Jump to content


  • Content count

  • Joined

  • Last visited

Posts posted by gdenby

  1. Hi, Gargamel,

    Please clarify. When you say  "border width of 4 pixels", do you mean the shape has a stroke width of 4? You may be getting confused by the apeearce of the stroke, which can be set to the outaside or inside of the shape, or centered on the shape boundary. The snapping is based on the shape, not the stroke. Try switch to outline view to see the base shape geometry.

  2. Hi, hebbardkennedy,

    W/o seeing what you want to work on, I think it is likely that there will be lots of problems.

    What is your source material? Is it a digital file that has been compressed? Those are likely to have lots of "artifacts" that have lots of grey noise at edges between B & W. "Jaggie" stair steps.

    Using a threshold adjustment will give you a B&W image. In my experience, most images w. fine detail change shape slightly depending on the threshold setting. Fine lines in engravings can go from scattered black pixel smears to solid shapes w. minor changes in the threshold.  Sometimes it helps to add a small amount of gaussian blur, about .5 pixel, to average out edge features before subjecting them to threshold.

    What you may want to do is clean up your pattern manually, and then change it to a vector drawing. Affinity does no do those, but the are good online free services for that, and some very inexpensive stand alone apps that do the job.

    The results will most likely need some smoothing by deleting extra nodes


  3. Hi, Keahi,

    Basically, the starting image is really "dirty." There are lots of compression artifacts which muddle the shape edges, but one can still see the patch's thread work, which adds more extraneous detail. 

    I tried a couple of different vectorizing settings, and did a little little image processing to clean it up for other tries, and got results much like PixelPest posted.

    One might start from there, and after lots of node tweaking, get a vector restoration that is a smoother and more accurate reproduction.

  4. Hi, Psych,

    I've used Painter, but its been several years since, and Krita a bit recently.

    AP is basically for photo development, adjustment, enhancement, compositing, etc. It has a good range of "brush" tools, as well as vector shapes. Unlike the above, it is not primarily a natural media emulator.

    For work being done w. both APhoyo & ADesigner, see here.

  5. 1 hour ago, JoergWe said:

    In any case, from the responses I gather there is no way for me to perfectly align my group of response bubbles. Thank you again for your replies.

    Not at all. Simple as pie. Set up your grid, w. snap to grid enabled. Draw a circle of desired size. Press "command" (Mac, don't know PC) and drag to next desired gird intersection. Press "command J," "Duplicate, and repeat the circle as needed. 

    If you then don't like the horizontal spread, move one end circle, and use the "distribute, horizontal spacing" widget. This will yield an equal spacing between al the circles.

    If I might suggest, stop thinking in AI, and start thinking in AD. Different methods for the same results.

  6. I've been sort of busy the past few days. Might have more time on the weekend. For now, some notes.

    The method starts w a straight line made up of evenly spaced nodes, a dozen or more. The line is duplicated in parallel.

    If all the resulting straight line curves are selected, and one switches to the node tool, one should see a grid of nodes. I prefer sharp nodes, because when the lines are bent, I find it easier to select columns of nodes connected by straight sections.

    Using the node tool, select a column of nodes from all the parallel lines. Then switch to the node tool selection box. This makes it easy to rotate, shift, skew and perform various transforms on the column. One may then, as desired, proceed thru other columns until all may have been twisted, expanded, contracted, etc. Then turn all the nodes into smooth.

    FWIW, one can apply a tapering pressure curve to one of the lines, copy the style to the rest. Then the series can be broken, making portions of the lines either wispy or thicker.

  7. Hi, abysan,

    Here are a few comments that may help.

    As far as I know, .eps files do not support transparency. They also formulate gradients different ways. Sometimes, AI rasterizes them. It may be that Affinity rasterizes them by default. Perhaps try exporting as .pdf.

    The problem with clipping probably has to the geometry of the layers. This is a common newbie mistake. All the vector objects define a 2D space. "Closed" curves can have a stroke placed on them. "Unclosed" will have a stroke on the defined lines. But if a fill is added to them, it fills from end point to endpoint of the defined curve.

    You may want to duplicate the curves that define the oval-ish shape. Take the original shapes, get rid of the stroke attribute, and the, using the pen and node tool, join the 2 shapes that are there. The gradient can be nested within that.

    Gradients are designed to be modified by the user to point in whatever direction the users wants, and have the end point placed wherever. 

    You might want to switch the view to "Outline." That way, you can see where the open curves are more easily.

  8. 15 hours ago, affinity4Christ15 said:

    I'm sorry, but the proposal to use the 3D FX as a suitable "workaround" is not only unsuitable, it is 100% unable to remotely achieve what is being asked here. It seems a "mod" would know better than that. That (cheap if you ask me) 3D tool can absolutely NOT achieve intricate realistic 3D lighting effects. It almost always looks amateur when used. Having the ability to curve gradients seems pretty much like a no-brainer feature that surely many others have also seen the obvious need for before me ???

    Reading the specs for .svg coded images, the gradients AD offers are those the standard supports, and which work w. vectors. Any gradient edge that offered a 3D look, whether from a shading routine like the fx, or a rendering from a 3D engine would need to be reproduced as pixels. The image then becomes larger (less suitable for web use) and will degraded w. scale changes. 

    Note, Photo does have the mesh warp which would sometimes aid in deforming a standard gradient.

  9. Hi, Affinity100,

    I think this will do for you.

    Make a rectangle, any size. Duplicate it and move it just as far to the side as it is wide. I often have a grid active, and w. snapping, can get even spaces.

    Select the group of rectangles. Open the transform panel. W. the anchor point positioned at the center of the bounding box diagram, enter identical values for H & W, what ever size you like. Everything will be in a square.

    W. the rectangles selected, use geometry/add. Then w. the node tool, select a row of points. Again using the transform panel, center the anchor, and enter what ever width you like to make a trapezoid.

    As practice using the pen tool, start snapping nodes the corners of the trapezoid line shape. If they are not quite right, switch to the node tool, zoom in, and tweak the positions.

    You can then nest the line shape inside the trapezoid, and the trapezoid's fill will block any layers lower in the hierarchy.


  10. "As few as possible. If you need to you can add nodes simply by tapping. Remember to use the edit feature to 'shape' your lines into curves. Try drawing circles using only 4 taps with the pen. Keep practicing with simple shapes and curves and before you know it you will be drawings cats ."

    Just 2 will do.


  11. If you don't want and embedded object, the saved .svg should be "opened" not "placed," But an .svg retains the vector structure, such as the layer hierarchy, and groups. AFAIK, vector images always retain the mathematical descriptions of the element sizes and positions, so that they items can be scaled and transformed. So each element has to remain and individual item within the structure.

  12. I don't have Publisher, but I'm supposing the color picker works the same as that for Designer and Photo. You can sample the wallpaper illustrations, and transfer the color reading from the monitor to the Affinity document. Also, the swatch panel allows creating a palette from and image.

    As far as choosing your colors from images found on the web, realize they are at best approximations of what you might see in real life. The "Seaweed" illustration is a .gif file, so the color range most likely has been reduced from the original. 

    Also, because printing depends on specific inks, different printers may or may not give a good representation of a CMYK image. Something that one might be expected to be a standard, simple black, can be quite hard to match. I worked w. a group trying to faithfully reproduce contemporary Chinese ink drawings, all monochrome, and it took at least half a dozen test runs to get the blacks right.


Important Information

These are the Terms of Use you will be asked to agree to if you join the forum. | Privacy Policy | Guidelines | We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.