Usually, I write these posts after I have worked my way through all the exercises and samples in the learning materials for the chapter. This time, though, I am coming up with so many thoughts and questions as I work through that it seems best to document the development process for my resolved sample as I go.
The story so far
I’ve reached the point where I am now working in a slightly different order to the notes. First, I played around with scans if some of my stitched samples from Chapter 11. Here (image 1) are a couple of designs using repeated scanned images based on the techniques on page 46 of the notes. While they are decorative, I felt that they represented a bit of a dead end, in that they look a bit more like a tile pattern than the effect I am after. Perhaps it’s the white space between the shapes. Anyway, not very promising.
Image 2 shows another example – this time, no (intentional) white space, and showing the shape growing in width:
I’ve been using CorelDraw for this. I’ve had it on my computer for years but never really learned to use it properly. I think this shows. At this point, I went back to my favourite samples from Chapter 11, and made a few decisions about how to interpret the brief (in a conceptual sense) and what techniques to use. I thought it may be useful to think of growth and disintegration in concrete terms – the example I decided to use was the process of star formation – coalescing from a gas cloud, evolving into a stars, then disintegrating, perhaps as a supernova (now that would be exciting!) or as a dwarf star. So my cross has just become a star.
The techniques from Chapter 11 I think looked the most promising were the ripple version of reverse appliqué and the seeding. Here (image 3) is my CorelDraw version of the star coalescing from the gas cloud, growing, then shrinking and disintegrating back to nothingness (OK, not astronomically correct, but I’m not designing a galaxy, and the star conceit was only meant to get me started).
In the top left-hand corner there are a few dashed lines. I was trying to work out how to show seeding or running stitch, without a great deal of success, but left the marks there to remind me. This time, I managed to produce a matching background. I thought that if I placed the growing and disintegrating crosses at random angles, and the fully developed cross square on the sample, this might reinforce the idea of the growth-and-disintegration process (oblique = chaotic).
The next step was to translate this design to paper, as per the ideas on page 45 of the notes. This is where I am up to as I write this post, so the rest of this post might be a bit ‘stream of consciousness’ as I develop my ideas.
The paper composition
I’m going to continue with my reddy-orange background, turquoise crosses colour scheme as I think this worked quite well last chapter. Looking at my papers, I think the printed orange papers are not subtle enough and are likely to ‘fight’ with the crosses. I want a printed pattern which will tone, not contrast with the paper, so as not to confuse the conceptual ‘point’ of the design. I still like the shape of my stamp but feel that the size of the design is too large, so I’ll make a smaller replica and print a sheet using a mixture of red, orange and gold paint.
Image 4 shows the result. So far so good.
I want to represent both fabric and stitch on this paper composition. I should be able to show stitch in opaque coloured pen or perhaps with painted abaca paper. Here goes (image 5):
Developing the resolved sample
I’m reasonably happy with the paper composition, so am now thinking my way through the process of translating it to fabric and stitch. Some of the questions which are arising can be dealt with as thought experiments; some will require trial samples to solve. I suspect that knowing the difference will be a useful design skill to work on.
How do I make sure that the red background doesn’t dominate, and doesn’t look flat and uninteresting?
- Use printed fabric
- Arrange turquoise crosses to cover most of the background
- Use seeding in turquoise threads to tie the whole composition together
- I will need to consider carefully how to stack the fabrics:
- Not too thick or it will be too difficult to hand-stitch
- Maybe two or three red and orange layers at the top to ‘grow’ the shapes out of, then a couple of turquoise layers to give variety in the crosses, then a couple more reddy-orange layers underneath to cut down to when disintegrating the crosses.
Can I machine stitch the ripple lines? Would this look odd if I add hand-stitch (running stitch and seeding) as design elements to form the cross shapes and suggest the ‘gas cloud’? Would the resolved sample be sufficiently robust if I hand-stitch the ripple lines? What if I machine the lines in matching thread then hand-stitch over them in contrasting thread? Work a trial sample.
Shall I use cotton homespun or bemsilk lining material for the background?
- The bemsilk has a lovely sheen but would it fray too much to support the cutting of slivers between the stitched ripple lines? Would printing it make a difference? Work a trial sample.
- Texture contrast in the background – maybe use shiny threads to stitch on a matt background, or vice versa.
What would be the best treatment for the parts of the design represented on the paper composition by abaca paper? I could:
- Use stitch – seeding, perhaps
- Include abaca paper as a layer in the stack of fabric – but it’s too dense, and I’m not sure the texture would be very compatible with the woven fabrics
- Use sheer fabrics or net layers in the stack
Probably a combination of stitching and sheer fabric would work. I can include strategically-placed sheer layers immediately above and below the turquoise layers to give a subtle transition, perhaps.
I worked four trial samples – the first (image 6) with printed cotton homespun as the top layer, ripples hand-stitched in running stitch in turquoise Danish flower thread, and seeding in the background in coton à broder, cut down to a layer of teal bemsilk; the second (image 7) as per the first but with ripples machine stitched in turquoise rayon; the third (image 8)with printed bemsilk as the top layer, with ripples stitched in reddy-orange coton à broder; and the fourth (image 9) as per the third but with the ripples machine stitched in matching orange thread.
What did I learn?
- The acetate bemsilk fabric looks too ‘plasticky’ as the top layer. I thought the sheen would work well but it actually detracts from the appearance of the samples. As well, it is quite fragile and, I think, would not stand up too well to the cutting process, especially if it were to be handled a lot;
- The cotton homespun performs much better when stitched and cut, although it has its problems too, as it has quite an open weave. It looks better than the bemsilk when printed, stitched and cut, so I’ll use it for the top layer;
- I thought the hand-stitched ripples would work better than machine stitching with the seeding on the background (and I am intending to stitch seeding onto some of the cut shapes, too). Having made the samples, I feel that the machine stitching does not look at all out of place with the hand stitching. The running stitch ripples, on the other hand, appear quite heavy and a bit clumsy – not a look I am trying to achieve. I’ll use machine stitching but, because the fabric is quite loosely woven, I’ll use shorter stitches, plus stitch a double row, which should increase the stability of the fabric as well as making the lines of stitching a little less formal without appearing heavy;
- Having tried stitching the ripples in contrasting and toning colours, I’ve decided to go with the contrasting thread because I’m looking to the stitching to define the growing and disintegrating cross shapes.
The other thing I learned relates to technique. If I am to have this sample depict growth and disintegration in the way shown in the paper composition, I think the technique will need to be a hybrid of contemporary appliqué and the ripple or contour effect (both Chapter 9). The reason I say this is because I won’t be able to get, for example, the clear large turquoise shape at the lower right hand corner of the paper composition if I stitch all the contours through the top (red) layer. So – I’ll need to stitch the first outline, then cut down until I reach the top turquoise layer, then stitch contours and keep cutting. If I don’t do this, I also won’t have any room left on the cross shapes to add any hand stitching.
stitching the resolved sample
The first step was to print the top layer of fabric (the reddy-orange homespun). I did this in much the same way as for the printed paper in image 4, with a mixture of red, orange and gold shapes. Then, I composed my stack of fabrics. The top two layers were reddy-orange: homespun on top, then a layer of fine red, sparkly tulle because I was trying to replicate the effect of the abaca paper in the paper model. Then, three layers of turquoise: a sheer glass organza, then turquoise homespun, then teal bemsilk. Under that, a layer of orange glass organza, then a final layer of reddy-orange homespun for the base.
The next step was to stitch all the outlines. I used shiny rayon thread, with a short stitch length, and stitched a double outline. I decided to use a paler turquoise thread for the shapes that were growing and disintegrating, and a darker thread for the shapes which were more or less fully formed, as image 10 shows:
At about this point I realised that I needed to cut down to a turquoise layer on the three largest shapes before stitching any more contours (I realised this only after adding a second contour to the top right-hand shape – fortunately, I was able to fix this by carefully disintegrating the orange layer within the outer shape). I also decided to make my shape grow and disintegrate in an anti-clockwise direction, and to convey this by cutting down from the top to a turquoise layer as the shape grows, and cutting down through the turquoise layer to the lower orange layers as the shape disintegrates (clear as mud!)
Here is the design with further stitching, and the first few shapes cut away (image 11):
As the stitching and cutting progressed, I found I needed to think logically about which layers I should be showing, and about how to ensure each cross shape incorporated a variety of surfaces for interest, and to convey the ‘growth and disintegration’ concept (image 12 – sorry, a bit of camera shake).
I thought the stage in image 13 was about the place to stop cutting. The balance between reddy-orange and turquoise appeared to be more or less right in terms of growth and disintegration of the shape.
The next thoughts were about adding hand stitch:
- to the shapes, to add interest and to reinforce the growth and disintegration idea – I decided to add some strategic seeding to three of the shapes; and
- to the background. The gold print on the fabric had turned out to be much more prominent than on the paper (image 4), so I wanted to knock this back a bit. I also wanted to add some stitching in turquoise to integrate the cross shapes with the background and deal with the impression that they had been ‘plonked on’. So, I opted for a powdering of seeding in a light turquoise coton à broder.
I stitched the seeding with the work in a hoop for a while, then realised that friction from the hoop was causing the fabric to fray a bit where I had cut away slivers of the top layer. I had to watch the tension of my stitches once I removed the hoop, but found that the thickness of the seven-layer stack kept the sample reasonably stable.
The hand stitching completed the resolved sample. Image 14 shows a photograph of the sample with a black window mount, and image 15 shows a scan. The colours in the photographs (especially image 13, above) are far more realistic than in the scans.
It is just possible to make out the red tulle overlying the outer turquoise layer on the centre and lower left hand crosses, and the sheer, shiny turquoise layers at the outer extremities of the three largest stars. It really looks more interesting texturally ‘in the flesh’, given the way the light catches the organza layers. One more try, with flash (image 16):