the year is wrong by a century, and for the kind of client who doesn't yet know to notice and would like to.

What this guide covers

• Why "historical" is a different brief than "fantasy"
• The era pre-conversation
• Medieval armor at portrait scale
• Viking and Norse, the helmet conversation
• Samurai eras at a glance
• Edwardian and Victorian, and the 1920s crossover
• Reference checking, how I verify accuracy
• Common mistakes I now sketch around
• Where to start your brief

Why "historical" is a different brief than "fantasy"

A fantasy character art commission gives me latitude. If a client briefs me a half-elf paladin in gilded plate with a sun motif on the breast, I can invent the silhouette, the rivet pattern, the cut of the surcoat. Nobody is going to email me to point out that the buckle style is wrong, because the buckles are mine. Fantasy gives the painter authorship over the world the character lives in.

A historical character art commission removes most of that authorship and replaces it with research. The world already exists. Someone built that armor in 1196. Someone wove that brocade in 1582. Someone owned that hairpin in late Heian Japan, and they had reasons for the length of the prong and the choice of lacquer. My job stops being "invent something that feels right" and becomes "find the right thing and paint it well."

The client base for historical work is different too. Fantasy clients send me Pinterest boards. Historical clients send me museum collection numbers. A reenactor named Sera once sent me a brief that included three Metropolitan Museum object IDs, a JSTOR article on twelfth-century surcoat construction, and a hand-drawn diagram of how she wanted the mail aventail to drape under the helm. That brief took me four hours to read properly. The painting it produced was the best historical piece I'd done that year.

Here is the trade. Fantasy gives the painter freedom and the client a longer rope. Historical gives the painter a research bill at the front of the project and the client a much narrower set of acceptable outcomes. Both are valid. They are not the same job.

The era pre-conversation, locking the year range

Before I touch a brush on a historical piece, I make the client and myself agree on a year range. Not a century. A range tighter than that. "Medieval" is a useless brief. "Western Europe, roughly 1180 to 1220" is something I can paint from.

I push this conversation hard because it prevents the Norman-knight-with-the-wrong-helm problem at the front of the project instead of the back. A fifty-year window is generous. A thirty-year window is better. A twenty-year window is what a reenactor will give me on the first email.

A few examples of how these conversations actually go. Quentin wanted a "medieval crusader." We ended up at "Frankish noble, Outremer, 1170s," which let me paint a surcoat over mail with an early heater shield, no plate at all, beard trimmed in the local Levantine fashion. Helene wanted a "Viking warrior." We ended up at "late Viking Age, Birka or Hedeby, roughly 950," which meant I could give her a tablet-woven trim on the apron-dress and a cap of a particular shape that wouldn't have existed a hundred years earlier or later. Olu wanted a "samurai." We ended up at "Sengoku-period retainer, around 1570," which is a different painting than a peacetime Edo officer from 1750, because the armor, the helmets, and the very shape of the man's silhouette are different by enough to embarrass me if I get it wrong.

The reenactors do this naturally. The TTRPG and book-illustration clients usually don't, and that is fine. Part of the job, on the first call, is helping a client narrow from "medieval" to a window I can actually paint. A period-piece book illustration brief starts with this same conversation.

The single most expensive minute of a historical commission is the one where you pick a year. Spend it before sketches, not after the painting is shipped.

Medieval armor, what actually reads at portrait scale

Most of my medieval clients want armor in the portrait. The honest news is that at portrait scale, with a head-and-shoulders or three-quarter crop, what reads is not the rivet count. What reads is the silhouette, the surface, and one or two correct details near the face.

Across the medieval window, those details change a lot. A few rough markers, because clients ask:

• Late twelfth century (Norman, early Plantagenet). Mail hauberk with integrated coif. Cylindrical or conical helm with a nasal bar. Surcoat starts appearing late in the period, plain or with a simple charge. No plate. The portrait silhouette is rounded shoulders under mail, with a soft drape at the neck where the coif meets the surcoat.
• Mid to late thirteenth century. Great helms arrive. Heraldry on the surcoat becomes loud. Mail is everywhere; plate is beginning to appear as small reinforcement pieces at the knees and elbows, but you won't see those in a portrait crop. The face is often invisible under the great helm in formal portraiture, which is a real compositional problem you have to solve.
• Fourteenth century. Transitional armor. Coat-of-plates over mail. Bascinet helmets, often with a movable visor (the "houndskull" or pig-faced bascinet shows up mid-century). Tighter, more tailored silhouette.
• Fifteenth century. The Gothic and Milanese plate harnesses, the things most people picture when they say "knight in shining armor." Full plate, fluted, fitted to the body. The silhouette is dramatic and reads instantly at portrait scale.

For the practical question "which armor do I paint at portrait scale," the answer is: paint the one or two details near the face correctly, and the rest can simplify. The great helm cheek-piece. The mail coif's gap at the throat. The cusp of the bascinet's visor where it hinges. Get those right and a knowledgeable viewer will trust the whole piece. I have a whole follow-up post on the five medieval armor mistakes I've made and fixed, which goes deeper into what to sketch from photo reference versus what to invent.

Viking and Norse, getting the helmet right

There is exactly one conversation that comes up on every single Viking commission, and I have learned to have it on the first email. Viking helmets did not have horns. Not in any reliable archaeological find from the actual Viking Age — not one. The horned helmet image comes from nineteenth-century opera costuming, specifically Wagner's Ring Cycle, where a designer named Carl Emil Doepler invented the look for the 1876 Bayreuth production. It has nothing to do with the Norse world your client thinks they want me to paint.

Most clients know this when prompted. A few push back: they want horns because horns are what their mental image of a Viking looks like. I tell them I can paint horns, but the piece will then read as fantasy or theatrical costume rather than as a historical Norse warrior. Almost everyone, once they understand the trade, asks for the period-accurate helmet instead. The actual options are roughly:

• The Gjermundbu helmet (named for the Norwegian find). Rounded iron cap, spectacle-style eye guards forming a half-mask over the brow and nose, sometimes a mail aventail at the back. This is the one credible, recovered Viking-era helmet, and it is the historically defensible default.
• A simple spangenhelm. Conical or rounded iron helmet built from segments riveted around a frame. Not exclusively Viking, but used across the Norse world and broadly correct for the period.
• No helmet at all. Plenty of period portraits in modern Viking art skip the helmet entirely. A character in a tunic with tablet-woven trim, hair braided or worn loose, a seax at the belt and a single arm-ring catching the light, reads as Viking faster than a wrong helmet would.

Beyond the helmet, the period markers that read at portrait scale are the brooches (oval tortoise brooches at the shoulders for women, penannular brooches at the cloak for men), the layered tunic-and-cloak silhouette, beard style, and the tablet-woven trim on the sleeves and hems. A good Viking portrait does not need armor at all. It needs the right brooch.

I cover all of this at full depth in the Viking era character art guide for Forbidden Lands and historical briefs, and there's a specific Viking portrait piece on helmets, braids, and period accuracy that walks through the specific trade-offs.

Samurai eras at a glance, Sengoku versus Edo

The most common mistake I see in samurai commissions, both my own early ones and most fan-art on the open internet, is conflating the entire span of samurai history into one look. The gap between a Sengoku-period (roughly 1467 to 1615) battlefield retainer and an Edo-period (1603 to 1868) peacetime officer is the difference between a Norman knight and a Napoleonic hussar. They are both "warriors of their country" and they look almost nothing alike.

The big distinctions, at portrait scale:

• Sengoku-period armor (ō-yoroi and dō-maru, then increasingly tōsei gusoku). Heavy lacquered plates laced together in distinctive horizontal bands. The kabuto (helmet) often features dramatic crests called maedate, which can be horns, half-moons, sunbursts, or stylized animals. This is the era of dramatic battlefield armor designed to be recognized from distance, and at portrait crop the kabuto and maedate do most of the work.
• Edo-period samurai. Long peace. Most samurai are administrators and bureaucrats in formal dress called kamishimo, with shaved pates (chonmage hairstyle), two swords at the belt (the daishō, katana plus wakizashi), and silks rather than armor. The portrait silhouette is square-shouldered from the kataginu (the stiff sleeveless garment), and the face is the focus, not the armor.

The other thing worth knowing is that women in samurai-class households also commissioned portraits, and the visual language is entirely different again. The onna-musha figures who fought, like Tomoe Gozen, are usually portrayed in armor, but the daily-life samurai-class women you see in Edo portraiture wear formal kimono with specific hairstyle conventions for marital status and age.

My rule of thumb: when a client says "samurai," I ask first whether they want battlefield or peacetime, then I ask whether they want a specific era within that. Sengoku and Edo are the two biggest buckets, and that single question saves everyone a repaint. The samurai character art piece on Japanese periods goes deeper, and the samurai portrait piece on kabuto, kimono, and era accuracy is the one to read if you want the specifics on a portrait composition.

Edwardian and Victorian, and the 1920s Cthulhu crossover

A surprising share of my historical commissions sit between roughly 1890 and 1930. Part of this is the broad popularity of period drama in the last decade. Clients come in wanting "their character if she lived in Downton" or "him in the era of the Great War." A larger part is the steady demand for 1920s Call of Cthulhu investigator portraits, which sit at the most active crossover point between my historical and horror work.

For this band of years, the era markers are tighter than people expect and the visual language shifts decade by decade. A working sketch of the era markers:

• Late Victorian (1880s–1890s). High collars, bustles in earlier years and slimmer skirts by the late 1890s, gentlemen in frock coats with stiff detachable collars. Hair is up for women, mustaches and full beards for men. The lighting language for these portraits is the gas-lamp era, warmer, more localized light sources, longer shadows.
• Edwardian (roughly 1901 to 1910). Looser silhouette for women, the famous S-curve corset, big hats. Men's lounge suits start to displace the frock coat. This is the Titanic era, and clients will often arrive with that visual reference whether they say so or not.
• 1910s. The war years. Military portraits are common requests here. Women's skirts shorten, hats simplify. The big change is the loss of the corseted silhouette.
• 1920s. Drop-waist dresses, bobs and finger waves, cloche hats. Men in three-piece suits with shorter, more tailored cuts. This is the era most Call of Cthulhu clients arrive at by default, and the visual language overlaps almost completely with horror character commissions set in the 1920s.

The Cthulhu crossover is worth its own paragraph. A 1920s investigator portrait is a historical portrait first and a horror portrait second. The clothes, the hair, the spectacles, the cigarette holder, the props in the foreground: all of that is being painted as if for a Victorian-society portrait, with horror coming in through the eyes, the posture, the props (a strange book, an unfamiliar artifact), and the atmosphere of the background. I have painted Cthulhu investigators in nearly the same compositional language as Edwardian portraiture, and the genre-cue is doing maybe ten percent of the lift. The other ninety percent is the period research. There's a full breakdown in the Edwardian and Victorian portrait commission piece, if a client is sitting at this crossover.

Renaissance courts, the era I work in less, and what I notice when I do

I take fewer Renaissance commissions than the other eras above, but they come in often enough that I have learned the basics. Italian Renaissance portraits (think 1480 to 1550) are about brocade, restrained silhouettes, and a specific lighting language pulled from Bellini, Botticelli, the early Titian. Northern Renaissance (German, Flemish) is darker, more textural, and built around the Holbein-and-Dürer tradition. French and Tudor court portraits sit in between.

The hardest thing to get right at portrait scale here is the fabric. Renaissance brocade is dense, gold-shot, geometric, and very expensive to paint properly. I charge a small premium on Renaissance commissions because the surface work is more than three times what a plain wool surcoat would take. Worth saying upfront so clients aren't surprised.

Reference checking, how I actually verify accuracy

This is the part of historical work that fantasy clients usually do not see, and the part that historical clients want me to be honest about. The research process for a historical commission has roughly four layers, and I lean on all of them on every piece.

The first layer is museum collections online. The Metropolitan Museum's open-access collection, the British Museum, the V&A, the Royal Armouries in Leeds, the National Museum of Denmark for Viking material, the Tokyo National Museum for Japanese material. These are the canonical primary references. If a client has sent me an object number, I look it up. If they haven't, I pull period-appropriate objects in the same category and use those as my visual baseline for the painting.

The second layer is specialist secondary literature. For European armor, I lean on Claude Blair, David Edge, and the Royal Armouries' published catalogs. For Viking material, the Birka and Hedeby excavation reports. For Japanese armor, the work of Trevor Absolon. I do not pretend to be an academic, and I do not need to be. I need enough of the secondary literature to know what is settled, what is contested, and what is myth.

The third layer is living history and reenactor communities. The serious reenactors who post photos of their kit, with notes on which pieces are documented from finds and which are reconstructed from period art, are an enormous resource. I have learned more about how mail actually drapes from a single Birka-focused reenactor's photo set than from any museum catalog.

The fourth layer, which I never skip on a paid commission, is a direct check on the client's specific references. If a client sends me an image they want me to draw from, I trace the image to its source. If it is a photo of a piece in the Wallace Collection, great. If it is a screenshot of a video game cutscene that someone has labeled "12th century knight," I will gently push back and propose a more defensible reference. Most clients accept this readily. The few who do not are usually clients I should not have taken to begin with.

I have a longer piece on the actual workflow I use for historical reference checking and period accuracy, which goes through a worked example end to end.

Research is a line item, not a hobby. On a historical brief, I am being paid to know what I do not know and to find it out before I touch the canvas.

A six-week historical commission, from brief to ship

A worked example to make the process concrete. Linnea came to me in late September wanting a portrait of an invented twelfth-century noblewoman based loosely on Eleanor of Aquitaine, a historical figure I have painted before in a different Eleanor of Aquitaine commission piece. Linnea did not want Eleanor herself. She wanted "a woman of Eleanor's milieu, around 1170, Aquitaine court, painted to live above my desk."

The kickoff call was forty minutes and we made four decisions. Year range: 1165 to 1175, narrow enough that the bliaut style is consistent and the headdress conventions are stable. Fabric: wool over linen, deep blue overdye, no silk on this piece because court-grade silk wasn't quite available yet at that latitude. Hair and headdress: two long plaits wrapped in silk ribbons, a circlet rather than a full veil, the noblewoman is at home, not at a public mass. Lighting: afternoon window light from camera-left, indoor, a private chamber rather than a great hall.

Week one was thumbnails and the research bill. I did three small compositions, sent them to Linnea with a half-page of notes on which museum pieces I was drawing the bliaut sleeve cuff from, and we picked the second thumbnail. Week two was value study and palette block. Week three started the rendering pass. Week four was the face, the longest stage on any historical piece, because the face has to read as a real person in a real century, not as a costume model. Week five wrapped the hands, the textile detail, and the small jewelry. Week six was final pass, a single revision (Linnea asked me to tighten the circlet to sit lower on the brow), framing prep, and ship.

The piece is in her study now. I have a photo of it from her on my phone, taken in the same afternoon light I painted it in, which I find unreasonably satisfying.

Common mistakes I now sketch around

After two years of historical work, here is the running list of things I check for on every brief, because they are the failures I have either made myself or watched other painters make.

• Mixing eras within a single character. A Norman helmet with a fourteenth-century surcoat. A Sengoku kabuto over Edo-period peacetime clothing. A 1920s investigator wearing a 1910s collar. These are all small individual errors that combine into a piece that reads as costume rather than as history.
• Letting fantasy bleed in. Spaulders and pauldrons borrowed from fantasy game design. Buckles that don't exist in the period. Cloaks that are bigger than any cloak any human has ever worn. The fantasy crutches creep in when the painter has not done enough reference work and is filling gaps from imagination. I have done this and learned to catch it.
• Costume drama clichés. The wrong things often arrive in a brief because of period dramas. Tudor-era characters in distinctly post-Tudor cuts. Vikings with the wrong braids (the braid styles in popular TV shows are not period-defensible). Edwardian characters in clothing that is actually mid-1920s. If the reference is a TV show, I treat it the way I treat a video game screenshot: as a starting point that needs to be checked.
• Wrong-period hair and grooming. Hair is the single fastest tell. A medieval European man with a modern haircut, even under a coif, reads wrong. A samurai with the wrong chonmage style for the era reads wrong to anyone who has looked at period prints. Hair takes more research than most clients expect.
• The horns. Already covered. Worth repeating, because it will come up on the first Viking brief every time.
• Anachronistic posture and expression. People in the past stood differently, sat differently, and posed for portraits with a different visual grammar. A 1920s woman doesn't smile for a portrait the way a 2020s woman does. A Renaissance courtier doesn't tilt their head like a contemporary fashion model. Posture is a period detail.
• Treating "historical" as a single genre. It isn't. A medieval Norman brief, a Viking brief, a samurai brief, and a 1920s investigator brief are four different jobs with four different research stacks. I rarely paint two of them in the same week, and when I do, I keep separate reference folders open on different monitors so I do not contaminate one with the other.

Where to start your brief

A historical commission is a longer conversation than a fantasy one. The good news is that the conversation is finite, the research is checkable, and the result, when it lands, sits on a wall for a long time without going out of style. Costume drama dates. A well-researched portrait of a 1170s Aquitaine noblewoman doesn't.

If you have a historical piece you have been carrying around, the most efficient way to get it in front of me is to drop a brief through the order form and include the year range you are thinking about, even if it is wide. I will narrow it with you on the call. The portfolio has historical pieces tagged by era, and if you want to see closest-comparable work for the period you are after, the gallery is the fastest way. The character work service page has the current turnaround and pricing for period-accurate single portraits, and the custom projects page is where book-illustration and multi-piece period work lives. Either way — write me a one-line pitch and a year range, and the rest we'll figure out together.