• The Disruptive Designer

Hackett Canvas Sample 2

Jan 2019

This sample is a development and continuation of the "Hacking a Hackett Jackett" sample. As each jacket only has two canvas's (left and right hand side) it allows for 2 samples to be created. In theory the two canvas's should be the same size and shape and therefore only one CAD design is required and flipped on the horizontal axis. (Assuming the two sides require the same design). As I am currently in the experimental stages of the project, it makes no sense embroidering two identical canvas's, therefore one will be a development of the other. It is an opportuntity to refine processes and attempt alternative techniques. From experience I am aware of the many different types of canvas construction methods, and therefore it is very unlikely that I will find identical canvas's in other makes of jackets. As a result it is impossible to create definite templates for embroidering a canvas, due to the differing sizes and shapes. Until the jacket is dissected and the canvas removed it is impossible for the artisan to know what the working area is. If designing a virgin jacket (New from scratch) the canvas can be designed to be part of a larger composition in its own right and not just a hidden structural component.

In order to embroider an upcycled canvas I need to devise a "process" rather than a "template" which can be applied to any canvas, irrespective of shape, size and design.

The first Hackett sample (Above and past post) was purely to ascertain if it was possible to embroider on a removed canvas and associated difficulties. Previous digital embroidery pieces over the years has given me a solid understanding on the machine limitations and available stitch types available to me. Due to this prior knowledge I was aware of the outcome if I left the machine to fully machine the design. Often the outcome is rigid and very stiff, looking as if it is forced and contains no "soul" or humanity. The first sample used the largest loop size (B) and was not concerned with placement of the design on the canvas. It involved one hooping and machine manipulation by using the skip stitch feature. Rhinestones were added afterwards in an attempt to create depth and create a more "luxurious" aesthetic.

The second sample involved manipulation of the canvas prior to embroider application in the form of recolouring (Marbling). A starch water mix was created following the instructions from a marbling craft box set and two acrylic paints.

Unfortunately, I misread the mix ratio and added too much cornflour to the water mix, creating a rather thick paste. The acrylic paint was diluted with water and added using a paintbrush to the paste in a tray. The canvas was laid on top, lightly pressed into the paint and removed. The canvas was left for a couple of days to dry. Outcome: Due to the cornflour and density, the canvas dried very thick and crispy. The paint however retained its vibrant colour, opacity and slight paint texture. The outcome is very unpredictable with an element of "chance" in the final aesthetic. For me the outcome is very "hippy dippy" and requires more control and direction.

The canvas was pressed flat ready to embroider.

For this sample I wanted to try and embroider the felted area of the canvas only, leaving the marbled canvas exposed. This required embroidering larger than the hoops capabilities and within a fixed shape and area.

A photograph was taken of the canvas and felt and uploaded to Adobe Photoshop where it was reduced in physical size, resolution and cropped to remove unwanted areas of the jacket.

Saved as a JPEG image, the canvas image was uploaded into Adobe Illustrator and locked as a Layer. On a new layer, the pen tool was used to trace the felted area only, creating a working area. A William Morris was chosen and applied to the work area as a clipping mask.

Image trace was used to turn the image into a vector. Once expanded the image was further manipulated in order to determine position, scale and removal of areas. Once happy the design was saved as a jpeg image (above) and imported to Janome Digitzer Pro, ready to turn into stitches.

Complication: Due to the machine software programming, it is not possible to export a design larger than the hoop size. (140x200mm). To over come this issue using the software I have access to, I need to turn the Jpeg image into stitches, ensuring the image fits into the size of hoop B.

Once saved as a .JEF file I am able to open up in another software programme (Embird) where I can rescale the design to its required size in order to fit in the felt area. Measuring the phyiscal canvas using a tape measure I am able to rescale the digital version referring to the digital tape measure or RESIZE features.

Once scaled to actual size, I can use the "Auto Split" or "Split" feature ensuring the maximium hoop area is (140x200). The software splits the design into several hoopings based on:

  • Physical size and stitch density

This design required 4 seperate hoopings in order to create the full design.

ISSUES Encountered

The biggest issue is hooping the canvas to ensure the machine embroiders in the correct place. A physical acetate grid was cross referenced with the digital equivalent, with the operator making a visual judgement on accuracy between the two. The same issue applies for the remaining hoopings when trying to pattern match. Any discrepancies could result is misalignment, gaps or overlapping.

As you can see in the above image, the 4 hoopings have been matched with no visible areas of join, however there is a significant misalignment taking place around the armhole area. The positioning of the first hoop was correct in matching to the bottom curved edge of the felt, however incorrect in degree rotation resulting in a triangular wedge of felt visible. The next sample needs a more robust system for ensuring alignment when rehooping. As the machine embroidered, I manually stopped the process in order to remove the trailing threads, slowing the process down considerably. After the 2nd hooping I was aware of the mis alignment, however if I correctly positioned the 3rd hooping the design itself would be affected. In this instance i could consider freehand machine embroidery to fill the exposed triangle felt.

The skip stitch feature was once again applied in order to reveal the structural stitch framework and thus generating a more looser "humanised" aesthetic in areas. As I am not aware of the visual result of the unstructured stitches until the machine actual embroiders them, it is not possible to "design" the image in advance. The artisan (operator) has to reactive instinctively and quickly during production in order to stop the machine and skip forward. If the operator hesitates or misses the opportunity the machine will continue to embroider the more decorative top stitches over the framework and the chance for intervention is missed (unpicking the stitches afterwards could damage the cloth and be costly in time).

Design decisions are carried out by the artisan:

  • Canvas and work area preparation prior to embroidery

  • Cad image design using hand drawings and Adobe software

  • Cad stitch and image design using Embroidery software

  • During machine embroidery production, adding fabric, leaving travel threads, skip stitches, thread selection etc..

  • Post embroidery production i.e. rhinestone, beading, hand embroidery etc.

There are many similarities between digital and traditional hand embroidery with the main advance of time saving experienced by opting to use digital tools. The skills sets are very different but both requiring high aesthetic judgement.