January 2015 Video Tutorial -- Tracing Is Not Cheating!

[Egg Bramhill]

Finally:

In reality, I must say, though, that Google analytics are the only measure that my bosses can go by to judge the success of a piece, and statistically, a lot of the audience moves on after 4 minutes, so the call has been to keep the tutorials down to 6 minutes, tops.

Serious users know that it takes more than 4 minutes to do a meaningful tutorial. What do they think, your teaching goldfish to draw Time lapse video's are fine for understanding that the author has great ability but don't teach anyone diddly-squat. I watched a time-lapse video of someone drawing a fantastic car in 3 minutes using only Paint, but I learnt nowt!

Also, I often revisit video tutorials (not just Xara Xones) and fast forward to the section I'm unclear on.

[angelize]

One of the things I took from this was the nudging trick. I'm wondering why I've never thought of that before!

I have some ideas for further distressing the watering can but I'm thinking that it might be taking the thread off topic so I'll play around with it tonight and I'll start a thread on distressing/ antiquing in the graphics chat area.

I also have an idea of something else I'm going to trace. Now I need to get some other stuff done before I think up any more ideas.

[Crow Haven]

Hey, I appreciate all the kind words about my effort this month, thank you. In reality, I must say, though, that Google analytics are the only measure that my bosses can go by to judge the success of a piece, and statistically, a lot of the audience moves on after 4 minutes, so the call has been to keep the tutorials down to 6 minutes, tops. This was a story that could not be told in installments and certainly not in four minutes. I think the speed-painting videos on youTube are fun, but they're not tutorials. So what I said at the beginning of the thread was a bonafide apology in advance in case anyone was expecting a 3 second gem instead of excavating a while in a mine.

I like the long tutorials! Most of the tutes I study are around 20 minutes or longer. Those longer ones explain not just the how-to but the why to do it this or that way, and I find understanding the why helps me remember the techniques much better. It's like learning to swim -- someone can tell you to just flail your arms and legs about and you'll likely sink like a rock. Or you can get the technique explained and practice...and learn to swim.

[Gare]

Hi Egg, and just briefly...

Here's an annotated illustration. Hypothetically, you're seeing the scene from a viewpoint in front, and my illustration is of what one might see from the side. The grey cans are what might be shown in the mirror plane as a reflection. What I was trying to suggest is that the left one is clearly wrong. At right is the correct projection of an object tilted toward the viewer and its mirror reflection, explaining why text is sometimes the giveaway of a straight copy and flip of the original. It's just a truth of optics. Sometimes you can cheat the angle while illustrating, most of the time: not.

-g

[Gare]

For the sake of our members, I've eliminated a lot of repeating Egg's text in the quoteback here, especially when we are in agreement with a technique but use different colors, he says stallion and I say horse, you know?

But we both tried to make this instructional reading, and it's a good aftermath to the video this month, so I honestly suggest you read our back and forths, because it helps explain a lot of stuff I had to gloss over due to time constrains on the Xara TV video:


Eric: I import the picture, scale it to about 500 pixels and then lock it. 500 px is the very minimum I'd use. The bigger the better. I often work in Xara at 20% zoom. I do the same but tend to have 2 copies of the original side by side, one on a locked layer the other on a lower unlocked layer, which I can move around if required. I trace over the locked layer whilst using the unlocked image as a reference for fill colours etc.


Gary: I chose 500 pixels as an arbitrary nudge value. Certainly you can go smaller or larger; you just should match the bitmap’s height or width in Options>Nudge distance, so (and the real point here has nothing to do with numbers), you can move a finished element out of the way (which will maske your view scroll sometimes to track the nudged shape, a minor inconvenience, don’t get disoriented, click last zoom value). Egg, I’m not certain that universally on tg, members understand the relationship between bitmap dimensions and bitmap resolution. I could, for example, give you a 300 ppi image that’s only 2” in width, and there (I think) would be sufficient resolution to make out the different areas and how light falloff makes a transition from area to area. But we are expressing the same thought here: the more visual information contained in the image’s pixels, the better you can trace.

Gary: I set the nudge value to about the same width, so I can work on a piece, nudge it over, work on another piece and nudge it over, so my view of the original photo can always be uncluttered and my tracings always align to each other, and to the photo.

Eric: Never do it this way. If I need to I just use Shift + Arrow Key and count the number of shifts. Once happy reverse the same number of steps.

Gary: You’re “power-nudging", but you’re still nudging a precise value to both move a traced shape and to then put it back into place relative to the original bitmap. We’re in synch here, Egg.


Gary: I use a white outline a lot so I can see the edges of areas to be traced, with no fill specified. I also move stuff over a little to see the color values and sample them, but them move the piece back so it aligns to the original.


Eric: I tend to use purple line colours and no fill. Rarely do images have predominantly purple fills but they often have black or whites and the line disappears. I get my fill colour values from the second image as mentioned above.


Gary:I use a white outline a lot, but not exclusively. Again, I absolutely agree with your method, Egg. Always use a contrasting outline colour that contrasts with most of the bitmap. You’re never goingh to get a color that contrasts with the entire image to be traced, so just go with unusual purple, white, hot orange, brilliant neon green. A dark photo demands a light outline, and although I trace using between 1 pixel and .5 pixels in outline width, sometimes I’ll goose it up to four pixels to be better able to see what I’m tracing from a zoomed out vantage point. And actually during the tutorial, the horizontal rails on the watering can lived for a while as traces, but centerline, not outline traces, I then converted to shapes I then filled.

Gary: I use Boolean Operations—Anrrange>Combine Shapes, and then the subtract and add and all those guys. Now you could just have the Arrange menu displayed as a toolbar, but what I do is Alt+drag the buttons I need off the button palette in the control Bars section and then try never to reinstall Xara! Seriously, you save time, like with that transparency fiece you first did if you just draw a coarse loop around the target area, CTRL+K a copy of what's directly underneath and overlapping, thebn select both shapes and choose Intersect (shapes). It's precise and takes very little time.

Eric: I slice & dice constantly. I find in more complicated renderings it's far easier to select the shape you require from sliced & diced shapes than a drawing with lots of shapes overlapping. I don't have the Combine Shapes on my toolbar, just use Ctrl + 1, 2 ,3 or 4. If I get it wrong Ctrl + Z and try again

Gary: I admire your memory. There are multiple programs I’ve written about over the past twenty years, and my brain holds exactly 7 keyboard shortcuts. I use and need Boolean (slice and dice) operations every day to make quick work of created new or complex shapes, so I Alt+drag rote commands such as delete, redo and undo off the Standard Bar to make room for the Booleans. If anyone reading this doesn’t know how to customize the toolbars and the workspace, just ask here…it’s key to tracing quickly. And by the way, if you look at my screen in the video, it’s not that I haven’t memorized Ctrl+Shift+O for Options, but that I use this so often to reset nudge values that I just put the Options button on the Toolbar. It takes two hands to perform Ctrl+Shift+O (comfortably!), so I’d prefer a one button solution—personal preference, many other programs I use offer bupkis in the way of customization. If it’s there, use it, make yourself at home in your favorite program.

Eric: I wouldn't say Xaras Combine shapes is accurate. It often produces gaps which become fairly obvious. I recall a thread re this shortcoming a couple of years back. What I tend to do as a workaround to remove the gaps is copy all, open a new file Ctrl + Shift + V, Ctrl + G, Ctrl + 1 to add the shapes. Then go into wireview and remove all the rogue nodes and often use the smoothing tool to remove excessive nodes. Then after this, with just one shape I Ctrl + Shift + V it back into the original file, send to back and give it a neutral colour to hide the gaps in the original. Sounds more complicated than it is!

Gary: I’ve not experienced gaps when a “slice and dice” operation is performed, but I’m usually unobservant, so it goes. One thing that creates a gap is having an outline width on any of the shapes to be processed. If, for example, each shape has a 1 point outline, when you do intersect shapes, the product will not have the precise intended dimensions. Always remove outline widths before Boolean-ing.

Let’s get back to tracing for a moment, though. After I’ve performed a Boolean operation or brought in an auto-trace that was done by a crap program and has superfluous control points, I’ll put it on the Guides layer, lock the Guides and manually trace the design. You ever turn a character into a shape, and the result is 50,000 unnecessary control points? Do this, lock the original on a layer, copy it to the Guides layer, and then with the Shape tool, simplify the misbegotten shape by lassoing it and using the slider on the Infobar. Now this will cause the node-reduced original shape to pull segments away from the underlying trace, but it’s easier to then correct them, than to have to retrace the whole character from scratch. I’ve donbe this to correct a lot of fonts that have interesting characters, that are poorly designed and coded.

Eric: Fills usually need multi-color stages because of the way light travels across curved shapes. A lot of times a linear gradient with about four control color points does the trick, but us Blend steps if the lighting gets too complicated. Also, don't be afraid to use transparency shapes in Stained Glass and/or Bleach mode, stacking them until you have a painterly look. I know Maya does that to get a pain quality to her illustrations.

Eric: I prefer to use various forms of fills rather than the blend option as it's more controllable. Using accurate fills requires the same/close number of nodes within the two shapes or it can quickly become a dogs dinner.


Gary: There’s a way to perform blends so the result doesn’t look like any pet’s dinner! Draw your start shape and way you like. Duplicate it (Ctrl+K) and then with the Shape tool, lasso all the control points on the duplicate and convert them to straight line segments. It is then a LOT easier to position the control points accurately relative to the originals, and when you’re done, you convert segments that should be curves back into curves. Here’s a fairly simple piece I traced from a model I’d created, and although, yeah, there are some multi-stage linear gradients going on, for the most part, I used Blend steps to get curving, uneven lighting.





+++++++++++++++++++++++++++++++++++

Does this resonate and help anyone get farther along in the Art of Tracing?

My Best,

[Gare]

his is for anyone who wants to try their hand at an equally challenging but different object to trace:

T

My Best,

Gary

[csehz]

Finding that Eric/Gary answer/answer section very valuable, it will be good to remember even later as so lot of experience cumulated there. Thanks for that summary in one place

[Crow Haven]

Welcome to TG, Bob!


Well, I had started to work on the jars in the caddy that was for Egg, then I saw the new jars Gare put out...and ended up making my own version of both, as there were things I liked about each, then I changed the caddy a bit and colors too...oh well. They turned out looking like jars at least.

[Egg Bramhill]

Hi Gary, always love a challenge especially if it involves creating vector traces. So here's my attempt at your sharp focus coloured jar. It took probably 20 hours but I'm pleased with the result. It could still do with a bit more tweaking and adding highlights that aren't on your original. I've also totally disregarded any jar shadow at present. Might add a Biscuits label also at a later date

Here are my results

Edit: forgot the xar file.

[Crow Haven]

Wonderful results, Egg!!!