Dale yet again I can not believe my eyes! I thought when I first saw sub and smooth -"no one will top this one!" yeah well wasn't I proven wrong!!!

Mate you are a breath of fresh air!

@plot-paris said:

I think, this is something to be implemented directly by Google rather than with a plugin...

Well I won't be holding my breath!

Just wondering if a script / function of the pushpull tool could be enabled or set limit the length of a push/pull action to that of the objects map size or a set proportion of it (= to the number of ie: brick repeats of the texture)?

This would be handy to model in material units!

Anyone? Just throwing the idea!

Adam for me I use it every day!! I find it helps to set a border to the likely size I want for output. So yes for some!

@solo said:

Great links, very helpful.

I see my website is even linked as an example of Vue + SU integration.

see mate your famous!!!!

@utiler said:

thanks mateo, mush have had something to do with me being signed in......
yes solo, it seems that takesh_h likes your work too!!!

There's quite a few regulars from SCF that frequent the bar also; these are my first two site in my favorites.....

I'm with you there mate!!

@pbacot said:

I'd say the amount of detail is already more than enough and small discrepancies would be unnoticed unless you were doing a shot from the POV of someone actually standing on the roof. But I suppose an OCD tendency is what drives this sort of craftsmanship.

This stuff is incredible. Like rocket science to a newb like me.

Do you have to just guess what sort of shape the 2d image will make, rendering it over and over until you get what you want? How do you also add color and texture or is it all in the map?

Mate you are probably right about the OCD!!!

It isn't the case of too much trial and error, although the capping was certainly a lot of that! For the second test the map took about 20minutes to hand craft. As what I wanted to achieve understood.

It is just the case that if you have a displacement map grading from 0 (full black) to 255 (full white) and set the displacement to 10cm (the white) you can actually make the height of any point relative as a percentage of the height = the percentage of the grey! Meaning in that case an area with a height of 5cm would be painted 127.5.

In the second test for example to gain the tile proflie I start in PS with three squares and using the gradient fill grey the first 50% to 100% top to bottom then the third square 0% to 50% top to bottom and the centre square diagonally from 0% to 100%. then just merge these and transform width to be half the ridge and half the valley of the profile, copy and flip to few times to fill the tile. Merge these and you have one tile then just copy to fill the map! Done!

The greyscale displacement map pushes the surface up and down the a colour map is used which drapes the tile colour onto this surface, I use the same map in that case as the bump map to add some fine texture to the tile finish!

@chango70 said:

Where the ridges meet looks a bit weird? Can it be resolved using diplacement alone?

Yes mate at the moment it is weird!!!!! I've got to generate another map for there! I will try to do it actually on the same map as the capping map and use components for the capping tiles that only use one part of the map. otherwise it will be another material.

@pibuz said:

Hi Richard!
Impressive tests, i must say!
But, tell us: i once tried maxwell's displacement capabilities, but the result was indeed very odd. Then i read that the displaced surface had to be subdivided into smaller triangles for maxwell to displace it properly. Has it been fixed, maybe? Or you simply subdivided you roof surface a little bit?
I think some SU screenshots would help a lot!
Thanks!

Thanks Mate!

Here is an image of the SU model - I have cut a bit of the roof away so you can see how the capping goes under the tile.

Subdivision of the surface is not a must but what it means is without the surface subdivided you will need really high precision levels to get the best outcome which will greatly increase render times.

To example - when I started this test I didn't sub div the surface and had to apply a precision of 2500 to get hard edges to the tiles. If I selected the adaptive option (which generates the best quality the res of your map will permit) the redner ran for an hour but still didn't start due to the calcs still running! Once I did some very simple sub div with adaptive option selected it cranked up in about three minutes!

@plot-paris said:

another marvellous displacement example, Richard! absolutely convincing.

is it possible with Maxwell to have layered textures with different UV coordinates within SketchUp, as you can do with Indigo?
or are the (truely beautyful) spots of dirt on your first image part of the roof tiles texture and will be repeated in a regual pattern?

Cheers mate!

And yes the option is there to offset or call upon a different repeat to any map used, unlike indigo maxwell can also utilse weight maps so the option is also available to place a large map over the whole roof and provide streaking or dirt where you want it exactly and then repeat the tile texture under that! And only one map needs to be present in the SU model!

I would never want to draw the short falls of Indigo as it has some amazing development but the containment of maps within the model and lack of weight mapping is a real bummer! I look at my low poly cars as an example and the same outcome can never be achieved with the current setup required within SU!

Another quick test with different tile profile! Just quick few minute render!

@plot-paris said:

impressive test, Richard.

especially the "up and down variations" seem a convincing reason for using displacement to me. adding "imperfections" surely is a lot easier than modelling it.

Thanks again PLot!!!

Yes slight offsets, rises and falls, chips and irregularities that all aid to increase realism at no additional modelling time is a real bonus toward the use of displacement. As well as the simple dealing with SU's poly limits.

With the last few renders I've done having the main context dwelling and a few surrounding dwelling with modelled tiles SU would crawl to a stand still once the tile layer was turned on! With just mapping and very simple geometry it will be a god sent!

Thanks all!! I can suggest lots of trial and error on this test!!!

Pete!

Mate I have till date modelled tiles though it is a real hassle given it would be easy if each could just be a component though it ends up that the end tiles often need to be trimmed back or they project out past the capping tile!

And obviously in SU the poly load from the tiles isn't handled well and not so bad with shingle tiles but once the tile has an intricate surface profile it can really blow out! In this case I just have to paint geometry!!! I found if I subdivide the roof well (just running a cookie cutter geometry through it and intersect with model) the rendertimes aren't significantly increased! Must say just painting a texture is really a far more simple process and where the roof becomes complex the saving really are great! And it also means I can paint dirt onto my large scale map and introduce some up and down variation in the tiles!

Hence in all the workflow is VERY simple!

As Frenchy points out the map is a simply devised greyscale map!

I made the maps in PhotoShop applied to very simple SU geometry and exported to maxwell.

I thought I'd have a go to see what results I could achieve myself with displacement maps!

This test as my first was a great learning curve (thanks to Mihai and KurtS - maxwell forum for some tips). The difficulty was all in the creation of the double graded map for the ridge capping tile and the offsetting of this map when displaced to intersect correctly with the roof tile displacement.

I have graded back the roof tile edge of the displacement map and put a solid colour anlong that line of the colour map to avoid the stripped edge to the tiles that would normally occur to the displaced edge! Similar was utilised to get the mortar bedding under the ridge tile!

Seriously it took about half an hour to do the roof tile map correctly and maybe over a day to get the capping tile map right as it was quite a test! Still need to work out the best way to do the capping ends!

@pbacot said:

I'm against the move for city commissions to require proprietary software for submittals, especially if it requires a "service" or the SW is not free. I understand this for larger cities, state school programs etc., but being "different" (Mac-user), it goes against my sense of fair play and free market. Anyway, some of these smaller cities get the idea to require AutoCAD files etc. and they are not really able to utilize the files.

On the other hand, the more they require, the more work for us, hopefully billable.

Well put and exactly my point. In the interest of fair play I think any city model should be so devised that a model for import can can be made in a bunch of universal formats and then converted easily for import. I also feel strongly that any town or city model should be available in a format useable by the public / designer to determine and test structures to the same level they will be assessed in this environment!!

Wow thats a disappointment! really!

I had high hopes that this would be a fast animation opportunity! Would be interesting though to read the license. It could be more in reference to charging for play??

Yes the direct export from SU would be welcomed! Though unless you can use it for commercial purposes then little to be gained!!!

Silver

Man this stuff is seriously wrong!! I'm amazed, the detail, the skill, the patience!!!!!

Bring it on brag all you like mate!!! When you can do stuff like this swing the d... as much as you want in my view!

Mate even Durant's big Sam loves it - so dam it's got to be good!

BTW 5 pages of posts in two days sounds like praise to me!!

Dave Nice image mate, good placement of the people. Might be nice IMHO to widen the camera angle and crop a bit of the ceiling out!

BTW you have neglected to show a few balls stuck in the ceiling here and there most likely between the portals and purlins!

Mate you have done it again! Super work and outcome!

Would love to see a higher res still image! Your grass is looking good, well from what we can tell from video! The tests I've done so far with this technique of two planes is coming out crap!

Would love to see an example of the trees too, and how the leaves are pinned. The foliage looks like it could be a bit denser and smaller leaves IMHO but if the scene had building as context the eye wouldn't be so drawn to this!

Great stuff mate! BTW how are you finding working with Fry???

Hey and your welcome on the idea of the two plane grass - hopefully your skills can push it to work a bit better than it is for me!

Cheers, Richard