Finished Animation

I think I have finally gotten my video to a point that I like it! It took some time; I ended up having to go back and remake a lot of the initial animations. Due to the way Photoshop exports them, there were black bars on the sides of some of them. My fix was to simply fill the empty space with more of the paper background. Simple yes, and also time consuming. But, I will definitely remember that for next time I am making animations.

Final animation

Talking with Margaret we came up with a couple ideas that appealed to me re: the exhibition. One was to bring all the posters I printed and have them as takeaways for the show. This is great, as my initial thought was to cut them all down into notebook covers and make notebooks from them to be given out at the show. However, leaving them as full sheets is a lot less work in a week already quite full of work. Plus, it will hopefully give people a sense of the scope of what I created better than a notebook might. Additionally, we agreed that it might be interesting to design some simple graphics to break down the labor involved in this project.

I set about attempting to total the labor up; I am guessing it’s probably a bit of an “underestimate”, but tentatively it looks like about 60 hours worth of work that went into a relatively simple looking 35 second video. The simplicity of the video actually reinforces the idea of the hidden labor of things; it certainly looks easy to make. Looking forward to hearing the reception to it.

The Labor of Love of Labor


It’s been a couple weeks! Some things were tried, some things were not successful, and some things were printed (of course). Last time in class, I spent some time with Alan looking for solutions to expedite the cropping process for the letters I am printing. We searched things like “image processing,” “auto slice,” and “auto crop.” There were a few duds, like the ones that required payment and relied on “AI” to automatically recognize and crop photos. There was ImageMagick, which seemed promising, especially when paired with Fred’s ImageMagick Scripts ( Fred had devised a couple of tools that seemed like they could do what I was looking for: scripts like ‘multicrop’ or ‘smartcrop’ might get me there.

That being said, I absolutely could not figure out how to utilize ImageMagick. I followed the instructions from their site, even installing HomeBrew using the terminal on my laptop (which was a new experience and slightly bewildering). But that is about the extent of what I was able to understand: I don’t really get how to use ImageMagick? Is it not just an application I can open? And then upon trying to utilize one of Fred’s scripts, I have no idea how to install them, or what I am supposed to do with all the text that comes when I download the script.

I had a slim hope that using the auto detect feature while scanning the prints would be the no brainer, easy solution to this, but after trying it, I was unable to get that to work, either.

I also explored GIMP, and attempted to utilize a tool called “Divide scanned images.” After spending 1-2 hours installing it, figuring out where the directory and files lived, and vaguely how it worked, I did some experimenting. While I was able to automatically crop the images, I could not figure out how to get the program to crop the image without squaring off the edges (see below). I tried a variety of setting changes, but quite frankly, after about 1 hour of either unsatisfactory results or being unable to locate where or if it was outputting the files to (do I just have 300 weirdly cropped images somewhere I can’t find? Or was it a bug?), I moved back to my first workflow, albeit modified.

It was modified in a couple ways. I harnessed some of Photoshop’s automated features in order to speed up the process. I can’t remember if I have mentioned that the paper size I am printing on does not fully fit on the scanner I have access to, so I have to scan each print twice and then piece it together in PS later. Well, I was able to speed this up significantly by using PS’s Photomerge feature, in which you can give it a number of files and it will attempt to automatically stitch them together into one panoramic image. Given the nature of these prints (black ink, white paper, lots of white space), PS did a remarkably good job with this, and saved me from having to do it myself. There were a few files that didn’t want to stitch together correctly, for some reason: they would merge horizontally instead of vertically for some reason. My workaround was to simply rotate the two images into a landscape orientation, and then photomerge them horizontally. This worked, though I am not sure why. Then, I simply rotated it 90 degrees after the photo was merged.

Not only did that save me time, but so did the ability to record and play actions in PS. Earlier attempts with this weren’t very useful, because of the closeness of the characters on the page as well as the variation between them. Adding a decent amount of white space all around the characters enabled me to record an action that would slice the page both vertically and horizontally. So now, instead of having to do all the slicing of each page by hand due to the closeness and shifting nature of the sheets, I was able to do it with essentially 3 clicks of the mouse. SUCH A TIME SAVER. Granted, these prints are still not perfect, and so the results are most certainly NOT 900 characters, all with the same amount of white space around them, cropped to the exact same dimensions. But it is 900 characters, cropped with SOME white space around them, done about 2 minutes after opening the photo in PS. This is a drastic improvement in efficiency from where I first started, which is pretty neat. 

The results of the actions. Made possible with 3 clicks of the mouse!
The results of Photomerge

This project has definitely been a push and pull between efficiency, artfulness, and technology. It has been fun to try to balance all these things in a way that still feels good to me and true to my practice and beliefs as a maker. It has enough inefficiency to feel “real,” enough “dumbness” to feel in line with my practice, but also enough technology to remind me that computers are merely a tool for me to use. In this case, I can use them to give me more time to print and to share the seeming insanity of printing 1000s of the same letter with a larger audience.


Turns out, about 2,000 letters takes a long time to crop out. Talking with Vic, he suggested that there is definitely a (pretty easy) way to have the computer do all the cropping work for me, assuming I can get things positioned consistently on the page. So, I added more space between the letters for this weekend’s print runs (4.5 pica between each letter and between each line). This project is interesting; it’s a push and pull between efficiency and visual interest and a love of labor. For instance, while these images might make it easier to edit and make into a further (digital) project, I am less entranced by them on the page than on previous attempts: they maintain their recognizability a lot more with the added white space. Which, of course, makes sense. Speaking of efficiency, another place to shave time would be to use a smaller sheet of paper, one that fits on the scanner bed so that I don’t have to scan both halves of the larger sheet. But, the size I have chosen (13 x 20) maximizes the amount of sheets I get out of the parent sheets I have (26 x 40). And of course, has the added benefit of being able to fit more letters per sheet.

So, I did two letters this week, to see how long it would take and if I could run out of ink, so I only inked at the start of printing. I learned that 1. it didn’t take that long – about 10 minutes to print 1 line of letters on 16 sheets of paper, and then to move it down one line on press. And 2., these prints don’t eat up ink as much as I expected and am used to; there is a slight noticeable decrease in coverage, but not as drastic as I would have guessed.

How I feel most Sundays

Tradeoffs are always present in any project. It will be interesting to see if we can figure out a way to speed up the “post-printing” process, because it will ultimately just mean I can print even more characters and have to worry less about how long it will take for me to crop them all. But even in this aspect I think it’s important to note that part of this whole project is the act of repetition; the time it takes to do things. I wonder if I will miss the hours of cropping the same letter? Will it affect my appreciation of the medium, or instead enhance it by allowing me to do more printing (read: less computer time), while still producing visually interesting work that can exist digitally. While I certainly use digital tools to facilitate my print practice, I haven’t worked with it as integrally as I am proposing here. And of course, this is hardly integral compared to what is possible, but as someone who spends most of their time surrounded by old, heavy, antique objects, it feels pretty significant!

Here is a video of the E, 713 individual instances of the character viewed at 30 frames per second.

Hidden Labor

Printing day today! I was waffling between choosing only characters from one font/character width, as it would create a cohesive look and feel. But a fellow printer pointed out that having variable character width is probably more visually interesting for this context, in which mostly non-printers will be viewing it. The variable widths will further enhance the idea of differences among wood type. In my mind, if I was printing a poster I would want the consistency, and so I was working from that mindset initially. But…I’m not printing a ‘designed’ poster per se, and so that can be less of a consideration. That isn’t to say that I am not printing a poster at all though. If a poster’s goal is to communicate something, then I hope they are communicating the message of hidden labor and machinery.

At any rate, I chose a narrow font for the next letter, H – it’s a 10-51, which means it is the same height (10) as the E, but narrower, since 51 is less than the 61 of our E. Also, added bonus there was a LOT of them, perhaps 25 or so. As for paper size, I bought some 26” x 40” parent sheet of Cougar 80#C in Bright White. Cutting this in quarters yields a 13” x 20” sheet, and this is a fairly common sheet size used in printing. I also did a couple different sized sheets, a 20” x 26”, and another one that I failed to write down. For this run of H, however, I stuck with the 13” x 20”; while the press can indeed handle a larger sheet, I was trying to be most economical with the paper I had, and the other sizes left scraps that weren’t at all useful for this project. At any rate, I had a full line of H this time, and the narrower width meant that I was able to both put kerning in between the letters and print a full line each time. This means that I would be able to print much faster. After using a Uline catalog to raise some of the characters to print better and positioning it correctly, I added a little more ink and then printed a line on 6 sheets of good, clean paper. Then, moved the line down, checked it on the makeready sheets, and repeated until the paper was full. I learned some things during this, which I had heard of happening but hadn’t really experienced yet. One: as the type moves further ‘down’ the press bed, it prints a little less well. I likely made this more apparent by not adding any more ink after I started the print run. Two, as you get towards the bottom of the page, the paper is harder to control and keep tight along the cylinder. The results are prints that are crooked, or not in their intended position, as well as ink dragging along the bottom edge of the paper. Now, I alluded to this in my previous entry, but I had never experienced it this severely: it actually required a bit of a sort of dance move to both hold the paper tight enough to get a decent print while simultaneously holding the peddle using my hip bone. There are two solutions to this: use a longer sheet of paper, giving you something to hold onto for longer, and then trim off the excess. Of course, I was trying to fill the sheet, and wanted to see what it was like to try to print a full one using the same routine. The other solution would be to turn the type upside down, move it back towards the ‘top’ of the press bed, and print from the other end of the paper. I guess I wouldn’t really need to do that for the letter H, since in this character set the crossbar seems to be pretty much in the middle. This is not always the case (fun type design trivia)! I didn’t do this either – even though it would have ultimately made an easier, ‘better’ print, for this project I am thinking about the repetitive aspect of printing, and this seemed to break up that repetition in a way I am not sure I was looking for.

So, each line had 20 H in it, and I was able to print 11 roughly evenly spaced lines per page, meaning there are 220 Hs per page, assuming they are spaced out enough to allow me to use them. Multiply that by at least 6 “good” pages and maybe another 2-3 worth of misprints, and we are looking at roughly 1980 specimens of 10-51 wood type H. A whopping 1 minute of video at 30 frames per second. Of course, this time I was able to work faster, so the time input : video length output ratio should be better. Or, will it remain the same as I have 3 times as many characters to cut out?

After spending about 30 minutes troubleshooting this silly camera, I think I was able to get it running, taking time lapse photos every two seconds. It remains unclear how I get the photos off the camera onto my computer to look at them – I think I need to hunt around my apartment for a microSD card adapter and then an SD card reader. Hopefully I can find mine! I am curious to see if the weird 360 degree photo looks or reveals anything interesting, or if it just is some bizarre looking, hard to decipher photos.

Printing the last line.
The bottom half of a sheet. You can see the lower 3 rows are skewed: this happens when you’re towards the bottom of a page and don’t have enough paper to hold on to as it wraps around the cylinder.


Finally got around to animating the E prints in class. It turned out to be a video about 24 seconds long, consisting of 713 unique, printed letter Es passing by at 30 frames per second. The effect is, well, mesmerizing, almost having the feel of an old film reel. I am not sure what I expected, but it certainly took me off guard. Reflecting on how it feels to boil so much work and manual labor down into 24 seconds. Really, if I count all the press time as well as scanning and cropping the photos and doing the animation, it probably is about 12 hours worth of labor distilled down into 24 seconds. Is something lost when the process is not apparent, not witnessed in real time? I believe so. This also serves to make me think about all the unseen labor that happens every day, just so that people can have anything they want, whenever they want it, and for the cheapest cost possible. What is the price we pay for that convenience?

These are just some ideas that come up when I think about my project. I am working on trying to figure out this camera – a Garmin Virb 360. Though apparently it uses a wide angle lens to give a 360 degree view of things. I wonder what a time lapse with this would look like if I do one of myself printing this weekend? Interesting visually, or a headache waiting to happen? Time to find out, I think. Heck, even this is interesting to think about: the camera is controlled through an app on my phone, it seems. What happens if the app stops getting updated, or I don’t have cell service? Does this still work, or is not essentially obsolete at that point? A piece of technology relying on another piece of technology to function. If one vanishes the other is useless. That being said, it’s not as though this is a new thing for modern technology, I don’t think. It’s just so much more widespread and at a much faster rate than the past that it is drastically more noticeable. 

Garmin Virb 360

As far as the process of digitizing goes, I definitely think it might be worth it to investigate Premier, especially if I end up printing more than 700. Even with this many, my computer had a slight struggle to deal with it. It might be more sensible to make several shorter ones, and then piece them together seamlessly. Someone also said this would be relatively easy to do in After Effects? So, maybe I will try those other two workflows to see if it is any faster. Ideally, I will space the letters a slight bit more in the next run of prints, so that I am able to create a photoshop action to cut them out. And, that the prints remain consistent enough for said action to work for all the scans.

Slices view in Photoshop.
The longest of timelines.

Some Research, Some Scanning, Some Math

Where is time going, and why is it happening so quickly? Class passed quickly, but productively I think. I was able to catch Margaret and Vic up on some of the moves I had been making towards the project. The main update was to the output; it seems to be moving more towards me as a machine and how I interact with the press (another machine), and what effects those have on the outcomes (prints, characters of the alphabet, the wood type).

Ryan and Annet again had a great suggestion of adding the element of time to the project. Examples might include: animating each and every specimen printed, printing as many letters as possible until the ink runs out, or even recording myself as I print, a bit of a performance aspect to it.

Frankly, I like all of these ideas, though I don’t like the thought of recording myself. However…I remembered that I was given a GoPro esque type camera, though I have never used it. I am pretty sure it has the capability to take photos at set intervals (or to record video, of course). But, I wonder if a timelapse of me doing all the printing played next to the rapidly moving type specimens might be a nice juxtaposition. Hours worth of printing sped up to a few minutes, meanwhile the type goes by even faster. Doing the math with Ryan: at 30 frames per second, if every character I printed the other day was considered a “frame,” I would have….20 seconds of video…So, about 4 hours worth of work resulted in 700 characters printed/frames, which boils down to 20 seconds. Kind of interesting, staggering, dismaying(?), but I’m into it.

This textured paper makes for an interesting effect.

Regarding paper, I ended up not having to order more from French Paper, as I was able to find a similar paper stock available locally! Thinking about sheet size, I looked up the specs for the various presses available for use. It appears that the SP-20 has the largest max sheet size (19.5” x 28), and fortunately it is what I used on the first test runs of this process. I think Annet has a good point and that I should try to maximize the sheet size on press, and really push the machine to its limits while I try to push myself too. An important note about cylinder proof presses like these. While the max length of a sheet might be 28”, oftentimes the printing gets less and less good/accurate the further down the press or the longer the sheet you use. So while the first several lines would be nice and consistent, there often is a fall off in quality as you progress down a sheet. There are workarounds, the main one I know is to print half a page, then rotate both the sheet and the form and print the other half. In this way, the type is sort of “upside down” when you look at it, but it is often a good way to get a more evenly inked proof.

Info from the excellent resource

Another side note: I didn’t include the press at Globe Collection and Press while researching, because I don’t have regular access to it. It is the main workhorse for that studio, and not really available for student use, at least not for the extended periods that I would likely need it for.

Next steps: figuring out this weird camera I have, and see what it is capable of. Also, start automating a cropping script in Photoshop to cut out individual letters in order to be able to animate them. And, of course, more printing. I think next might be the letter ‘Z’, which is the least commonly used character in the alphabet and thus theoretically will have the least wear and thus print the best? Time will tell.


I suspect this might develop more after meetings again in class this week, but so far I know that I need to:

ORDER PAPER (at least more white paper)
Talk with Annet and Ryan about findings, thoughts, next steps
Scan and begin to digitize? Bring scanner to class.

Have more things printed between class periods
Continue digital processes in class

Have more things printed between class periods
Continue digital processes in class
Build random character selection tool? (if this is a route I end up taking)

Have more things printed between class periods
Continue digital processes in class
Figure out how/where to implement random character selection?

But still try to print a few more things 🙂

Try to have a ‘product’ or something to show. As for what this will end up being, I hope to gain clarity upon further meetings with the professors.

Have presentation ready to go.
Practice presentation, have index cards with short notes.

Thoughts, Experiments

We presented our research questions in class. I think it went relatively well, though presenting with a time limit and with a mask on was a LOT different than the artist talk I gave over Zoom earlier in the day. Definitely a noticeable contrast between the two. At any rate, we also had time to talk with others, think about potential collaborations and the like, which was useful. Upon talking through my idea with Alfie a bit more, she suggested that even I was to make something computer generated or machine based, chances are I am still going to be more interested or end up doing it the “dumb” way. Dumb here is definitely not being used pejoratively, just to clarify. Honestly, she is totally correct: I really would rather do it by hand, even with all the extra work that might entail.

When I expressed this to Ryan and Annet, as well as my concerns about feeling, quite frankly, overwhelmed by the thought/task of learning enough programming, etc., to do what I was seeking to do, they were (amazingly) understanding. Annet even proposed or talked about how using my own body as the machine instead of a computer, which is something I hadn’t considered. Could I push this idea, and maybe my own limits, and see what was generated? Pretty intriguing!

I did a test run of an idea over the past weekend, and definitely learned a few things. First, I will describe the set up: fill a page up with the same letter, printed repeatedly. I chose the letter E because it is the most commonly used letter in typesetting, and thus felt like a natural first choice.

California Job Case layout. Note that the lowercase ‘e’ has the largest size container.

Granted, wood type is a bit different than lead type, but the point remains that E is a very commonly used letter. So, I examined what sizes I had available to me in the letterpress studio, looking for something that was roughly a square size, thinking this would help to maximize the area on the sheet, as well as making my math easier as I moved things around. I settled on a “10-60″, which in Globe Press jargon means it is 10 line, or a bit over 1.5” tall. The 60 is actually a catalog number from when they ordered the type, but it serves as a guide to relative width. For instance, a “10-51” would be narrower than “10-60,” and “10-71” would be even wider than “10-60” is.

As is the case with these things, there was a limited number of specimens available. While I originally intended on printed just 1 letter E, I decided to use as many as I could find, in this case 4. This proved to be smarter, as the process took a lot longer than I could have anticipated.

Initial lockup. Hopefully other printers won’t judge it too harshly!

Upon locking up the type, I did a few proofs to get it positioned in the way I wanted, again, to maximize the number of characters I could print on a sheet. My sheet size was 12.5″ wide, and so I was able to get 8 Es per line. The process from here was mostly the same:
1. Print 4 Es
2. Move the Es over and print another 4
3. Move the 4 Es down, print
4. Move the 4 Es over, print
5. Repeat, repeat, repeat.

An initial proof.
First full lines, with earlier proofs above! Ignore the mess.
Here I am on Step 3 of the process, where the letters have been moved down one line.
Some proofs mixed in with good prints mixed in with Globe prints.

I had about 4 hours to work on this over the weekend, as that is what my schedule allowed for. I thought it would be no problem to fill a whole sheet (12.5″ x 19″) with Es. I was definitely wrong! In the end I only got 4 lines done. But, when you do the math, that is 32 per sheet, and if I did about 12-15 good sheets, that is still 384 Es, which is nothing to scoff at.

Now, I probably could have been a bit more productive had I taken more care when selecting the type. It wasn’t until I was moving it around that I noticed some problems with it that I would end up having to troubleshoot every time I moved the line around. Mainly, this letters weren’t square! Thus, when I tried to lock them up on press, they naturally wanted to spring upward and form a small arch, which can lead to disaster while printing, damaging the type, the press, or both. I eventually figured out a solution that worked, but still took time to reset it each time I moved the line.

Side view of the type. Note how some of the sides seem to taper inward, especially on the light colored block. This meant that when applied sideways pressure, it wanted to spring upwards, acting almost as a ‘key block’ in an arch.

It wouldn’t be printing without some troubleshooting and having ink all over your hands (or at least in my opinion). While it wasn’t as productive as I thought, it was still a valuable learning experience, and I can take what I learned and apply it moving forward, both in this class and future printing endeavors. As for next steps…maybe it is scanning these and beginning to figure out a workflow to cut them out? Or is it fully fill the sheet and see what that looks like. Do we even recognize it as an E anymore? What if the character was narrower, or wider? What if it was a word instead of a single letter?

Starting to look more like a pattern than a letter!

Presenting and Speed Dating

A bit of a different class this week. Margaret gave the class a presentation on, well, presentations (and how to be a successful presenter), in preparation for class next week. Next week, we will be giving short, 6 minute presentations talking about both ourselves and our research topics. Next we did speed dating – though there was some confusion about the order in which we should be changing seats.

At any rate, we were all asked to write down 5 skills that we have, 5 skills that we need, and what we think our research question might be. Then we had 3 minutes with each other, which managed to fly by at an alarming rate. It was great to hear more about others’ interests and skills, and often we found ourselves running out of time. It always fascinates me to learn more about people’s interests, so 3 minutes definitely did not feel long enough! 

My list.

Later meeting with Ryan and Margaret and Annet was helpful, and they encouraged me to potentially seek out a collaborator for my project, as it requires a few tech skills that I decidedly do not have! I think this week of presentations will help to give a more in depth insight into people’s topics, and hopefully a collaborative partner might emerge.

More Research Thoughts

Attending office hours with Ryan today seems to have helped crystallize some of the thoughts/interests I had been having regarding the rest of the semester. With a printmaking background, I definitely seem to trend towards the “open edition” method of printing – that is, I keep producing them until I am tired. It takes me longer to get tired because they all have small differences that keep my eye interested. Due to this, I tend to accumulate a lot of prints, and textures. And of course, the characters from the alphabet are no exception.

Plenty of specimens to work with.

I was thinking about my work for a record cover that was seemingly simple: the word “SHUTOFF” repeated and filling up the entire cover. There are filters and effects out there at simulate wood grain/printed textures, even entire letterpress “fonts” that mimic the look. The problem with a lot of the digital solutions is the lack of variety; that same texture often appears every time that letter appears. And once you see it, you can’t unsee it. So for this record cover, I printed 20-30 specimens of the word, which were then digitized. Then, letter by letter, the final artwork was constructed digitally. In this way, each appearance of the letter was unique because each original print of the letter was unique. Then, to imbue a little more texture to it, the albums were then screen printed on uncoated paper. Sure, there is an easier way, but is something lost along the path?

The final artwork.

So I guess the question becomes, well, what if the computer could simulate/generate unique textures? Could I input a given number of versions of the letter S, and then the computer could output even more, uniquely remixing/texturizing them each time? It would be pretty neat (though possibly resource intensive) if every time you typed the letter S, one was generated at random, based on our previous inputs. Plus, I enjoy seeing how a computer might mimic or map something handmade/printed. Will it be so good I can’t tell the difference? The implications are interesting.

There is a little bit of precedence in this lane: However, they seem to be relying truly on scans here, so while alternate forms do appear, it is only possible if it was a form that they originally input, as opposed to one being generated, as I am proposing. Secondly, this font seems to rest a little bit on the variation, leading to what has sometimes been referred to as the “ransom note” style. While charming, it lacks utility, and quite frankly just isn’t my vibe.

Plymouth Press font.
Great texture, but would like something more cohesive.

Create your website with
Get started