How do I learn to ”thinking in vector”

[Jenny_li]

Hi,

I’m trying to learn vector design from scratch by learning Affinity Designer for iPad. I guess this is a good thing and a challenge: I don’t have any prior knowledge that I need to ”unlearn” (except years of photo editing in Photoshop), but then again – all the tutorials and resources I can find are very focused on explaining the Affinity Designer software features, whereas I hardly know how vectors work in the first place. (I do, but only to a certain, insufficient, degree.) 

Now, I started out trying a design I thought was really simple – a line based, circular process chart I’d sketched out on paper, and figured I should be able to vectorize and then use as it is and develop further.  But after experimenting I found I can’t merge layers, and realized that somehow I’ve managed to create pixel layers instead of vector layers – without even knowing. My conclusion: This is me, not knowing how to ”think in vectors”.

I am taking an online course on Affinity Designer for iPad, through Udemy, but it’s very tool-oriented, and I don’t think it’s going to cover the ”thinking”-part that I need. So my question is... Could anyone give me some tips on resources for basically understanding vector graphics. I think I may have gone about breaking down my paper-sketch to vector-shapes an ”inefficient” was - thinking in lines rather than shapes.

(Optimal for me would of course be if the Serif team would make a series of general tutorials for vector beginners – how to think USING the affinity designer interface.)

I hope this question makes sense - and that I can have some tips on how to learn from scratch using the Affinity Designer - I am still confident that I will be able to do everything I want. I just need to get over this threshold...!   

[SandraM]

Maybe this link will help you

https://medium.com/coding-artist/a-beginners-guide-to-vector-graphic-design-815cb1cb4d70

[Alfred]

That beginner’s guide is great for users of the desktop version of AD. Unfortunately, as in many graphics tutorials the author uses hexadecimal notation when specifying the colours to be used, and the iPad version of AD currently only allows decimal input. So instead of being able to enter #25A9FC in a single box, you have to not only break it down into the separate RGB values #25, #A9 and #FC, but also convert each component to its decimal equivalent. Fortunately, online resources can help with this: I simply used the search engine DuckDuckGo to get the information I needed.

https://duckduckgo.com/?q=25A9FC&t=ipad&ia=answer

Quote

Hex: #25A9FC

RGBA(37, 169, 252, 1)

HSL(203, 97%, 57%)

CMYB(85%, 33%, 0%, 1%)

 

[gdenby]

Hi, Jenny_li

Here's a good basic definition:

http://www.dignitasdigital.com/blog/what-is-vector-graphics-an-introduction/

Maybe think of this way. It is a little like making a collage from paper. The difference is that the shapes are flexible. And they can change color as you like. Typically, the shapes will be laid one on top of another, tho' sometimes one might fit the edges together. So, most of the time, the order that the shapes are stacked on on another is very important.

You can make the shapes w. "connect-the-dots" tools like the pen or pencil, or use the built in shapes, all of which are extremely variable, tho' they will always be definite geometric shapes.

[Jenny_li]

Hi all,

Thanks for your responses – all interesting, although I really like the collage metaphor gdenby offers. I think I figured out last night that my ”thinking problem” has been thinking in lines when I need to think in shapes. I haven’t had time to read the whole article you posted, but it looks interesting and I look forward to reading it. Thank you!

This is both inspiring and a bit... overwhelming. There’s so MUCH I need to learn and figure out to be able to do the stuff I want...! Where do I even start...?

Anyways. An other question: Given that vector graphics requires shapes which can be rendered in different ways - how would you experts go about vectorizing the enclosed model...? (Crappy shot, but you get the picture...) I ended up actually doing it with the donut-shape, but it acts kind of weirdly, and don’t even get me started on the layer-weirdness... 

But is there a ”smart way” to make a model like this? How would you do it? *curious*

//Jenny

[gdenby]

Here is a sample.

I haven't used the iPad version much. Often can't remember where various control panels are. I have fairly bad arthritis, and so the touch interface is hard for me to use. So the sample took quite some time, probably 4 - 5 times longer than doing it on the desktop. 

What you will see is 6 parametric shapes, and 1 hand drawn triangular "curve" layered in a specific order. On the side, a rough figure sketch made w. the pencil tool, w. an ellipse subtracted for the face. That was duplicated, and made darker. Then a rough shape was drawn over the darker copy, and subtracted. At this point, I had spent as much time as I could. So I transferred it over to the desktop, tweaked some of the node positions, copied the figure shapes, and grouped those. I moved it to the approximate position, and scales it down to fit.

I don't do much work w. text, but suppose fitting the words into the portions might be quite fussy. 

thinkvector 2.afdesign

[Jenny_li]

Hi, 

Sorry about the late response – just had one of THOSE weeks... 

But gdenby – wow...  You just did exactly what I was trying to, only way better than I thought possible (after my own fumbling attempts...). And using the cog? I didn’t even think of that... I’ve fiddled a little bit with your image to figure out how you did that, and it’s just a beautiful solution. Thank you! 

I guess I have a long way to go learning vector drawing (and thinking) in general and in Affinity designer specifically – it certainly takes some getting used to...! :-)

One thing – IF I should want to fill the fields of this model, how do I go about doing that? Can I merge the layers to get ”closed” shapes to fill with colours? Or would I then have to start over and ”think again”...   

Also: your model has the smooth, perfect lines I thought were THE thing of vector design. But in some of the documents I’ve been playing around me, I get lines that look like a saw-blade. What am I doing wrong...? I’m guessing it has to do with the document setup, but I still can’t figure it out. See enclosed image for exampla of ugly lines – they look like too-low resolution pixel graphics, but this is supposed to be vectors... (Never mind the image itself, I’m just playing around trying to trace/vectorize whatever paper sketches I have lying around...!  )

Again, thank you for the image you sent, gdenby. It’s most appreciated! 

//Jenny

[gdenby]

My guess is that you may have the "view mode" set to pixel, instead of vector.  See attached.

Designer's method of handling vectors is that the area enclosed by a curve can have both a fill and a stroke assigned. This applies to both open and closed curves. The difference is that the break between the ends of an unclosed curve will have no stroke. If you switch the view mode to outline, you will see the wireframe w/o those attributes.

The curves/ layers cannot be merged like layers in Photo. The curves/layers are subject to various algebraic operations, usually called "booleans" after the mathematician who defined the logic. These operations are grouped under geometry. But in terms of filling, the operations automatically close open curves, while converting the shapes, but giving them the attributes of the lower level. In the example file I posted, the cloak figure was a hand drawn closed shape. Then I pulled out an ellipse where the face in the cowl was to be, and subtracted that from the cloak form.

Trying to put open curves together can be quite complex, as the auto complete behavior will most likely change shapes. Typically, the curve strokes are "expanded," forming thin closed shapes, which can more easily be added, subtracted, etc. The down side is that the expansion routine can generate large, or even huge numbers of nodes, which can make continued work hard to manage.

 

Note the more evident anti-aliasing "jaggy" edges are somewhat more evident. The difference is not as great as it would appear within the application. This is when I made the screen grab it rasterizes the vector display which doesn't come out wuite clean in the browser window.