44th Annual Meeting, Textile Session, May 15, "The Creation of a Digitally Printed Reproduction Sleeve for and Eighteenth-Century Painted Silk Dress", by Miriam Murphy and Alexandra Barlow

When I heard the title of this talk, I must admit images of laser scanners and super computers came to mind.  And while the technology is doubtlessly impressive, I was relieved to learn that it is far more accessible (and much less sci-fi-esque!) than I had originally imagined.
The dress at the center of this talk was a silk taffeta Robe à la Polonaise, circa 1780, that had been donated to the Metropolitan Museum of Art in 1976, sans one proper left sleeve.

Because the silk was hand painted, the best option for recreating it in the 1970s was to hand paint a new sleeve as well.  However, after four decades, and three exhibits, the reproduction had become damaged.  Before being featured in the 2015 show China: Through the Looking Glass, it was decided a new sleeve would have to be made.  For the 2004 show, Dangerous Liaisons: Fashion and Furniture in the Eighteenth Century, an entire reproduction petticoat had been digitally printed, so it was not a technology with which the Met was unfamiliar.  Digital printing has also, in recent years, been used to create fabric for mounts, upholstery for chairs, and a dress for a historical interpreter at Colonial Williamsburg.
The first step in the process was to get a good digital image of the fabric, ideally through scanning.  As luck would have it (or so they thought) a flat pattern piece of the same fabric was at the Cooper-Hewitt, Smithsonian Design Museum.  Unfortunately, it was soon discovered that there were subtle but definite differences between the two.  Because of the 3-D nature of the dress, scanning was not an option, so digital photography had to be used instead.  The petticoat was selected for this process, rather than the present sleeve, because it offered a large, flat expanse, and the majority of the pattern repeat.  The image was taken at the Met’s photography lab and sent to Dyenamix textile printing, for image manipulation. This proved more difficult than expected because the undulations of the petticoat, visible in the photograph, had to be removed, and the full pattern repeat stitched together.  Another stumbling block was what part of the repeat to print?  Should they go with a mirror image of the extant sleeve, or could it be assumed that such expensive fabric was used conservatively, without an attempt at symmetry? Ultimately, the former path was selected.
The next step was to find a suitable base fabric.  Although the fabric texture can be printed, the best results are achieved when a close match is found, which was a challenge since 18^th century taffeta was finer than that available today.  When a good approximation was sourced (from Manhattan’s Garment District), both white and ivory was purchased.  Dyenamix first sent the fabric to Jacquard in California for pretreatment, to prepare it to accept the ink.  When the fabric returned, it was ready for the printing process. The printer prints ¼” at a time, going over each section eight times.
It took many attempts and many weeks to get the colors exactly right. Most of the early attempts were too bright and saturated, qualities prized by industry clients, but not by conservators. Once the fabric was printed, a pattern was taken from measurements and multiple mock-ups were made before the final sleeve was constructed. All-in-all, the process took six weeks and $2,400 (not including the time of the Met staff) to complete. Interestingly, roughly half of that seems to have been because Dyenamix did the image manipulation, judging by a similar project done in the Met’s upholstery department.

Dyenamix had claimed that their inks were archival and because the dress needed to be ready for exhibition, no further tests were performed beforehand. However, after the exhibition closed, the fabric was given the Oddy test, which it failed. In response, the sleeve was removed from the dress and stored separately.
I would be interested to know what caused the sleeve to fail the Oddy test: was it the pretreatment done by Jacquard to ready the silk for the ink, or the ink itself? This process is a great tool for any conservator to have and (as always!) more research is desired!