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

kirkt

Members
  • Posts

    440
  • Joined

  • Last visited

Reputation Activity

  1. Like
    kirkt reacted to R C-R in John Francis   
    Because the details of how the Affinity apps store data in the native file format are considered proprietary trade secrets, the developers have only given us a few clues about how it works. One of them is that images are not saved conventionally as is typical of other apps but instead in part dynamically as serialized data, & that when certain unspecified conditions obtain, some (or all of?) the serialized data is moved into the unserialized part of the file & the cycle begins again. (This is my interpretation of what they have said, pieced together from different posts, so it may be inaccurate or too simplistic in some respects.)
     
    What exactly this means has never been made clear but empirically it seems to mean that trying to relate file size to the number of layers, adjustments, or other changing states of the file usually won't make much sense -- the size will grow or shrink depending on how much data is serialized at any given time & how much of it is redundant.
     
    They have also hinted at why it is done this way, mentioning (again without disclosing any details they consider to be proprietary) that this facilitates quick saves & loads, improves memory (RAM) efficiency, & reduces the amount of code that needs to be rewritten when features are added or improved that require changes to the the native format.
     
    All that said, it still seems apparent that something must be wrong somewhere if a RAW file developed in Affinity Photo results in a file many hundreds of times larger than the RAW original, but I have no idea what that might be.
  2. Like
    kirkt got a reaction from Alfred in Affinity performance in general and specifically in Focus Merge   
    Of course my reply was predictable, but it looks like you got a response.  I'm not sure why you think AP should have all of the tools and performance that Zerene or Helicon Focus does - it is a general purpose pixel editing application.  But hey, that's one of the things that this forum is for - requesting features.  Try going on the Adobe forum and posting your plea there and see what you get.  Also, consider that, regardless of the tools, the Windows version of AP is months old - enhancing the features and speed of focus stacking may not be a top priority on the Windows development map.  In the interim, consider using Zerene or Helicon Focus, they are both top notch.  You can also use enblend/enfuse (if you do not like the command line, there are GUI's for it).  Photoshop's focus stacking tools generally are not good.
     
    My standard and expectation of this forum is derived from experience with it and the software.  While I do not know the developers or the forum members personally, I have found them to be helpful and I know that they are not a legion of help desk folks awaiting pleas for help 24 hours a day, but a small(er) operation trying to maintain an unrealistically high level of innovation and incorporating thousands of user suggestions, feedback and bug reports.  I am impressed beyond expectation at what they have managed to accomplish in such a short period of time at what are likely unprecedented levels of growth and demand from the users.
     
    kirk
  3. Like
    kirkt reacted to GaryLearnTech in working with zones in B&W images   
    > bottom: click this icon to place a sampler and readout the color under the sampler.
    > BE AWARE! - the sampler will be placed at the top left corner of the document -
    > you can then drag the sampler to the desired location on the image.  
    > The top left corner is where every new sampler is created, unlike PS where the sampler gets placed on the location you click.
     
    @kirkt  There's a small trick here you might be missing.  Instead of clicking on the colour sampler icon to have the sampler appear at the top left, the tool has another option.  Drag-and-drop the sampler onto the image.  You'll be given a small loupe tool which can aid your initial positioning (although I sometimes find it actually gets in the road and makes the task fiddly).  Release the mouse button when you have the sampler in position.  The sampler's position can then be fine-tuned in the normal way by dragging (or by going back and drag-and-dropping from the Info panel again if you find the loupe tool useful).
     

  4. Like
    kirkt got a reaction from anon1 in Gaussian Blur - Live Filter v Filter performance/speed   
    I read your other posts and I agree with you assessment. Flattening a few adjustment layers on top of the two I mentioned above makes the flattening process even more measurably longer. 
     
    Something with live layers needs to be optimized maybe. 
     
    Kirk
  5. Like
    kirkt reacted to anon1 in toggle off layers - lag   
    if you have a pixel layer (photo) and some adjustment layers (like 5 curves, HSL, vibrance) and the underlying image is a pixel layer only
     
    toggling off the layers (shift select, toggle off) above should instantly reveal the pure image below 
    it takes some time, half a second or so, though 
     
    in theory Affinity should not have to do any processing, just straight display the image below 
     
    this would make before/ after comparisons much more enjoyable/ faster 
     
    cheers 
     
    PS
    this may be related
    https://forum.affinity.serif.com/index.php?/topic/44671-snappy-popups/
  6. Like
    kirkt reacted to anon1 in working with zones in B&W images   
    The exact same topic is also described here with solutions for affinity photo 
     
    hope that helps 
  7. Like
    kirkt got a reaction from anon1 in Gaussian Blur - Live Filter v Filter performance/speed   
    Hi Folks,
     
    This issue appears in both the current commercial release as well as the current beta (1.6.5 B5), so I am guessing it may just be the way it is.
     
    I have recently been experimenting with some tonal balancing operations that bring highlights down and shadows up in an image (tonal range compression).  The quick and dirty description is that I start with my contrasty image and produce a grayscale from one of the channels that represents that contrast best.  I make this grayscale image a new layer above the background, invert it and set it to Overlay mode.  This will produce a very flat, mostly awful looking image that has lost much of its local contrast too - however, if you then apply a Gaussian blur of suitable radius to the overlay layer, local contrast is restored.
     
    In trying to devise a strategy for automation via a macro and for producing a non-destructive workflow, I have experimented with using a Gaussian blur Live Filter Layer nested in the Overlay layer with an initial radius set - the user can adjust on the fly to get a good balance of local contrast restoration without halo.
     
    Here is the rub - if all I do is set up the overlay layer with the Live Filter Layer doing the Gaussian blur and then flatten the image, it takes AP many many seconds to perform the flatten operation.  If, however, I apply a Gaussian blur filter destructively to the overlay layer, the effect is nearly instantaneous - if I then perform a flatten operation, the effect is nearly instantaneous.
     
    Here is an example: 5760x3840 px Canon 5DIII image.  If I use the Live Filter Layer method and nest a 100 px radius Live Filter Gaussian blur to the overlay layer, and then perform a flatten of the two layers (the Overlay layer with the live filter applied, plus the background) it takes **24 seconds** to flatten on my machine; if I perform the same, but destructive, operation of blurring the overlay layer and then flattening, each operation in that sequence is nearly instantaneous.
     
    This is on a Mac Book Pro Retina Mid 2015 model, 2.8GHz i7 4 cores, 16GB RAM, etc.  OS 10.12.6.  OpenGL acceleration enabled.
     
    Is there any explanation for why it appears to take AP excessively long to render the Live Filter layer when flattening?  The actual drawing to screen when I adjust the Live Filter is fast, pretty much real time.
     
    Seems odd.
     
    Thanks,
     
    kirk thibault
    berwyn, pa usa
  8. Like
    kirkt got a reaction from bobski2 in Linear colour space handling in photo?   
    Does your rendering application tag the linear output file you generate with a particular color profile?  If not, then AP has no way of knowing how to interpret the color in your image, so it interprets it in the working color space (which is probably not linear).  You can use an OCIO adjustment layer to tell AP how to display the image (non-destructively) so that you can work on linear data while displaying it in some other gamma encoding.
     
    The OCIO config set for vfx and animation are here:
     
    http://opencolorio.org/configurations/index.html
     
    and will probably give you the transforms you are looking for.  The ACES 1.0.3 config set is particularly useful.
     
    If your image is 32bit, you can use the same transforms in the 32-bit preview panel.
     
    Unfortunately, you cannot set up AP to use a linear color space as its working color space for files lower than 32bit per channel and you cannot assign a lower bit depth (16 or 8 bit) file a linear profile.  If your images are 32bit then you are good to go - just set up the color preferences properly and apply the OCIO transform you want for display in the 32bit preview panel.
     
    You can also use a LUT if you have one that does the specific transform you need, again as an adjustment layer.
     
    There are a few tutorial videos that cover processing of 3D rendering images in AP - check the video tutorials list, especially the HDR, OpenEXR and Color Management (OCIO) sections.
     
    Just a note that assigning a linear profile to the image will only change its appearance, but it will not convert the numbers to a gamma-encoded form.  That is, the data remain linear but the assigning of the linear profile help with display.  This is one important role that OCIO plays in AP, as an adjustment layer.  Same goes for the 32-bit Preview panel.  Both approaches help visualize linear data in a gamma encoded manner while maintaining the linear nature of the data for compositing, etc., where data need to add linearly to preserve light physics.
     
    It sounds like you CONVERT your linear files to sRGB in PS, which will change the data to sRGB gamma encoded numbers - are you sure this is what you want?  The data are no longer linear after conversion.  If that is what you want to do, you can use OCIO as an adjustment layer to do the transform and then flatten the result and assign sRGB as the color space to the document.
     
    Anyway, I hope this helps.
     
    kirk
  9. Like
    kirkt reacted to Patrick Connor in So, so, so many layers & Ugly interface   
    Then it's a good thing it didn't truncate at 30 characters!
  10. Like
    kirkt reacted to MikeFromMesa in Why can I not do this?   
    My mistake. I re-read your instructions and this time, when I did what you wrote, it worked. I am not sure why I misunderstood your post before, but I have it now.
     
    Thanks for following up.
  11. Like
    kirkt got a reaction from Alfred in Why can I not do this?   
    If you hold down the SHIFT key while using the freehand selection tool (the "lasso") you will constrain it to straight lines between clicks of the mouse.
     
    Remember that when you select a tool, there are helpful tool modifier hints along the bottom of the image window, including the one I just described.
     
    kirk
  12. Like
    kirkt got a reaction from F_Kal in Cropping - am I going crazy?   
    You can also use the marquee tool to select the "crop" area/dimensions of your layer - then invert the selection (SHIFT+CMD+I) and hit delete to get rid of the pixels on that layer outside of the original selection (the inverted selection will persist until you change it or deselect).  In Photoshop, for example, you can crop with a selection by making a selection and then using the Image ... Crop menu item.  Here, it appears there is no "Crop" command for a selection.  It may save some time to make a macro that inverts your selection, deletes and then deselects.  Then you can crop with a selection in one step.
     
    It would be nice to be able to assign a keyboard shortcut to a Macro to make this kind of operation a quick keystroke.
     
    kirk
  13. Like
    kirkt got a reaction from verysame in Levels   
    That is not constructive.  Perhaps suggest some features or current limitations that you have observed, and what the devs could do to fulfill your needs.
     
    kirk
  14. Like
    kirkt reacted to terry47 in Manual or Tutorial   
    Many thanks for the link Kirk.
  15. Like
    kirkt reacted to jorismak in linear profiles only in 32bit?   
    Thanks for the tips ! Been using luts in videoworld , never thought about using them in photo work. Stupid :).
     
    If the 16bit conversion is indeed done in working space, that means I can work around it which is a huge help already
  16. Like
    kirkt got a reaction from LilleG in Request - Vignette as Adjustment Layer (video demo)   
    This technique is a very clever use of AP's ability to use both pixel-based and vector-based content seamlessly in a single working environment.  Very nicely done and the video makes it very easy to understand the concept and all of its features and powerful adjustments.  This technique could be extended to make any number of dodge and burn vector-based masks that would complement the available raw developer gradient adjustment, without the need for brushing in an edit.  A library of mask shapes could be made with each having its own macro to generate the initial shape.  The user could then modify the shape to their liking for the specific application.  
     
    The example in the video of the vignette being converted to a curve to conform to some of the landscape features is a good example of taking, say, a gradient adjustment (a graduated neutral density filter, but in post) and adapting its effect to the landscape.  Etc.
     
    Have you tried creating a Macro to build the basic structure of the vignette, with some of the controls exposed to the user?  It may be an ad-hoc way to provide the tool in a semi-automated manner to users.
     
    Nice work!
     
    kirk
  17. Like
    kirkt got a reaction from ronnyb in Grain Effect using Equation Filter   
    Yes, you can scale the effect in the equation with a constant - you could even make one of the three parameters (a, b or c) the scaling factor and expose that also, allowing a user driven range for the slider max!
     
    I sort of forgot about this thread I was so busy last week.  Glad I stumbled back into it!
     
    kirk
  18. Like
    kirkt got a reaction from anon1 in [APh] Version 1.6 - suggested additions   
    I responded in the thread you linked to in your above post.
     
    Kirk
  19. Like
    kirkt reacted to kirkt in curves in/ out min/max   
    To experiment with this feature of the Curves adjustment, try the following:
     
    Make a new document - 32bit. This gives you an unbounded (not constrained between 0 and 1) working document. Now, use the rectangle tool, or whatever, and make a box and then use the color chooser to make boxes with various exposures. I have made one for you - you can download the .aphoto file here:
     
    https://www.dropbox.com/s/8l13orpca6xr8x1/Unbounded.afphoto?dl=0
     
    The example file has a large rectangle that is 0EV and four rectangles below 0EV (-1, -2, -4, -8) and four above 0EV (+1, +2, +4, +8). In 32bit color, you can express these values in "intensity" (like log photographic exposure or stops) or in floating point (FP) values. Normally, a bounded working document has all values between 0 (black) and 1 (white). An unbounded document can have any intensity values - presumably even negative ones.
     
    The 32bit document is linear, so doubling the FP value doubles the intensity, equivalent to increasing by 1 photographic stop, or EV. The Color Chooser in AP takes on a special form in a 32bit document, permitting you to choose "color" (RGB values) and intensity (in Stops). You can also dial in both using the FP values. The relationship between EV and FP values is:
     
    FP Value = 2^(EV).
     
    To visualize values beyond the display range, you can use the 32bit preview tool (Studio > 32bit preview). Slide the Exposure slider to adjust the range of values displayed in your 32 bit document. You will see that the current document shows you the rectangles that are in the range between 0 and 1. Dial in negative exposure and you will see the upper +EV rectangles appear.
     
    In the document I provided, there is a curves adjustment layer at the top of the layer stack. You can use the min/max input values to specify what range of FP values you want to comprise the lower and upper bounds of your curves adjustment. Try setting the max input to 0.25 and pulling the white point node down to the horizontal axis (0 output). Notice that only the squares that have a FP value <=0.25 are affected. Etc.
     
    This way you have a true 32bit curves adjustment that you can control the range of input values that get affected in 32bit mode.
     
    Totally cool!
     

  20. Like
    kirkt got a reaction from anon1 in curves in/ out min/max   
    To experiment with this feature of the Curves adjustment, try the following:
     
    Make a new document - 32bit. This gives you an unbounded (not constrained between 0 and 1) working document. Now, use the rectangle tool, or whatever, and make a box and then use the color chooser to make boxes with various exposures. I have made one for you - you can download the .aphoto file here:
     
    https://www.dropbox.com/s/8l13orpca6xr8x1/Unbounded.afphoto?dl=0
     
    The example file has a large rectangle that is 0EV and four rectangles below 0EV (-1, -2, -4, -8) and four above 0EV (+1, +2, +4, +8). In 32bit color, you can express these values in "intensity" (like log photographic exposure or stops) or in floating point (FP) values. Normally, a bounded working document has all values between 0 (black) and 1 (white). An unbounded document can have any intensity values - presumably even negative ones.
     
    The 32bit document is linear, so doubling the FP value doubles the intensity, equivalent to increasing by 1 photographic stop, or EV. The Color Chooser in AP takes on a special form in a 32bit document, permitting you to choose "color" (RGB values) and intensity (in Stops). You can also dial in both using the FP values. The relationship between EV and FP values is:
     
    FP Value = 2^(EV).
     
    To visualize values beyond the display range, you can use the 32bit preview tool (Studio > 32bit preview). Slide the Exposure slider to adjust the range of values displayed in your 32 bit document. You will see that the current document shows you the rectangles that are in the range between 0 and 1. Dial in negative exposure and you will see the upper +EV rectangles appear.
     
    In the document I provided, there is a curves adjustment layer at the top of the layer stack. You can use the min/max input values to specify what range of FP values you want to comprise the lower and upper bounds of your curves adjustment. Try setting the max input to 0.25 and pulling the white point node down to the horizontal axis (0 output). Notice that only the squares that have a FP value <=0.25 are affected. Etc.
     
    This way you have a true 32bit curves adjustment that you can control the range of input values that get affected in 32bit mode.
     
    Totally cool!
     

  21. Like
    kirkt reacted to paolo.limoncelli in DAUB Mosaic   
    Well... (update...)
     

     
    A set of 6 simple stupid tools to vent your obsessive-compulsive disorder  :lol:
    These brushes offer a ramp shaped to jitter hues and luminosity of the colour you're painting with.
    Download here
     
    Happy mosaic!
  22. Like
    kirkt got a reaction from jbartley in [AP] Soft Proof Adjustment Layer - Gamut Check   
    If it is possible, it would be terrific if the "Gamut Check" feature in the Soft Proof adjustment layer provided a false-color overlay/map that gave an indication of how far OOG the pixels were, say in deltaE, instead of a gray overlay that does not indicate how far OOG each selected pixel actually is.  
     
    Even better would be a preference to map the false color overlay to different dE values - that way, the user could specify the level of dE discriminated in the OOG overlay.  For example, green for a dE <2, yellow for dE between 2 and 6, red for dE > 6 or something like that.
     
    This would be useful in assessing OOG colors for print, for example, and what adjustments would be necessary to bring the current document into gamut for the intended output space.  Sort of highlight and shadow clip warnings in a typical raw converter.  I can envision making a saturation adjustment or a Lab curves adjustment and watching the OOG indicator map change from lots of red and yellow to a little bit of yellow and calling that adjustment sufficient for the conversion into the destination color space.
     
    Color Think:
     
    http://www.chromix.com/colorthink/pro/pro_worksheet?-session=SessID:47E61C221d6c420F6BuomWB1C0FC
     
    provides this kind of gamut map, but is a professional color tool that is expensive and requires a lot of additional knowledge to use, and is not an embedded tool in an image processing application.  The attached example shows a particular blue in AdobeRGB and how it would map to an Epson printer profile, with the "gamut map" showing how far OOG some of the blue areas are.
     
    Thanks!
     
    Kirk Thibault
    Berwyn, PA

  23. Like
    kirkt got a reaction from anon1 in [AP] Soft Proof Adjustment Layer - Gamut Check   
    If it is possible, it would be terrific if the "Gamut Check" feature in the Soft Proof adjustment layer provided a false-color overlay/map that gave an indication of how far OOG the pixels were, say in deltaE, instead of a gray overlay that does not indicate how far OOG each selected pixel actually is.  
     
    Even better would be a preference to map the false color overlay to different dE values - that way, the user could specify the level of dE discriminated in the OOG overlay.  For example, green for a dE <2, yellow for dE between 2 and 6, red for dE > 6 or something like that.
     
    This would be useful in assessing OOG colors for print, for example, and what adjustments would be necessary to bring the current document into gamut for the intended output space.  Sort of highlight and shadow clip warnings in a typical raw converter.  I can envision making a saturation adjustment or a Lab curves adjustment and watching the OOG indicator map change from lots of red and yellow to a little bit of yellow and calling that adjustment sufficient for the conversion into the destination color space.
     
    Color Think:
     
    http://www.chromix.com/colorthink/pro/pro_worksheet?-session=SessID:47E61C221d6c420F6BuomWB1C0FC
     
    provides this kind of gamut map, but is a professional color tool that is expensive and requires a lot of additional knowledge to use, and is not an embedded tool in an image processing application.  The attached example shows a particular blue in AdobeRGB and how it would map to an Epson printer profile, with the "gamut map" showing how far OOG some of the blue areas are.
     
    Thanks!
     
    Kirk Thibault
    Berwyn, PA

  24. Like
    kirkt reacted to smadell in Affinity Photo Customer Beta (1.5.0 - RC2)   
    I can confirm that the Gamma slider within the Levels panel doesn't seem to be functional. Tested on AP native file, JPG, and TIFF.
  25. Like
    kirkt reacted to Peter Werner in Affinity Photo Customer Beta (1.5 - RC1)   
    First off, hats off for a spectacular release, this is absolutely fantastic!!!
     
    Sorry for bringing in my feedback at this point in time and in this form, but unfortunately I haven't had the time to take part in the Beta this time around. The following is all based on looking at James' excellent video demos, not on trying the features in practice, so my apologies if I'm overlooking things that are already there.
     
    Here we go:
    Having to activate the clone tool before double clicking a source seems kind of redundant. It would be useful if double clicking a clone source would also activate the clone tool as well if it is not active yet. In the Equations filter (which looks amazing by the way!), it would be useful to have a flyout menu that just offers to add the supported functions until a full expression builder is available. A button that takes the user directly to the corresponding help page with all the documentation might also come in handy. It would be nice if the Equations filter allowed to save equations as presets. This would also allow you to ship a few examples to get everyone started/inspired. Same situation for Apply Image. Macros are of course a workaround, but they are not as convenient if you want to use existing presets as a starting point for your own equations. When creating a new macro category, I think popping up a dialog asking for a name would be useful since pretty much everyone would likely want to name their categories right away. Those who don't want to could always just press the return key. I think the Pixel Art resize filters could be built into the regular image resizing dialog with an option to switch between regular resampling and pixel art resampling, and then changing the other options based on the choice. As the number of features increases, there will be lots of menu items, so this would keep the main UI less cluttered. It's also quite likely that a new user looking for the pixel art resizing feature would look for it unter the "Resize" command, and not think as far as to search for a separate command. Integrating it into the pre-existing dialog would also allow for readouts of the dimensions that, say, "4x" would result in (just re-use the existing input fields). For 360° editing, I think it would be infinitely more intuitive to add a 360° navigation tool (similar to the pan hand or zoom tools). In the current form, it's really hard to discover how to get back into the tool for someone who hasn't seen the video tutorial. The options bar could include an additional checkbox option for "Enable equirectangular 360° projection" for quick access (would be possible even in the current form). Since 360° projection viewing seems to be a per-layer setting, it would be useful to display a corresponding indicator icon in the layers panel for each layer that has the setting enabled. Or a separate checkbox column that can be shown and hidden via the flyout menu, or shown only when 360 projection is enabled for at least one layer. For macros, you are using the eye icon to indicate visibility of settings, but for layers in the layers panel, you are using a checkmark, which is a bit inconsistent. Might be wrong, but it seems to me that it is not possible yet to do HDR and panorama merge in one step, which would be cool since it would eliminate the need to manually export out the HDR files to EXR and re-import them to merge them to a panorama. For Apply Image, a small thumbnail in the dialog box would be useful to see what the source is currently set to and/or if a drag and drop operation yielded the desired result. In Apply Image, just like in Curves and other parts of the software (Channel selection in Adjustment Layers etc.), the color space choice is a drop down field, which, compared to, say, an NSSegmentedControl, requires an additional unnecessary click from the user. Drop downs are better suited for long lists like blend mode, but if there are only very few options, direct access buttons are much faster to use. This is really annoying in Photoshop as well. Also in Apply Image, the S in the variable names seems a bit superfluous – as far as I can tell, recursive operations don't make sense there, so just using r, g, b for the input color values would suffice I believe and make things slightly more readable. And if destination values should really be accessible, they would be differentiated by the D already, eg. DR = R + 50; DG = DR; (i.e. copy the result of the calculation for the red channel)  DB = B; In the video, EXR loading seems to occasionally take quite a bit of time – visual feedback in the form of some kind of progress bar would be useful, especially on slower computers. Seems to me that the histogram panel only shows values from 0..1 for now It would be cool if the Lighting Effect filter could access normals imported from EXR files for interactive relighting Tone Mapping: Need to play with this myself, but on first impression, it seems like an "Auto" button for Tone Compression would be useful. Simply look for the maximum value in the image and set the slider value so that all information is included in the result without any clipping. The first thing James does in every video is find this value manually, and then start from there. So if a button could do that work for us, that would save some time. I know it's too late for this release, and my apologies for just dumping all this in one solid block, but I hope it's still going to be helpful for future development.
     
    Again, congratulations on entering the RC phase and thank you all for your hard work!
×
×
  • 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.