Stair Tool - Pain Points

[PVA - Admin]

The point of this thread is to collect, collate and discuss needed improvements to the stair tool. For the purpose of this discussion, it should be assumed that all changes are possible, up to and including rebuilding the tool from scratch. However, we will leave the decision on how go about it to engineering, instead we will focus on what exists that needs improvement, what exists that does not work at all, and what the tool is not yet capable of that it needs to be in order to satisfy our architectural users specifically. Everything from the creation UI through the final resulting geometry is on the table and mutable.

 

I will collect the items here, and if one or more spread into a more specific and complex discussion, I will split those off into their own threads as needed, but we can treat this overall as a single massive feature request which is not how we normally do wish list items. I, and I’m sure many here feel this issue is important enough to bend the normal forum rules at least for now.

 

The below are in no particular order. Please try to avoid “We shouldn’t do X until we do Y” or “Don't listen to Bob, it should be done the way Charles said!” types of deadlocks, we will let the engineering managers handle implementation and scheduling. What we should focus on here is what is needed conceptually.

Section and Clip Cube Appearance:

(These issues have been identified and plans have been made to correct them, so we can leave this aspect out generally for the time being.)
 

Length/Height Limitation:

It is not possible to create a greater than 2-story or 3-”sided” stair. Users must stack multiple stair objects on top of one another. This is felt to be a regression since Custom Stair was capable of simply adding more platforms and flights to accommodate. Users need to be able to create a single stair objects that can span as many vertical stories as their building has.

 

Railing Geometry:

The generated railing geometry is very often incorrect. This is most easily seen on a default U configuration stair, the inner bends twist and warp in an unacceptable manner. Rectangular handrail profiles seem to suffer the worst in terms of bending and breaking randomly near corners, but even round profiles that manage the bends properly are faceted and not actually round which looks particularly bad in renders when using reflective textures. This is compounded by the fact that even when these railing objects are correct enough to be used, it is not possible to apply a directional (usually wood) texture to them as the mapping is warped and bent. Users are limited to solid colors or textures without any patterns at all.

 

To go further on railing geometry, if users choose to disable the rails and handrails built into the Stair tool, they then turn to the Railing/Fence plugin object, which does not acknowledge or interact with the stair tool, making even manual clicking on each tread and platform location to draw out a railing impossible without excessive editing. This makes what would appear to be the most obvious workaround available even harder to work with and  both of the above result in users creating extrude-along-path objects with their own custom rail profiles, which of course have to be configured manually each time the stairs core parameters are altered.

Users should be able to add their own 2D symbol as the profile for handrail/guardrail/post geometry right in the Stair UI. We could never hope to preempt every possible configuration, so the ability to fully define it from scratch should be handed over to the user.

 

Often, railings are mounted not to the stair itself but to the nearby walls. However, the current stair forces the user to pick one, regardless of whether walls are adjacent to each flight and platform. This is far from the most pressing issue, but the ability of stairs to detect adjacent walls and allow users to override railings on those sides is needed.

 

A decision needs to be made: Will Stair objects contain their railings as part of the Stair plugin, will we offer a Handrail tool that is capable of linking to and intelligently adapting to a Stair object, or will we offer both for different scenarios?
 

In-place edit-ability:

Users want and need the ability to select and edit at least some parameters, if not all, while in the 3D drafting environment. Currently, all input must be done via dialog boxes or OIP fields. OIP fields would even be acceptable if it were possible to click or right click on a stringer, tread, platform, or railing of a stair and simply choose “Edit” and be taken right to the appropriate field in the appropriate palette or dialog.

 

Visual Stair Creation:

Designers are visual people. There is a reason they don't do data entry as their career. They understand what they can see and touch above many other ways of thinking and expressing ideas. Stairs all now have to start from a predefined simple template, when the opposite is often needed. A drawing mode that allows users to mouse click between the starts and stops of flights, modifier keys to insert platforms, winders or curves, or even multiple click creation where the user defines the depth and angle of each tread would be welcome improvements. The best example I could find of this UI was in Chief Architect where this is the default method of creating stairs.

The current methods of adjusting portions of a Stair are limited to distances and angles and it isn't always clear where each value is in the dialog and what it controls unless you confirm the change and hop back out of the dialog which it time consuming, when this method of defining a stair would probably be the second or last choices of most designers.

 

If we want to predefine many of these stair modes, they would best be stored as a Style resource instead, for stamping down standard stairs that only suit standard situations. Not forcing users to always start from these points. We want to encourage creativity and this does the opposite.

 

Preview/Viewing Resulting Changes:

It is very difficult to get an idea of how applied changes in the Stair Settings dialog box actually appear. Yes it’s possible to view the stairs in all of the standard views, and this is an improvements over older versions, but it needs to allow free rotation, pan and zoom via the cursor in the Settings dialog. Especially when attempting to correct railing detail issues, a perfect orthographic projection is often not helpful as it obscures the areas users need to focus on.

Not only should there be more free view control in this dialog (or, perhaps the removal of the dialog entirely and have changes appear live in the drawing?) but the preview itself should be expandable all the way to full screen as it is in the Image Effects dialog.

In either case, there is far too much need to bounce in and out of the Stair settings dialog repeatedly when making changes.

Documentation:

There are a number of very commonly needed construction drawings required as output for stair filing. Plan view with riser counts, top front and side views with full dimensions, section views, etc. These could be created automatically for users on a defined sheet from a button right on the stairs object info palette and linked so that if the user made updates to the stairs, these viewports would update themselves automatically, or at least allow the suer to update them when ready, but keep them linked so that a change to a single stair object didnt require a lot of manually editing its associated viewports as it does currently.

 

Wrap-around Stairs:

There is no clear provision for creating wrap around stairs as are commonly seen on decks, patios, porches and main entrances. These types of stairs are common enough that our default Stair tool should be capable of handling them.

 

Custom Treads:

A method of overriding any given tread, especially those at the top and bottom of a flight, needs to be included. Users need to be able to define the shape of a tread manually to allow for not only artistic customization but also to accommodate existing construction during renovation. This would also allow for the aesthetically pleasing gradually-flared bottom treads very commonly seen in high end construction.

If a user needed to, they should be able to manually define the plan profile for each individual tread themselves. This would of course necessitate that they be able to disable all automatic limitations that force stairs to adhere to a code. The designer should have this freedom.

 

Stair Types:

The current possible configurations are U, O, L, Straight, and Circular. These do not allow for stairs with a continuously variable curve, flared stairs, or S-style stairs that require at a minimum two inner and outer radii. Users are currently forced to cobble together multiple separate stairs to accomplish this and this makes layer and level height changes significantly more time consuming.

 

Texturing:

It needs to be possible adjust the texture mapping on rails, posts, treads, platforms and stringers separately, not simply allow users to change which item gets which texture as can mostly be accomplished now with Class Attributes. This is a broader issue that also affects Windows, Doors and more complex Wall setups, but it bears listing.

 

Underside ceilings:

It needs to be possible to add a smoothed ceiling to the underside of stair objects. The geometry generated currently on the bottom of stairs is often very polygonal and mesh-looking, and users must manually add the intended flat ceilings to these areas, and manually update them if they change stair, layer or level heights. This is more important than ever now that we have interior panoramics exports and walkthrough/webviews that will have viewers staring right at this geometry as they walk around.

 

[PVA - Admin]

If I have missed anything glaring, please speak up and I will add it. Once we have mulled this over a good bit, I will schedule some meetings and get some significant balls rolling. 

[zoomer]

Uh, that is a quite comprehensive collection of all we could wish for .... 🤔

I could sign that.

 

I'm just a bit concerned if that all (beside the Railings) would go together

in one Tool. And so far I am quite happy with a strict parametrical approach

where you can "hard code" your inputs, opposed to a Archicad like AI Stair Tool

that looks like deciding most for you.

But that are just concerns and possibly diminish when you finally see and get

used to such options.

 

Maybe an addition to above said would be a Stair awareness by Slabs so that

cuts are done dynamic and automated/linked.

And of course things like Fire Stairs or Industrial Stairs on Oil Platforms that span

more than one Story, as a single controlled Object would be nice.

 

 

Edited September 14, 2018 by zoomer

[jnr]

Awesome list dude. Thorough and comprehensive. 

 

Priorities:

1. STABILITY. 

2. SPEED. Tired of waiting around for the menu with a thousand data fields to open.

2. FLEXIBILITY-sometimes you need a dumb stair, sometimes you need a really complicated one, sometimes one that's in between (i.e. pet peeve: current stair won' t allow a landing as first riser-old custom stair did)

 

If votes matter, would strongly support a visual graphical creation method as a first means, level with an option to tweak with data fields. Zoomers idea that a stair hacks a hole in a slab would be a time saver too.

 

 

[line-weight]

I'd like to be able to deconstruct into dumb, editable solids - rather than what @zoomer described as "mesh salad" in the other thread. There are some situations (eg for presentation drawings) where the quickest approach is to use the parametric stuff to get you 'nearly there' and then do the final tweaks manually.

[zoomer]

1 hour ago, jnr said:

sometimes you need a dumb stair,

 

I think that is a good point.

So something like the Basic Editing Mode that German Windows have.

You can set them up quickly and increase LOD and Features at any point later.

Edited September 14, 2018 by zoomer

[line-weight]

The tool needs to be flexible enough that we could model *any* of the staircases below, without having to resort to bodges and directly-modelled add-on bits that VW doesn't know are part of the stairs. I think that means full control of geometry on an individual tread and riser basis if necessary.

 

I'm not that bothered if the tool can't work out everything automatically for me; it would be too much to expect it to know what to do with the end of the handrail in the spiral-ended one for example; what we want is maximum flexibility in what we can model and still have understood as a staircase element.

 

Conceptually I don't think that a choice between concrete, timber, etc should be a starting point, and neither should 'U type' or 'L type'. That's how the current tool feels, you start off down one of these predefined paths and then try and bash it into something vaguely similar to what you want. Instead these should be nothing more than 'useful presets' or styles. The starting point should be basic geometry - top and bottom, rise and going, then you fiddle around with flights/landing combinations, something like that.

 

 

Edited September 14, 2018 by line-weight

[zoomer]

14 hours ago, line-weight said:

The starting point should be

 

I wish for a customizable fixed Stair Origin - from Start, End or Middle

(or Custom)

to better control the direction a Stair expands or shrinks,

for Story height changes with bounded Stairs

Edited September 15, 2018 by zoomer

[line-weight]

1 hour ago, zoomer said:

 

I wish for a customizable fixed Stair Origin - from Start, End or Middle

(or Custom)

to better control the direction a Stair expands or shrinks,

for Story height changes with bounded Stairs

 

yes, good point.

[zeno]

There are some points in where vw need to add interaction between objects. Like a windows in a wall, the same behavior should have the stairs with slabs and walls. That means: if i add a stair, the hole on slabs must be generated automatically, so like if i modify a stair, the hole follow it.

If i decide to model a specified stair in detail, the software should allowed to modify, i should be able to choose wich stair's part are connected with a specified slab compoent, even with a temporary 2D interface to detail managment.

And yes: it would be really important that users can change in 2d every single step, even if they want to draw a non sense stair (we are architects..).

Every part (construction parts, railing or external railing) must be visibile in 2D and 3D and completely edited. But not like now. Actually is too much complicated to manage the stair aspect in every form with the attribute dialog. The good way is allow to manage it like the 2d or 3d symbol parts.

 

I think it is essential to think of a whole new tool, please don't try to fix little parts of actually stair logic tool.

 

Totally agree with other points on this thread.

Edited September 15, 2018 by Zeno

[jbtroost]

On 9/15/2018 at 1:20 AM, line-weight said:

The tool needs to be flexible enough that we could model *any* of the staircases below, without having to resort to bodges and directly-modelled add-on bits that VW doesn't know are part of the stairs. I think that means full control of geometry on an individual tread and riser basis if necessary.

 

I'm not that bothered if the tool can't work out everything automatically for me; it would be too much to expect it to know what to do with the end of the handrail in the spiral-ended one for example; what we want is maximum flexibility in what we can model and still have understood as a staircase element.

 

Conceptually I don't think that a choice between concrete, timber, etc should be a starting point, and neither should 'U type' or 'L type'. That's how the current tool feels, you start off down one of these predefined paths and then try and bash it into something vaguely similar to what you want. Instead these should be nothing more than 'useful presets' or styles. The starting point should be basic geometry - top and bottom, rise and going, then you fiddle around with flights/landing combinations, something like that.

 

 

Your first image would be awesome to do parametrically. The second and third I do manually. They are too specific/custom to have in a tool, I guess. 

[line-weight]

37 minutes ago, Jan-Burger TROOST said:

Your first image would be awesome to do parametrically. The second and third I do manually. They are too specific/custom to have in a tool, I guess. 

If we were free to model each tread, riser, balustrade segment, etc directly with whatever solid form we liked, I don't see why either couldn't be 'parametric' even if all that the tool really supplied us with was a definition of heights at top and bottom, going length, and number of risers. So, taking the second one you'd model each tread individually but they would be attached to that basic geometry. This would allow VW to understand the object as a 'stair' but also, it would mean that if, say, the difference between floor levels was adjusted slightly, then a situation where each riser needed to be extended by 20mm could be achieved instantly without any redrawing. For a non-parametric object that kind of thing can be very tedious and time-consuming to adjust (and inevitably it then gets reverted back to the original at some point later in the design process!)

[jbtroost]

19 minutes ago, line-weight said:

If we were free to model each tread, riser, balustrade segment, etc directly with whatever solid form we liked, I don't see why either couldn't be 'parametric' even if all that the tool really supplied us with was a definition of heights at top and bottom, going length, and number of risers. So, taking the second one you'd model each tread individually but they would be attached to that basic geometry. This would allow VW to understand the object as a 'stair' but also, it would mean that if, say, the difference between floor levels was adjusted slightly, then a situation where each riser needed to be extended by 20mm could be achieved instantly without any redrawing. For a non-parametric object that kind of thing can be very tedious and time-consuming to adjust (and inevitably it then gets reverted back to the original at some point later in the design process!)

Wow, if that could be achieved! I now hold my students from using the tool, since it seems counter productive the way it collects all parameters, but your suggestions would help explaining it to them in a clearer way!

[twk]

Thanks for this initiative VW( @Jim Wilson). Agree with everything already said.

I'd add +100 to the LOD comment. That would be awesome to able to start with a basic stair and the ability to built upon that later(parametrically) if needed.

 

On 9/15/2018 at 9:57 AM, zoomer said:

So something like the Basic Editing Mode that German Windows have.

You can set them up quickly and increase LOD and Features at any

point later.

[herbieherb]

We need more detailed stair to landing connections. Best as with editable solids an adjustable 2D basic form. At the moment I don't use the stair tool at all. It has very strange mesh geometry, it is very complicated to adjust everything, you are limited with textures, there is no possibility to apply a levelling layer (plaster) to the bottom view. Basically, it is easier to build a staircase out of editable solids. Because I need them more detailed later than is possible with the current tool, I build them from solids already in the pre-project phase.
Here is an example of the stair to landing connections of our 1:50 plans:

staircase bearing.pdf

[zoomer]

46 minutes ago, herbieherb said:

Here is an example of the stair to landing connections of our 1:50 plans:

 

The VW 2019 Stair Update is thought to solve your Stair Start and Ends Details.

(Haven't tried that so far)

[Asemblance]

All of the above.

 

It would be great if somehow a 'follow me' type tool could be implemented for stairs. This would be especially great with any curved spaces. I'm imagining a polygon as the object the stair follows, with straight segments and curved segments wherever the user wants them.

 

The tool would guide the staircase to essentially follow the drawn object, ensuring a minimum tread depth occurs, and widening treads where curves occur. Once its got the geometry of the curve/stepped wall/whatever correct in 2D plan, you can then worry about adjusting heights of treads and numbers of stairs etc. Once this is dealt with, some way to customise treads either as a group, or individually (with a 'reset' button incase we screw things up..).

[Jim Smith]

I've not loaded 2019 yet, but I'm totally from Missouri on this tool. I check out each iteration but gave up on it years ago & only model stairs as a 3D symbol. If the stair needs to change I basically scrap the symbol & start from scratch. I see the need for the tool but there are just some things that are too complex for a tool. My feeling on the start tool can be summed up with this  Jack Handy quote: "If you ever drop your keys into a river of molten lava, let 'em go, because man, they're gone".

[Joe-SA]

Stairs that transition from open treads to closed treads such as the very simple residential stair in the attached image are nearly impossible within a single stair object.  The Custom Stair tool allowed me to get close by configuring a landing at the transition point but that didn't work very well.   

-  What we need to achieve this is the ability to have two independent straight runs of stairs with the ability to 'offset' the centerline of one run in order to get the desired alignment on one side despite the different tread widths.

- We would also need to control the side tread extensions per run depending on the open vs closed condition on each run.  Independent left and right side controls.

- Give us the option to wrap the nosing controls to the side of the treads in open tread conditions.  Independent controls for left and right sides and by stair run.

- We should be able to align the stringers that are created as part of each of the two runs of stairs...which should be possible with independent control of the tread extensions.

- There should be a distinction between the rough frame stringer object and the exposed open stringer trim you see in this photo.  The PIO should be able to make both with each located at the desired offset.

 

If a tread extension was desired just at the first tread or two...this could be achieved with a series of short stair runs in direct succession.   

 

We don't often create finished interior stair images so we use the Custom Tool to get what we need for plans and building sections.  Anything beyond that is custom built 3D.

 

Edited September 19, 2018 by Joe-SA

[JMR]

Here are my ten cents:

 

 

This applies to small and midsize public buildings, where the stairs are mostly simple in geometry.

 

1.       Stability and predictability. Currently the stair object sometimes generates ghost stair shapes to another layer and is, in general, very buggy.

 

2.       Clarity. It is currently difficult to exactly decipher which settings affect which geometry and how, until one exits the dialog window.

 

3.       The lock settings should work more reliably, currently it’s a bit difficult to predict what happens after a particular lock. The other tabs don’t always update as they should.

 

4.       There are error messages with no apparent solution offered.

 

5.       Stair run/rise is the usual starting point. The rise cannot change within a stair flight, also the run should remain constant unless the stair is winding around a curved landing. Usually the architect knows beforehand what the elevation difference between the stories is, what kind of rise/run is possible for a stair of that particular usage situation (normal/egress etc.) use, and this with the available space determines the resulting size and configuration of the rest of the stair.

 

6.       Ability to lock top and bottom z, as suggested in this thread. For an architect, it is very important to remain confident that all kinds of elevation values stay put. Currently this can be affected too easily.

 

7.       Ability to lock bottom step vertical edge and (2nd) top vertical edge location. Most often we know beforehand, how much space there is for a stair available.

 

8.       Ability to separately input and lock landing dimensions. The landings usually must be of certain size to allow for stretcher use. There are also other regulations regarding landing sizes.

 

9.       Ability to adjust multiple line types/thicknesses at once. Now they have to be picked one by one, AFAIK.

 

10.   Stair step/landing elevation display would be nice option, easy to check landing heights etc.

 

11.   Railings: Currently these are almost unusable. Railings are laborious to draw by hand, therefore a proper geometry would allow for generation sections/elevation views of the railings. Currently the railings jump vertically at a landing, a condition never seen in real life. Also the abovementioned issues with geometry.

 

12.   Railings: The architect is interested in the top/bottom heights of the railings, and in the spaces left between the posts and possibly between the stair and the railing. There are regulations for the maximum space between the posts, vertical distance between open steps, “holes” between the railing construction and the step. All these should be lockable and editable. The regulations vary by country. Eg. horizontal divisions that allow climbing are not allowed in some countries.

 

13.   Perhaps a possibility to use custom geometry for the railing profile as in extrude along path.

 

14.   Ability to easily control what is shown above and below the break line, the current method is almost there but I’ve had difficulties with railing visibility setting regarding the break line.

 

15.   Ability to add “friction” strips on steps, or visual cues for the visually challenged. Perhaps by a custom symbol or geometry.

 

16.   Circular stairs: Currently the structure options are way too limiting, as are the 2D graphics. The circular stairs are mighty difficult to control as it is. There are regulations as to the minimum run at a certain distance from the inner edge, also for another distance at the inner edge, eg. 900mm. The stair minimum width is calculated from this inner minimum run, not from the inner edge. Additionally, the inner edge must not be zero. These regulations vary by country. Therefore; to be able to set minimum runs at least two different distances from the inner edge and display these for building permit approval.

 

17.   Separate central column controls and outer structure (eg. cage) controls. Ability to get rid of structure beneath the steps if the stair is supported from the centre and the outer ring.

 

18.   Ability to stretch individual steps in plan, if desired. Perhaps also solid operations as suggested.

 

19.   Ability to produce all necessary drawings for the stair by using current elevation/projection/plan views.

 

20.   Ability to get rid of the top/bottom structural joinings if necessary; Often the architect draws a solid stair solidly connecting to the top/bottom floor, then the construction bid winner selects a prefabricated stair fitting to the particular dimensions and the engineer makes a detail drawing of the connection suitable for that particular manufacturer’s stair.

 

21.   Stairs spanning multiple stories. This really is a must.

[bpsabatier]

No need to add any more... seems everyone has all my wishes covered.

 

However, one point from the last entry... the custom geometry for the railing profile as in extrude along path.  This needs to be an option in multiple tools.  Most significantly, the door and window trim options.  As of now, we can add trim to the doors/windows, but it is a simple rectangle with no options to add a backband or other profile.  While we can always add it as an extrude along path, can it not be included in the door/window trim options?  Seems like a simple path can already be defined in the door/window, and give us the option to extrude a profile of our choice.  

 

Jim, maybe another thread for these tools is appropriate.  Your original thread mentioned Chief Architect's stair tool.  Check out their door/window tool also, as this option to add a profile is available also.

[Boh]

As well as being stable and providing good presentation control the biggest thing for me is that the tool has to be a DESIGN tool. It has to be quick and easy to edit/customise the stair. Keeping the tool interface simple and intuitive is important. The current stair tool is so overwhelmingly complicated that you need to take a deep breath before delving into it.

 

I like the suggestion of being able to choose between different complexity level modes in the tool so that it can be used as a basic "sketch" design mode so that general layout and headroom etc can be quickly checked/modified, then a "developed" design level mode of complexity so that the styling and look of the stair can be customised, then a "detail" design level of complexity so that all the finer details can be incorporated.

 

This process of sketch through to detail design would reflect how an architect actually works. Many concepts are quick and sketchy and only need the first two modes.

[FBernardo]

I have stopped relying on this tool and use a website (uk based) for designing the stairs and is so easy that even offers a good 3D visualiser. The simplicity of this website is in giving values when we set the floor height that then will generate accordingly the design that we will see in Plan and then in 3D, we can then add accessories and so on.

I would suggest on trying to contact these people and see how they've built this tool and use similar features.

 

Also i would suggest a better "rules" management on the stair, as i've setup the UK Building regulations in the stair tool and i never know if it's actually using it or not.

 

The possibility to actually change the drawing shape of the geometry offered by the tool, the existing one is just impossible to work with.

 

I hope this helps and let's improve the stair tool!

 

 

I've attached the stair tool vs the website i've mentioned and it's the exact same values and the geometry created is nothing alike. (the stair values in vectorworks are rounded up, but are the same)

Edited November 1, 2018 by FBernardo
adding the stair details on the website

[cberg]

I pretty much agree with everything else on this long list. 

 

Sadly, I wish we could turn off the construction tab entirely for the stairs.  That way I could use the tool for general building layout purposes only tread thickness and floor to floor.  As currently implemented, the handrails and construction settings are so inflexible and architecturally incorrect as to be completely not useful.  

 

On my wishlist: 

• Stringers that can use architectural and structural metal shapes/profiles or at least a custom profiles, as it was in the legacy custom stair tool.  Why did the channel stringer option go away in the "improved" stair tool?
• Pickets, balusters, and handrails should have an option to link to metal profiles as well.  Steel tube sizes.  Aluminum extrusion profiles.  Make sure to include aluminum profiles with have square edges.  Not to mention decorative extrusions. Or custom turned wood profiles. 
• Tread material customization.  The other day I was searching for grating or checker plate treads, to no avail in the texture library.  Many times treads are made out of a different material such as wood, or precast.  Each with their own attachment requirements. 
• Ability to customize the risers with angles, mesh, infill materials.
• Newell posts. 
• Panels that incorporate mesh.  Sure this could be a texture.  But meshes require exterior supporting structure.
• Railings that can be made out of cable or rod rigging. 
• Custom nosing profiles, again this was in the custom stair tool and it was never put back into the revised stair tool.

Or if we can't build this any of this into the program, we need to attach 2d view options to each of the 3d elevation and section components so that it can be accurately represented in 2d.

 

 

 

 

 

 

 

[rgcn]

This is a fantastic list Jim.  And a lot of folks here have commented on wish list items I would like as well as some I hadn't thought of.

 

Rob