Skip to main content

Syncing Layers

Comments

6 comments

  • SFA
    Top Commenter

    How would you propose to ensure that the unwanted layers were matched to the new "wanted" layers for replacement?

    Do you envisage enforcing some sort of naming model requirement for layers?

    0
  • Cassaundra Bergman

    I imagine it being similar to what Lightroom allows. I'm not asking to reinvent the wheel. I just want the option to overwrite, sync, or merge layers. None of those options exist currently. Basically, all that happens is Capture One duplicates existing layers a 2nd time. I cannot imagine a use case where that is the least bit useful. Similar choices to what happens with "copy and apply adjustments" would be nice. In face, any dialogue box that allows a choice to be made would be an improvement. 

    1
  • Walter Rowe
    Moderator
    Top Commenter

    If multiple selected images have one or more layers with the same names, perhaps a Layer Sync might identify the common layers by name, let you choose which of those common layers to sync (similar to layer copy-apply), and let you choose which adjustments to sync (with adjustment clipboard).

    • Image 1 Layers: Sky, Lake, People
    • Image 2 Layers: Sky, Lake
    • Image 3 Layers: Lake, People

    Image 1 and 2 have common layers by name of Sky and Lake. Image 1 and 3 have common layers by name of Lake and People. Select all three images with Image 1 as the Primary. Sky and Lake could sync from Image 1 to Image 2, and Lake and People could sync from Image 1 to Image 3.

    The Layer-Sync UI might look similar to the Create Style UI. Choose which layers and which adjustments to synchronize. It might have a checkbox for whether to include the layer masks or only the adjustments.

    0
  • SFA
    Top Commenter

    Cassaundra,

    The original layer synchronisation functionality, as I recall form some versions past, simply copied selected layers (default = all) from the source image and replaced any layers on the target images. 

    In some use cases this would be entirely acceptable. In others, it would be entirely unacceptable.

    So popular demand resulted in a "safer" change where existing layers were retained and the synchronised layers added so that users had more control and important layers were not lost accidentally.

    The current approach seems to be the better option for most people in most situations if we were to take a majority vote. Neither is ideal for all users.

    If it was acceptable to enforce a named layer design, such that all layers had to have unique names and could only be associated with certain types of adjustments for certain areas of images, an improved synchronisation would be possible.

    To take Walter's example, if AI could predetermine what was Sky, Lake, Subject (People?) etc., despite those users who have no inclination at all to provide (consistent) content-related layer names, then the potential for automated synchronisation that satisfies nearly everyone would be much greater.  It would not automatically ensure that all synchronisations would match the user's expectations but could go some way towards that.

    Synchronisation across a small number of variants should be achievable even without any special assessment assistance since one might simply expect the person editing to employ some simple rules for their process that would make the task easier in any circumstances. But one cannot be certain that people will automatically work in that way or accept the potential consequences of failure to do so.

    How one might present the possible solution - something like that suggested by Walter sounds very logical and reasonable - if hundreds or thousands of variants might be involved (perhaps with some selection errors?), is something I think could be difficult to define and agree especially if the expectation of the user is, say, a digitech working on a shoot rather than an enthusiast working on a family album on their retirement leisure time.

     

    And then there is the question of Layer Masks and how they should be assessed and applied ...

    I have the impression that the newly released functionality in V16.3 may be part of the key to making mass layer synchronisation a truly viable proposition at some point.

    AI might offer some potential to assess whether "Layer1"  in image A has the same sort of content as "Layer1" in image B and therefore actively decide, based on some user-selected rules, whether to replace or add an additional "Layer1" to image B. 

    Walter's proposed "Layer-Sync UI" tool could then list the actions/results of the process and provide the user with direct connection to any images they choose to check based on possible unwanted outcomes.

    All of that assumes that the arrival of new technology to the process does not automatically create new ways of doing things that change the user's natural approach to editing in ways that fundamentally bypass the existing ways that we seek to edit images one by one. We might, for example, take a batch of similar images, use AI to select and adjust Sky values across all of them immediately, with commonly named layers created by AI, and bypass the naming convention issue completely. Maybe the mask issue to. 

    That would be somewhat similar to the approach used in Smartphones and similar devices or styles and adjustments for jpg files in cameras. It's an approach that seems to have wide acceptance around the world. Might it also become appropriate for batch adjustments in the professional and enthusiast photographer worlds?

    0
  • Class A

    I believe the term "synchronisation" is making the requested feature seem more complicated than it is.

    For each layer that is listed in the Adjustment Clipboard, one simply needs to be able to choose between "additive" or "replacing".

    Name equality is sufficient to match layers to each other. It might be helpful to enforce unique layer names but the suggested feature does not depend on it; various resolution strategies could be employed to deal with ambiguous cases.

    The legacy "only replace (and delete)" approach had its use cases and the current "only add" approach has its use cases as well. Neither, however, are sufficient to cover regularly occurring use cases.

    If there were an option to batch-delete layers from a set of images, one could at least manually achieve the "replace" mechanics, unfortunately no such functionality is available.

    0
  • SFA
    Top Commenter

    Replace (entirely) should be relatively easy to achieve. 

    Create the layers one requires and copy to clipboard, select the images one wishes to apply them to, then create a batch of new variants. Apply the contents of the clipboard to the new variants.

    Delete the old variants. It may help if they have previously been tagged in a way that made them especially easy to reselect.

    The problem, as I see it, is that I doubt this process would always satisfy all users. Or even most users.

    If it did they would have found it and the subject would not arise much in the forum.

    That's why I think that any development related to the request will, ultimately, need to be more complicated than either you or I feel is necessary. If it is not "fully engineered" you can be sure that the initial response will be a stream of negative posts in the forums and other places.

    Creating the function with known limited abilities and making that obvious (using context-sensitive advisory messages for example) might fend off such criticisms. However, to be able to do that one has to discuss what the expectations may be from the widest possible user base perspective, whether realistic and rational or not.

    Also to adopt the "do no harm" principle, especially where batch processing is involved.

    0

Please sign in to leave a comment.