ERC EU logo


Natural Purples on Silk – Discovering an Early Modern Palette  

By Kirsi Mantua-Kommonen

In 2022-23, the ERC-Refashioning project carried out a historical velvet waving experiment, in collaboration with the Fashion History Lab at Aalto University, Finland.  The purpose of the experiment was to test what a high-quality purple silk velvet might have looked like in the early modern period and to record the processes and experiences of both the weaving and dyeing. The experiment was designed and carried out by Professor Paula Hohti, myself, and the weaver and MA student Elise Piquemal. My role in this experiment was to dye the silk samples in various shades of purpl.

The Dye Process

For the purposes of this experimentwe created a palette of exquisite purples using natural dyes available in Early Modern Europe. Furthermore, we had the luxury of working with the highest quality mulberry silk grown and produced locally in France. We selected the following natural dyes: carmine extracted from cochineal (Dactylopius coccus), indigo blue from local woad (Isatis tinctoria L.), warm reds from brazilwood (Caesalpinia L.) and madder (Rubia tinctorum L.), and a purplish shade from logwood (Haematoxylum campechianum L.). Combining these we achieved the palette shown below. 

Next, I will briefly introduce the colours and natural dyestuffs shown in more detail.

Blue with Finnish Woad

Woad (Isatis tinctoria L.) is an indigo-producing plant that can be grown across Europe. For this reason, it was the most important source for blue for textiles in Early Modern Europe, while also imported indigo (Indigofera) was sometimes used.

The magical process of indigo dyeing is intriguing, because the colorant is not water-soluble. Therefore, it requires first making a dye vat, the colour of which is not blue, but yellow or greenish. The dye is reduced into water-soluble form, and when it is ready, the surface of the liquid seems metallic, like purple-coloured quicksilver! In the middle it has formed fluffy foam or bubbles, poetically called ‘the flower’.

                 Dried woad balls.                                                               Woad powder.                                                    Indigo vat ready for dyeing.

However, getting a uniform blue on our delicate silk yarn was not easy, because the dye only fixes to the surface of the skein. We carefully dipped the yarn skeins into the solution for a few minutes. We gently moved the skeins around in the vat to facilitate an even dye result, but avoiding letting oxygen into the liquid. As the yarn was lifted from the dye vat, it was first yellow and then immediately turned green. Within only a few minutes, the colour transformed again, now into blue, right before our eyes. This process is fascinating every time!

                                                  Preparing the indigo vat.                                       Observing the colour transformation.

On a less desirable side, the prevailing pungent odours of the fumes emitted in the process, which used to involve fermenting in urine, nowadays replaced with ammonia, are far from pleasant or harmless. This may have been one of the main reasons for the dyers’ quarters being located further away from those of the elites in Early Modern towns.

Even today, woad is an important dye plant grown in several European countries, including Finland. Woad is a biennial plant, which means it has a lifecycle of two years: On the first year, the plant germinates and grows its leaves that contain the indigo compounds, and on the second year it blooms and produces seeds, which are collected for sowing again the following year.

                                             Mulberry silk yarn (left) and industrial silk fabric (right, still damp), both dyed at the same time in an indigo vat.

When this indigo blue was over-dyed with red dyes (details follow below), a palette of purples began to emerge

Purple with Logwood

 Logwood (Haematoxylum campechianum L.) is a tree species originating from Central America, and coveted for the dye extracted from its hardwood. The beautiful purple colour it provides is not lightfast. With certain mordants no longer recommended due to their toxicity (e.g. copper and tin), it provided lustrous shades of blues, purples and reds, which were also inclined to fade. Consequently, already in Early Modern Europe its use was strictly restricted. However, it was more successfully introduced as a compound for dyeing beautiful, colourfast blacks when combined with madder and tannin, and mordanted with chromium, also still commonly used at that time. The consequent high interest towards this new dye resource meant chopping down entire forests in areas of modern day Mexico and Belize, and introducing the tree to further areas to keep up with the demand.

Certified, sustainably grown logwood may nowadays still be an option as a base for dyeing natural black. Even today, the main colour compound extracted from logwood, hematoxylin, is significant as a histochemical staining agent – it is used as a contrast agent in pathological diagnostic analysis safely applicable for all organs.

Like the following colorants in our experiment, also logwood is a so-called mordant dye. This means that the dye and fibre fix better with the help of a mordant, creating a more colourfast result. The least harmful and therefore concurrently most used mordant is alum, which we applied before dyeing.

Opening the lid of the logwood kettle during the long extraction and dyeing processes released lovely aromas of herbs, hay, and hints of sweet caramel, adding to the delight of our natural dyeing experience.

Similarly to indigo, also logwood can be combined with other dyes for purples. This even increases its colourfastness. For purples we experimented with combining logwood with madder and cochineal as depicted below.

Carmine from Cochineal

 Cochineal (Dactylopius coccus) is a tiny scale insect, which feeds off the red sap of a certain species of cacti, Prickly Pear (Opuntia phaecantha), in subtropical South America. Indigenous peoples, first in Peru and later also for example Mayas and Aztecs (Mexico), had perfected the skill for cultivating cochineal. This New World dye charmed Europeans with its brilliance, unmatched by the ones previously known in Europe, and became the dye for royalty.

Nowadays, the production of this traditional survives in Peru, Mexico and other South American countries, and also on the Island of Lanzarote of the Canary Islands. There, one family continues the work of their ancestors, passed on from the era of conquistadors, who finally got to smuggle some cacti and cochineals with them and succeeded in local cultivation of this valuable dye source, with a very high dye content. This dye is used for textiles, but also for colouring food, drinks, and cosmetics, such as lipstick. Rather than saying carmine acid, the package is likely to have the code E120 – this is cochineal!

For anyone enjoying shades of hot pink, fuchsia, and carmine red, dyeing with cochineal is a treat, since such bright and bold shades are almost impossible to achieve with other natural dyes. We first ground the dried insects in a mortar to a fine powder, and then extracted the colour from it by cooking. Achieving bright pink is easy, but the heat needs to be kept quite high for the colour to fix properly with the fibre. Combined with indigo or logwood, it provides exquisite, colourfast purples.

Madder and Brazilwood Provide Warm Reds

Dyers’ madder (Rubia tinctorum L.) is a plant that produces a multitude of red dye compounds such as alizarin, purpurin, and rubiadin, in its roots. Madder is the oldest source of red dye found on historical textiles in many areas in Europe, and it is native to the Middle East and eastern Mediterranean area. Later its cultivation spread west and north as a perennial, and as an annual plant even further north.

Dyeing with madder is fascinating but, if a particular shade is aimed for, also demanding: changes in extraction and dye temperature influence the resulting red colour, ranging from yellow-oranges to reds to brownish reds. The best tones of red are achieved at temperatures below seventy degrees centigrade, and even cold-dyeing method works well with madder. Higher temperatures provide brownish tones. At first the roots extract yellow, and over the course of history many ways have been developed to “get rid of” this yellow or dye something else with it first to secure the purest possible red from this dyestuff. Even the soil and area where the plant was grown have an influence on which compounds it has produced and, consequently, on the dye result.

Brazilwood and related redwood species (Caesalpina ssp) are trees, which produce similar shades of red, as does madder. However, these dyes are not very colourfast, which is why it was often combined with other red dyes, or used for the warp or other less visible parts of the textiles. The dye from brazilwood is sensitive to changes in acidity/alkalinity, which is rather typical to the less colourfast dyes. Therefore, to support the colourfastness it would have been important to use soap with similar pH-level as that of the dye bath if the product needed to be washed, and to store the piece protected from light.

Also, similarly to logwood, very large quantities of redwood shavings are required for extracting enough dye to obtain rich colours for fibres, typically ten times the weight of fibre. Therefore, for modern day experiments it is important to check the certification of the material producer.

Dyeing with brazilwood is an enjoyable experience not only due to the beautiful reds it affords, but also because of the lovely scents for exotic flowers and fruit that seem to have followed the dye from the place of origin of these trees.

General Remarks

Many of the above dyes; cochineal, madder and brazilwood in particular, were used for purple also on their own by modifying their colour with metal mordants, copper, chromium, tin, or led. Consequently, in our experiments we did not use these, since the fine silk we worked with would likely have been damaged by these treatments. Furthermore, these substances are nowadays known to be dangerous for the dyer, the environment – due to effluent waters – and even the user of the dyed products, so their use is not recommendable if not absolutely necessary for the particular experiment. Iron mordant is less toxic, but it could also have been harmful for our delicate silk yarn.

The human eye perceives colours from natural dyes very differently in changing lighting conditions, and the reflections on silk amplify this phenomenon. The photos below show the difference between a photo taken without or with a flashlight. As candlelight would most probably have set the atmosphere for observing such materials, we can only imagine the dramatic effect these kinds of silks had for the people whose vision had not yet been polluted by synthetic dyes.

To see what these colours look when woven into velvet and how they further changed their appearance, see Elise Piquemal’s post on the weaving process in the Refashioning the Renaissance blog.


For more information on the Refashioning weaving experiment, click here



Boutrup, J., & Ellis, C. (2018). The art and science of natural dyes: principles, experiments, and results. Schiffer Publishing, Limited.

Cardon, D. (2007). Natural dyes. Sources, tradition, technology and science, 268.

Kirby, J., van Bommel, M. R., Verhecken, A., Spring, M., Vanden Berghe, I., Stege, H., & Richter, M. (2014). Natural colorants for dyeing and lake pigments: practical recipes and their historical sources. Archetype Publications Ltd in association with CHARISMA.

Rodríguez, L. C., & Niemeyer, H. M. (2001). Cochineal production: a reviving Precolumbian industry. Athena Review2(4), 76-78.

©Photos: Kirsi Mantua-Kommonen, Elise Piquemal/Fashion History Lab



Representing Renaissance textures and colours digitally

By Maarit Kalmakurki

In this blogpost, I discuss how textures and colours are created as digital garments and how beneficial digital animation is for testing different Renaissance fabrics and colours. I joined the Refashioning the Renaissance team in 2019 with the aim to build a digital reconstruction of an early 17th century male doublet. This practice-led research task was highly interesting to me as it enabled me to combine my knowledge in historical dress, pattern cutting and garment making in general, as well as my theoretical and practical knowledge in digital costume design and animation. The initial aim was to build a digital doublet, present many of its complex inner layers and structures, and make an animation or video of the doublet on an avatar in a virtual environment. The nature of practice-led research allows us to discover new working methods and to make decisions along the way, which was one of the most interesting and exciting parts of this process.

The final rendition of the digital doublet was achieved through a combination of many source materials and decisions along the animation process. One of the inspirational source materials for the digital doublet was a post-mortem inventory of a Florentine water seller Francesco Ristori. The inventory described one garment; a black doublet made out of stamped mockado. The School of Historical Dress in London made a physical version of this doublet and I used the same paper patterns for the digital creation. I also studied other 17thcentury doublet patterns, sewing techniques and materials in order to make changes that were more appropriate for the digital animation. One inspirational and eye-opening learning experience was the one-week course of doublet making at the School of Historical Dress. This enabled me to understand why certain materials and sewing techniques are implemented in doublets and my task later was to re-create such features digitally.

The starting point of the digital animation is based on patterns. In the case of a doublet which includes several layers of fabric and additional supportive materials sandwiched between the lining and main fabric, I multiplied the patterns to gain the same amount of pattern pieces as there would be in a historical doublet. Once the patterns are multiplied, they literally fly in the digital space (Figure 1) before assembling them on top of the avatar body. The patterns are colour coded according to what material they are made from. Digital material behaviour and texture are interesting as everything can be adjusted in the animation program. Fabric thickness, weight and sheen can all be controlled in quite some detail with a click of a mouse. This aspect proves to be one of the most beneficial aspects of digital animation.

Figure 1. This image shows how digital patterns literally fly in the digital space before they are assembled next to the avatar body and simulated as a garment. Screenshot by Maarit Kalmakurki.

Another beneficial aspect of digital animation is the choice of illustrating material textures. Fabric surfaces in digital garments are actually photographic images of texture, inserted on top of the patterns. This means that it is relatively easy to change material textures and colours in digital animation. My early experiments with material texture show how different garments look depending on what kind of texture image is applied on top of the patterns. The appearance of a historical doublet is achieved with the right kind of texture image and colour. Figures 2-3 illustrate one of my first examples of the doublet and its texture. The doublet material is a photograph of a stamped velvet I made during the doublet tailoring course at the School of Historic Dress. Stamping was made by using a fork heated over a metal hotplate.

Figures 2-3. In one of the first doublet material experiments I used an image of a stamped velvet made by using a fork heated over a metal hotplate.





For the final digital reconstruction of the waterseller’s doublet, I chose to photograph the stamped mockado that was custom woven and dyed for the Refashioning the Renaissance project’s doublet experiment. I adjusted the thickness of the fabric first in Clo3D to resemble the thick, wool material. This also helped to refine the sculpted form of the doublet. The photograph of stamped mockado, however, appeared fairly dark in the digital environment even though it appears dark blue in the material reconstruction. Colour is easy to adjust in digital animation, consequently, I changed the blue tone slightly to make the stamped pattern more visible in the wool (Figure 4).

Figure 4. Refashioning the Renaissance digital animation experiment. Screenshot by Maarit Kalmakurki.

Because Francesco Ristori’s post-mortem inventory was one of the important starting points for the entire doublet experiment, I wanted to further use the benefits of digital animation and decided to make one version of the digital doublet in black (Figure 5). Black colour was difficult to achieve in the early modern period and very often dyed in hues of dark brown, red or green, depending on the dye recipe (see Ortega Saez and Cattersel 2022). The Refashioning doublet made at the School of Historical Dress was dyed in deep blue colour because expensive black was traditionally made by dying cloth with red and blue tones. Before the discovery of logwood, there was no dyestuff that made true black. In the case of the Refashioning doublet, the dyer made a very deep blue. Details, however, are difficult to perceive in any dark colours, so the stamping and ribbons almost disappeared in the black doublet. However, quite effortlessly, it helped us to achieve one of the goals in visualising what Francesco Ristori’s doublet might have looked like.

The method of implementing digital animation in dress history research proved useful in the Refashioning project. It offers new possibilities to test, experiment and illustrate historical garments, their construction, fit and materials. Alongside with hands on making, digital reconstruction helps to test textures and colours with just click of a mouse.  

Image 5. Digital doublet reproduction in black colour. Screenshot by Maarit Kalmakurki.

See the full animation by Maarit Kalmakurki on our You-Tube Channel:
Digital Reconstruction of a Renaissance Doublet

Knitting a 17th-century silk stocking for the Refashioning the Renaissance participatory research project

By Liisa Kylmänen

Taking on a challenge

I’ve been knitting since I was a teen and loved all kinds of needlework all my life. I also love historical clothes so when I heard that the Refashioning the Renaissance participatory research project was trying to replicate a really fine, old silk stocking from the 1650s, I enrolled as a volunteer. Challenging myself with anything hand made is always worth a try!

In early spring of 2019 our team of about 35 volunteers met for the first time. We were given the option to make three different kinds of socks, and I volunteered to knit the finest type, a replica of an old pair found in a tomb in the Turku cathedral. Before we got to work on the actual stocking we had to find out how the original pair had been knitted. Our pattern team divided into groups to study the close-up photos of the Turku stockings, counting stitches and trying to work out a logical pattern. It was quite demanding to keep track of what was happening as the socks of the pair had not been made exactly the same way and we only saw one side of the stocking at a time from the photos we had. We ended up with 11 pages of text, photos and charts, plus an excel sheet of all the rows of the stocking and what was happening on each row.

Enlargements of the photographs of the Turku stockings with my markings and the completed pattern

Testing ourselves and the materials 

The next big question was how to achieve the right gauge or stitches per cm. At our group’s first meeting we experimented with fine wool tread. I started knitting a tiny little swatch with 0,7mm needles and the thinnest wool yarn. The smallest needles I had knitted with until then were 1,75 mm, and the difference seemed enormous. The needles (HiyaHiya steel double pointed) felt really flimsy and bendy in my hands, but the result was a fabulously thin and fine piece of knitting with just over 10 stitches per centimeter! I learned to knit the continental way but on the purl rows it was impossible to open up the stitch to catch the yarn. For the first time in my life I tried the English style of knitting which I’ve always considered very cumbersome. With the tiny stitches and reduced strength of the needles, the English method was the way to go for the purl side.

             The start of my mini sock with 60 stitches and the first finest swatch with 30 stitches

For my next swatch I decided to knit in the round – like we would for the actual stocking. I used some slightly thicker yarn left over from an Estonian shawl and the same 0,7 mm needles. I ended up knitting a miniature stocking using the directions of an old Danish sock that was introduced to our team as one of the socks to be knitted by the volunteers. I often fell into trouble with a slipped stitch, a mistake to correct or just the furry yarn going into an unidentifiable mess. I had to dig out a loop to pick up the stitches from the heel flap for the gusset.

Here’s the mini sock modeled by an old action man doll. It has all the features of the 16th-century stocking: the garter, shaping by increases and decreases, a purl decoration in the back seam, a clock decoration at the ankle, a folded heel and a long gusset

In the summer we finally got to try out some silk. It was in fact much easier to knit than I had expected. Unlike the wool, any silk strands left out from the needle were easy to spot as they formed a visible loop on the outgoing yarn. We tried different kinds of silk, some still containing  their natural sericin and other softer yarns from which it had been washed out. Both had pros and cons: the sericin was harder on the hands, stiffer and hence stayed together better, whereas the washed silk was easier on the hands, more supple and also split much more. I made little swatches of many different yarns, practicing increases, decreases, purl decorations and picking up stitches at the same time. The gauge we were aiming for was around 8 to 9 stitches per centimeter, and with most yarns I achieved that with 1 mm needles.

Here is a swatch with single silk yarn where I tried out some twisted stitches, picking up from
the side, different types of increases and decreases

Getting to work on the actual stocking 

The silk we eventually received for our project was a bit of a compromise and so after all my experiments aiming for the same quality as the original, I ended up starting my sock with a double thickness of yarn.  The needles I used were HiyaHiya 1 mm steel needles and I was VERY happy that I didn’t have to take up the project with the thinner 0.7 mm ones!

Casting on and knitting the first row was very slow.

The beginning was – as always – the hardest part! It took me nearly three hours to cast on the 288 stitches and to knit the first row. I used a cast on method where the stitches were crocheted with one thread (in my case double) around the knitting needle. The knit was quite dense due to the double thickness and hence quite cumbersome to knit. I needed to tighten each stitch, especially the first couple stitches of each needle. After a couple of rows I made good use of a thimble to push the needle along! Also because of the double yarn thickness, the needles were so “full” that I managed to push only a couple of stitches forward at a time, making the process even slower.

Taking notes of time spent on each row

The stocking starts with a garter of 8 stitches knitted and 8 purled. After that it continues with a plain stocking stitch and an additional zigzag decoration on both sides of the center back. With the 1 mm needles I was near the 8 mm/cm gauge. The stocking continued with decreases down to the knee and then increasing again to make room for the calf. On both sides of the ankle there was a decorative clock which was made with purl stitches. At that point I noticed my knitting was getting looser because the purls were so hard to tighten. Compared to the rest of the stocking, the ankle remained a little looser even after washing and blocking, which was a bit disappointing…

Trying to keep the purl stitches of the clock tight enough was challenging

The heel flap was knitted down in one continuous piece with some decoration on the side. This was a very common way to make the heel. The flap was then folded in half and closed with a three needle bind off. The most tedious part of the whole knitting process was picking up the stitches from the side of the heel flap! The knit was so tight that I really had to fight with every loop picked up, using my 0,75mm crochet hook to help with every stitch, not just the extra ones picked up after every three loops. The next row was the second toughest part of the whole knitting. Not only were the stitches extremely tight but also the silk was totally split up from being worked so hard. Luckily, each row was easier and shorter after that!

The heel flap before it was folded and knitted together

Picking up the stitches from the heel flap was the toughest part of the knitting process

For the first 120 rows my pace was roughly 30 minutes per row. As the stocking progressed, slowly but surely, I managed to pick up a little speed. Normally I’m a fairly fast knitter, but with the bendy needles and tiny stitches it was impossible to knit without concentrating all the time. No “flow” like with a normal sock, just toiling away focusing on the knitting. My progress was spurred on by milestones like reaching the knee, starting the clock, and casting off the heel.

Decreases at the gusset. The last part of the stocking and each row easier from here to the end

Fortunately, I didn’t make many mistakes because unraveling the double thickness of thread was a real headache! We had noticed that the knitters back in the 17th century had not been too fussy either, so we decided with our volunteer group to accept minor mistakes, such as a decrease one row later instead of backing a whole row. Moreover, I imagine no one would have noticed anyway!

The agony of picking up a dropped stitch! Often, I had to go down several rows and with the thread being worked on many times, it ended up in an indistinguishable mess

Finally getting to the end after 260 hours 

The stocking has over 900 rows, 733 from the top to the bottom of the heel flap, plus 177 rows on the foot part. It measures 61 cm from the garter to the end of the heel. The clock or decoration on both sides of the ankle is 101 rows high. The weight of the stocking before being washed was 96 grams.

260 hours later, getting there!

The total time it took me to knit the one silk stocking was about 260 hours. I knitted on 105 days which makes an average of 2,5 hours per day. The longest I knitted on one single day was just over six hours. I have no idea what the working hours were in the 17th century, but certainly longer than that! Good lighting is imperative, so summer time would have been optimal. (Unfortunately, the silk arrived in the autumn, so I started my stocking in November and finished in April.) Other questions that came to mind: How many people would have had access to glasses, which at my age are essential? What kind of needles would the knitters have used? I knitted with what I imagine are the best quality modern steel needles and even they had a hard time keeping up with the task. I also tried a swatch with modern carbon needles, but they produced a gray stripe where I had used them and the ends of the needles seemed to disintegrate, as that was where the strain was the biggest. As a result, the tips were “frayed” and started to resemble a paint brush!

I used two sets of needles. One broke in two on row 160 and for the rest of the way, I kept rotating them. Here are the survivors 

After it was finished my stocking was handed over to the project, well over a year ago. Since then it has been washed to get rid of the sericin and dyed by the Refashioning team with fustic, which produced a beautifully deep yellow.  The stocking was exhibited at the Aalto University campus in Stitching History! Making and Wearing Early Modern Knitted Stockings, along with all the beautiful things that the research group has discovered, documented and reconstructed during the past five years.

I saw the original pair of the historic stockings at the exhibition in the Turku Castle just before starting to knit my version and remember well how daunting it seemed to try to achieve something so refined. I considered it an honor to be a part of this participatory research project and strove to make my stocking as uniform and as beautiful as the original ones. The perfectionist in me is not completely satisfied, but I am quite happy with the result. After this experience, I totally understand why a pair of stockings was such a luxury item that it might have cost you a horse!

Photos of the finished stocking

You can clearly see the needle change running through the middle of the stocking despite my efforts to minimize it. The ankle is quite wide and so is the foot part

The sole is also very short and will need a lot of stretching to fit a normal foot

The sole also has shaping to compensate for the decreases on the top side

The clock decoration at the ankle was made with purl stitches