What would YOU change about the Poser UI???

  • @eclark1849 Poser's UI design mechanisms are already very flexible AND user accessible (to a limited extent), if you're prepared to hack the visible resource XML files, which derived from the early Apple Macintosh design concepts of extracting language and regionally customisable resource components so that applications could be more easily adapted to use in areas where the original application language (and script direction) were not the default.

    If one runs any of the image resources through an image editing program, Poser's entire colour scheme can be modified, without affecting how it runs. Think "Browser Skins" for a parallel concept. Don't like the way your browser presents it's menus and tabs? Just apply a different skin. The same concept is applicable to Poser. (But don't forget to keep backups of the original ;-))

  • @shvrdavid said in What would YOU change about the Poser UI???:

    To me, the best thing that could be done with the UI is make everything in it accessible thru python.
    And a pane for custom UI's

    Then anything in it could be customized to ones liking

    That sounds like a plan that would work for anyone, because they could set things up the way they want. The one thing I absolutely hated when I first started with Poser 5 was the inability to move anything where I wanted, which is one of the reasons I didn't stick with Poser.

    Then again, Poser 5's Runtime structure was another nightmare. They finally got past that, so it would be nice if they could do the same for the UI, and PLEEEEEASE someone get them to agree to load the Hierarchy Editor with EVERYTHING COLLAPSED!! I'm sick of wasting time collapsing everything when I load up items in a scene. That's the one feature I always liked about DS 3A. I didn't have to collapse all items in the Scene Tab, it was already like that when first loaded into the scene.

  • @anomalaus OR include several layout styles which you could customize even further if you like. Apparently, I'm the only person who thinks the tabs make the most sense, but fine, then give me the option to make that MY customised layout. BTW, That would also qualify as a feature in my opinion.

  • @tburzio said in What would YOU change about the Poser UI???:

    @eclark1849 No, but then again, the other programs aren't called Poser and get posing wrong... :-)

    Sorry but that is nonsense.

    Poser is NOT a modeller. NOT a Content creating tool. It's a
    tool for posing existing content and it does it VERY WELL.

  • @anomalaus said in What would YOU change about the Poser UI???:

    @trekkiegrrrl & @vilters the thing I'm aware of not being able to do with the current UI is to delete or rename folders one has created, without having to switch to the OS and then reload Poser's library. I'm also desperate for a Python API feature to force the library to reload it's icons, whether for a specific folder or entirely (like the refresh button), since I have scripts which save Animation Set filtered poses, but can't refresh the library to display those new items.

    Perhaps another new organisational tool which would be useful for the library is to have Poser create the usual suite of sub-folders in a new Runtime. I'd really like some way of tagging all related items in different runtime folders for a particular installed product. That way, once could search for that tag, (whether it be an XML tag or implemented some other way) and then drag all of the related items to another Runtime.

    OK That's no biggie for me, personally. I'm constantly switching back and forth between Windows and Poser, and reloading, without giving it a sencond thought.

    But I'm all for a Python option to do it from inside Poser

  • @trekkiegrrrl No, actually it pretty much sucks at it, sorry. Try posing a figure wearing clothes, 95% scaling. First the standard poses won't work, the figure is all messed up. Then every time you grab a limb you get the clothes. Motion for a character, having a figure run, bleech. It's awful! Sure, it eventually works, but, grrr... Problem is, it's not all that far off, but sheesh! Posing is the point. In this case, a very dull point... :-)

  • @tburzio

    Your Poser is working very differently than mine. Every character I have is not to standard scale, yet I have no problem posing a fully clothed figure (or in my case up to 13 fully clothed figures in a scene).

  • JamesCasual, 95%, running pose applied. It's been like this on the PC and the Mac for several decades.


  • This post is deleted!

  • Since the system won't allow me to correct...

    You have applied a Ryan Pose (Poser 8 Male) to James Casual (Poser 6 Male). Since the figures are rigged differently, you will have to adjust James..

  • @tburzio said in What would YOU change about the Poser UI???:

    @trekkiegrrrl No, actually it pretty much sucks at it, sorry. Try posing a figure wearing clothes, 95% scaling. First the standard poses won't work, the figure is all messed up. Then every time you grab a limb you get the clothes. Motion for a character, having a figure run, bleech. It's awful! Sure, it eventually works, but, grrr... Problem is, it's not all that far off, but sheesh! Posing is the point. In this case, a very dull point... :-)

    I find using the dials is the best way to go once your figure is clothed.

  • @eclark1849 even if they have added extra features they have remained consistent in they way they look Case in point is the Bullet Physics palette which looks like several of the other ones so you do not have to relearn what new looks mean. I almost never directly click and drag on interface elements except for the camera movement one which I use a lot instead of the dials, because I can move in two directions at the same time (or use my 3Dmouse). Even though I don't use the Light Control to move lights I do use it to have a visual guide as to where my lights are positioned and a basic idea of how bright they are. If you move something like the Light control to it's own tab what would you gain? If I have to change tabs to move the lights then discover I need to turn the figure by a couple of degrees that will require me to jump back to the Pose tab to move the character and then you get into a tennis match of jumping tabs. As has been pointed out above you can reskin the current UI elements. Also if you really wanted to you could recode them to change location of elements, since all they are is a button with a picture and as long as you have a button with the same GUID number then it will (should) work no problem.

    As far as gaining more work space I would suggest either a bigger monitor with more pixels or two + monitors. I personally do both. My setup consists of two monitors both 21x9. The Preview window is in the smaller upper window since it is 2560x1080, which gives me a full HD size view of the movies I work on. It also has a bit of space to the side to hold some of the seldom used palettes. The larger one is 3820x1600 which holds all my animation palettes and graphs. That way I can have a lot open and still not be crowded.

  • @richard60 I find it rather interesting that you guys don't realize that you're countering your own arguments. You point out that Poser's interface is customizable, but argue against moving any particular UI element from it's default position. Seriously, all I'm saying really is put the Editing tools under tab that says "Editing tools" out of the way. The Camera Controls are just redundant. You have those over the viewport. And I fully admit I'd like a total redesign of the Lighting widget.

  • @eclark1849 Try doing this. Close all the windows you don't want on the main Posing page. Then when you want to use that widget go up to the windows tab and re-open it, but reduce the size of the Pose window to a size you would get in something like the Material room. If you can do that and not get sick of the tennis match as you toggle back and forth then maybe you have a point. However I program control systems to operate the Audio Video equipment in classrooms and part of that is to make touch panel layouts and try to minimize the number of buttons and pages a user has to use. The problem is the more complex a system gets the harder it is to keep the most likely to be used controls in reach. Place a control used a fair amount of time and make a user click two or three times and they get mad. I can tell you the fewer clicks the happier people are. Case in point how many people want the Hierarchy control to open collapsed?

  • 0_1534506011564_Poser 12 concept2.jpg

    I made this concept a while back. It includes:

    • scripts organizer, favourites, etc.
    • extra keyframing functions, insert gap, remove redundant etc...
    • cached previews of frame shows with mouse over
    • three new tracks controlling which camera is active, what events take place (kinda like the categories now), and a comica panel track
    • frame information per frame
    • Fitting room removed and is now merged with setup to make "Rigging room"
    • New tab called Story, where you can organize your frames, render sizes, comic panels, speech bubbles, and integration into moho or other SM stuff.
    • Extra options in the hierarchy panel to collapse stuff, remove a portion of the scene from the memory stack (to free up resources while scene building), add item to a group
    • parameters panel now has checkboxes with the dials so you can do your selecting there instead of scrolling through HUGE menus.
    • Selections can be saved
    • Bulk delete of selected parameters
    • (not added but definitely needed: search dials!!!)
    • selected dials can be saved as a pose, or an injection

    Funny enough this inspired me to actually make my own control panel and this is what followed:


    NOT user friendly at all but I cannot live without it now.

    Here's an example of what it can already do:

    • Poses load instantly now (compare: 3-6 seconds > instant)
    • load rotations only of poses
    • load partial poses using presets
    • load poses without cycling IK and without losing IK detail (also really fast)
    • move and rotate figures and props with preset snap values
    • zero certain axis in translation as well as rotations
    • move relative to camera or actor
    • use a slider to set a value of a parameter before entering it (avoiding lag)
    • analysis of body control dials and zeroing them converting it into a true pose
    • update rotations for JCM / point at control across given timeline (incl quick one)
    • symmetry buttons simplified
    • place before camera
    • quick pick selection of difficult bodyparts
    • get parent of selected
    • collect keyframes from previous frame without going there
    • more IK options
    • move towards target prop/figure by factor
    • bring near to something else
    • partnering figures for erotic poses
    • make selected face target
    • point anything at anything else without committing to point at
    • define a loop
    • quick movie preview of loop
    • set up beats per minute for music videos and skip forward at set frames
    • various trick strips like loading items on actor's origins or passing rotations onto other actor
    • make predefined locations to use to place actors there quickly
    • in development: storyboarding with summaries and dialog per frame using 'custom_data' in a locked prop in the scene
    • actor groups and loop information, render folders and other files associated with the scene, story chapters, example images
    • supporting filetype that records the scene assets and cr2 locations (including info in previous two points): a project file basically
    • Keyframe control, bake all parameters or just the XYZ rotations and translations, of either figure or prop or the entire scene
    • refactor parameter values in given loop to absolute, relative or cumulative values.
    • most buttons have tool tips
    • Shows you what is still conformed / parented to something
    • make actor get rotations of previous keyframes


    • autofocus
    • pre-organize poses based on storyboard
    • new parameters palette with more control and selection options incl search function
    • new hierarchy palette with more control
    • 'Combination open' of figures. Combine items like hair, clothes, materials, before committing a figure load.
    • Texture temp gauge to prevent a crash
    • multi renamer of props
    • light manager
    • fake IK without activating IK
    • walk simulator with set waypoints (that actually starts and stops and works with all figures)
    • previews of all frames in storyboarding section
    • text balloon inserts

  • @erogenesis this won't be a nit-pick of your entire list, so don't sweat ;-)

    Would you care to share the technique that gives instant loading of poses? Is it simply a matter of preventing runtime hierarchy searching, or something more devious/inventive? :-)

    Anything I create in Python scripts could, ultimately, be compiled into C dynamic libraries to accelerate the code execution, but will ultimately still depend on what the developers have exposed in the Python API. Some omissions can be worked around (e.g. new mesh group creation via filesystem writes and OBJ loading), but many cannot (like python access to zone falloff curves). Can you load a character or scene via script, complete with every detail specified in the files, without simply calling the routines which do that as a whole, voiding any possibility of pre-processing or filtering?

  • @anomalaus what I did with the poses is simply read the pz2/p2z file and read in all the rotations of all the bodyparts. Then just set the values on the selected figure. When its an IK figure I have two methods, both take a little longer, but still MUCH faster than Poser's default loader. 1. I use dummy bones I make on the fly to temporarily replicate the skeleton at the given points and calculate the translations of the goals from there. 2. I load a makeshift skeleton in the scene, copy over the joint info, load the pose to that skeleton, set IK on, and then transfer the rotations to the selected figure. This stops Poser from cycling the IK of the selected figure which destroys all the poses in the scene of that figure (if you have limits on), and is MUCH faster.

    I have no idea why the default Poser method takes so long, they might indeed be trying to be smart about something, like scanning for differences in figure setup or something. But as per usual with Poser, when they try to be smart, things go wrong, take long, and/or crash. They need to keep things simple and (gasp) optional!!

    I will hand over this script to them so they can sort it out, or someone that wants to debug/make it user friendly. No way I'm going to do that.

  • Conjecture on my part but Poser maybe trying to read the pose and apply the universal fit to it. Another thing @erogenesis script can be fast on his system, but totally fail on someone else's due to whatever differences there may be between the systems. You can get fast with no error checking on a single platform, however look at how many different systems people here use and if you don't put in error checking you will run into problems. Even with error checking you can have problems.

  • @richard60 I highly doubt that because all I'm doing is accessing python's file read utility as well as a gzip utility which should work across all platforms, and is a very basic routine. Once the text is read, I extract the actor names, parameter names, and values. Those then get assigned to actors and parameters of the same kind, and that's all very basic Poser internal stuff. That should also work universally across all platforms. There is hardly any magic involved.

    But what Poser does, yeah maybe that is dependent on certain systems, but across all my PCs (eight of them), Intel, Asus, HP, Windows 7 & 10, Poser Pro 2014 and 11, its all very very slow. I would be surprised if there would be any difference elsewhere. So Poser is probably doing something etra we don't know about... and I really think they should make that optional.

  • @erogenesis it's just occurred to me that perhaps their parser is slower than yours or mine because it properly supports all of the current and legacy syntax of Poser files. If you're just looking for the transform parameter animated or static values in pose files, but not testing the current piece of file syntax for things you haven't implemented, there may be some performance hit, though I'd expect internal Poser code is fully compiled, rather than interpreted (this may be an invalid assumption, but I've never had any discussions with devs past or present about how Poser is coded).

    I've just run into the need to re-use some old poses saved before the existence of masterSync and before I extracted all of the legacy/deprecated linkParms from my figure. What a pain! Just about every master parameter on the Body has it's dial value replicated on the applicable body part's slaved dial, essentially doubling the effect. GRRRRR! I'm having to dig up the code I used to distil the duplicated dial values, but there's no Python API to cycle the masterSync setting on the slave so it's value can be zeroed, without affecting the master parameter! It can only be done currently by removing the slaving valueOperation (but only when the master and slave parameter names are identical, which is part of the assumed (and undocumented!) prerequisites for masterSync operation), resetting both the master and slave dials to zero, then restoring the slave valueOperation and setting the master dial only! Phew!