hrontos

For me the digital and real entries are equal. I do not like when competition is about size of someone's budget. MINI contest was good because the part limit was low so virtually anybody could participate. So for me it is. completely ok if potential C-model would allow digital entries. EB stuff can decide if it will be theme month or a competition. It will be fun and challenge anyway. And again a good source of nice models for those who own the set. I somehow understand that in traditional competitions the digital entries are not allowed as stock of digital bricks is not limited and gives unlimited possibilities. But in C-model competition parts are given so digital entries have no advantage.

POV-Ray has some feature that protects some files from being read. It can be deactivated. If it is not that case, it would be good to share more details as it quite difficult to guess what can be the reason for such messages. Is it possible that there are some transparent parts in the area being rendered at 47%? They are really slow to render.

Congratulations to the winners. Well deserved. It was pleasure to participate and we got many inspirational models that we can build without any big budget. Than you also to you Jim for smooth organization and really quickly published results.

If I remember properly, even it was already mentioned in that entrry topic.. Creator mentioned, that he removed 3 parts since it was published.

This is also a good idea. The C-model idea is interesting for me mainly because it somehow pushes people from being a collector to being a creator. And shows that also with 200 parts model there are many possibilities. I remember a topic where somebody made 10 C-models out of one roughly 200 parts set. I wonder if whole comunity could cover whole aplhabet.

Now you got me. I hope exponential format will get you. Like 0.005E3.

And use some international decimal separator like comma instead of dot so that Jim will have to make full international suppport in his script.

Your expectations are standard, as this should be a simple task these days. It is ok to ask about it when a man cannot find correct menu item in the application. Your question is good and legit. It was just not clear, if you want to use custom decals directly in LDD or you would like to do some pictures (renderings) of your models with custom decals. Many forum members asked this question and most of them want custom decals directly in the LDD. Some of them even presented links to the internet with ideas how to do it. As Calabar said, no answer was meant to accuse you. Technically it is possible to add things to LDD. But on this forum it was decided that even it is technically possible, it will not be promoted or supported, since it is a violation of LDD EULA. If you found my answers too strict or offending, please accept my apologies.

The big difference between LDD and any other 3D software is that in any 3D software you can add your textures and materials and even model objects. In LDD not. It comes with library of objects, materials and textures and you cannot extend it. He clearly asked how to add it in LDD (I assumed that he did it based on some experience with other 3D software). And that's why the answer was that it is not possible without modification of the application's library which is considered violation of EULA. There are many people asking how to add some missing part or decal so as a question it is quite common and answer is always that it is not allowed. When this question is asked nobody assumes that people are looking for some 3rd party tools to add decals to their model images. Everybody want to have it in model (LXF) file, use it during building proces, share it with friends etc.

You do not need such high tipping angle, so try shorter beams. You will get smaller tipping angle, but more force.

F1 voting system or any similar voting system is strongly bound to the nature of the competition. They are not random nicely looking numbers. They are designed to somehow express how difficult it is to achieve a certain place. In F1 you get 25 points for winning, only 15 points for being third. But value of 15 was chosen, because somebody found out that it is more difficult and therefore more valuable to be twice 3rd than once 1st and second time not to score at all. For technic contest to find such numbers can be quite tricky and therefore I like the freedom of the standard voting we are used to.

If you want make renders of LDD models start with LDD to POV-Ray. Then follow the tutorial about custom decals on LDD renders.

There are so many great entries in this contest. Since the size of the models is within the range my kids can build, we were watching the entries quite with interest during whole contest. And so I have decided to let them vote. Both of them were given 8 points and this is how they allocated them: 2: 1 4: 1 6: 1 7: 1 16: 1 19: 1 28: 2 40: 2 42: 4 47: 1 55: 1 3: 1 19: 1 29: 1 31: 1 I was quite satisfied with their selection. Only 4 points left for me so I added my last 4 lines. Thank you for organizing this contest, it was fun for whole family. I especially like models with LXF files, since we can build them.

This is great. I like how you make your C-models. Making a C-model reminds me the time 25 years ago. In former Czechoslovakia LEGO was something hard to get, so 600+ pieces Technic model like 8860 was a huge amount of bricks and the creations a man was able to create only from this set were a real challenge comparing to these days when you can order any parts you wish.

There two global settings: max_trace_level - it is number of surfaces light ray can hit during tracing. When light is not reached within this limit, pixel will be black. Higher might be better, but takes longer. adc_bailout - minimal amount of change each reflection must bring - when change in computed pixel color is less than this value, tracing of the pixel is stopped and pixel is drawn - each pixel can have 256 levels of RGB components so minimum change is 1/256 = 0,0039, value of 0.01/2 = 0,005 is a bit more. Higher value means earlier stopping of tracing and less colorfull image. Sometimes the max_trace_level is useless, if changes in colors are minimal - adc_bailout helps to end before max_trace_level is reached, to speed up rendering while keeping possibility of deep tracing some pixels when it makes sense. Try to render without conventional lights. Only HDR and your 2 eyes lights. HDR can be made brighter for example by changing assumed gamma of the HDR. image_map { hdr "foo.hdr" gamma 1.8 } The key feature of the HDR format is the amount of image info you have there. HDR contains much more lighting levels than standard PNG. HDR much more behaves like human eye - for light parts of the scene eye consumes less light, for dark part more light. Standard digital camera takes some average and this is stored in JPG or PNG. As a result, dark part of the image is not lit enough, light part seems to be overexposed. With HDR you can nicely make a picture of a person standing in a room in front of the window. You will get all details on the face. With JPG or PNG, face would be too dark, window too bright. When HDR is used as light source, light source seen on the image really behave as light sources on your scene.

It looks like your radiosity settings are not good enough for this image. Skyspheres showing just skies are not so sensitive to radosity settings, that's why your render is more or less ok. But the first skysphere contains many colors and light sources and that makes it more sensitive to errors and quality settings. Also your global tresholds adjustments may help, since chrome is very reflective and require higher max. trace level.

I am just not sure how to tell people, which standard model to buy, without telling them too early. May be a color schema match would not be obligatory to allow building from in-house bricks or digital entries acceptance would resolve this.

I think the key reason for longer period are the Bricklink orders. Since it is necessary to build the model physically in the right colors, people are using BL to get the parts and it takes some time. It is not that much about the size of the build, missing micro model part will arrive after the same number of days as complete flagship. Of course, it probability of missing part on micro build is lower. It would be nice to have some speed building contest - something like 24hours contest. What can you build in 24 hours from 200 parts. Or what C-model can you design from parts of some standard model in 24hours.

It is may be not that important, if we use "F1" system or traditional "distribute 10 or 20" system. Statistically with that many voters results will show nicely using both systems. Distribute 10 or 20 seems to be more flexible, since one can decide if he clearly selects his personal winner or rewards more creations. Looking at the voting results, it looks like it is difficult for people to decide and so they reward more creations. "Distribute 10 or 20" also somehow eliminates slightly voting for friends or more famous builders, since you are not forced to decide which is the best. When somebody is forced to decide, he will do it based on many factors, not only model itself. Since many people cannot decide and they distribute the points among many entries, it looks like it is more important to eliminate huge difference between min. and max. number of allocated points. The reason is simple: when difference is too big, one "decisive" voter may actually completely change results of contest by assigning max. number of points. Taking into account these points, "distribute 10" seems to be best for me, may be with higher minimum (like 2), to force people to choose max. 5 entries. So even entry on the 5th place would get 2 points. This is also 2.5times less points than the 1st place, which is the same as in F1 (25 for 1st and 10 for 5th).

You can try to define custom values for definitions from ldd_finishes.inc, which allow you to define material properties. Position of lights in POV-Ray are absolute in scene. But LDD2POVRay use lights relative to camera. That's why there is "transform { ldd_camera_transformation }". Lights are moving with camera. Yes, it is possible to put light into object. Each brick in POV file contains 12 coordinates which represent transformation from origin. The same transformation can be used for lights. See LDD2POVRay talk for more info.

LDD is closed source SW, you cannot add anything. Moreover it would be a violation of EULA.

Yes, I like the design, so I will try it. Thank you for sharing good ideas.

Since you already managed to wait for the first pass(collection of the radiosity data) to finish, the second pass (rendering) will not be faster without radiosity. Actually rendering takes almost the same time with or without radiosity, collection of radiosity data is the extra time. And renderings with radiosity look better, especialy with some skysphere lighting image.

LDD2POVRay by default uses radiosity which implies two pass rendering. That's may be why you see 70-75%, then start from 0%. Or may be I am completely wrong and it would be useful to see your settings.

This looks as a very nice and compact solution. The only thing I am not sure about is the bracing of the 8 and 16 tooth gears. Will two half bushes and weight of the vehicle be enough to prevent gear skipping?