jdaniel715 Posted January 5, 2021 Share Posted January 5, 2021 how can i change a piece of art from png to make it high resolution or into vector Quote Link to comment Share on other sites More sharing options...
v_kyr Posted January 6, 2021 Share Posted January 6, 2021 For vectorization either vectorize/retrace it manually, or using some third party bitmap to vector tracing tools. See for example ... Some forum threads about bitmap tracing/vectorization: Image Tracing in Affinity Designer? Image Trace to Vector Path Converting Pixel drawing to Vector? Auto Trace In AFFINITY DESIGNER Best Image Vectorizer for Mac with Affinity Designer ... and so on ... Vectorization and autotracing software for Win + Macs: Super Vectorizer 2 (commercial, Mac) TracedLines (commercial, Mac) Intaglio Vectorize (free to use, Mac) DragPotrace (Mac) + Potrace (free, Win + Mac) Potrace (free, Win + Mac) AutoTrace (free, Win+Mac) Inkscape (free, Win + Mac) MS Expression Design 4 (win, nowadays free) Image Vectorizer (commercial, mac) Vector Magic (commercial, win + mac) Online tracing tools: autotracer vectorizer vectorization Vectorize Raster Images Online (online Photopea, free) ...etc... Online centerline supporting tracing tools: Rapid resizer fConvert Online Vectorizer ...etc... Minus44, NotMyFault and emburke 2 1 Quote ☛ Affinity Designer 1.10.8 ◆ Affinity Photo 1.10.8 ◆ Affinity Publisher 1.10.8 ◆ OSX El Capitan ☛ Affinity V2.3 apps ◆ MacOS Sonoma 14.2 ◆ iPad OS 17.2 Link to comment Share on other sites More sharing options...
John Rostron Posted January 6, 2021 Share Posted January 6, 2021 10 hours ago, jdaniel715 said: how can i change a piece of art from png to make it high resolution? What do you mean by 'high resolution'? You can increase the dpi (dots or pixels per inch) or you can increase the overal image size by interpolating extra pixels. Both of these are done via the menu item Document > Resize Document. To simply increase the dpi, change the dpi in the box provided. There is a drop-down list, but you can enter any dpi you like. You must untick the box labelled Resample. To increase the actual size of the image, you must leave the Resample box ticked and supply a new size in the Width or Height box. By default the Aspect Ratio is preserved. You can enter any units you like (px, mm) or use a percentage (200%). You need to select a resampling algorithm. For photo-images, Lanczos 3 non-separable is probably best. Even with Lanczos, there are limits on how much you can enlarge an image. For diagrams you might find nearest neighbour or bilinear might be better. See the Affinity Turorial on Resizing and Resampling here. John Quote Windows 11, Affinity Photo 2.4.2 Designer 2.4.2 and Publisher 2.4.2 (mainly Photo). CPU: Intel Core i5 8500 @ 3.00GHz. RAM: 32.0GB DDR4 @ 1063MHz, Graphics: 2047MB NVIDIA GeForce GTX 1050 Link to comment Share on other sites More sharing options...
wonderings Posted January 6, 2021 Share Posted January 6, 2021 It is very simple, you look at your computer and say "enhance" and it gets sharper! Of course this is not true. While you can increase the resolution of a picture you are not going to be able to magically make the image itself more clear or sharp. If the image is bad or low res that is what you get. Converting a PNG to vector again is not something you do with a click of a button. You can try doing an "Image Trace" in Illustrator but that has mixed results. I do not think Designer has tracing features yet. Basically you would need to reset/redesign to get a true vector file. Quote Link to comment Share on other sites More sharing options...
v_kyr Posted January 6, 2021 Share Posted January 6, 2021 3 hours ago, wonderings said: Converting a PNG to vector again is not something you do with a click of a button. It always depends how you implement and route application interactions ... ... when I open a bitmap in my own Bitmap Tracy app it starts and automatically traces that one according to the last pre adjusted settings. When you alter a setting, it immediately performs a background retrace process, since settings have changed. So you always have an actual/updated vector representation of the bitmap in memory. When I press the "Open in Designer" button, the in memory vector representation is exported (temporary stored as an SVG) and a process calls Affinity Designer to open that file (instead I could also copy/paste between the apps). - So in my case it's more or less just a click of a button. Fixx 1 Quote ☛ Affinity Designer 1.10.8 ◆ Affinity Photo 1.10.8 ◆ Affinity Publisher 1.10.8 ◆ OSX El Capitan ☛ Affinity V2.3 apps ◆ MacOS Sonoma 14.2 ◆ iPad OS 17.2 Link to comment Share on other sites More sharing options...
wonderings Posted January 6, 2021 Share Posted January 6, 2021 5 hours ago, v_kyr said: It always depends how you implement and route application interactions ... ... when I open a bitmap in my own Bitmap Tracy app it starts and automatically traces that one according to the last pre adjusted settings. When you alter a setting, it immediately performs a background retrace process, since settings have changed. So you always have an actual/updated vector representation of the bitmap in memory. When I press the "Open in Designer" button, the in memory vector representation is exported (temporary stored as an SVG) and a process calls Affinity Designer to open that file (instead I could also copy/paste between the apps). - So in my case it's more or less just a click of a button. That would fall under image tracing, I was talking more that you don't just magically open a PNG as a vector, it does need tracing, be it by the app you are using itself or another one like what you are using. That being said, how do you find that application works for tracing? I don't do a lot of it but now and then I use Illustrator and don't find it the greatest. Is that a free or paid app? Quote Link to comment Share on other sites More sharing options...
v_kyr Posted January 6, 2021 Share Posted January 6, 2021 1 hour ago, wonderings said: I was talking more that you don't just magically open a PNG as a vector Ah Ok, well that should usually be logical, since a bitmap/raster graphics file is no vector format file. 1 hour ago, wonderings said: That being said, how do you find that application works for tracing? ... Is that a free or paid app? It works good, but it's not a free or commercial available app, instead it's an own developed app for my own quick turnaround purposes. It is based on code algorithms from Potrace. There are already a bunch of quite good tracing apps, some freeware most commercial (see above lists). The best tracing apps of those, in terms of the quality of their generated vector output, are probably ... Super Vectorizer 2 (MacOS, commercial) Vector Magic (Win / MacOS, commercial) Inkscape's build-in tracer (Win / MacOS / Linux, freeware, also based on Potrace and for centerline tracing AutoTrace) Quote ☛ Affinity Designer 1.10.8 ◆ Affinity Photo 1.10.8 ◆ Affinity Publisher 1.10.8 ◆ OSX El Capitan ☛ Affinity V2.3 apps ◆ MacOS Sonoma 14.2 ◆ iPad OS 17.2 Link to comment Share on other sites More sharing options...
iconoclast Posted February 23, 2022 Share Posted February 23, 2022 On 1/6/2021 at 12:33 AM, jdaniel715 said: how can i change a piece of art from png to make it high resolution or into vector It is a classical misunderstanding to believe that you can upscale a small pixel image to a big high quality image by converting it into a vector graphic. Pixel images consist of pixels. Pixels are small squares. Each single square contains only one color. So pixel images in fact are tesselated images, mosaics. If you convert a pixel image into vectors, you will convert pixels into vector. And if you upscale that, you will upscale the pixels, so that the quality will get even worse, the more you upscale (aliasing). Another point is that pixel images can't contain details that are smaller than the pixels. If you could upscale a pixel image without aliasing, you would have to perceive that the result has an irritating loss of details, because where should they come from if they can't be on the source image? So, it's true that you can scale vector images lossless, but you can't fool reality by converting a pixel image into vectors to upscale it lossless. If you think about it a little, you will see that it is irrational. Quote Link to comment Share on other sites More sharing options...
Catshill Posted February 23, 2022 Share Posted February 23, 2022 Affinity’s predecessor Serif Draw is Windows only but has an image trace function. I’ve used it a number of times. Quote Link to comment Share on other sites More sharing options...
Alfred Posted February 23, 2022 Share Posted February 23, 2022 1 hour ago, Catshill said: Affinity’s predecessor Serif Draw is Windows only but has an image trace function. I’ve used it a number of times. DrawPlus isn’t really a predecessor of Affinity Designer. Its code base is completely different, and since it remained available for purchase for a long time after AD was released it didn’t predecease it in that sense, either. Quote Alfred Affinity Designer/Photo/Publisher 2 for Windows • Windows 10 Home/Pro Affinity Designer/Photo/Publisher 2 for iPad • iPadOS 17.5.1 (iPad 7th gen) Link to comment Share on other sites More sharing options...
v_kyr Posted February 23, 2022 Share Posted February 23, 2022 2 hours ago, iconoclast said: It is a classical misunderstanding to believe that you can upscale a small pixel image to a big high quality image by converting it into a vector graphic. Pixel images consist of pixels. Pixels are small squares. Each single square contains only one color. So pixel images in fact are tesselated images, mosaics. If you convert a pixel image into vectors, you will convert pixels into vector. And if you upscale that, you will upscale the pixels, so that the quality will get even worse, the more you upscale (aliasing). This depends on the pixel data of an image, whether pixels are scattered individually, or represent a connected line, area etc. The algorithms used in good tracers recognize connected color pixel arrangements, interpolate, smooth and transform them into vector line segments. 2 hours ago, iconoclast said: Another point is that pixel images can't contain details that are smaller than the pixels. ... Since a pixel represents the lowest common denominator optically and physically on a monitor device, where should even smaller ones do visually come from? Quote ☛ Affinity Designer 1.10.8 ◆ Affinity Photo 1.10.8 ◆ Affinity Publisher 1.10.8 ◆ OSX El Capitan ☛ Affinity V2.3 apps ◆ MacOS Sonoma 14.2 ◆ iPad OS 17.2 Link to comment Share on other sites More sharing options...
iconoclast Posted February 23, 2022 Share Posted February 23, 2022 9 hours ago, v_kyr said: Since a pixel represents the lowest common denominator optically and physically on a monitor device, where should even smaller ones do visually come from? If you make a photo with your camera, it is like laying a raster over whatever you are photographing. What details of the reality you are shooting can be displayed on the photo, depends on the size of the details in proportion to the resolution of the photo. Exemplarily hair, branches and other fine stuff like that. Depending on the proportion they will be displayed as aliased lines with diffuse blended colors - the pixels are a mix of the color of the hair and of the background, because the single pixels interleave both. Or the single hair, branch etc. can't be displayed at all, if it is too small. Logical? Quote This depends on the pixel data of an image, whether pixels are scattered individually, or represent a connected line, area etc. The algorithms used in good tracers recognize connected color pixel arrangements, interpolate, smooth and transform them into vector line segments. If you vectorise pixel images, you always get more or less posterised images, depending on the threshold settings you choose. The point is that to get a vector image of a photographic quality, converted from a pixel image, you would need to vectorise each single pixel. And this would only work in the same quality if you forgo scaling, because otherwise the vectorised pixels would become visible. And the resulting file would be very complex and much bigger than the pixel image. So it doesn't make sense. Vector dates are dates like points in a coordinate system, angles, curvatures... Vector graphics are made for graphic images, not for photographs or paintings with many details. To think you can upscale photos lossless by converting them to vector images - because vector images can be scaled lossless - is a fatal fallacy. If that would work, it would be more like witchcraft than physics. Nontheless it is not really the first time I have this discussion. As I said, it's a classical misunderstanding. By the way, sorry for my bad english. Hope it's understandable anyway. Quote Link to comment Share on other sites More sharing options...
v_kyr Posted February 23, 2022 Share Posted February 23, 2022 41 minutes ago, iconoclast said: If you make a photo with your camera, it is like laying a raster over whatever you are photographing. What details ... - the pixels are a mix of the color of the hair and of the background, because the single pixels interleave both. ... All fine and good, though my point was more with how you had initially wrote it up here. - And what I meant is, since all you can get is physically/technically limited to a computer screen's representation (aka the smallest displayable unit is a pixel given at some coordinate x, y) it looks technically like this. Unlike raster/bitmap graphics, vector graphics are not based on a grid in which each picture element (pixel) is assigned a color value and a coordinate, instead more mathematically on an image description (via a markup language) that defines the objects that make up the image/graphics. Vector graphics consist of paths defined by a start and end point, and certain geometric elements. For example, a circle can be fully described in a vector graphic using the position of the center, the radius, the line thickness and the color. Only these parameters are saved and In contrast to raster graphics, this can be changed and transformed easily and without loss. So the strength of vector graphics in general is their resolution independence, i. i.e. they are suitable for reproduction (screen, print) in any resolution. However, the later (screen, print) always requires device dependent a technically complex rendering of the vector graphic into a raster graphic. All in all the strength of vector graphics lies in representations that can be satisfactorily described as a collection of graphic primitives, such as diagrams or company logos. They are not suitable for scanned images and digital photos, which by their very nature are captured as raster graphics and cannot be converted without loss. Vector formats also reach their limits with complex rendered images, which are also calculated directly as raster graphics. Quote ☛ Affinity Designer 1.10.8 ◆ Affinity Photo 1.10.8 ◆ Affinity Publisher 1.10.8 ◆ OSX El Capitan ☛ Affinity V2.3 apps ◆ MacOS Sonoma 14.2 ◆ iPad OS 17.2 Link to comment Share on other sites More sharing options...
iconoclast Posted February 23, 2022 Share Posted February 23, 2022 28 minutes ago, v_kyr said: All fine and good, though my point was more with how you had initially wrote it up here. - And what I meant is, since all you can get is physically/technically limited to a computer screen's representation (aka the smallest displayable unit is a pixel given at some coordinate x, y) it looks technically like this. Unlike raster/bitmap graphics, vector graphics are not based on a grid in which each picture element (pixel) is assigned a color value and a coordinate, instead more mathematically on an image description (via a markup language) that defines the objects that make up the image/graphics. Vector graphics consist of paths defined by a start and end point, and certain geometric elements. For example, a circle can be fully described in a vector graphic using the position of the center, the radius, the line thickness and the color. Only these parameters are saved and In contrast to raster graphics, this can be changed and transformed easily and without loss. So the strength of vector graphics in general is their resolution independence, i. i.e. they are suitable for reproduction (screen, print) in any resolution. However, the later (screen, print) always requires device dependent a technically complex rendering of the vector graphic into a raster graphic. All in all the strength of vector graphics lies in representations that can be satisfactorily described as a collection of graphic primitives, such as diagrams or company logos. They are not suitable for scanned images and digital photos, which by their very nature are captured as raster graphics and cannot be converted without loss. Vector formats also reach their limits with complex rendered images, which are also calculated directly as raster graphics. OK, but why this discussion? The basic point I was talking about in my first post was that it is a misunderstanding that one can upscale a pixel image in a better quality if he/her converts it into a vector graphic. Vector graphics are well known for the opportunity of lossless scaling (SVG="Scalable Vector Graphics"). But that is the source of the misunderstanding, because the source image still consists of pixels, and they will not disappear wonderously just by autotracing. They will also be in the vector graphic. But vectorised. So you will not get a poster in high end photo quality from a small image you downloaded from web. That is what I was talking about. Because it is a wide spread misunderstanding I already was confronted with several times. And I thought that possibly this was what the threat starter wanted to do. Quote Link to comment Share on other sites More sharing options...
v_kyr Posted February 23, 2022 Share Posted February 23, 2022 25 minutes ago, iconoclast said: And I thought that possibly this was what the threat starter wanted to do. Honestly, I don't know what the OP meant by "... a piece of art from png to make it high resolution or into vector" specifically, so if maybe some sort of drawing with a limited color set, or instead some captured cam color photo (?). - If his intension is to get a lowres captured photo image into a highres one via tracing, then it's not really possible this way in a flawless looking manner. Quote ☛ Affinity Designer 1.10.8 ◆ Affinity Photo 1.10.8 ◆ Affinity Publisher 1.10.8 ◆ OSX El Capitan ☛ Affinity V2.3 apps ◆ MacOS Sonoma 14.2 ◆ iPad OS 17.2 Link to comment Share on other sites More sharing options...
Catshill Posted February 24, 2022 Share Posted February 24, 2022 22 hours ago, Alfred said: DrawPlus isn’t really a predecessor of Affinity Designer. Its code base is completely different, and since it remained available for purchase for a long time after AD was released it didn’t predecease it in that sense, either. I am happy with my choice of words and I believe they accurately describe the release of the two vector applications produced by Serif. We are not going to agree on this. Quote Link to comment Share on other sites More sharing options...
Alfred Posted February 24, 2022 Share Posted February 24, 2022 3 hours ago, Catshill said: I am happy with my choice of words and I believe they accurately describe the release of the two vector applications produced by Serif. We are not going to agree on this. Quote Alfred Affinity Designer/Photo/Publisher 2 for Windows • Windows 10 Home/Pro Affinity Designer/Photo/Publisher 2 for iPad • iPadOS 17.5.1 (iPad 7th gen) Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.