Model of Venice

[Wildsided]

4 minutes ago, robcat2075 said:

Bitmap Plus does have a "normal" option

Learn something new everyday. 

[Pitcher]

Wildsided, you have nothing to be sorry about. I just don't have nearly enough background to understand what is probably simple for a lot of people. My level of training is that I'm mainly self-taught, except for the many times, I have been totally stumped and had to call for help from Rodney, robcat2075, fuchur, Fae Alba, and many others who commented on questions I raised. I think I probably need to sit down and go through a bunch of tutorials. The only thing about that is that I forget what I learned unless I apply it almost immediately to one of my own projects. Thanks for trying to help me. It sounds like you have a great method, and I really liked your example.

 

[Pitcher]

I guess my level is needing to know what button to press, what number to put in as a setting, and in what order. I am much better at figuring out what is going on after I see what the process is or if someone explains what the computer is trying to do. I have learned about computers by operating them for about 25 or 30 years. I haven't taken classes in computers, except for short in-service classes on how to operate certain applications.

 

[robcat2075]

Give me a chance to sleep and we can look at the options for some canal water. 

[Pitcher]

Thanks. I'll look forward to it. This is a real treat for me to learn how to jazz up my pictures.

[robcat2075]

If your water plane is a grid with fairly square patches, the easiest way to use the normal map is to select it in the modeler>RMB >Add Image and choose the normal map from the drop down list.

This creates a Group of your selection and puts a copy of the image on every patch in the Group. (This is called a "Patch Image")

In the properties for that image set Type to "Normal"

If the image is facing the wrong way you can select the Group>RMB>Rotate Images to turn it in 90° increments, or you can rotate your water plane.

If the pattern is too large, you can shrink it in multiple by setting the "Repeat" values to 2,3,4...

Or you can remodel your water plane to have smaller patches and do the patch image over.

[Pitcher]

Here is a picture of the water plane. It is a group called Water in my Land and Canals model. I'm not sure what you mean by RMB. My image is in the Decal folder.

I think I figured out what you were directing. I am now waiting for it to render so I can see how it worked. I'm not sure if I put enough repeats or too many. It's still rendering, but I can see it now. My grid is so uneven that the seamless water texture is extremely different in the small canal than it is in the Grand Canal. This won't work as an even distribution.

 

[robcat2075]

2 hours ago, Pitcher said:

My grid is so uneven that the seamless water texture is extremely different in the small canal than it is in the Grand Canal. This won't work as an even distribution.

 

I figured you had just a simple plane underneath the land areas, so...

Anther solution is to decal it once on that whole mesh (not the Patch image method), then use the repeat values to skinny the tiles down. I'm guessing X and Y repeats in the range of 10-30 might do.

 

[Pitcher]

I'll have to give it a rest for a little while. I tried two or three different ways today. I have another way in mind that might even out the mesh, but it is rather work intensive. I think I'll take a break for a day or two and then go at it again. Next time when I have something like this to model, I have an idea how to do it where I will have a more even grid for the water. I think the patch image approach would work very well in an ocean scene, but if there is land and canals or rivers, the modeling needs to anticipate that type of approach. Thanks very much for your ideas and help. I think that the picture has improved immensely.

[robcat2075]

Tell  us when you are back!  There are other ways we haven't tried yet.

[Pitcher]

Hi! I restructured my model of the land and canals by cutting away the water and making a plane with an even grid to intersect the new model. Originally, my model was 2 level on the large canals and 1 level on the small canals. My restructure made the entire land model 2 level so that each canal would have banks. After doing that, I tried robcat's approach again with the patch image. I tried a little reflectivity and a little transparency. I left the transparency, but the reflectivity was a little too much, so I removed it. Here is the picture:

 

 

[robcat2075]

You won't need the transparency. At this distance the canal water is pretty much opaque. You could set the surface color of the water to a dark greenish brown as a first gambit.

But you do need the reflection. That is most of what makes water look like water.

You can reduce your testing time by turning off the AO and going back to conventional lighting long enough to see if your reflections are looking watery

[Pitcher]

I tried the reflection at 2% and then at 1%. With all the pattern of the water and the reflections of the building, it looked like a confused mess to me. I stopped part way through the renders. I'll go ahead and do a complete render with 1% and show it to you. I used a very dark color as the base color for the water (0,0,128). Whenever I use this decal, it gets light like this. I'm attaching a screenshot showing what the picture looks like on the Shaded setting.

[robcat2075]

Here is a shot of Venice in "From Russia with Love" that is very similar to your scene's view point. Notice the refection of the buildings near the water's edge.

For now, take the normal map off and see if you can find a reflection amount that approximates that amount of building seen in the water.  I'm guessing the reflection level is somewhere between 25-50%.

Notice how the amount of reflection decrease as we look more directly into the water at the bottom of the frame. Later on we'll make a gradient that does that for us.

 

 

[Pitcher]

Here is the picture with no transparency and 1% reflectivity and an even water pattern.

[robcat2075]

We must be talking about some other parameter because 1% reflectivity is next to no reflectivity at all, and there is way more than no reflectivity there.

I am also doubtful that the normal map has been set to type "Normal" . It should be distorting the reflections like waves would, but the reflections are still undistorted. notice how the lines of the columns are still straight.

[Pitcher]

I'm sure you are right. I will check all of that. It seems that no matter how many times I try to save the settings, they go back to default. I know that when I did the patch image, I set the patch to Normal. I checked about Normal again and found that it had gone back to Color. I also know that on the properties of the water group I set the reflectivity to 1%. There seems to be more than one place to make settings sometimes, so maybe I am making my settings in the wrong place. Also, on the reflectivity, I have been reading lots of science books. According to them the amount of reflectivity has to do with the position of the light source, the position of the viewer, the position of the object, the intensity of the light source, and whether the viewer has his/her eyes open.

I just started the render again at Normal and this time with 1%, and it was drastically different. I now have turned up reflectivity to 35%. We'll see what that looks like. I'll post it here in a few minutes. I think this is the best yet. I probably could increase the reflectivity a bit, but the water finally looks more like the real water. Thanks for your help, robcat2075!

 

 

[robcat2075]

4 hours ago, Pitcher said:

Also, on the reflectivity, I have been reading lots of science books. According to them the amount of reflectivity has to do with the position of the light source, the position of the viewer, the position of the object, the intensity of the light source, and whether the viewer has his/her eyes open.

🙂

That is true. We can create the change of reflectivity due to the "angle of incidence" with an "edge gradient" material. The rest we can do with judicious color and lighting choices.

 

On one of my contest entries I used an edge gradient to make the reflection of the horizon line stronger at shallower viewing angles...

[Pitcher]

robcat2075, you are always ready for anything. I really like your explanation and contest picture, and I learned something.

[robcat2075]

If you look at the movie pic I posted above you can see that the dark parts of the waves are where the surface faces us more and has less reflection so we mostly see the color of the water.

The lighter parts are where the surface is tilted away from us and reflects the blue sky and buildings more.

[Pitcher]

I deleted a picture here accidentally and no longer have it.

 

 

 

 

[robcat2075]

👍

There is probably a Material solution we can pursue to get a less striped appearance.

However, at this distance and scale, representing the individual stones may not be what we need.

What do you have for reference for those areas?

 

[Pitcher]

I don't understand what you mean by "reference for those areas". I really don't care about having individual cobblestones from this distance, but when I move closer, cobblestones would be nice. I always have the idea of using the same models for close-up and distance shots like you would if you were in a real life space. I agree that the stripes are not ideal.

 

[Pitcher]

I'm trying it with a different texture image. I'll post it here when it renders.

 

 

 

[robcat2075]

"Reference" would be any pictures you are using as examples and models to follow

[Pitcher]

I used the following:

 

[robcat2075]

Do you have some pictures from Venice of these stones and streets?

 

[Pitcher]

Here are some pictures, but they are not very close-up. You can also see the painting on page 1 of this topic.

[robcat2075]

It looks more like tiles than cobble stones

[Pitcher]

You're right. It's also like a mosaic. I was just trying the cobblestones to see if it might look old. It seems that even in 1600 or so it might have had large tiles laid out with geometrical designs. It always has to be complicated. I just found some charcoal gray tile. I'll post it here when it renders. I probably should note that the reason I am adding a land texture is really focused on the views inside St. Mark's Square. The texture will spill out into the view from the Grand Canal, however.

I did the tile. It looks almost like the land I made with no texture. It's not really worth the bother of rendering. In the first couple of passes, about the only difference was some striping due to the uneven color in the texture image.

I'm making this note sometime later. The tile is really starting to "grow on me" in the picture of St. Mark's Square from the Grand Canal. Later in this strand, robcat2075 says something about using a different strategy in large spaces and small spaces. I might start thinking about using the tile for this view and the cobblestones for the view inside St. Mark's Square. I really like the pattern of the texture inside the Square, but just the hint of the stripes might be enough pattern for the picture below (and it is a little lighter and shows the shadows better). I'll have to consider whether that is the kind of thing artists do, or is that just wild and crazy?

 

[Pitcher]

Here is the version in the square with regular lighting and cobblestones:

 

 

[robcat2075]

I like the diagonal look.

It will probably be necessary to pursue separate strategies for near and far to get favorable results with either.

[Pitcher]

What is a strategy?

 

[robcat2075]

strategy... a plan for getting something done.

[Pitcher]

I guess it's no surprise, but other than producing some semblance of Venice that looks like a reasonable cartoon background, I don't really have a strategy. I try to make buildings that have the general form of the originals. I put them together on a model of land and water in a similar way to how they appear in reality (although I understand that there are many, many concessions to limits of my skills and computer capacity. Then, I make adjustments until I can't stand to look at the model anymore. Then I leave it alone for a long time and probably come back to it and make many more adjustments after I have learned a few more techniques. Since this is my hobby, I don't have a lot of time pressures. My projects get a little more complex as I go along. I started out with a one road town with buildings that are similar to those found in Brenham, Texas. There were other small towns. Then, I made an imaginary modern (but not completely modern) city. Then, I made the Giza plateau (simplified), Medieval Athens (not really authentic, but more authentic than a normal cartoon), and now Venice. When you said "strategy", I thought you had in mind some sort of approach learned in art school or computer school.  Other than art lessons I had at the Museum of Fine Arts-Houston as a prize for winning fifth place in an art contest in fifth grade, I never took art. (It occurs to me that I did have a semester of art history in college.) Other than teacher in-service on how to do a few things on the computer, I never took a computer science class. I guess my "strategy" is trial and error, with a huge emphasis on error.

[robcat2075]

I just meant strategy in the most forward-looking terms. You don't have to have it all figured out yet.  🙂

[pixelplucker]

Background you can use an image. I like the model you have so far. Looks great.

[Pitcher]

When I started out in using Animation:Master, part of it was because I liked the movie Shrek and wanted to learn how to do 3-D animation. Another part of my motivation was making a movie to tell a story. I can't afford to hire actors, buy cameras, etc., so I thought it would be great if I learned how to build the actors, sets, and props and use the cameras and lights in the software to make a movie. One of the things I really like about 3-D is building the environment or set and then using it like a real location. I have experimented with using flat pictures in a rotoscope fashion, as background but as I moved my characters, the background did not look real (in the sense that it looked like a flat picture background). I took a small free course in visual fx offered by Hitfilm (about 6 lessons) online, and one of the techniques was to place what looked like cutouts in different locations in a 3-D environment and project a 3-D motion visual onto the flat cutouts. That seemed to give more of a 3-D feel to the movie, but there seemed to be quite a bit of set-up required to make it work properly. I thought that building the 3-D environment itself might be better in the long run. 

Once I have the environment, I try to place my characters, cameras, and lights in a variety of ways to see how it looks. I guess I have a poor imagination, because I cannot figure out how it's going to look until I do all of this. I have a general idea beforehand, and I may even have sketches, but without moving characters, lights, and the camera into various positions, I don't know which I'm going to like the best. The pictures I have made so far to illustrate my books are in choreography files, and can be used as starting points for camera and character movement and zooming in or out and panning and various transitions. I hope to one day, if I live long enough, go back to my choreography files, improve the quality of pictures (even though they might still have a cartoon style) and create animated movies. I started trying to animate my own stories so that I would have some content to provide motivation and experience that would call forth the need to learn more technique. All of this is to explain that although I am very slow in getting to know a lot of the ins and outs of the animation program, I do have a general strategy in mind for accomplishing a goal.

I really appreciate the huge amount and quality of help I have received on the Forum, and although I might not use every lesson in helping me to approach my goal directly, I am learning techniques that I will be able to apply in some small way to my project. For example, I have admired for years the way artists use texturing, and I am very impressed with how closely 3-D artists can simulate reality, especially in special effects in movies. I have been very glad to learn what I have about using textures, even though the style of my work is mainly cartoons. For example, I made a picture of dirt and rocks that had fallen down a cliff. The dirt did not really look like dirt enough to communicate what it was. The rocks did not really look like rocks. I plan to use some dirt texture and some rock texture on those items in my picture even though the rest will remain cartoony. I may even do something to the texture image in Photoshop that pulls it back slightly from hard reality to make it communicate substance without looking photographic. It's just a style choice for these projects, although, as I said before, I really admire the ability of 3-D artists "to make it look real."

[Pitcher]

Thanks, pixelplucker!

[robcat2075]

Let me get a few desperately late projects behind me and I'll be able to take a close look at some solutions for you. 🙂

[Pitcher]

Thanks, robcat2075. I'll look forward to learning more. I'll go ahead and leave a question. For whatever reason, the program had the cobblestone pattern going one way to the left of the canal that the Bridge of Sighs spans and another way on the right side. I could not rotate the pattern to the right until I split it off from the one on the left and rotated it. I gave both the same properties, but for whatever reason the texture on the left ends up much darker than the pattern on the right. Here is the example I am describing:

 

 

 

[robcat2075]

Can you bring this problem to Live Answer Time?

[Pitcher]

Done.

[robcat2075]

1 hour ago, Pitcher said:

Done.

Tomorrow at Noon?

[Pitcher]

I'm not sure where I'll be on Saturday at noon. I'll try to be there, but my schedule is very fluid. That's one reason why I leave messages usually, rather than chat.

 

[robcat2075]

I'm not sure what "done" is, then. 😀

[Pitcher]

I posted it on the link you provided. Here is where it is:

 

[robcat2075]

oops, sorry, I thought today was Friday!

I meant Noon on Saturday... "Live Answer Time"

[Pitcher]

robcat2075, don't worry, I understood what you meant. The problem with the picture is solved now, and I know that flipping normals is important for 3-D printing and texturing, too. I guess next time, I'll try flipping normals as one of my "go-tos" for solving problems. It will be like rebooting a computer. Thanks for all your advice in how to use the patch image method and how to rotate images, and so many other things. Thanks to Fuchur again also. If you haven't looked at the 3-D printing page in a while, I printed the model of the Sphinx at the library one more time and took pictures. There still are some imperfections, but I think it turned out a lot better. I also made significant changes to the model. Have a good weekend!

 

[robcat2075]

Is the librarian tolerating you better now?