[Plugin] ShadowProjector
-
@newvector said:
Hello Friend.
You made a very good plugin. But I have problems in using it. I get artifacts with a shadow. Can you help me with this problem?
I installed this plugin in 2017
Please explain these 'artifacts' more fully.
Use a screen-shot or even a small SKP example... -
Hi
I regret the inconvenience. That's my fault. I reduced the value of the variable 've.length' too much. This led to the appearance of artifacts. Now everything is fine, I changed this value to 0.00001.mm
The value chosen by you 0.25mm is not very good, it generates additional points when the scene is exporting. -
This tool is incredibly useful... thank you very much TIG!
Do you plan to update it? I'm using it in Sketchup 2018, and it's very slow, even though the geometry is simple.
Neither it is able to create the percentages.Best regards!!!
-
i just want to leave my thanks
this helped me to get the sketch-up shadow over into autocad
still needed a bit trickery but still a step closer to getting all these programs closer together
keep on tig! -
Great plugin for shadow analysis!
If I may suggest a few improvements:
- ability to project shadows on grouped faces, so if you select a group with faces, shadows are also projected there and you don't have to explode groups,
- applying the projecton wouldn't reset the selection,
- make an automatic text tablel (outside the shadow face) only informing at what time of day a projection was taken.
-
I also noticed that the instance ID of a projected shadow-face shows wrong hour (the difference is always 2 hours). As in example below: I projected a shadow at 17:00 and it says 19:00.
-
Thanks for the report on the time-stamp error.
There was also an error with the % formatting in the newer SketchUps.
Here's v6.0 https://sketchucation.com/pluginstore?pln=TIG_shadowProjectorI've fixed the issue with those incorrect UTC time-stamps, and the error with sprintf %. It's also been re-signed.Apparently the incorrect time-stamp from the UTC AND ignored DaylightSavings when reported by SketchUp's ShadowInfo have been known for some time. But I've only just learnt of the workaround when extracting the time from the data, which I think now fixes it...
-
Wow, you did this faster than I could think
Thanks, it's fixed!
Are you perhaps going to further develop this plugin? Implementing these three points would be so cool! -
Here's v7.0 https://sketchucation.com/pluginstore?pln=TIG_shadowProjector The original Selection is now kept on completion.
The Time & Date [...@hh:mm-mo-dd] is now added to the Shadow group's name for clarity - so it's easier to find and select in the Outliner.
It has been re-signed. -
Thank you!
Would it make a lot of work to also code projecting shadows onto grouped faces when a group is selected? -
Yes, it is much much more difficult - since the group's transformation must be accounted for when deciding the 'hits' etc.
I have already looked at it briefly, but I've decided to do nothing in the immediate future, since I have other more urgent fish-to-fry...
A workaround that avoids exploding a group, is [after processing any faces within the model's context] to edit that group, and then select any faces within it to process.
Any objects outside of the group [in the model's context] should still cast shadows on the face[s], making the Shadow@... group[s] as normal.
You can then Select any Shadow@... groups inside the group's edit [use the Outline if needed] and then use Edit>Cut; immediately exit the group-edit, and use Edit>PasteInPlace to relocate the Shadow@... group[s] into the main model's context onto the surface of the group... -
Sure. Anyways, your plugin is already a game-changer!
-
Hi TIG,
I am a beginner under SketchUp
I've tried with the project below to calculate % shadow on EACH window panel
If I understand, I've to degroup then modify group but I've ... Bug Splat
Thanks for help
Laurent
-
Please explain your requirements more fully.
You have reset the axes, this will not affect the sun direction.
Either rotate the building as needed or ?If you are trying to calculate the shadows falling on a particular window's glass at a particular time then you need to edit down into the windows entities context and select the glass faces.
BUT, considering the way the windows are built up as components, then I suggest this workaround... draw simple rectangles over the windows and group them, and give that a layer.
Then select that group's face[s] and do the Shadow Projection...It'll be accurate enough for your purposes ?
The many small shadows created by the slatted handrails might well cause issues with geometry and splats...
Try making a much simpler version, of massing blocks/faces...
-
Hi TIG !
Thanks for your help
I need to know how much shadow I have for EACH window panel … so group them seems to me not be the solution.
I do not understand why to draw simple rectangles over the window : is it a tip to group them (it is NOT what I need) ?
I agree with you : the many small shadows created by the slatted handrails might well cause issues with geometry and splats BUT … I'm not the architect and I've to go with them
What do you mean by "Try making massing blocks/faces" ? How to do that ?
Many thanks for your help
Kind regards
Laurent -
Hi TIG !
I'v degroupped objects but the file grows up to 12Mo ...
I can now easily pick the face I want to analyse but ... it still Bug Splat !
How to succeed ?
Kind Regards
Laurent
-
I little more in depth analysis...
The model has not been Geo-located - you must do this in the Model Info settings.
Otherwise the shadow results will not be accurate - they vary by geographical-location, date and time.
Currently it has the template's default of Boulder (CO) USA, 40.01...N 105.28...W
Which I guess is NOT what you want ??Once that's been resolved it'd be a good idea to set up some scene tabs for different dates/times - see below***...
The next thing you need to establish is the building's real orientation.
The axes have been changed to be square to the building [useful for modeling], but if you reset them to the model axes the building is at an angle in plan.
The sun will cast shadows according to the model-axes, irrespective of the current axes.
You can also get a tool from the EWH to reset 'North', thus changing the shadows without affecting the building's drawn layout - however, I don't think this is needed in your case ?Next I suggest you edit the glass material ["VITRAGE"] in the model's materials so it is not transparent [i.e. 100% opaque] - that way it'll accept shadows, making it much easier for you to see which of the windows need analyzing.
There's no need to calculate those windows which are fully shaded - only those in the sun will get also some shading from their frames and reveals, and the timber deck/roof.
This will save save wasted effort.In passing - I also notice that the topmost floors' windows have no 'brise-soleils' above them in the center parts, but in the outermost wings they do - which is illogical - you either need to shade windows or you don't - a 'half-way-house' is either wasting the client's money or not giving him a properly worked out design. Remember that brise-soleils are not 'aesthetic' they have a function !
***You also need to establish the time of day and the month needed for these shadow studies.
Usually it is around noon at the two solstices and also at an equinox, and possibly earlier and/or later in the day depending on the requirements.Are you wanting to find the % shaded ?
Or are you looking to find the % getting sunlight ?*Because of the complex geometry of the slatted parts, to be honest, it'd probably be far simpler for you to overdraw faces onto the glass around the shaded areas [or the lit areas*], the get the area[s] of selected face[s] from Entity Info...
I note that [using the more-than-likely 'wrong' Geo-location] at noon in June [and March] only the top few windows get any light at all - the rest are shaded 100%.
Also because the form of the building several windows/shadows sets are repeated, so there's no need to process every one of them ?
An noon in December they will all get some light, but again there are many 'repeats' of those sets.
Tip: switch off snapping in Model Info > Units when over-drawing...If you go down the manual drawing route, then exploding everything could also be counter-productive...
It'd be easier to draw over components/groups !
PS: I also note that you have raw geometry on various layers - probably because you exploded several component/groups etc.
Raw geometry [edges and faces] should be assigned to Layer0: otherwise madness can result !
To revert them use a tool like my LayerWatcher, or simply delete all of the confusing layers in the Layer Manager and let everything revert to the default [Layer0]...PPS: Here's a screenshot of the left-one done with ShadowProjector and the right-one drawn manually+EntityInfo+Calc+3dText [somewhat faster!].
Note that the actual pane's area 1.65m² was rounded to 1 d.p. by SP as 1.7m².
I also happened to use a different font for my 3dText...
-
Hi TIG
Many thanks for your detailed answer
I had planned to orient my model: the file was sent so that you realize the complexity of modeling and understanding why SP crashes
In fatc, I realize a PHPP (PassivHaus : energetic standard for buildings) study which asks for EACH bay leaf a correction factor (one for winter and one for summer) which indicates how much of the solar radiation hits the window surface relative to the radiation that would occur without this shading element (hence the complement of the % SP).
I thought I would take the average between the most favorable and the worst of each period (winter / summer)The many small shadows created by the slatted handrails are too complicated to manually calculate this % : taht's the reason I pass through SketchUP + SP !
The rounding of 1.65 to 1.7m ² is against a real problem for the precision of the result : can SP be more precise ?
I do not understand when you explain : *"it should probably be simpler for you to get over the surfaces of the shaded areas [or the bed areas ], get the area [s] of selected face [s] from Entity Info"
The result will not be the same as the % given by ShadowProject ? Why do it if SP gives it to me directly, knowing that the model is too complex to draw manually ?Sorry, I'm trying to understand at best ...
Thanks a lot for your help to a beginner (but probably your answer can help anothers who wonder how SP can answer to ...)
Laurent -
Before you proceed...
Get your model Geo-located properly.
Set up your Scene-tabs for the relevant times/dates.
Set the glass material to be 100% opaque - so you can see the shadows.
You only need to consider windows with some light falling on them.
Next, you can use SP on individual panes of glass to get % shading.
However, unless there are lots of panes with different 'slat' shadows, then a manual fix would be quicker/safer...
Many windows' shading repeats so you do not need to do every one...Note my comment about simplifying your layers, so all geometry has Layer0...
If I understand you correctly, you need to do two time/date studies [with some 'averaging?] - if you need 'lit' versus 'shaded' %s then a manual method might well be quicker ?
-
Thanks for all TIG : I'll try ...
What I still do not understand is that you stay on a manual method (quicker according to you) BUT, as already said, the many small shadows created by the slatted handrails are too complicated to calculate this % by drawing manually : that's the reason I use SketchUp + SP !
Is there something that escapes me ?Last question : how to take into account diffuse radiation ? Because it can count for more than 50% of total solar radiation in Central Europe, and up to 100% for surfaces in the North !
Do SU and SP take this into account ?Kind Regards
Laurent
Advertisement