Running length canvas shrinkage INTERMITTENT (not everything)

[Josh Wise]

Hello, everyone! I work in a print gallery where we create a lot of large and small girl canvas prints and gallery wraps. I have a read a lot over various forums about canvas shrinkage, and how it only runs length. We had issues with this, and have tried several remedies, such as adjusting image size, length compensation (in DTG Mirage), and paper feed adjustment. I have seen a ton of posts, but none seem to cover our PARTICULAR problem... it now only runs short on length on CERTAIN prints, and there is no discernible difference between the files! I will recap all our settings... I am using an Intel iMac that is a few years old. I use photoshop CS6 and DTG Mirage RIP to an Epson 9900 printer with IJ Premium canvas and OEM HDR inks. We set up a lot of custom settings throughout the years, such as always running our paper feed adjustment for canvas at +.70 and using our own profiles. We used to have problems on our large canvases falling short, but the problem we have now is that certain canvasses that are medium to large are perfectly sized, and some SMALL canvasses come up 1/4 in short. Traditionally, the bigger the canvas, the bigger the pull from tensioning, and the shrinkage should increase in relation to the overall size. This problem defies that trend! We use the same profiles and canvas and TIF files with same color mode, dot gain, etc. I even looked at plate separations, thinking maybe the shrinkers are using more of one color that shrinks more.... however, the images look equally balances in terms of rough color amounts per channel. I have had to set up a second profile the same as our first canvas profile, but with compensation turned to 1.01; this compensation makes our large print TOO BIG, and they are a 1/4 in over. Soon... I don't know how to predict WHICH images I have to use compensation on, and it is a waste of supplies. Anyone else had this SELECTIVE and INTERMITTENT canvas shrinkage happen to them?

[Landscapes]

So first, my experience is with a Canon iPF6400, so there is certainly a huge difference here.  I know I can also run the adjustment, but it seems like this will not be dependent on paper type, and i don't want to mess with this for my satin paper.  It seems like this same adjustment is used to dial in either banding or feed amount, so if you see banding, I guess it feeds ever so slightly less to eliminate the banding, and so if I adjust the feed for the length to make it longer, it might actually create banding.  So far, I have just stuck with stretching the image.

Anyway, I don't have a solution unfortunately, but I did want to say that based on what I read and my own experiences, I really don't think its shrinkage.  I know that a canvas which is made from at least 50% cotton will shrink, which is a great trick to tighten it up once stretched if you spray the back with water, but I just don't think that the amount of ink that is used can affect it.  I mean given a typical 24x36 canvas, there is only about 6 or 7ml of ink used, which is spread out over the whole canvas!  Imagine how much water you need to put on the back to shrink it!  Plus, the solvent in the ink, meaning the water for aqueous inks, pretty much evaporates right away.  So I really don't think its shrinkage from ink as the canvas is barely ever "wet".

So in my opinion, its only that feed adjustment that has an affect.  I'm not exactly sure how the printer measures this, but I think its the textrure of the canvas that fools it ever so slightly.  I'm not sure if textured paper has this same problem, but at the same time, textured paper is much stiffer, whereas canvas can bunch up or be microscopically compressed if that makes any sense.  Also, I have seen a difference when using different canvases (Epson vs. Simply Elegant), so this makes me further suspect that it all has to do with the texture and how the printer is "reading" the advancement from the surface.

I also know what you mean that length is a factor.  If I was printing out a 72 inch canvas, I would need to stretch the image to about 72.5" to get it to print to about 72 on the dot.  But if I only needed to print a canvas that was 36 inches long, stretching the image to 36.5 was now too much.  So certainly, its more of a percentage thing, but it sounds like you have tried to work your numbers already for each different size.

The canvas has a nice stretch to it, so you can always get at least 1/4" extra out of it if you stretch it tighter, and since I use black tape around the side, any amount of image that wraps around the side is ok to lose under the tape, as long as there is enough to wrap around.  So I'm with you in that consistency is difficult, but I always err on the side of having a little more and it works for me.  Sorry I don't have a better solution.

[Josh Wise]

Thanks, Landscapes! The part that is vexing me is that a percentage helped other people fix theirs', and also the bigger the canvas, the more it was off. In my case, the bigger the canvas, it prints right on the money! A smaller canvas is off by 1/4 of an inch. It is like we solved the problem for big canvas, but not for small... and this is using the same everything... same file type, profile, ICC Profile, RIP, printer, inks, dot gain. Everything is the same, but the size is only off on small files and not big files, which is the reverse of the problem everyone else is having.

It is a crazy situation for shrinkage because I have heard a ton of reasons, etc. Some say it can't be shrink because the un-inked canvas would pucker, some say that it is the weave pattern, some say tensioner bar, etc. But they all have a problem that increases linearly as does canvas size, but mine does not. Certain Tifs are wrong size, but some with exact same settings are not, and larger files don't seem to have the problem.

Once again, Thank you Landscape for the reply!

[Josh Wise]

Bump