The English book printed project: production report 2

TL:DR; Almost perfect if colors weren't way too bright.

To give a bit more context, since the first article published two month ago, I continued to work very hard on the book project. What started as a: "it will take just two weeks of works!"

...It took me months and I'm still on it. For a reminder, it is my self-published comic book project, 200 pages high quality color, big size hardcover and produced from scratch with 100% FLOSS software and released under a permissive CC-By license. To give you an idea, the book weight 1.3Kg, PDF file-weight of the page inside 2.8Gb and the cover is a 6K pixels large CMYK files combining vector and raster elements. It's probably one of the most challenging book Scribus saw and opening the file takes 6min on a 8 core machine and rendering takes almost 30min.

I have today the first book in my hand but the colors are wrong. So wrong my plan to sell it on the e-Shop for Xmas is now ruined. This rendering will never get my approval for being pushed into production and I must say, I felt very bad about all of this work for this result because I followed all the specifications to the letter to specifically avoid that situation. Here is a video of my hand browsing the first printed proof so you can get an idea of the object:

a 5.4MB self-hosted video illustration in MP4 format

The color issue can't be really seen on the video but the comic pages are like 10% too bright; this even sometime merges bright colors togethers. Here is a comparison, look at the cloud or the city in background:

left: my book, right: Glénat version

As you can imagine, these colors does not justice to Pepper&Carrot as I painted it. The cause? It could be a lot of things, but the gamma/brightness rings a bell to a bug I found with a suspicious gamma issue. It was declared "resolved upstream" when the main dev of LittleCMS put the blame into the color profile of my printer. Really? accusing one of the only 23 official CMYK profiles of the ICC International Color Consortium produced by X-Rite while I have other books from my printer made with the same profile and Adobe InDesign that manage gorgeous colors with it? So, I come to doubt the profile is borked especially given how big my printer is ( www.onebookshelf.com). I also discovered later on the process of making the book way more bugs, workflow bottleneck, crashes and other things that I'll gently skip here because they belong to bug-trackers...

left: Breton version by Ar Gripi publishing, right: my book

Fortunately, this first printed proof also has revealed plenty of positive aspects. The fonts in intro are good, deep black and at the right size. The composition of the pages are really good with very good respect of bleeds and large comic page that fill the area. The thickness of the book and the glossy hardcover is ideal. The thin grained smooth paper is non-glossy and gives a traditional aspect to the artworks. Also, this is the most hi-resolution printed version of Pepper&Carrot I saw to the date. So, I have hopes, once I fix this color issue it will be perfect. I'm sure the color profile in itself has this tendency to be bright; but not probably as bright as that.

"Books are great!"... :-D Oh the irony.

Unfortunately, I can't ask my printer to solve the color issues for me, this is not like the little printer enterprise of a local city but a large service of POD with -I assume- machine operators that just put the file into the printers and have too big volumes to daily print (and pack books) to care about my non Adobe InDesign PDF files. So, I'll have to experiment and learn with more trial and errors: I'll keep sending on their server more files with various color strategy and wait for new printed proof until I can validate one. When I'll find the best workflow, I'll publish how I did. This project cost me already four time more than what I expected. Benefits of the books will never cover the months I did put into it, and the printed proof will make just increase the cost to a point where making this book with FLOSS already cost me more than buying a random Windows 10 laptop and pay a one month subscription for InDesign. But I'm doing it full FLOSS to reveal those issues, report them and aiming for quality always had a high price.

So that's all, expect new updates for the book project on 2020 when I'll receive the new printed proof that require around 20 days to be generated. Waiting for that, I'll focus with the Pepper&Carrot team of contributors on the next Pepper&Carrot episode!