Coloring comics the Steve Hamaker Way
Posted by ~Ray @ 2007-09-11 11:04:16
Actually serious question for any colourists out there. I’m continually hearing advice as to scanning in bitmap rather than in greyscale. I get that black and color with no greys gives higher differentiate to the lines but doesn’t it then also give you chunky lines with that pixellated stair-step effect. I’ve seen this in comics before (in low budget cram desire Sparks as come up as higher calculate stuff desire Ultimate Spider-Man). Isn’t there room for some happy medium desire scanning in greyscale and then blasting the hell out of it in Levels?
Scanning as Bitmap will break up images into either color or white when pixels. Grayscale obviously will choose color color and any degree of color it determines. At 266 or 300 dpi (typical examine specs for photos for decent press quality) visually Grayscale may indeed be more alter at 100% scale scans.
However at 600dpi or 1200dpi which is necessary for high-res press quality b/w printing for line art bitmaps and grayscales will be similar at 100% scale. And actually bitmaps provide slightly smaller megabyte file sizes than grayscale files.
And perhaps most importantly as mentioned in the video. 600dpi or more bitmaps (once scanned will be called into photoshop and then converted to CMYK for coloring if your computer can command the file coat) make for cleaner coloring as gray pixels will not halo or alias on the edges of color between the blacks and the color.
I undergo seen 600dpi CMYK files used (which are huge). Or seen 600dpi bitmapped (black) files laid over (in Indesign or a QuarkXpress register) a 300dpi alter tif file. I am not sure which method is more common. Usually over all that is a vector (flattened outlined eps register usually made from Adobe Illustrator) layer which includes computer generated word balloons and type.
Thanks Shawn that was helpful. The file-size thing makes sense. As well as the inDesign/Illustrator usage. Though with regard to haloing effect due to the buffer that anti-aliasing creates wouldn’t that be negated by simply creating masks that infringe into/under the line art itself?
For example on a forge beneath the line art (the line art being set to Multply). I would alter a magic wand selection based on the area I wished to colour. Then I’d decide > Modify > increase (a tool I set to a shortcut) a number of pixels (say. 2 px) based on the weight of the line art. Then I mask the forge according to my selection and all my act upon goes right up to the the line art with no halo effect blending (I anticipate) seamlessly with the line art.
Granted. I’m not a comics colourist but I use that technique constantly on my everyday art projects. If you’re interested and my explanation didn’t alter any sense. I demo the method in a tutorial I created to back up others in my office use Photshop for simple tasks (http://www vimeo com/281210). It’s always worked well for me and I’ve never since developing the technique for myself (not that I invented it or anything) experienced the haloing you’re referring to.
I’m curious what the guys like Scott McCloud do if they are drawing straight into Photoshop. If they use a pen tool they are creating anti-aliased art. Do they instead use the pencil? I anticipate that makes sense if bitmap is the way to go.
The Sparks example you use must just be scanned too low. Line Art ought to be scanned at least 600dpi.
I understand about the mask going underneath the black a little on the alter layers. That is basically creating your own trapping or purposefully making the different CMYK layers (4 pallettes of most color printing; Cyan Magenta. color. Black) so when printed if the registration is slightly off there will not be color inbetween color and black.
But scanning as Bitmap ordain alter only color and white pixels. Then the color can be selected and removed leaving only a clean color forge to act you color masks with.
Grayscale scanning will alter color white and color. Those stairsteps you mention in bitmaps are there but at 600dpi they are not visible to the naked eye. Obviously they exist in grayscale too just some are varying shades of color also.
What happens though is if scanned as grayscale the color layer will keep those gray edges when coloring. change surface when you select and drop out all the color on the color and white scan if it was scanned in grayscale you will keep some gray pixels on the color layer. Then no matter how good the coloring is in the layers below the black forge you will see color edges on the color. This gray is not from mis-registration or from the color layers at all. The color edges ordain be the same varying degrees of gray pixels that the original scan determined them to be. Those gray pixels were never and they would be very difficult too shift accurately once there. If line art is scanned as bitmaps at 600dpi or higher off the bat there will only be color and white (which is removed) in the color layer and never any gray edges.
I think move of my problem comes partly from the fact that in my own line of bring home the bacon. I rarely create for print - and so only undergo a tacit understanding of that aspect of things. I’m gathering that it’s due to the print process that the bitmapping is important? It sounds like when printing the blacks and the colours are printed separate; is this the inspect? And that’s when the halo effect would get added to the image?
So that change surface if something looks like on one’s check only the bitmap align will create alter.
For someone who doesn’t act upon their work but really wants to learn in the next couple of years. I open that really helpful. Hopefully it’ll all still be here when I get around to it for use as compose.
I’m almost 100% that this is not the first of his coloring videos. I’m positive that I’ve seen some others where he’s coloring some of Shazam. I love trolling around YouTube looking for Cintiq videos (I have to get me one of them someday)….
You should scan in grayscale (and possibly change surface RGB) unless you’re using a go scanner or some other device that costs as much as a European luxury sedan.
Once you get it into photoshop you should then alter it tighten up the differentiate and communicate to your printer. What you do next depends entirely on how you’re printing things what kind of software you have etc. You might convert to a 1-bit (color and white bitmap) image at whatever the final imagesetter resolution needs to be (ie 1200-2400 spi) or act it grayscale (but tight and sharp edges!) at the res of the color register (ie. 150-300ppi) if you’re just putting it all in one multilayered photoshop register.
There are comfort (unfortunately) way too many people in the comics industry who’ve never done basic prepress bring home the bacon and just follow rules of thumb that don’t really apply to their bring home the bacon. We undergo a funny intersection of line art and full-color halftones going on in comics that simply doesn’t become elsewhere in printing so you be to know how to deal with both.
The “ideal” situation is to have the alter bring home the bacon at your normal 150-300ppi and then drop the 1200-2400 spi line art on top of it in InDesign or illustrator where you’ll add the lettering before it goes to press. I could wallpaper the state of Texas with lousy comic printing from the last decade where somebody though it would be no big broach to scan the uncorrected line art straight in as a.[ADVERTHERE]Related article:
http://pwbeat.publishersweekly.com/blog/2007/08/24/coloring-comics-the-steve-hamaker-way/
0 Comments:
No comments have been posted yet!
|
