Friday, December 31, 1999

designer-related woes

[Originally posted March 10, 2016, at about 8:00PM.]

My previous "frank" post showed off some of my cover-design work for a book that, alas, has been shot down by our CEO. I've been working, over the past month, on churning out the second of a projected nine-book textbook series that concentrates on grammar and vocabulary while also building students' writing skills.

Yesterday involved a paroxysm of proofreading effort—the final heave before the manuscript (I'll call that an "ms" from now on, as pro writers do; the plural is "mss") was to be hand-carried to Paju, where our designers are now based. I was at the office in Daechi-dong until almost 10:30PM last night, and I griped to my boss that the Korean designers were wasting opportunities with the photos I had gathered for them.

Let's back up a bit. The way it works is like this: I write most or all of the ms for every chapter in our ten-chapter textbook. Chapters are 15 pages long; we brought in a British guy who writes reading passages and reading-comprehension questions for each unit. He takes care of four pages per chapter; I do the other eleven—grammar sections plus a slew of exercises. I also go online at Shutterstock.com (our company has a Shutterstock account) to find photos and vector (basically, scalable) illustrations; I download these images—Shutterstock lets you download 25 per day—and populate each chapter with several (maybe eight to ten, usually one image per page). All of these chapters are done up as MS Word documents. All the Word documents together constitute the textbook's ms. The ms files are bundled up as a ZIP file, then uploaded to the unfortunately named "Webhard" (online hard drive; think of FTP space or Cloud storage, i.e. a physical place to store data that's not on my personal or office computer but is readily accessible from any Net-connected computer). The design team downloads my ms data and puts together their own draft ms, incorporating their own design elements. Fonts are changed; colors are added; text gets scrunched, pulled, and stretched like toffee to fit the constraints of the book's anticipated final physical dimensions; graphical elements like colorful borders, tiny cartoons, etc., are inserted to spice up each page, and the Shutterstock photos that I've chosen are integrated into the overall design as well.

What began to piss me off, however, was that, when I flipped through the designer's draft ms a couple days ago (they had sent their file as a PDF that we color-printed out in the office), I saw that the Shutterstock photos—on every page, without fail—had been reduced to the size of a postage stamp and tucked into spots where there was no text. It was the same dull, bland, uncreative procedure for every single photo: shrink-and-tuck, shrink-and-tuck. There had been no attempt to integrate the images with the text, no attempt to make the images pop for the reader—nothing. It was a phoned-in performance that bordered on insulting, almost as if the designer were saying, "Fuck this shit. This is beneath me. I'm not going to expend any thought on this." The pattern of behavior, increasingly visible as I flipped from page to page in the ms, was ruthlessly uncreative.

I had the impression that different designers were working on different aspects of the ms. There was someone who did the page borders, for example; someone else did the cute little cartoon characters that appeared every now and then. I was particularly enthralled by whoever had worked on the reading-passage title graphics: that person had very cleverly redesigned the letters in the title to create mini-illustrations that perfectly fit the topic and theme of each passage. Whoever had done that had been awake, in my opinion, and I told my boss via text that that person was the one who needed to be on Shutterstock duty, not whichever idiot was currently in charge of formatting and placing my photos.

Surprise, surprise: according to my boss, there was, in fact, a small design team, but the people doing the cartoons, the borders, the title graphics, and the photos were all one in the same lady. I was flabbergasted. How could she be so creative with her title designs and so disgustingly uncreative with the Shutterstock material?

I had relayed my reaction via text to my boss while he was out in Paju with the designer lady. Later in the afternoon, when the boss had returned, he told me that she had tried to stretch the Shutterstock photos out, but they had become too fuzzy and unusable. I smelled bullshit: the photos I had downloaded from Shutterstock were all 1000 pixels wide. Assuming 72 dpi (dots per inch)—which is screen resolution* for a given image—the images would be naturally about 13.9 inches wide. A standard A4-sized sheet of paper is about 8.27 inches wide, so that's a natural spillover of over 5 inches: the pics, at normal size, are too big for an A4 page. This means the pics need to be shrunk, not stretched, and shrinking a comparatively low-resolution image normally improves the quality of the image (the designer lady is right: stretching a non-vector image, like a JPEG, does lower the image's quality). So, yeah: bullshit. Besides, the images wouldn't all have to be stretched to fill an entire page, anyway: there are creative ways to incorporate slightly enlarged, or un-stretched, images into a textbook's page.

So with the designer offering a bullshit excuse for unimaginative work, all I can do is take some Shutterstock images myself and show my boss what an imaginative designer can do. I might work on that tomorrow (i.e., 3/11/16); we're now in a lull period because the ms has been turned in, the designer draft has been proofed and corrected (in Paju, my boss took my manually proofed copy of the designer's ms and went over all the corrections, page by page, for 180-some pages), and it's all in God's hands.

But will it be worth my while? Even if I convince my boss that it's possible to be an order of magnitude more creative with the Shutterstock images I downloaded, I think it's already too late for this book. Adding insult to injury, the designer told my boss that we'd need to download much larger images from Shutterstock (you can opt to download images at various sizes) if we want bigger, bolder page designs next time, almost as if her inability to work creatively were somehow our fault. Someone needs to be fucking slapped.



*Screen resolution is fine for things appearing on your monitor; many, if not most, monitors have a resolution of 72 dpi (more recent monitors have finer res). That dpi isn't so great, however, when it's time to print the image onto paper. For paper images, you normally want to shoot for about 300 dpi (much tinier pixels, which produce a finer-grained, cleaner-looking, generally sharper image). You could go up to 2400 dpi, which is about as fine a resolution as the human eye can discern, but that would be a waste of memory: 2400-dpi images make for huge—wastefully huge—data files.

While we're on the subject, I should note that "dpi" and "resolution," which I sloppily use interchangeably above, are not really the same thing. I could, for example, create a hi-res image with low-dpi pixels. Let's say my pixels are the size of a balloon, but my "painting surface" is the size of an entire football field. With balloon-sized pixels, I could create a very detailed, crisp-looking image. It would look amazing if I were floating several hundred feet in the air and looking down at my work. This football-field image would be high-resolution, given the fineness of the detail, but the dpi would be awfully low: in fact, with balloons, you'd have to talk about inches per pixel instead of pixels per inch.


_

3 comments:

Charles said...

I'm a little confused here. In your footnote you do acknowledge the difference between screen resolution and print resolution, but you base your calculations on the screen resolution of the images. At the minimum print resolution (300 dpi), a 1000 pixel wide image would be around 85 mm wide, or roughly 40% of the width of an A4 page. If they're going for 600 dpi (which is entirely possible), then those images would be around 42 mm wide, which is rather tiny (although not quite postage stamp size). Or am I misunderstanding you?

Whatever the case, the designer/printer should have given you specific instructions in advance as to what they needed in terms of images. Sounds pretty sloppy to me.

Kevin Kim said...

C,

I based my calculations on screen resolution because I'm pretty sure that that's the resolution of the typical Shutterstock photo, i.e., that's the resolution that I had to work with. I could be wrong; I'd need to check. In fact, if I'm wrong and the Shutterstock dpi turns out to be 300, then double-shame on the designer for acting helpless: you can stretch a 300-dpi image to 150% of its original size without truly ruining the quality of the image—an issue that matters even less if the image is there to serve as some sort of backdrop.

Just as an aside, I've taken 72-dpi images, reformatted them to 300 dpi, and resized them with little loss in image quality. Raising the dpi makes a low-quality image more scalable.

Charles said...

If I may play devil's (or designer's, I guess) advocate for a bit more...

When you say that you have "reformatted" 72-dpi images to 300 dpi, what you've actually done is resampled them. That is, you've changed the number of pixels in the image, in this case adding more pixels (more than four times more, in fact). These pixels are interpolated by the software, which means that you are reducing the quality of the image. This is not noticeable on the screen because a) the software is pretty good at interpolating the new pixels, and b) screen resolution is still only 72 dpi, so you can't really see the difference. But were you to print that out, the difference would be noticeable. The quality difference between the 72-dpi image and the resampled 300-dpi version would be roughly the same as the difference between an original image and that image increased in area by four times (so, double the height and width).

(I'm assuming here that when you say you have resized such resampled images with little loss in image quality, you're talking about what is displayed on the screen, not what ends up being printed.)

In other words, a low-quality image is always going to be a low-quality image no matter what you do to it. If your software is good enough, you can certainly make improvements, but it will never become a high-quality image. I think the problem here is that print designers generally think in much higher resolution and only use the screen as an approximation. So you may be thinking that it looks fine on the screen, and even in print it may look OK, but the designer may be unwilling to leave image quality up to software interpolation, and the images that she is getting are not up to her standards.

Which goes back to my original point, and what I think the true problem is: Designers have very clear standards and guidelines regarding the resolution of images, etc. The fact that these were not communicated to you shows a lack of professionalism (or at least thoroughness and/or experience) on the part of the design team. I know you're convinced that the images are good enough and are trying to defend that here, but I think you might be fighting the wrong battle. Rather than spending your time trying to show what can be done with those specific images, I think it would be more fruitful to just ask the design team specifically what they need and then give that to them. Like you said in the post, this woman has done great work with other graphical elements; I find it hard to believe that she would suddenly become lackadaisical in this one aspect.

Again, I'm not saying you were in the wrong here--it was definitely the design team's responsibility to convey their standards to you, and they failed to do that. That's the real problem here.

(Of course, it's completely possible that the design team will do a horrible job of conveying their standards even if you do ask. Designers (and anyone, really) can be great at what they do but horrible at communication and interpersonal relations.)