Author Topic: My wishlist - conclusions regarding Lumion  (Read 17034 times)

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« on: January 04, 2011, 06:53:10 am »
January 04, 2011, 06:53:10 am
First of all congratulations for your wonderfull software. I am an architect and i've been loking for something like that for ages. I understand this is the first version so i would like to suggest some improvements as this software can be so much greater.


The terrain tools need a slight improvement and that would be the cut option (straight not circular) so you may cut straight the terrain around the placed building. This is very useful for buildings placed on hard grounds.


You need more water types and by that i mean water for fountains and/or rivers. A wave effect (texture) in areas near the shores would be very spectacular.

Car number is ok but cars with people in them driving are definitely a must. It is just weird to see cars moving around with nobody in them.

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #1 on: January 04, 2011, 06:56:01 am »
January 04, 2011, 06:56:01 am
People in more positions are a must. I would suggest people sitting on (chair) and layed down (on a bed let's say)


More work is needed on the import object settings. First of all there is a problem with the placement of a building on a spot. If you already have modeled the terrain in that spot the buildingwants to get on top of the terrain and is very hard to position on the right spot. I think that it would be way better that imported objects do not reaction to changes of the terrain. I mean when you place the building is should stick on the ground but when you make changes to the terrain it should stay at previous level. Many times when you try to build the terrain around the building it lifts up.


There are also problems with reloading materials after changing in sketchup (it does not always place the materials as it should). Also, it somehow changes the position of the building to where it was first. I think the software should recognise the new position of the imported building and replace the modified one in the same place, so you do not have to place it again after reloading.


The interior renders are ok, but sources of lightning should definetely be added. I don't know if it is possible with this engine but global illumination or some kind of v-ray like render for interiors should be added. If not other way possible it should be great just for static renders. This will not be a complete software until high quality interior renders are implemented. 

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #2 on: January 04, 2011, 06:56:51 am »
January 04, 2011, 06:56:51 am
The material queue for a imported object should be bigger as for interior renders you need lots and lots of materials.


A bug i've found is when saving after placing moving objects in a scene (people and cars) when you reload the scene and play the movie objects are messed up, trees are moving instead of cars or people etc, so you have to redo the movement all over again. This should definetely be fixed.


Some fishes would be great for underwater environements.


For effects i would add a bubble like effect for underwater scenes. Also, a must in my list would be the introduction of a text with a tv-like effect. One similar to windows movie maker would be great. If not, there is no problem, it still can be made in post production.


This would be great for on-going description of the project during the movie.


A basic fire material effect should be added for that great fireplace scene. Smoke may also be a nice addition.

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #3 on: January 04, 2011, 06:58:26 am »
January 04, 2011, 06:58:26 am
Tools for accurate terrain modelling are also a must. I have used you sistem with bitmap import and it was ok, but not very very acurate. There was however a slight modification in height as you have 255 shades of grey in most softwares and you should have 200 shades for each meter. A different sistem should be created. Keep in mind at all times that your software should be for architects, do not lose focus.


Architects do need accurate terrain (not artist, game like work), google earth import of large terrain areas would be great. In its current configuration the terrain maker is more aimed at gamers and it was hard to me to make something close to reality with it. The option to add CGIAR-CSI geospatial images from google would be great, as the plugin and google earth are free, everybody can use them easily.


A street maker should be made. It should be intuitive and easy to use. You should draw lines after the real autocad files (splines or straight lines on the ground) and then the software should stick them on the ground. If you push them up the software should automatically design a bridge or elevated highway. A city engine like approach would be the best in my opinion. (but it should be even easier to use) In my opinion this software should be used for creating ambiances around the buildings for visualisation and virtualisation. Therefore the proffesional design is made in other softwares like allplan, archicad, revit, etc.


What lumion should do is create quick, fast representations of ambiances around the central object that is the building. This is why other software failed, because architects should already learn very hard software and  learning gruesome softwares like 3Dmax just for creating the ambiance around the building is not something you should follow. Therefore this software should keep it simple, fun and easy to use. This attributes really can beat all competition because let's face it, after a lot of work on the building design who needs 2 weeks of work just to animate some people and trees and make the environment around the building... 

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #4 on: January 04, 2011, 06:58:57 am »
January 04, 2011, 06:58:57 am
Moving wheels for cars should be added. Also cars and people should have pathes (curved) to follow.

The cameras have a little time between each other, they do no flow one after another. Although this is not necesarily bad, there should be the option to make them flow one after another with no interruption.


An option to place a collection of plants in a brush like style would also be great. Anoying thing is also the placement of trees on top of each other when placing forests. There should be a option to stick trees only on the ground.


There should be a phisics system that recognise people and cars position and places them correctly, as far as i;ve seen if they move and the terrain changes height they stay in the air. There should definetely be interaction with the ground.


In the future, effects like rain, snow etc should be great additions, but they're not a must yet. Still they would make some very nice presentations. Also, painting snow, dirt, etc (or an automated process for this) on imported objects would be great.

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #5 on: January 04, 2011, 07:01:25 am »
January 04, 2011, 07:01:25 am
Exterior light, clouds, water effect, sky, sun and fog are simply perfect. I have no comment on them whatsoever. I do feel sorry for the interior lightning system though. Texture ilumination baking and such is such a no-no for me. At least an internal system for that should be added. Let's say for example i add light sources to interiors, and lumion calculates and applies the textures for illumination as required. This should get good results and stay in the "fun-to-work" area.


For the object library i would add flying vehicles, helis and planes. Also, more boats like large ships (they would be great in the background in ocean scenes) A nice effect in the future would be that boats recognise water surfaces and move slightly on the water. There should be some people just standing still and moving just a little. The animated people provided just keep moving like crazy people so you must put them toghether so they do not seem nuts but just talking to each other. (joking) 



Thank you so much for this wonderfull software and keep improving it. You are so close to perfection with this gem. Please keep up the good work and don't mess it up. Stay focused on the main purpose of this software that is architectural presentation, spectacular and simple to be made, fast and easy to render.


PS: the competition (twinmotion) totally sucks :) It doesn't even import my huge models, and lumion does it in seconds. Stay ahead of them and you will soon have the market. That should not even be very hard :)


My wishlist - conclusions regarding Lumion
« Reply #6 on: January 04, 2011, 09:28:07 am »
January 04, 2011, 09:28:07 am
We are already creating more water types. From ocean with shore to fountains.

 

We also want people in the cars as well as rotating tires.

 

Sitting people is indeed important.

 

You can already lock a building so that it doesn't lift up when painting the terrain!

 

We already have a reload that works like you describe and I am not sure if you used the update but that already fixes an issue we had with materials.

 

We will add lights but we won't add VRAY type renders. For that you need other software. We believe our fast renders are our niche if we try to get better quality then VRAY we our out of our league.

 

The update already has more material slots.

 

If you know how to reproduce the moving bug please let us know.

 

We are working on fishes right now.

 

We had a text overlay before but it wasn't good enough for release. We will bring this back this year of course.

 

We already are working on smoke and fire.

 

We where thinking of adding geotiff support but any advice on how architects get in correct landscape is welcome.

 

One problem we think we will have with roads and sidewalks is that we won't be able to recreate it like all possible combinations in the world. So we wonder how ussefull a road builder is since you can also do this in skechtup.

 

We are thinking of letting people paint a square and we just animate people in it. The same with cars.

 

There already is an option to place trees on our ground. This is done by pressing [G] while painting.

 

We will add rain and snow but it won't 'snow' on the ground just in the air for now.

 

We will keep adding content to our library and all suggestions are welcome. We don't know what people need but we are only interested in things more architects will need.

 

Thank you for trying Lumion and lets hope we stay focused :)

Aaron

    Reputation: 80
My wishlist - conclusions regarding Lumion
« Reply #7 on: January 04, 2011, 09:35:17 am »
January 04, 2011, 09:35:17 am
Trainz Simulator 2010 ED seems to have some nice features that share a high similarity with Lumion.

http://www.ts2010.com/videos.php

If a track/road editor like featured in their movies could be integrated into Lumion that would be very sweet.

My wishlist - conclusions regarding Lumion
« Reply #8 on: January 04, 2011, 11:08:54 am »
January 04, 2011, 11:08:54 am
Hi Gabriel, I just wanted to say thanks a lot for taking the time to post your detailed feedback on your first few weeks with Lumion. Posts like these are extremely important to us as the points made often confirm the concerns that we (and other users) have regarding the major issues which we would like to address in future updates.

Early adopters like you, Aaron, Simoni, Cpercer, Sliner and all the rest of you put up with the rough edges and motivate us to keep going with your valuable constructive criticism – and of course your cool entries in the user gallery.

We'll do our best to make your projects shine in 2011… and beyond Cool

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #9 on: January 05, 2011, 12:54:49 am »
January 05, 2011, 12:54:49 am
Yes, trainz simulator seems pretty close to what a road creator might do. I saw a image in their movie creating a railroad, that is the way it should work, simple &fast. Indeed roads are easy made in sketchup, archicad or allplan. BUT and there is a big BUT here, it is very hard to make them leveled to the ground, at least with the softwares i am used to work with.

Most of the time you make streets in 2D and the ground is flat, and that is fast and ok. You draw the street borders, the streets, etc. For example i have allready made settings for the heights and colors of street with textures and etc. After that i just draw in 2D the roads, street borders and give them all their heights. Allthrough this works well for flat terrain it al becomes tedious for leveled grounds. So in my opinion this is where lumion should come to help. 

So for the road maker i suggest the following (from my perspective as an architect). 

In the first place there should be a method of tracing the streets after the real ones from a dwg file of urban planning. Usually they come as a vectorised project or bitmap (jpg, tiff, etc) or they may be traced after images from google earth. I would love to have a layer for example in which to put the file with the road shape (be it vectorised (on scale) or not, this should be the two methods, and a scaling option for bitmap should be provided)

 

The true greatness of the software would be a procedural road. First of all let's face it. It IS VERY HARD to make intersections for roads with traditional software (and harder to make changes to them for that matter). In my opinion the roads do not necesarily need to be from different countryes, they should have just the bigger parameters implemented, and after that different texures and objects applied to them would make them look more american, european etc.

 

By main parameters i mean the section of the road. This is what urban planners use to define a road. This is definetely something you should take into consideration. I think what you have in mind is more like a puzzle combining the roads. Different types of roads at a size and you put them one after another like you do with fences or bushes now. I think this should not be the approach.

 

You should be able to control the main parameters of the road like: size of one lane, numbers of lane, size of sidewalk, size of middle ground (for large boulevards or highways.) Some basic aditional options should be provided like let's say if there is or not a safety belt for the highway, or other accesories on the side of the road. I would recomend 2 different types, one for roads (highways) outside of city and one for city roads. (because outside roads do not have sidewalks) What matters most is lane size, and sidewalk size. Borders and other accesories should be added automaticaly. A great relief would be adding automatic textures for borders (which are very hard to produce for curved roads, and I never really had time for this)

 

After you set the main parameters of the road you should be able to just draw a line on the terrain (curved or simple) and then the software does the rest with your desired section of the road. After that slight adjustements should be made possible from key points (to move a intersection for example) .Pedestrian crossings should be added where you need them by dragging a square.

 

The key is that after you make the road modeller you may give the apearance of let's say paris, or london, or new york with aditional objects that you place, in my opinion just the main parameters of the road (which are the same for all countries) should be automaticaly made by the software.

 

So in conclusion the following steps should be followed:

-a way to insert bitmap from google earth and/or vectorised scaled file (dwg) to draw road after

-a scripted simple road that follows section modifiable by the user

-section of the road easy to change (a image of the road section where you drag sidewalk and each lane of the road to desired length)

-roads that follow the level of the ground

-aditional objects to create ambient for american, european or custom road

-possibility to change road texures with the ones you desire

-simple way to draw markings on the roads (pedestrian crossing, priority, etc)

My wishlist - conclusions regarding Lumion
« Reply #10 on: January 05, 2011, 09:29:05 am »
January 05, 2011, 09:29:05 am
Thank you for your insight. We will be looking at roads at some point. We already had some road tests here but nothing we wanted to put into the release for now.

My wishlist - conclusions regarding Lumion
« Reply #11 on: January 05, 2011, 10:50:20 am »
January 05, 2011, 10:50:20 am
What you're describing is not impossible, but is certainly a massive undertaking for any company as infrastructure projects can be extremely complex.

Apart from the section-based "basic" roads you mentioned, throw in underpasses, overpasses, bridges, multi-lane roundabouts, all of which need to be able to join up seamlessly as well as working traffic lights from multiple regions and reasonably "intelligent" motorists and pedestrians and so on, and you've really got your work cut out! Smile

It sounds like you already have some experience with designated infrastructure design software such as InRoads, MXROAD, AutoDesk Civil 3D, Cadian, Carlson Civil etc. If so, it would be great if you could elaborate on specific reasons why the software you tend to use fails to meet your needs (e.g. you mentioned that it was hard to create intersections using "traditional software"), so that we have a better idea of how Lumion might fill that gap in the market place (assuming there is a gap we can fill). Is it to do with the ease of use, lack of core features or …?

Thanks again for your valuable feedback!

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #12 on: January 05, 2011, 11:04:46 am »
January 05, 2011, 11:04:46 am
As for the terrain the building is on, it usually comes as a tiff or pdf file with level curves or height points. The software i am using lets me insert height points one at a time, specifying their height. Then the software produces a triangulated surface of the land you introduced. While this would be hard and pointless for lumion to do, what is should do is continuing from there. 


For example i make the house and the land on the property, but let's say i make a presentation with all surrounding area, ocean, mountains etc. Therefore i use the ground material from lumion for the layer of the terrain parcel the house is on and then i need to integrate it seamlessly into the big terrain provided by lumion. (the terrain area required for a aerial presentation may be as big as 10x10 km, as the horizon line is at 8 km in real world)

 

Now there are 2 options. I either make the big terrain myself or I import it from a real source to be as realistic as possible. As far as i'seen there are 2 available possibilities. One is importing data from SRTM here http://srtm.csi.cgiar.org/SELE.....tCoord.asp They should provide satellite height imagery, but I really do not know how to use the tiffs there, all I have obtained are B&W images of the region. 

 

The other option which is way simpler to use is importing terrain in sketchup, they have a cool option there. It would be great if lumion could take that terrain... exported as collada from sketchup. After you import the big chunk of terrain into lumion you may import your model with his own parcel and integrate it seamlessly into the big terrain took from sketchup.  There is however a limitation as you may only import maximum 2 km x 2km squares as far as i've seen. The results are good as far as i've seen and maybe you can stick together more of them squares.

 

The lumion import bitmap feature is good, but is better for computer gaming and such, where realistic land is not necesarily a must. It is way more creative, but is hard to produce realistic terrain with it, although i have managed to do it (but hard and not so precise, and needed more afterwork in lumion). It should import collada files in my opinion and transform them to lumion terrain.

Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #13 on: January 05, 2011, 11:44:50 am »
January 05, 2011, 11:44:50 am
Morten said:


What you're describing is not impossible, but is certainly a massive undertaking for any company as infrastructure projects can be extremely complex.

Apart from the section-based "basic" roads you mentioned, throw in underpasses, overpasses, bridges, multi-lane roundabouts, all of which need to be able to join up seamlessly as well as working traffic lights from multiple regions and reasonably "intelligent" motorists and pedestrians and so on, and you've really got your work cut out! Smile

It sounds like you already have some experience with designated infrastructure design software such as InRoads, MXROAD, AutoDesk Civil 3D, Cadian, Carlson Civil etc. If so, it would be great if you could elaborate on specific reasons why the software you tend to use fails to meet your needs (e.g. you mentioned that it was hard to create intersections using "traditional software"), so that we have a better idea of how Lumion might fill that gap in the market place (assuming there is a gap we can fill). Is it to do with the ease of use, lack of core features or …?

Thanks again for your valuable feedback!


Well, i'd say what you described is waaaay out of what i had in mind. It would be incredibly advanced and wonderful if you could realise something like that but it sounds like a huge amount of work. No, actually i had way simpler things in mind.

I am an architect and i do not have experience with the software you mentioned. I have 10 years experience with Allplan & Autocad. I just played a couple of times with the demo of city engine, a good software but pretty useless for me so far. Also, hard to use to get desired results. You may see here how their road maker functions http://www.procedural.com/city.....reets.html Their sistem is however pretty advanced. I think something way simpler will do for now.

 

I will post some example of my work, to show you how traditional architectural software may work and look like. Right now, i make the streets as you see in the following pictures and import them in lumion together with the building. After that I make the terrain around the street to fully integrate it. Problems arise for non-flatten terrains… where roads as i made here (simple designs) are a tough nut to crack…


 






Gabriel

    Reputation: 0
My wishlist - conclusions regarding Lumion
« Reply #14 on: January 05, 2011, 12:11:56 pm »
January 05, 2011, 12:11:56 pm
This is the result in 3D. In my opinion streets like these are more than enough, if they have the option to "stick" to the ground. In this example however, the street that goes to the building should've rised one and a half meter. As this task would've been tedious to manualy do in Allplan, i just made it flat. The side terrain as you see is terraced, and lumion's terrain should mold to that.

 

As for what you said earlier with cars and people in a box... I dunno.. look at my example... the sidewalks are curved... I really would need a curved path for each person walking. Not even in the park you may not put a square with persons... as you may see there is a fountain (in the circle) and it would be unwise to have people walking through the fountain. Also, people appearing out of nowhere in smaler boxes would also be unrealistic.

 

So in this case a box would be a total no-no.  I say the sistem is ok as it is, it just needs a curved path for persons (which shows when you chose the move object effect) After all there is no need for thousands of persons, i think for most scenes 20-30 persons are more than enough in my opinion, unless you use to design subay stations at rush hours :))))))