Spotlight: DriveThru Comics’ Print-On-Demand service

In the past month, Icarus has been participating in the beta for DriveThru Comic’s print-on-demand service, in partnership with Lightning Source.  On the 26th, we received our first proof book, the POD edition of Comic AG Digital issue 00 (ironic, isn’t it?).

This is a quick and dirty review based on the proof (you’ll have to forgive the poor lighting and aperture on the photos, as I had to do this rather quickly).  The focus is primarily on the production; how the book actually looks and feels.  In this area, POD has always been much maligned, and rightfully so.  Early POD, just like the e-readers of today, were concerned with reproducing text, not complex graphics.  Manga presents an especially difficult challenge to POD; modern printing takes an analog source (the original drawing), converts it to digital data (we scan instead of shoot film nowadays), then reproduces it on an analog medium (ink and paper).  As such, manga is a perfect test for POD, because the screentones commonly used in it are completely refined for analog technology.  It does not reproduce well unless the scans are high-resolution, and the printer can output at a high resolution.

But hardware is only part of the equation.  I think it is also true that most early POD operators simply did not have the proper experience with graphics.  Comics were not properly optimized for POD, and the printers were not pushed to their technical limits.  One common example of this is when high-res line art are down-sampled to 300dpi grayscale.  This is absolutely the wrong thing to do, and in the case of manga, can produce results that are unreadable.  Has POD improved, both in technology and expertise?

I.  Exterior finish quality

The covers for Comic AG 00 POD Edition are printed on a nice card stock, with a smooth varnish.  The square binding appears solid as well.  The trim is even.  The book, at 108 pages, is thinner than what you’d expect.  Here’s an image comparing the cover of Comic AG issue 00, and the same illustration as it appears on the back cover of Aqua Bless.  The POD is on the left:

The offset cover on the right appears more vibrant (despite what this photo shows, the offset cover is also slightly darker/richer, especially the blues).  But the color accuracy is acceptable.

Lightning Source accepts 300-dpi images for color (CMYK).  300 is pretty standard for grayscale and color files; it’s what’s most often used for offset printing.  However, a direct comparison between color offset and color POD resolution cannot be made, because color POD is based on inkjet technology.  Instead of four uniform color screens used in offset, the dots of cyan, magenta, yellow, and black in POD are arranged in a diffuse pattern.  POD sometimes wins out on color, because the lpi used for offset color isn’t always very high; there’s a limit on how much ink paper can hold.  In this case though our offset printing was done in Asia on heavy stock, so the quality edges out the POD.  But to most readers, the difference is negligible.  The cover looks and feels good.

II. Interior finish

The interior is printed on a white, matte stock.  The paper is thin, so there is minor bleed-through and the book overall feels insubstantial.

Lightning Source takes 600-dpi line art, with 106 lpi output for grayscale.  This is actually quite impressive, and what really prompted us to test out the system.  As I mentioned earlier, other POD printers, even some that specifically handle comic books, don’t always accept line art.  In printing, grayscale images are essentially converted into a large screentone.  What was sharp line art and lettering become fuzzy.  At lower lpi, that fuzziness may become pixellation.  And with manga, you get moire: the kind of unsightly patterns that result from layering one screentone on top of another.  So the fact that LS can handle line art at a decent (not great) resolution is exciting.

At minimum, manga is scanned at 600 dpi, *if the source art is very large*.  Once “shrunk” to the actual book size, the equivalent resolution is really between 1200 to 2400 dpi or higher.  So 600-dpi files, while sufficient for line art, is really not enough for manga.  The problem, once again, lies in the screentones… low dpi also creates moire.  But there are ways to optimize manga for low resolutions, and that is what we did with Comic AG Digital ( it involves converting line art to grayscale, softening the image, doing the necessary resizing and resampling, then converting it back to line art using the diffused dithering method).  There will still be some patterning, but they will appear much softer.

The various stories in Comic AG Digital 00 come from three different sources: 1. scans from original art boards, 2. scans of photocopies, and 3. art created entirely on computer.

This is a closeup of a page from Midara, which was scanned from original art.  The results are extremely impressive.  Because most of the screentones used in Midara were pretty large to begin with, moire is almost non-existent, except in some of the more tightly-packed background shading.

This is Wish of My Sister, which was toned digitally.  Somewhat counter-intuitively, moire is an even bigger problem for digitally toned comics, *if* the comic is resized.  That is because of the exactitude of digital screentones: if any error is introduced, that error is uniform across the entire screentone, creating a neat but undesirable plaid-like effect.  But through the diffuse dithering method, we were able to eliminate almost all instances of moire.  The reproduction here is so close to offset, you wouldn’t notice unless you were looking for the differences.  This is the most impressive section of the book.

Also notice the text.  It’s very sharp, but no fonts were embedded in our output PDF.  All fonts were rasterized.  600 dpi is enough to reproduce fonts and line art faithfully.

This is a close-up from Blue Eyes, which was scanned from a photocopy – a perfectly standard practice for manga.  Until this decade, photocopies were how most mangaka turned in their manuscripts.  Here we begin to see the limitations of working with 600-dpi.  Its simply not enough to cleanly reproduce the very tight screentones used in Blue Eyes.  The shading looks patchy, although some of it is due to the age of the manuscript.

This is the contents page, which was originally in grayscale, but we converted it to line art using 106 lpi, as instructed by Lightning Source.  It may not show up too well in the photo, but there were banding issues with this page.  The problem here is that 600 dpi is not really compatible with 106 lpi, at least when it comes to bitmaps (the dots would be spread from each other at intervals of 5.66 pixels; obviously, you can’t have a fraction of a pixel.  Moire is the result of that imperfection).  It may be best to use 100 lpi, or simply leave grayscale images as is.  Another issue apparent in this page is that the image is too dark.  This is due to ink spread, the way the paper stock absorbs the ink.

For comparison, our offset books use 175 lpi for grayscale images, so there’s really no contest here.  But 106 is still decent, higher than most newspapers, and in fact rather close to our printed Comic AG magazines, which have an lpi of 120.

This is an ad page which suffers from the same problems as the image above it.  We’re not concerned with the art in the background, but the areas of solid tones.  The ink spread is causing blotches.

On the whole, I was pleasantly surprised by the quality of the interior B/W pages.  Shocked, even.  This is one of those rare instances where I’ve seen a POD service reproduce manga relatively well. And for the most part, the results were predictable; the problems we see on some of these pages are the result of having to sample them to 600 dpi.  We knew what they would look like.  Lightning Source reproduced those files perfectly, warts and all.  Considering I could actually see jpeg compression artifacts around the lettering on some early POD comics, the technology has come a long way.

However, I have one major reservation, and that is the alignment of the pages.  To be blunt, the bleed was completely botched.

See the white gaps between the art and the edge of the page?  That shouldn’t happen.  The art files we provided to LS extended 1/8th of an inch past the book size on all four sides.  Not only where the pages carelessly aligned, but the bleed area does not appear to have been printed at all.  This is really unusual, and I can only chalk it up to comics with bleeds not being the normal kind of job LS handles.  Alignment on text-only books need not be very exacting, but for comics, this looks amateurish.  This defect appeared on almost all of the pages.  It may be possible that the inclusion of printer’s marks in our interior files threw off LS’s alignment, or created more work for them than they were used to, so I will reserve final judgment until I see some more proofs.

(Edit: That’s a knee, not a penis.  Settle down now.)

III. Closing thoughts

This is not a review of the complete POD service as provided by DriveThru, because the program is still in beta.  POD isn’t just a type of printing, but a distribution and retail system rolled into one.  In order for POD to be a viable option for individual creators or even publishers, it has to succeed on all of these fronts.

In comics, POD was supposed to be the great equalizer between self-publishers and traditional publishers, but that has not quite been fully realized.  The quality of the product was just one obstacle.  Few comic PODs out there offered true single-copy printing; most had sub-500 print runs that still required a not-insubstantial upfront investment, at a very high cost-per-copy.  The creator/publisher still had to deal with inventory.  In short, POD was no different from offset printing except in scale.  That’s why it wasn’t really competitive.

Should everything remain the same at launch, DriveThru would have done much to address these issues and close the viability gap between POD and traditional publishing.  The comics, barring the bleed issue, look nice.  DriveThru offers true print-on-demand… all the creator does is upload the print-ready files.  There is no print run to set, because there is no print run.  Lightning Source handles the production and the fulfillment.  Pricing per book, again with the caveat that they may change when the program is officially unveiled in January 2010, is fair; it’s good enough that some self-publishers may see the same or better rate of return on each copy sold as they would have on an offset product moved through traditional channels, but not good enough that these comics can be sold to retailers at a reasonable discount, unless some pricing concessions are made (that’s the holy grail, for the books to be cheap enough to include brick and mortar retailers).  As it is, DriveThru POD may be perfect for small press, self publishers, and doujinshi/group efforts.  Whether DriveThru will be able to attract enough customers to completely supplant offset publishing and retail distribution is debatable, but that’s not the only role it has to play; Comics212 lays out how POD may be complimentary to more traditional strategies.  And yet, for others who may not want to publish professionally, but merely want to dabble in the world of print, POD may be the only option… but it is one that has improved a great deal.

But some may wonder, is POD a little late?  Why not simply do webcomics?  Online distribution has been getting most of the fanfare lately, but books occupy the space between fine art valued as artifact, and the pure content of video media… the corporeal nature of books will always be valued.  So even as the world becomes more virtual, and the internet brings nearly-free distribution to everyone, the idea of print will still be a significant part of that world, even if traditional book distribution may see a decline.

One last thing… DriveThru POD will launch sometime in January 2010.  All issues of Comic AG Digital will be available through the program, including the one you see in this post, for $7.99.  Buy it if you like porn manga.  But, buy it if you are curious about the print quality yourself, as photos are really not an adequate gauge for moire or print clarity.  Since this was done merely to satisfy my own curiosity, we might not include future issues of Comic AG Digital in the POD program… unless we see a really, really strong demand for it.