Spent a couple of days attempting to replicate a single motif from one of the pillow covers (T81-1924). This is the object I have a high resolution scan of – a massive consideration as I’ve been running into difficulties working with my own low resolution photos of other objects (I’ve been copying a coif (T12-1948) tracing the stitches over a photograph using my iPad and I found myself having to refer to close-ups and macros I had taken as the quality of the larger photograph made it difficult to see the details properly).

The purpose of doing this single ‘replica’ is to look at ways I might approach and present the stitch decay. I chose an area of the original that has extensive decay but just enough of the original pattern to be able to reconstruct it. I also chose a relatively simple pattern to make things a bit easier for myself.

I began by digitally tracing the stitch pattern onto the photo – using a blue lines to draw the existing stitches and then pink lines (on a separate layer) to draw in the missing stitches, using the holes as a guide. I noted that I had to make a decision about the edges of the motif, as the design is outlined in a thick braided stitch which has been applied over the pattern fill. I traced three outlines – the inside and outside of the braid stitch and an estimated middle line between the two which I decided to use as the likeliest edge of the motif drawing. Drawing the patterns up to this imagined middle line (even though the stitches themselves were not visible), I took the decision that those extending from remaining stitches are probably still there – in fact you can just see the ends and edges of some – and so drawn in blue, while those that would have extended from the decayed areas are drawn in pink (even though they are probably still there under the braid).

• 

• 

• 

• 

What I have created here is a multi layered digital diagram – a working guide for myself. By drawing the different elements on separate layers I can view or hide, I’ve made a document that is hugely useful for reference and (I think) clearly shows the observations I have made. Of course, this doesn’t take into account the texture or pliability of the stitches – but that’s not what this is for – it’s a diagram rather than a drawing and there is a whole lot of stuff in the difference and overlaps of drawings and diagrams – this is a limitation of language. But, for my purposes, I refer to a drawing as a diagram when it is made for the purpose of illustrating and describing particular technical things (like stitch patterns), even if it’s not strictly accurate in terms of texture or anomalies or distortions. On the whole, I tend to draw diagrams when I’m wanting a guide for making an embroidery or studying an existing one – it’s a way of describing the way is thing is made.

What I want use this diagram for is experimenting with possible ways to show decay and what remains. So, instead of drawing the motif outline and filling it with the pattern arbitrarily, I decided to copy the stitch pattern fill by counting the stitches as they are drawing in the diagram. I began with a total fill, stitching both the existing and decayed stitches – it should be noted that this complete embroidery is in itself a guide for further experiments as it gives me a stitch scale to work from. Let me explain that better.

I do not know the thread count of the linen of the original embroidery, nor do I have a way of getting that information at present and, even if I did know, it’s unlikely I have a similar fabric (or a way of getting any). So, I have to work with what I have and, at the moment, the finest with fabric in my stash is a 60 TPI calico. This means the scale of my copy is different to the original (even though the stitches are worked at the same thread count of four per stitch) but, by stitching a complete pattern, I can then scale the motif outline accordingly and apply it to further reconstructions. 

At least that was the idea.

What has actually happened is the pattern fill I have stitched, matching the stitches of the original at exactly, doesn’t match the motif outline! I can only assume this is due to differences in materials, differences in tension applied by myself and the original embroiderer, or distortions that have occurred over time and through use – perhaps it’s more likely a combination of all three.

What this means is that it is going to be impossible to make an accurate replica that matches the motif outline. It’s not the aim of this research the stitch an exact replica – but it seems really relevant for this particular exercise – I’m trying to explore what is there and what is not. It seems disingenuous to make up or estimate the fill to fit the outline – but I can’t make the motif outline match the pattern fill I have stitched.

It’s a quandary.

I have tried to scaling up and rotating the copy (photo) but it doesn’t line up.

Instead, I have laid an acetate onto the embroidery I have made and drawn the outline, matching the points where the pattern fill goes beneath the braid stitch on the original. The resulting outline is quite different to the original. So I am left with a lot of things to consider.

Accurately copying the pattern fill does not match the motif outline, while stitching a pattern to fit an outline is not likely to result in the same stitch pattern, and by extension, the patterns of decay.

I think, for the purpose of this set of tests, I’m going to ignore the outline and just work with the counted thread patterns as they translate to the scale of my cloth – these are just test pieces.

But this is going to need more thinking about. Perhaps the impossibility of accurate stitching means that drawn and digital reconstructions are the only possible solution?