Skip to main content

Universal Brush




  • Tim Brune

    I agree with everything said. This clone heal tool needs major improvement and soon.

  • Class A

    I've added the following observation to the original description:

    Note that it is good practice to name one's layers. This is not encouraged by the new automatic layer creation approach because the layer is not created before one actually applies a brush (after having selected the brush). So this means, using the new system, one would choose a brush, then do a preliminary stroke, interrupt the brushing to name the layer, and finally go back to brushing.

    For users frequently employing multiple layers, i.e., those who need to name their layers, it is more natural to, first create the layer, then name it, and finally start brushing. However, with the new system, one needs to match the brush type to the type of the layer one has created. Previously, one could just choose the universal brush and it would exhibit the right behaviour (depending on the type of layer).

  • Adam G

    Agree with above concerns.  I vastly preferred the previous universal brush, and the previous clone / heal layer system.  Managing multiple heal layers? No problem, because it was simple and intuitive enough, especially with good naming practices.   This 'updated' version has really thrown a wrench into my workflow.

  • BeO


    On first sight, the new automatism and paradigm looks nice. But as with most automatisms, it favors one specific workflow over  others.

    The new pradigm is cursor tool centric, not layer centric, but uses layers to implement the cursor centric idea, e.g. it is not possible to heal and clone in the same layer, or in any layer, and that is probably because the mask pixels are still "universal". Don't get me wrong, I don't want to have mixed layers...

    The new cursor system is not even consistent in itself, because we still have the universal Eraser!

    (Don't get me wrong, I don't want to have a heal eraser and a clone eraser.)

    The new system is only slightly advantageous for workflows with one layer of type heal or clone and only if one doesn't use the layer tool to add a new layer, or the viewers layer selection box to switch to an existing one, and use the respective cursor with the mouse instead of keybord shortcuts.

    You listed the disadvantages which outweigh the advantages by a lot, imo.

    They could have only implemented the multi-source point enhancement, keep the universal brush, copy and feather masks function. Better value imo.

  • SFA

    I think the multi-point selections and associated masks would have been much too technically complicated (just like the spot tool) to be used through the multi-tool layer system. I can't imagine writing a test plan that would have been effective. Nor running the tests.

    Even worse for the majority of users ... probably.

    But it's just another example of why it is so difficult to keep all of the people happy all of the time.

    If, at some future point, it is decided to merge the functionality because a way has been found to do so  ... so be it.

    For now it is much better, IMO, to release it to the wild and the varied skills of the user community in a form that keeps the component parts self contained and so tests the approach for each use type from a usability POV without stressing users excessively.

  • Class A


    I have no idea what complicated schemes you are talking about.

    Just imagine the old layers and the old single brush with the only difference being that one can have multiple source points in heal and clone layers.

    What would be complicated about that?

    No one is talking about mixing brushes in one layer; that has never been possible.

    The new scheme could be easily fixed by adding a "chameleon" brush that just relieves the user from manually matching the brush type to the layer.

    This could very easily be implemented and testing wouldn't be hard either.

    Think of it as an added "brush auto select" functionality, if you like.

  • Class A

    I don't know what you mean by "worse for the majority of users" either.

    I submit that the current design has objectively bad usability.

    Why let a user allow to make a layer selection, if the latter is then overriden if the currently selected brush type doesn't match the layer type?

    Note that it often makes sense to select another layer, i.e. in those cases where the brush type still matches.

    So the user is encouraged to select layers and tweak the masks in them but has to keep track of whether or not the layer type and brush type are still matching. This is bad.

    Either one has different brush types which work in all layers (in combination, i.e. using a single layer type) or one has different layer types and a single brush type that works in all of them (as it was the case in previous versions).

    The only scheme that doesn't make sense from a usability perspective is the current one where both different layer types and different brush types exist AND the onus is on the user to match them.

    Now this is a complicated scheme.

  • SFA

    Not complicated if people manage their layer activity by naming. More complicated (many seem to suggest) when no grouping is encouraged and anything can end up anywhere. So by listening to a group that insist they have a better way of working the product managers will have so me newly happy people (perhaps grudgingly happy and still complaining that hundreds of other ideas they have have been ignored)  at the same time a whole new set of people who do not want to work that way are now upset. It has always been that way.

    But this is one of those topics that will divide opinion anyway when humans are involved.

    Given that the code for the new heal layer, in particular, seems like is may have some relatively clever things going on and, for what it is doing, seems fast I think it is a great idea to insert it as a stand alone feature (in so far as possible) and so give it a good test in the field with real users but without making its functionality or that of the other approach cross infect each other at this time.

    In fact in my opinion and based on my experiences in the software industry over many years, I would say it is an essential approach for anything vaguely different and potentially performance centric.

    That it looks so simple that doing it differently becomes something people wish to suggest within days of its availability may well be a sign of a complex job well done.




  • BeO

    I fully agree with you Class A except that the universal brush, which we have had all the years until 13.0 incl., should not be an additional brush, but the only brush.


    this was and is easy, consistent, non-complicated, fits every workflow, if the user knows how to create a layer, which he must know anyway if he really wants to work with layers, and David is showing this intensively in his webinars.

    This forced jump to the top most layer is a counter productive gimmik, as it doesn't account for any organized layers way of working, except for the occasional use. This forced selection to pick the correct brush is complicating and interupting layer workflows unnessesarily.

    Having spent a lot of time in the former forum, I don't remember any voice asking for a special healing brush instead of the universal brush which you have to choose if you are already in the correct layer and already have the universal (now degraded to an adjustments) brush.

    And if you work with keyboard shortcuts, well, my keyboard is already filled with custom shortcuts, and remembering two additional ones is not helping either.

    C1 starts to overload with cursor tools as PS and Affinity, say good bye to a slim slick neat easy interface.

    A good compromise could probably be allowing the adjustment brush also working in any selected heal and clone layer. Like the Eraser brush.

    Then one could automatically create the first heal layer or auto jump to the topmost using the new brushes, if one would like to, and still work the old way with only universal brush plus eraser, B and E.

  • Class A


    you wrote "anything can end up anywhere".

    I have never suggested anything like that. I'm puzzled what makes you think that.

    You also wrote "I think it is a great idea to insert it as a stand alone feature (in so far as possible) and so give it a good test in the field with real users but without making its functionality or that of the other approach cross infect each other at this time." 

    I agree 100%!

    However, instead of just enabling multiple source points, they "cross infected" this improvement with 

    a) adding a proliferation of brush types that the user manually has to match to layer types, and

    b) taking away functionality from heal/clone layers (-> mask refinement). 

    So they did exactly what your experience as a software engineer told you they shouldn't have done. 


  • Class A


    you wrote "the universal brush, which we have had all the years until 13.0 incl., should not be an additional brush, but the only brush."

    In principle, I agree.

    I just thought that it it wouldn't be realistic to achieve that. Someone obviously thought that having these different brush types is a great idea so I don't see them going away. 

    For some very casual users who only use few layers and don't revisit them much, the new system could indeed be regarded as a simplification.

    That's why I was asking for the old way to be enabled again, instead of having it replace the new approach.

  • SFA

    Perhaps the problem is the focus on the "brush" identifier as a common meme.

    In fact we have different tools that happen to use some "area identification device" of a type that has some to be described universally, but not really accurately, as a "brush" for want of a better set of words the link common understanding.

    The functionality of the Heal tool, the Clone tool and the Dust and Spot tool are really not the same at all despite sharing some basic appearances and none of them are really the same as the "area markers" that can be used for the other tools.

    Thus for all that they look like "brushes" with a universal on screen appearance they are no more "Universal" than is a an artist's paint brush to a house painter's brush to a street sweeper's brush.

    Or indeed, in all cases, brushes for different types of application.


    Which is why the TOOL function is the thing we are selecting, not really a "brush" at all.

    In turn that makes some sense of automatically selecting (or creating) the layer based on choosing and deploying the "brush" for a specific tool. If more then one layer for that tool then the latest added (presumed to be the one at the top of the layer store (?) ) seems the most logical one to auto select and auto select seems like a sensible choice in the circumstances.

    People that understand what they are doing will soon adapt to the change in approach. People that have no clue being new will, hopefully, be protected from excessive confusion and in any case be guided to the correct tool whether they are newly creating an instance of it or adding to what is already there.


    One potential area of additional confusion is the Spot tool. Not a Brush since it is not a Layer enabled tool BUT in the tool bar it is grouped with the Heal and Clone tools. And one can use it whilst seemingly working in a layer, mask on display.

    Moreover it is 2 different forms of a "healing" tool but NOT entirely interchangeable with the heal tool - old or new version. In fact not really interchangeable at all if one considers its long time design intent.


    For universal tools action the erase brush is a better candidate. Maybe just call it an eraser. Or a wiper. But even that tool has some special characteristics that complicate simplistic decision making.

  • SFA

    a) adding a proliferation of brush types that the user manually has to match to layer types, and

    Ref my post above - tools not brush types

    The way to make the tool do what people were asking for it to do and to do so without limits (apparently) and with good performance meant redeveloping the tool. The term "brush" is just misleading in that context.


    b) taking away functionality from heal/clone layers (-> mask refinement).



    Not a feature I have used much and never with heal and clone activities. How pertinent would mask refinement really be for a heal activity?

    I could perhaps see it being more relevant for certain types of clone activity but nothing springs immediately to mind.

    Still, clearly at least one person has made use of it in their workflow so it would be interesting to look into it a bit more.


    That said  - yes eliminating the previous approach is probably not a great idea and is even a little unusual for C1. I wonder what happens when working with earlier process engines.

    Of course most of the time C1 does indeed support older versions - but not necessarily the option to be able to undertake new edits in a new  version using old and likely no longer available routines.

    How much use such backwards compatibility actually gets might be an interesting question. We all pay for the retention of functionality one way or another. I hope at least a few people enjoy it.


  • Class A


    let me start by pointing out a current inconsistency in the new design:

    If one selects a clone/heal layer, all tools that cannot operate on such a layer become disabled. For instance, one can no longer adjust the "exposure", if a healing/cloning brush is selected. So far, so good.

    If the UI design were consistent, then the two brush types that do not match a selected layer type should also be disabled. Furthermore, the radial and linear gradient masking tools should also be disabled.
    It does not make sense to suggest the use of a tool that isn't applicable for the selected layer.

    The other way of making the current design consistent would be to automatically select the background or topmost adjustment layer as soon as a user starts using a tool that does not work on a selected layer. I hope you agree with me that this wouldn't make any sense at all, yet it is the approach that is currently used for dealing with incompatible brushes. Select a clone layer and, as a new user, attempt to define the target area with the radial gradient tool. What happens? The layer selection changes! If this isn't confusing then I don't know what is.

    FWIW, if you argue that healing, cloning and adjustment mask brushing are all different then it follows that there should be three different types of erasers as well. That this doesn't make sense, tells you that there is much more similarity between the brush tools than you are currently admitting.

    In general, since version 11, C1 became layer-centric. In other words, layers were given a much more pivotal role in the overall design. This was fitting, because

    - the use of layers should be common, replacing background editing in many cases.
    - layers share a common interaction pattern, regardless of their nature (healing vs cloning vs adjustments).
    - layers determine what is possible and what isn't.

    So it made complete sense to have the type of layer determine what the effect of brushing was and to further determine what operations are still possible or not. Hence, the selection of the layer should come first and only then one can consider which tools are left.

    One can use the reverse approach and choose a tool first but what is the point if one then discovers that the layer one intended to work on is not compatible with the selected tool?

    The fact remains that all types of current brushes share a common way of painting a mask (using size, hardness, opacity, flow, drawing straight lines via shift, etc.) Currently "Auto Mask" is not a shared feature but why not? Why shouldn't it possible to get some assistance when drawing a cloning target for instance? Same holds for feathering and refining masks.

    Two of the brushes also feature source points but the commonalities are larger and remain regardless.

    In my years of following discussions on C1 forums, I have never ever once heard anyone complain about the old "universal" brush system. Not once. In contrast, the new system already has some detractors. I can see where the design of the new system was coming from, but the current implementation does not cut it yet. It introduces inconsistencies in the user interaction and forces more work on many users. It might ostensibly be more intuitive for beginners and people who don't use layers extensively won't be hurt by it, but the design just doesn't scale up to serious use.

  • Class A


    you wrote "I wonder what happens when working with earlier process engines."

    Great question.

    I haven't tried to work on any old images because I'm still debating with myself whether to accept the update for my editing machine or not.

    I cannot see how the new design is worth the problems it created.


  • Adam G


    "Not a feature I have used much and never with heal and clone activities. How pertinent would mask refinement really be for a heal activity?

    I could perhaps see it being more relevant for certain types of clone activity but nothing springs immediately to mind.

    Still, clearly at least one person has made use of it in their workflow so it would be interesting to look into it a bit more."...

    I am definitely that one person, and though I can speak only for myself, I suspect that there are others.  I switch between heal/clone layers somewhat interchangeably, and I use the refine/feather functionality almost every time I touch-up a portrait.

    A typical example for me would be removing creases/ circles under someone's eyes:

    1. Create new layer.
    2. Rough-sketch the mask to the area being re-touched using a hard, opaque brush.
    3. Refine or feather said mask until it blends smooth.
    4. Set layer to Heal
    5. Set source point that achieves desired effect (for eye circles, generally just above the cheekbone).
    6. Adjust layer opacity until it is natural and smooth.

    Works great, and only takes me a few seconds-- it is mostly muscle memory by now, and I generally even know what 'number' to use for the mask refinement step.  I've been doing it this way for about three years.

    All of this functionality was lost with the recent release, and I could not find a satisfying workaround.  This wasn't a theoretical interface inconvenience to me-- it was in fact so disruptive to my workflow that I reverted to the prior version, and submitted a separate feature request:

    My apologies if I ended up splintering the 'universal brush' conversation into two threads, because, as pointed out, they are essentially two sides of the same coin...



  • SFA


    With respect to Healing I thing the intent of the Heal tool is to take you straight to your step 5 and allow you to work there.

    I selected an image at random last night - one that happened to be in session I had open - and played around with it based on your process description and what the newly revised tools do as they are.

    This was not studio subject - a 30 something male shot in bright sun outdoors with some extreme bright and shade on the face. Wrinkles and lines aplenty and and a typical mottled skin colouration.

    I was not expecting much but the heal tool alone did a remarkable job, even for large areas, although I disagreed with half of its automatic source selections. The final blend step - basically after one has stopped tweaking stuff, seemed especially skilful.

    I am mindful that my subject matter might in fact represent an easier heal target DUE TO the inconsistent skin colouration and that a well made up and well lit model may in fact be a greater challenge. However I tried another example from my test image collection  - again a random selection that happens to have some lighting extremes but at least studio lit this time - and felt the results achievable with very little effort were fine.

    What I will try, hopefully later today, is a return to 20.0.3 and then follow you step by step process in order to understand to differences and the benefits for the particular use case you mentioned. (Or as close to it as I can get.)


    Finally I will just  mention that although I was using 20.1 the image of the male was using the V9 processing engine and the variant image of the female is using V12.

    If the processing under the masking was doing anything very different to V20 it was not something I could see on the screen.

    What I did not look at in detail is the effect of working with heal and clone layers from earlier images. At least not yet in a specific test.



  • SFA

    This video from Pratik Naik is interesting in the context of the latest healing brush functionality.


    Having also taken a deeper look at the Clone tool  - where I see the potential for using masks modified by refinement as having potentially greater use than in a healing setting - it seems that there may be some missed opportunities but I'm not at all convinced that the way it has been engineered as a tool suggests that a Universal Brush offers a useful enhancement.


    Heal and clone are, in processing terms, two similar but different tools. They happen to share a somewhat similar UI. Perhaps they shouldn't.

    At the very least it could make more sense to allow the option, for cloning, to pick the object or area on wishes to clone using the mask drawing facility. Then pick where is to be cloned to. Dragging the mask would seem to be a good way to do that. Perhaps also retain the selection point although I could also envisage an option to see the effects of the mask on the target area as the mask is created and then just invert.

    In that context, somewhat unlike healing at least in my perception, the concept of refining the mask edges (less so feathering perhaps?)  could have a purpose for making a cleaner selection faster.

    Feathering would likely have a purpose for integrating the copied area with its new background - basically a form of edge healing.

    But I still think the basis of functionality here is for different tools for different purposes and so the idea of a universal brush paradigm may not be the best proposition for future enhancements.





  • Class A


    you wrote "pick the object or area on wishes to clone using the mask drawing facility. Then pick where is to be cloned to. Dragging the mask would seem to be a good way to do that."

    You are describing how the patch tool works in Photoshop.
    Indeed, I'd agree with you that the patch tool is a different tool compared to a clone or healing brush.

    That does not mean, however, that using a brush metaphor to paint the mask for cloning/healing target areas does not make sense. Lightroom, Photoshop and many other editors use the same metaphor.

    You wrote: " the idea of a universal brush paradigm may not be the best proposition for future enhancements."

    So are you saying that the old C1 approach to cloning, healing and adjustments (all using the same brush metaphor) was flawed?

    Are you saying that even the new brush design, which still uses the same "brush" metaphor for all these functions, is flawed?

    You are welcome to your opinion, but as long as the tools use the same metaphor (that of a brush) and offer the same user interaction then I don't see why there should be artificial segmentation between several brush types when a single brush tool would do very nicely.

    Remember that a "universal brush" was what we had before and nobody complained. Remember that before we had the ability to use auto-masking, mask refinement, copying masks between layers, etc. all uniformly independently of the layer type.

    Now we have exceptions and limitations. Please explain to me how that constitutes an improvement.

    P.S.: I realise that the new brush design brings some improvements but these could have been added without breaking anything.

  • BeO

    A healing, clone and adjustments brush do exactly the same, they brush pixels with a certain intensity (determined by opacity and flow, as well as a feather/falloff/hardness). Thats the basis of all, the mask pixels, and it is hard to justify a different shortcut or icon. The right mouse click menu is also exactly the same.

    Depending on the layer type the mask is used for, healing process is triggered or the clone process or adjustments. But that is the only difference, and the universal brush has been existing in previous versions too. The multi source points have nothing to do with the masking action itself. The brushes look basically identical, a center circle and an outer circle. 

    Whats new is the creation of layers or auto jump to highest layer. C1 could keep all this, including the new icons and auto jump, but allow to draw the mask with the B brush, in any selected  layer type.

  • Class A


    "C1 could keep all this, including the new icons and auto jump, but allow to draw the mask with the B brush, in any selected layer type."


    I supported your "Autoselect the correct brush type when selecting a layer" request ( and would love to hear your response to the question I asked.

  • Mihael Tominsek

    Exactly. I was unhappy after upgrade that I pressed B out of  habit on heal layer and suddenly nothing worked, not more healings... only to see new adjustment layer was done. 

    I REALLY LOVED the system you had to name the layer 1st. If I did not want to name it I just pressed enter. Yes, sometimes I pressed B too soon (not after enter) and I had plenty of layers named "B", "BBB", "BBEBEB", checking wy the brush do not work. But that could be prevented. Create new layer and offer naming straight after (As it was), but if mouse is pressed to brush, close naming and leave "Layer #". 

    I did not know ALT reversed brush action. Anyway this is now end. What I would like is to give right click the opposite tool. If B is active, than left click = brush, right click = erase and vice versa. It is more intuitive to use left-right click than always pay attention to have finger on proper key to alt the action. I would rather have menu on ALT key (or user picked one). Right click "erase" will allow to use (generic) pen without any customisation and is easier and faster too on mouse. 

    MENU should not be WAITING to fade. At least give it 0 ms (zero delay) to cancel out. It's time hog. If you happen to click something a bit too soon it is not registered. Or cancel it when ever something is pressed. 

  • BeO

    I don't think changing the right click behaviour is a good idea, right click works for every cursor tool to bring up some options, since long ago, and also is standard at least in Windows (context menu).

    But I think you can overwrite mouse click behaviour in your system, or even depending on the application.

    For example, I have a gaming mouse (only for the use with C1), and configured it as shown below. Until I bought a Wacom, I used these mouse buttons and still think this is faster than with the Pen (but less fun):


  • Mihael Tominsek

    BeO I hhave same G-software for 2x G500, G500s and now 2nd G403 mouse and G910 keyboard - pile of crap, dying ours after warranty or even sooner (also repeating keystrokes, scroll whheel skips, etc). But yes, we can make all sorts of tweaks on buttons. But that it is not it. We can do shortcuts also in C1... It is not the same, since it is permanently changed behaviour of button. 

    If something is done in software internally it is other deal. I have few such tools on Windows that uses right click to reverse the brush function and works fabulously. Right click to get menu roulette is software dependant. Starting in Windows already it is different where you click. If you click on item it is item options, if on desktop, it is desktop options, ... etc. It is really dependant on software how they choose. I specifically am against external software hacks here, because it adds extra layer of delay, might work or not, burden computer with usually bloated and power hungry software (as G-console is), and why lock down to certain vendor. Any pen without any proprietary software loaded emulate right/left click. 

    And I would like to have user optional in preferences:
    - use right click for toggle the tool: yes / no - (no = default)
    - use holding down key for menu: *** (if left empty default is right mouse click)

    So no user will be harmed if it will not know or want this new behaviour. That way Adobe SHOULD make the changes. Instead they silently made change undocumented (I pushed update Premier Pro and went on lunch break, only to came back and spent 2 hours on web wondering what was wrong, until learned this was official "improvement" - it not happened only once). Than they make option after a version or two, after users complain. Yuck, I left Adobe for this.   

  • BeO

    I have the G402 since 4+ years and apart from the middle click (mouse wheel click), which stopped working, it is the best, smoothest mouse I have ever had. But where mechanics is involved, one or the other defect will come sooner or later.

    Anyway, this is not a accessories quality thread, so let me reply to the C1 stuff: From a development point of view I find it problematic to change application behavior ever so often somebody has a nice idea. From a user perspective I like to see more configuration options, and configurable mouse events in addition to the keyboad shortcuts would certainly be a handy improvement. (But you still need a crappy mouse then, right?) :-)

    This having said, I would like to see more menu items / functions  configurable for the Windows version, e.g. Keyboard shortcuts for a specific workspace, " Edit with " a specific app, etc. (There is a workaround built in the Mac system which is not available in Windows.)

  • BeO

    I forgot: create a specific floating tool with a shortcut...

  • Mihael Tominsek

    BeO, I already repied for C1 stuff: 

    - "I find it problematic to change application behaviour ever so often somebody has a nice idea" -> this I hated at Adobe, so I always advocate it has to be added feature only and user selectable, with default set as it was before, not to mess all unaware users, but to please those who benefit from it. Time tells than when u can dispose old stuff or not. 

    - Addition: In C1 I can't understand why only crop ratios have no dedicated shortcuts, while everything else does. It would make it easier.

  • Class A

    I just watched "Capture One 20 Live: Talks | Edit with Joe McNally" on YouTube and noticed that Joe McNally was about to encounter a hiccup due to the new brush design when David Grover reminded him just in time to switch the brush tool before he started brushing.

    Joe had just used the healing brush and was about to start brushing on a freshly created adjustment layer when David Grover had to remind him to switch away from the healing brush to the adjustment brush before he started brushing on the new layer.

    Note that this hint would not have been necessary in the previous brush system, where only one universal brush existed that worked on all layers. In my view, Joe's intuition was completely justified. He was using a brush on a layer and then wanted to use a brush on another layer. The old system would have allowed him to do just that.

    With the new brush design, Joe had to manually switch from the healing brush to the adjustment brush, otherwise the layer selection would have automatically switched back to the healing layer, which I'm sure would have not been welcomed by Joe.

    In my view, this was a real life demonstration why the universal brush should be brought back and may even be considered the default choice. 

    @Lily, is this feature request going to be considered?

    P.S.: Creating the new adjustment layer with "B" would not have worked because there already was an existing adjustment layer. Instead of creating the layer for Joe and selecting the adjustment brush, pressing "B" would have just selected the adjustment brush and would have selected the already existing layer.

  • BeO

    Please give us back the universal brush.

  • Mihael Tominsek

    O would rather say UPGRADE the system to recognise and auto-switch to correct brush. And put that change into preferences "auto brush switch: yes no". Universal brush would mean yet another 4th brush. Already too much icons there. 

    HEALING brush is superb, since it allows super fast dots, specks, lines, correcting. It is better than spot removal and previous healing brush combined. It must stay. 

    ADJUSTMENT BRUSH need to be introduced to differentiate the task. I like it separated and I'm willing to adjust to that. Why: so many times I wanted to dodge/and burn and I forgot to click on proper layer. So I ruined healing. Or vice versa. So it really is not that different than before. ONE switch needed now or before. This was improves simplicity (one or just few layers instead of plenty) and speed of editing. It is easier to switch brush with "Q or B" and it goes where it needs to by itself, than it was finding layer tab, scrolling for proper layer and click on it. Before a layer naming was a MUST for any healing project, now for the most uses there is one layer and "healing layer" name is OK. 

    CLONE BRUSH - same. I would only like to be able to switch CLONE to HEAL and reverse. Sometimes heal can't match hue and produce weird pink colour, so before I just switch it to clone and it was done. Now I need to redraw whole mask. 

    I'm more concerned that before I could drag node outside frame and do crazy good background replacements, now I cant. Really limited. 


Please sign in to leave a comment.