Jump to content
You must now use your email address to sign in [click for more info] ×

Herbert123

Members
  • Posts

    230
  • Joined

  • Last visited

Posts posted by Herbert123

  1. It's difficult to answer: I process each image individually for web use. After processing I scale down with either Mitchell-Netravali or Catmull-Rom resampling algorithms. Catmull-Rom in particular keeps small details clear and sharp. I NEVER EVER use Bicubic, Lanczos 3 or 8 for downscaling images: Lanczos works well for upscaling, not so much for the other way around - especially when dealing with sharp-edged illustrative artwork.

     

    Read up on it here: https://pixinsight.com/doc/docs/InterpolationAlgorithms/InterpolationAlgorithms.html

    Scroll down for visual comparisons. Catmull-Rom and Mitchell-Netravali just result in better down-sampled images. I also found that scaling down sharp-edged artwork with Lanczos may introduce artefacting or halos between dark edges and light fills. In any case, Lanczos results in too soft a result - as if blurred a bit. 

     

    Unfortunately Affinity Photo or Designer do not offer either Catmull-Rom or Mitchell-Netravali as a resample algorithm (yet?). That is a real shame, because it does make quite a difference. Anyway, you can always download ColorQuantizer, which is free, and does support these algorithms (and many more!). Then save the result as a lossless PNG, and open it in Photo for further processing.

     

    http://x128.ho.ua/color-quantizer.html

     

    If your content is illustrative sharp-edged artwork, PNG works best. For best final compression and quality control again ColorQuantizer bests every single other tool out there. I never rely on anything else at this point.

     

    Regarding final web output formats: for photos I never us PNG - instead I would suggest JPG at a higher quality (up until the time that we can finally say farewell to JPG and use WebP instead).

     

    If you prefer to keep using PNG as a final web output format for photos, I would advise you to do the final compression and optimization again in ColorQuantizer.

  2. We now have Designer for Windows, Photo in beta for Windows, Publisher promised 'last quarter of 2017'. Any prospect of Affinity Web? I still find that Dreamweaver is the only game in town, but I would like to make the final break from Adobe before it's too late for me.

     

    John

     

    Dreamweaver is a train wreck. It becomes worse with each new release, and the latest one (2017) is a real mess with that half-hearted Brackets integration. I left Dreamweaver years and years ago. Most coders did the same: DW does not even show up in any of the user statistics anymore (for example: http://stackoverflow.com/research/developer-survey-2016

     

    Adobe should just bury it - leave it to Adobe to destroy good products (Freehand, Fireworks, Director, Dreamweaver).

     

    I work predominantly in Netbeans myself, with Notepad++ and Atom on the side for certain tasks. And PineGrow for a more visual environment that is still very code friendly.

     

    To replace Dreamweaver, have a look at the PineGrow <-> Atom combo: bidirectional real-time updates while you work. A much better workflow compared to DW. It's pretty awesome. https://pinegrow.com/

  3. It is by design I guess... It has been working this way since earliest builds of AD.

    In order to make a vector mask gradient you need to create a rectangle, fill it with a gradient and use opacity of gradient stops instead of grey ramp.

    Obviously move it to nest as mask, in this case the gradient widget will be there.

     

    That is too bad. I am used to non-destructive bitmap mask gradients.

  4. I mentioned this before when I tested Designer, and unfortunately the same problem persists in Photo: the 3d emboss layer effect with a custom profile yields unsatisfactory results - quite visible banding spoils the effect, and it becomes unusable.

     

    Here is a comparison between Photo and a competing product (bottom one is Affinity Photo):

     

    24qn61j.png

    No tricks are used, similar settings. No scaling of assets. A smoothing of 1 is used in the competing product (which is negligible) - applying any kind of smoothing in Affinity Photo produces a too soft look, and still does not remove the banding. Both files were produced at 1024x1024px. The only difference: Affinity Photo's file is 16bpc, while the other file is 8bpc. If anything, I would expect better results because I created the file at a higher bit-depth in Affinity Photo.

     

    Here the 3d emboss is applied to a letter (different profile, because Photo crashes the instant I press "linear"):

    eprxom.png

    Last time I mentioned this the developers came to believe I had used tricks in the other application, and that is just not true.

     

    I hope this will be resolved in a future update.

  5. Cloning objects and layers is awesome. Competing software has it, and it allows for mirroring, cloning layer masks for re-use throughout the project (and the layer masks can be based on a clone of the original image), vector clones can be transformed and adjusted with live adjustment layers and effects... The workflow is brilliant. And real-time updates of the clones when the original is edited is truly useful as well.

     

    Once you get used to this, it is hard to work in Photo - it really limits the workflow. At least Affinity does offer symbols - but it is not quite the same.

  6. That isn't a new technique - I have been aware of it for years. I taught that in Photoshop classes myself (granted, with some variations). It is also quite handy to avoid sharpening those pesky JPG artifacts, and still improve the overall texture.

     

    The trouble in Photoshop is that layer masks cannot be cloned or used as a smart object, unless you resort to clunky clipping layers. And adjustment layers cannot be cloned or put in a smart object and still affect the main document.

     

    In Affinity Photo, as you state yourself, layers and masks cannot be cloned or instanced either, nor are smart objects available (yet?). Without the option in either application to virtually clone/instance layers and recycle those as layer masks, it is going to be impossible to create a non-destructive option - unless the developers implement a dedicated tool for this type of functionality - which is not the right path to be taking, in my opinion.

     

    Here's how you would do it in a competitor that does support cloned layers (fully non-destructive, and very controllable with the outline and gray mixer adjustment layers).

     

    Notice how the unsharp masking layer mask is an instance of the original photo layer. When the background is replaced with a different photo, the virtual copy that creates the mask updates automatically. 

     

    6ghcw1.jpg

  7. Affinity would have to compete with the likes of DaVinci Resolve and Fusion - a tough nut to crack indeed.

     

    DaVinci Resolve is a brilliant non-linear video editor, and the industry standard colour grading tool.

    Fusion is a nodal compositor for visual effects - superior to After Effects for that type of work. Not as good for motion graphics, though.

     

    Both are used in feature film production. Can't get any better than this, because the great thing is: both are free to work with up to ultra HD footage!

     

    Get them here:

    https://www.blackmagicdesign.com/ca/products/davinciresolve

    https://www.blackmagicdesign.com/ca/products/fusion

     

    I would advice Serif to stay away from trying to compete in the professional video business.

     

    I would, however, very much welcome a Flash type competitor - but even in that segment the competition is quite tough: 2016 has been very interesting for animators: Moho, ClipStudio, and of course the open sourced OpenTOonz (which is getting better with every new release). Krita also includes animation now, and is finally Mac ready!

  8. Well, when we are discussing system requirements - I remember the days Deluxe Paint IV ran in 2MB!

     

    I installed Photo on an older i5 system with 4GB ram, and it runs quite slow indeed. Adding a curve adjustment slows things down to a crawl, although that might be a beta issue. The competitors all run at an acceptable up to good pace in comparison on the same machine. Then again, Photo is still in beta, and it is still a young fledgling compared to the competitors.

     

    One more thing I noticed is Photo's (and Designer's) heavy installation footprint: around 600GB per application, and that is excludes the .NET framework required to run both. The installation files are a hefty ~250MB. I am aware storage space is inexpensive nowadays, but still.

     

    Adobe's competing products are much worse though: a ridiculous 3GB of space is required for each. Insane. I read somewhere that a major reason for the size increase over the years is the GUI: GUI frameworks require a lot of resources.

     

    Which is why I find it surprising that one other competitor's installation file weighs in at a paltry 22MB, and only requires 50mb for its installation - and can be run from a portable pen drive. It even runs on Windows XP(!) and MacOS 10.6. Yet is on par with functionality.

     

    I blame all those heavy (GUI and other) frameworks developers tend to rely on nowadays.

  9. @svicalifornia: Thanks for the clear explanation: the table grouping example is a good one. I am not against an isolation mode - I just do not like the way it is implemented in Illustrator. If the shortcut key could be modified (with an option to use a modifier key + single click), and the screen darken effect could be turned off as well, I would certainly welcome such an isolation mode.

  10. Refer to the image below. Compared to other Curves implementations, I like these things:

    • the ability to work directly in HSV and HIS mode. Super handy. A simple saturation curve is quite powerful and simple to apply.
    • the option to open any curve in a scale-able window that can be resized to any size - as big as the screen, if needed. I just do not understand why most image editors will not allow the user to do this. Photoshop's curve palette is tiny, even in expanded mode! Curves are so important, and this is one of my pet peeves.  
    • thumbnails of curve presets.
    • the option to apply curve presets to specific channels.
    • various curve types: spline, langrange, bezier, line, text input, etc.
    • a preview option with a split view of which the split can be move to the left and right 
    • the option to quickly create a stepped curve.
    • the option to invert a curve.
    • the option to move the entire curve left, right, up, and down.

    My main issue with Affinity Photo's curves is that it seems quite heavy on processing. On an older i5 Windows tablet (EPE121) the curves choke that machine, and adjusting the curves is almost impossible. I do not experience that issue with other image editors and curves adjustments are snappy and responsive.

     

    Anyway, I feel it is a good idea to compare the various curves implementations, and learn from them.

     

    I do like the alpha channel option in Photo.

     

    dptgsi.jpg

  11. If I understand correctly, Affinity Photo is supposed to run Photoshop plugins. So I want to mention that Filterforge 5 does not run.

     

    Try this: in the Filter Forge.config file change "false"

    <IndependentUI value="false"/> 

     

    to true. That allowed it to run as a Photoshop plugin in other applications for me. I haven't tried it in Photo myself, though. It might work.

  12. I'm not sure that a stabilization would be the solution... that works great for actual hardware pen/screen jitter.... But I believe this is mostly a programming thing. I mean, when you paint in very zoomed out, if you later zoom in to check what you painted, you totally see a grid. This I bet is a big clue. Is like if zooming out, it looses the actual resolution, and instead, treats the image as a very lower resolution, to handle the whole scene fast.  It should force the pen to draw over the actual pixel density or a better approximation, average of the real pixel density that is really present there.  You see how the lines almost are drawing over a very coarse grid (not doing diagonals, but bordering grid cells) of very big quad cells. So, IMO, is more a matter than when is zoomed out, and the pen gets to action, it should paint over the actual pixels existing there, not over a very low res canvas, probably even lower res than what you see as the whole canvas in reduction, even. Because the grid shown while painting with those lines at that zoomed, seems to be of lower resolution than a bitmap of that zoomed out canvas. This does not happen in other packages.

     

    The stroke Stabilizer resolved the identical issue in PhotoLine. Even in Photoshop, despite the interpolation algorithms used while drawing, thin strokes wind up looking odd when painting on a high resolution canvas at a high zoom out level. It also happens in Windows Paint, ArtRage 3,  

     

    IMO, the bad aliasing can be compensated by decreasing hardness up to 50%. Works well for me. I'm more frustrated by the pixelating/trembling pixels that happen all over the line strokes and dots while just even moving the cursor. Distracting to a point of non usability.

     

    I feel we should not have to change the softness of a brush to compensate for Photo's flawed screen anti-aliasing. On the other hand, the developers of Krita spent an inordinate amount of time and energy on getting it right: quality screen anti-aliasing is not as simple as it sounds to develop. 

     

    I strongly disagree (peacefully) about the "no use" point...That's what I thought, totally after discovering the painting issues. My disagreement is because you can actually paint in other tools, like Mypaint, or Krita, which are amazing and free, and very professional for what is actually painting, and use Photo for what I can see will excel : An actual photo/image preparation tool, the needed tool to finish and prepare your illustration to industry standard specs. Because there's a huge issue if you work professionally as an illustrator, with the non high end cost softwares : Simply no tools handling cmyk color profiles in depth or at all (among many other pro features for this). Of those you mentioned, only Krita supports cmyk mode while painting ! (but no color profiles support for cmyk, so...) This is HUGE, as a lot of serious printers work with offset, and even some digital will require your files in an specific cmyk profile. As indeed, is the only fully accurate way to send an illustration to print, other than if using a professional photo lab.

     

    Absolutely agree - up to a point. I would never consider doing a digital painting in Photo, PhotoLine, or even Photoshop at this point: Krita and ClipStudio are my two favourite applications for that task.

     

    But often I just need to do a simple quick sketch or line drawing while working on a comp, or write text. That is currently not an option for me in Photo: the kinks and squigglies are too distracting, and unusable even for a simple rough sketch. I would have to add that I generally work at high resolutions (for example A4@600ppi, 16bpc), and zooming out is essential while sketching.   

     

    Gimp support of cmyk is through a plugin, not while painting. You NEED to be able to fix the color variations and loses once you convert from RGB if you worked in RGB. Or like, me, paint from the start in CMYK, with the exact target cmyk profile. Also, Gimp painting capabilities are clearly inferior to Krita's, while Gimp is superior (though krita 3 is a huge jump) in what it allows in image editing, but then again, still lacks of the important pro part for printing. 

     

    Yep, agreed: I prepare my prepress work in an image editor. Since I also work with comics since a couple of months, I require support for creating a press-ready PDF that consists of a 300ppi colour art layer, and a 1200ppi monochrome 1bit black and white layer for the line art. Support for that is hard to find, though.

     

    As far as I can tell from my testing, Affinity Photo does not support this - similar to Photoshop the base resolution of the document decides the resolution of the layers, which is a shame. In Photoline any layer can have any dimensions, bit-depth, colour mode, and each layer can be individually colour managed. It also means conversions between colour spaces are non-destructive: the information in a layer is kept intact: a colour space change from RGB to CMYK to grayscale, and back to RGB retains all colour information.

     

    Problematic is the lack of support for monochrome 1bit in Affinity Photo and Design. Even if I wanted, I cannot use either one to prepare 1bit line art output.

     

    Hopefully these limitations will be resolved at some point. Perhaps Serif intends this to be possible in the upcoming Publisher, though - it would make more sense.

     

     

    Clip Studio / Manga Studio, yep, that one has great painting features, and CMYK good support. Still, not of the depth in photo and image editing that Photoshop has in some areas, and even Gimp has. Photoline, I believe it has some support of cmyk... Not sure till what extent. Lazy nezumi is only an external line stabilization utility. Which is good, but maybe not the best. Perhaps, from what I have seen, internal stabilization of Clip Studio, and SAI could be a bit superior to Nezumi's. But you can't beat nezumi's price and the fact that works with  many Windows painting applications (not with all).  Of those you mentioned, you included one that can make the perfect combo which am certainly going to use: I will totally purchase Affinity Photo (and Designer), AND will use Krita for painting (as I am already doing) . That kind of joins the best of two worlds. Meanwhile, Affinity will keep improving every area, eventually painting, too. The color management and editing capabilities of Photo (if people realize this is an uber young application growing at full speed) are very on par of what Photoshop does, at least for freelancers. What I need is what is implemented in Photo, just needs usual beta polishing and the room for improvement several years (imo, probably just one year or less) will make to it. My 2c at least, opinions about workflows are infinite...

     

    Krita and ClipStudio are great for drawing and painting. Love both! I feel image editors ought to avoid competing with dedicated painting and drawing applications. Even Photoshop can't compare to these.

     

    I agree: as far as workflow goes, use whichever software gives you the best result. In particular as freelancer (me too) we have more freedom of choice.

     

    Photoline, I believe it has some support of cmyk... Not sure till what extent.

     

    Very good support - about on par with Photoshop. Fully colour managed, and even on a per-layer basis if needed. Linear colour workflow supported in 32bpc. ICC profiles can be set, and conversions are also possible. Tools work directly with CMYK channels if required. Spot colours are supported as well. Also works on Linux with WINE - an alternative CMS (Little CMS) is available that is compatible on Linux.

     

    Custom transfer curve is not available like the one in Photoshop - but we should not mess with those anyway. 

     

    I'm not sure that a stabilization would be the solution... that works great for actual hardware pen/screen jitter.... But I believe this is mostly a programming thing. I mean, when you paint in very zoomed out, if you later zoom in to check what you painted, you totally see a grid. This I bet is a big clue. Is like if zooming out, it looses the actual resolution, and instead, treats the image as a very lower resolution, to handle the whole scene fast.  It should force the pen to draw over the actual pixel density or a better approximation, average of the real pixel density that is really present there.  You see how the lines almost are drawing over a very coarse grid (not doing diagonals, but bordering grid cells) of very big quad cells. So, IMO, is more a matter than when is zoomed out, and the pen gets to action, it should paint over the actual pixels existing there, not over a very low res canvas, probably even lower res than what you see as the whole canvas in reduction, even. Because the grid shown while painting with those lines at that zoomed, seems to be of lower resolution than a bitmap of that zoomed out canvas. This does not happen in other packages.

     

     

    IMO, the bad aliasing can be compensated by decreasing hardness up to 50%. Works well for me. I'm more frustrated by the pixelating/trembling pixels that happen all over the line strokes and dots while just even moving the cursor. Distracting to a point of non usability.

     

    I strongly disagree (peacefully) about the "no use" point...That's what I thought, totally after discovering the painting issues. My disagreement is because you can actually paint in other tools, like Mypaint, or Krita, which are amazing and free, and very professional for what is actually painting, and use Photo for what I can see will excel : An actual photo/image preparation tool, the needed tool to finish and prepare your illustration to industry standard specs. Because there's a huge issue if you work professionally as an illustrator, with the non high end cost softwares : Simply no tools handling cmyk color profiles in depth or at all (among many other pro features for this). Of those you mentioned, only Krita supports cmyk mode while painting ! (but no color profiles support for cmyk, so...) This is HUGE, as a lot of serious printers work with offset, and even some digital will require your files in an specific cmyk profile. As indeed, is the only fully accurate way to send an illustration to print, other than if using a professional photo lab.

     

    Gimp support of cmyk is through a plugin, not while painting. You NEED to be able to fix the color variations and loses once you convert from RGB if you worked in RGB. Or like, me, paint from the start in CMYK, with the exact target cmyk profile. Also, Gimp painting capabilities are clearly inferior to Krita's, while Gimp is superior (though krita 3 is a huge jump) in what it allows in image editing, but then again, still lacks of the important pro part for printing. 

     

    Clip Studio / Manga Studio, yep, that one has great painting features, and CMYK good support. Still, not of the depth in photo and image editing that Photoshop has in some areas, and even Gimp has. Photoline, I believe it has some support of cmyk... Not sure till what extent. Lazy nezumi is only an external line stabilization utility. Which is good, but maybe not the best. Perhaps, from what I have seen, internal stabilization of Clip Studio, and SAI could be a bit superior to Nezumi's. But you can't beat nezumi's price and the fact that works with  many Windows painting applications (not with all).  Of those you mentioned, you included one that can make the perfect combo which am certainly going to use: I will totally purchase Affinity Photo (and Designer), AND will use Krita for painting (as I am already doing) . That kind of joins the best of two worlds. Meanwhile, Affinity will keep improving every area, eventually painting, too. The color management and editing capabilities of Photo (if people realize this is an uber young application growing at full speed) are very on par of what Photoshop does, at least for freelancers. What I need is what is implemented in Photo, just needs usual beta polishing and the room for improvement several years (imo, probably just one year or less) will make to it. My 2c at least, opinions about workflows are infinite...

     
  13. I can confirm the issues with drawing smooth lines on Windows. I tested Photo on a Asus EPE121, which has a Wacom built-in, and the lines are squiggly and show kinks. 

     

    I have experienced this behaviour in other bitmap drawing applications that run on Windows, and it always turns out to be interpolation problems between the canvas (pixels on the canvas) and the screen resolution (physical pixels of the display). 

     

    The relation is easy to test: create an A4 document at 300ppi, and zoom out. Draw with a 1px brush. Zoom out more. Draw again. You will notice that the more the canvas is zoomed out, the worse the interpolation - which results in abysmal looking lines with kinks all along the strokes when you zoom in to view the results at 100%. I believe this problem does not exist on Macs - this is an issue on a Windows OS level.

     

    Surprisingly enough, I noticed that this effect is even visible at 100% in Affinity Photo (1px canvas equals 1px screen resolution) - which is the first time I have experienced the kinks appearing at 100% in any bitmap drawing application on Windows. Ordinarily the interpolation should be fine at a one-on-one px drawing. But in Photo it is not, which is strange.

     

    Other applications have solved this issue in various ways. The best method (in my opinion) is to use a stroke stabilizer algorithm - with parameters, something like Lazy Nezumi. Many brush engines in alternative bitmap editors have this built-in nowadays - excepting Photoshop and Affinity Photo.

     

    As it stands, this beta of Affinity Photo is unusable for digital drawing for me. Suggestion to the developers: add a stroke smooth function with weight settings. Look at Krita, Gimp, Clip Studio, Photoline, and Lazy Nezumi for inspiration.

     

    Unfortunately, a second issue exists: the screen anti-aliasing quality of the drawn strokes. Thin strokes partly break up at certain zoom percentages, and zooming out to 25% on an A4@300ppi with 1px strokes results in a pixel mess. It is by far the worst of any digital drawing software that I have used so far on Windows. The BY FAR best on-screen anti-aliasing is on offer in ClipStudio - nothing comes close.

     

    I realize this is a beta version. I hope it will be resolved - in its current state Photo is unsuitable for drawing (for me at least).

     

    Tested on a EPE121 1280*800px Wacom pen Tablet PC.

  14. In this case I am going to break my own rule (not to mention other competing applications here anymore) because this time the app in question will not compete directly with Affinity since Affinity will not run on Linux (or WINE).

     

    A professional alternative for photo editing for Windows that DOES work in WINE for Linux exists, and is even actively supported by the developers to run in WINE.

     

    PhotoLine works without any issues in WINE on Linux. A number of users work with Photoline in WINE for their work on a daily basis - and the developers even added a Linux compatible colour management system alternative (Little CMS) that can be activated in the preferences. PhotoLine also happily supports vector editing, and will link to InkScape and Krita for a round-trip editing workflow.

     

    Full CMYK and Lab support, mostly non-destructive workflow, non-destructive Raw editing, and 16 and 32bpc is supported. If you are working on Linux look no further for a great Adobe alternative.

     

    Perhaps in five years or so when more users will have made the switch to Linux (I expect to do this myself in a couple of years) Affinity will be made available on that platform as well.

  15. I have a couple of IxD related question for the Affinity developer team:

    • Is an experienced Interaction/UX designer part of your team?
    • Did you perform usability tests with test participants right from the start using prototypes?
    • Do you currently perform usability tests?

    Just curious about your development workflow in regards to usability design - this is not meant in a negative manner. When I design GUIs, games, and sites I almost always perform usability tests with test participants. These may either be formal or informal, depending on the scope of the job at hand.

×
×
  • Create New...

Important Information

Terms of Use | 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.