Wednesday, January 23, 2008

Scanning Tutorial for Clean Line art

When I do a finished pencil drawing, I typically keep the drawing as clean as possible while drawing it. This saves in clean up time during the scanning process.

I do that by starting my under-drawing with a very light (hard) pencil lead. 4H Mechanical Lead Holder. After I have established the structure, perspective and loose details of the drawing I go over the drawing tightening every thing up with a darker (softer) lead. H Lead. In the past I've used 2H, but ended up drawing harder to get a nice dark line and was pressing grooves into the page. This just makes it hard to ink. Anything darker than 2H will be likely to start smearing as you rub your hand over the drawing, resulting in a general gray "haze" over the whole drawing.

TO AVOID SMEARING AS I DRAW, I place a piece of typing paper under my hand to cover portions of the underdrawing while working on it. Sometimes for pages I will even mask off, or tape pages together, leaving a window of just the panel Im working on open to draw. Then as I move to the next panel I re-tape my "mask" pages to leave the next panel open.This light underdrawing is the equivelant to using a blue line pencil or lightboxing a sketch from another page.

After working with Randy Green he had shown me that going directly to the board with a light pencil, it saves alot of time and helps keep the energy from your sketch directly on the page.So after I have used a darker lead and finished the pencil drawing, I scan in the drawing into photoshop.I do not adjust the levels in the actual scanning program. I just set the mode to gray, and typically scan everything in at 400 DPI. I scan everything so high, because even sketches might end up in a printed sketchbook later, and you can always downsize something. You cannot improve the quality of the scan though with out rescanning. So every file I have is usually a high-res and a "interenet-ready" size that is more manageable.

So I have the scanned image open in Photoshop. I enlarge the image to 12.5% or 25% to get a clear view of the line work. I then duplicate the background layer, so that I have two layers of the line art. I change the mode of the top layer to MULTIPLY. This will automatically make your drawing darker. And not just darker but you've doubled the range of grays in your drawing.

Here is the theory behind it, when you change the contrast or levels of your drawing, you are taking away the number of grays and making the line art darker and the white of the page lighter. When you start from your original scan you only have a certain number of gray levels throughout the file. When you mulitply that file, you are doubling that number of gray levels in your drawing. Basically creating twice as much information to start from. So after the top layer is set to MULITPLY, you flatten the image, and your drawing should look alot darker and the gray "haze" will increase. Now when you adjust the gray levels you are starting from a darker image and you loose less quality when you make the page lighter to eliminate the haze.

Generally I open the Level control under Image/adjustments/levels. or Cntrl L. I move the black arrow to the right and set to around 50. and move the white arrow to the left to around 225.

These numbers are very approximate, while I slide the arrows of contrast I am looking at the image to see the change that is happening. When moving the black arrow I look for the point when the line art is getting to a nice "readable" dark and when you go too far, making the smaller details too muddy. I move the white arrow looking at the white of the page and adjusting to the point where the gray "haze" is dissappearing and when I go too far and start to lose rendering details. So the numbers I mentioned above are subject to the these conditions of each drawing.

After I set the levels I have a more clear pencil drawing, sometimes nearly black line art.


-R said...

Multiplying the sketch layers does indeed double the number of greys in the image, as long as the original values were no darker than 50% grey. If they are darker than that, you'll end up crushing some of your greys into black, but I'm sure it's not objectionable.

Just wanted to mention, though, that if you have a Photoshop CS2 you can change the image mode to 16-bit (or 32-bit, I think), multiply the layers, merge them, and then convert back down to RGB 8-bit to preserve some grey values.

Though now that I'm writing this out I realize it probably wouldn't make a noticeable difference anyway. So disregard! Just a compositing habit, I guess.

Devour Vitality said...

:D Cool, I will have to try this. Although my photoshoping expertise is that of a slug. (slugs don't have hands so its very difficult for them to use the computer)

Aveline said...

Interesting to know.