Saturday, September 28, 2019

When 2916 became 235

I have alluded to this story before - how from the outside, the dye lots of threads seem to wander or change over time and what is behind it.  So after over a year of work, color 2916 is now 235.  What does that mean?

Back of a 17th century stumpwork picture and one 'draft' of a color palette
Go back to around 2011.  I am working again on the color palette for the original course, Cabinet of Curiosities.  This is the work that will define all thread making going forward, a set of around 40 colors that will replicate the color families on the back of 17th century stumpwork.  By this time I have 'inventoried' the colors from the works, counted about how many shades in a color family were used and looked at motifs and thought about what would be the minimum number of shades in a family to effectively embroider these motifs.  That was then combined with our knowledge of natural dyes and many samples of wools or other threads dyed in natural dyes and compared to the back of the embroideries.

Pre-2014 Au Ver a Soie color card.  Everything is in order
by number.  1033 looks a little out of order there!
So out came the color card for Au Ver a Soie.  What no one understands is that color cards are a snapshot in time.  They are really hard and expensive to make with little pieces of cardboard wrapped with a few inches of every color in existence.  But there are several colors introduced every year, custom dyed for companies like Hermes for their needs and the excess of the order are added to the line to sell it out.  They will be one-time colors.  Sometimes they are on the color cards - only if they were in existence when the card with made.  Some colors are gone.  When you got the color card stock of them existed but doesn't any more and they will not be remade.  Why?  Well, perhaps it was not a 'barn burner'.  Meaning that it was made 15 years ago and it took that long for that color to sell through one manufacturing run.  Not a profitable thing.

So if any of you have seen the old Au Ver a Soie color card, you will know that the colors are in numerical order.  Sometimes that corresponds to a color family.  Sometimes it doesn't.  The color numbers have been assigned over 200 years.  So yes, there are some things that just don't make sense as they have been assigned by so many people who had different reasons for choosing that number.  So putting color families together is a bit of a challenge.  You might find three in a row that look well shaded but you want a darker version and it doesn't have one or you know that in natural dying the undercast color shifts towards something as it gets darker or lighter and the color family in these synthetic dyes don't capture that nuance.

So I started grabbing things from other parts of the color card and in the other silk thread families.  The colors used for soie perlee, soie d'alger, soie paris, etc. are different.  There are a handful of colors that are used in each (the heavy sellers) but in some, such as soie paris, the colors are defined by the distributors or companies that special order that thread.  You are getting it not because it was defined as a thread for the hobby hand embroidery market but as it was a need for finishing scarves or making buttonholes and the colors have built up over time to be a product line based on their custom orders.

As an aside, Soie Paris was a thread developed because Access Commodities asked for it.  Lamora knew that a need in the American market was a stranded filament silk thread which was the same weight as the soie d' alger series.  Once Au Ver a Soie agreed to make it, she picked the colors that the largest buyer of silk threads at that time used in their sampler designs - Shepard's Bush.  So a group of about 10 thread colors they used were dyed in this new Soie Paris thread.  That was the entire line.  As designers tried it, Access would order new colors to be dyed to fulfill credible requests.  And slowly the line grew.  Each time the cones were emptied, a decision had to be made.  If Access request it to be dyed, they have to buy the entire batch.  If Au Ver a Soie sends a dye pot list for February that they are dying in X, if the thread type is compatible, you can throw in a kilogram of your thread type to be dyed and only have to buy that smaller amount.  So colors can slowly grow that way or never happen again.  It's business.  Not some bible of colors that must exist.  
So I grabbed things like 2645, 945, and 2916, 2914, 2012, and 1011 to make a red family.  We dug into those colors on the shelf and sometimes found that there were cones which had a brighter or duller cast to them.  We had little spools marked 'old 2914' and 'current 2914' on them.   These were all slight differences in dying over a decade.  Some of those minor changes were due to changes in environmental laws about water use and discharge.  If you have ever seen any lecture on natural dying (I just did at the MET), the water use is enormous and dirty.  I was shown a lecture about how an entire village picked up and moved to another region because an earthquake shifted the ground water to a seam that released more iron into their water and thus their natural dyes didn't take the same way.  So they moved the village!

So I would embroider on the white, cream and neutral backgrounds that my students would work on.
Trying the red family out on fabric with multiple types of thread
One of the things I had learned over the years was that med-pink is a big wild-card in embroidery.  A pink that looks amazing on the spool and you swear will look wonderful on the fabric will take on the cast of the fabric and change its looks.  Almost always it shifts to barbie doll pink.  So undertones of brown in a pink will make it look more madder in look when it is stitched on our neutral sampler backgrounds.  So I was extra careful with the pinks.  That is why 2914 became 741.  Lamora and I determined that we needed to use the 'old 2914' in the color family and not the new 2914.  So the first packages contained the last of those identified 'old 2914' cones on the shelf while we commissioned a dye bath of it - with strict instructions to match the old 2914 and not the new one.  So of course, it needed a new number because we couldn't call it 'old 2914'.  It was assigned "741".  I have no idea why that is the number - don't ask me.  Now if you were to run around the warehouse for Access Commodities, you would sometimes find little sticky notes on cones that say 'Tricia's cone' or "Don't fulfill Tricia's orders with this'.  I have to just hug the lovely ladies who fill orders for all their care in identifying these little details on when we have a shift in color happening and which cones are 'approved COC' ones.  It does get a little exhausting for all of us to remember many of these things.

Ok - so I am set.  I have a color line and we are speeding along with it, sending soie ovale to others to get other threads made to match, etc.  Meanwhile, back at the ranch in Paris they have their own initiative underway regarding the silk colors.  The business has been shifting from custom runs for fashion companies and increasing for hand embroiderers.  Hand embroiderers love shading.  People making buttonholes on high end fashion don't need that.  So there was a move to reorient the color cards into shaded families so embroiderers, stores, designers, etc could more easily choose colors and in that process some 'fixing' would go on.  That means that the families might have a color jump in it. Two colors that were too close together so they were a bit indistinguishable or too big of a difference in between.  Those places were identified over many years before the new color card was to be launched and they were fixed by shifting the next dying to the right place in the gradation.  
The 'old' 2916 seen on the old AVS color card

OHHHHHHH - I hear many of you understanding why such and such color is not 'the same dye lot'.  You are starting to see why over a period of 20 years, no color is set in stone.  There are so many forces on them.  People like me, the way business goes, the length of time it takes to get the manufacturing run sold, how long it has been in that drawer at the store or on the shelf at the distributor, needs to shift colors to satisfy current businesses, etc.  

So 2211 was the next victim.  It had been a darker color and now became a lighter version of itself.  I happened upon this when I put the Goldwork course out again.  That was a 2007 course and yes, 2211 was now not what it once was.  So I ended up with about a hundred tubes of something too light.  And I needed to follow up replacing it with 2212.  So I had a batch dyed... 

And then came 2916.  The original 2916 was a lovely bright pink that came after 945 in the series.  But the new color card had it shifted to be almost 945 in its own family that needed a darker version to make the family
The new 2916 in its color family on the new AVS color cards
The new color card organizes of the families instead of in
numerical order.  So much easier to figure out what
threads you want to get for shading.
correct.  See the old color card and the new color card here.  
 So now they were almost indistinguishable and there was a big leap between 2916 and the 741 color.  (You may ask, why didn't I use the 2915 that was already in the 2910 family?  Well look hard at it in the picture of all the threads on the pink silk above.  You will see a skein of 2915 laying near the top on the right.  It's undertones are blue.  Not a natural madder/cochineal look.  These undertones are something that we really pay attention to.)

So when Lamora and I traveled to Paris last October, this was one of the things on our list to get fixed.  I needed a 2916 that was like the old one.

This is when I got to see 'the drawers'.  I had brought my new stitch samples and a tube of what I wanted it to be.  Mark pulled out the historic 2916 color family drawer.  That is a drawer of the last 200 years of samples from dye batches that happened.  Wow.

It was exciting to see and to figure out which one was exactly what I wanted.

Some of the 2916 family of threads the master samples
Working on the new 2916 color - see my stitch sample at the bottom with the old and new 2916 stitched into it like a cross
So we finally decided upon the change and the spool was set aside and the decision was made that the 2916 in the new color card would stay and the new 'Tricia's 2916' would be dyed in all the threads again and given a new number :  235

So now you know.  There is quite a lot of hand wringing behind every color.  On that piece of linen in the picture above is another issue - 703 and 710 are too close together.  There is a brown issue we are working on.  2125 and 2126 have need of attention.  It goes on.  All about making and keeping a big color palette that can be used to do 17th century embroidery.  

Now knowing all of this background, you might realize that before this Cabinet of Curiosities effort, there was no way to switch between threads like Soie Paris to Soie Gobelin in the same colors.  You couldn't all of a sudden decide to pick out a figure and use Soie Perlee instead of Soie Ovale.  It wasn't built into the Au Ver a Soie line because that isn't how the line has developed.  Lamora and I sometimes sit back and marvel about the achievement of having a full color line that has portability across soie trame, soie paris, soie perlee, soie gobelin, soie ovale, 100/3, silk gimp, three silk purl sizes, silk scallop, crinkle silk, silk soutache, silk lacet, Soie de Tresse 1/6, and so on (Have I forgotten one?).  

That is an achievement in itself outside of the caskets.  It makes creative embroidery more fun.  


  1. Wow. I had NO clue about how the colors come and go over the years based on fashion orders, that color cards can be outdated, or why the color cards have changed from numerical to color family order. I did know dye lots can change, and that some colors of cotton flosses (no experience with silks) changed over the years due to dye changes for environmental reasons. I'm glad Au Ver a Soie is paying attention to the embroidery world.

    So, in short, if stitching an old pattern, don't be surprised if the called-for colors aren't right, or not available?

  2. That is 100% right. Old charts used colors that were 'there and that color' in that year. Never to be 100% the same.