The next step for my Last Supper project was to add hair to the models. I have very little experience with modeling/rendering hair in Blender, so this step was also more involved than I originally anticipated. My intent was to create a rough draft of the hair, with another iteration of styling/coloring once the near-final lighting for the scene is complete. So, from that point of view, I was successful, but I still think the current result looks like a toupee factory exploded. :)
The project to 3D print the mission in San Luis Obispo worked out pretty well, so I decided to create a model of my wife's childhood home. Just to keep things interesting, this time I decided I would make each floor a separately printed piece so they could be added and/or removed after production. Other than having to model the interior of the building, I didn't think it would add additional complexity. You know what they say, 'Go big or go home'...although apparently they also say, 'Ignorance is bliss!'. :)
I'm still pretty new to 3D printing so to me it is still magic. Create something electronically, send it off to a web site and in a week or so I can hold my creation in my hand...that's magic. This project was to recreate the historic mission in San Luis Obispo.
The good news about about having to scrap and rework everything you have just done is that, eventually, you get fairly quick at whatever you are doing. In this case, I have gotten much quicker at creating a model in Makehuman, exporting/importing to Blender and posing. :) This week, I was able to create initial clothing and re-pose each of the characters. Some mistakes I made were part of the learning curve; others still have me scratching my head. As someone once said, 'All movement is progress!'
I used MakeHuman to create Jesus and the disciples. There was a small delay after installing MakeHuman, apparently due to a conflict with python libraries. I found that if I ran MakeHuman with Administrator rights, I avoided whatever the problem was. The learning curve was fairly steep but at the end I created a custom character, fully rigged and imported into blender in about 20 minutes. Interestingly, with Blender set to imperial units, I needed to export the MakeHuman models with the units set to 'meters'. Below is the current scene with bodies roughly positioned. (CAUTION: Partial Nudity - May not be appropriate for children under 53 years. ;) ).
I dual-boot my computer with Windows/Linux because there are still a few programs that I haven't gotten working on Windows (and vice versa). This week, I took the plunge and installed OpenFOAM (in a later post) and MakeHuman. I had been nervous about installing MakeHuman because of the python warning on the web site.
I'm sure some configurations will cause problems, but in my case, I haven't experienced any problems. Once I finished, I realized 'Installation' is not accurate. Download the zipped files from the web site and extract them wherever you want--no installation is actually required. The entire process takes less than 5 minutes (YMMV).
I get a chuckle out of the nudity warning. In our litigious society, I'm sure there is a reason for the warning; it still makes me laugh.
The Image Of The Month (IOTM) for January was based on the painting by Robert Finale, Home For Christmas. It seemed like a good idea on January 1st. In retrospect, I could have started with a month or three of paintings not quite as complex. However, the good thing about challenging yourself is that even if you only get halfway there, you are still making progress.
I borrowed heavily from the last two vehicles for inspiration, but the air stair chassis was created new this week. Unlike the luggage conveyor, I used the Subdivision Surface modifier on the chassis for the air stair which helped the model look cleaner, particularly around the wheel wells. Normally, I start from the ground up. This time, I got the stair extension/retraction mechanism working first before attaching it to a vehicle.
It is always difficult to get going in January, so I am going to consider the week of 12 January the first week, for this month's Image Of The Month (IOTM); unfortunately the deadline is still the end of the month.
The first week consisted of framing the shot and attempting to match the artist's 'camera' and perspective as closely as possible. I ended up cutting the people out of the original and pasting them into a single file. This allowed me to scale each person/group so they were relatively equal, which, in turn, allowed me to export the images as separate files and import them into Blender. Once their size was established, the only way to visually scale the people larger/smaller was to move them nearer/farther from the camera. This was very helpful in establishing the characters', and associated buildings/props, relative distance from the camera. Using that information, I created rough approximations of the major objects in the scene.
To the left is the scene setup. The near and far houses, church, large trees and street lamps are represented. The bridge, being a major component in the scene, has the most detail. A screen is set up to represent the background, but textures have not been applied to any items, at this point. I spent a considerable amount of time, playing with the lens position and size to perfectly match the artist. Eventually, I realized part of the beauty of art is that, although it is quite good, it is not perfect. I am mentally coming to grips with the fact that 'close enough' is the goal. :)
To the right is an initial render. You can see the images of the people that were imported to help with relative positioning. The large 'mountain' of snow' in the background is the cylinder that will eventually contain the background texture. I may use terrain and trees to form the background, but I suspect I will need at least a week tweaking the lighting and materials, leaving me very short on time for modeling. The good news is my Model Of The Week is providing good lessons, allowing me to finish early. That should free up a little extra time for this project. If I stay on schedule, the modeling will be finished by the end of this week. That will leave me a full week to tweak materials, lighting, camera and the worst modeling SNAFU's. What could possibly go wrong? :)
MOTW history was made this week (a full two weeks into the challenge): the model was finished on Thursday. The MOTW followed the 'airport support vehicle' theme that started last week. This will be the theme for a few more weeks until all the support vehicles are one big happy family. This week was a tug and luggage carts that form the train responsible for getting your luggage from the terminal to the airplane.
After struggling to model the wheel wells last week, I decided a Subdivision Surface modifier is a requirement for nearly all vehicle models. That worked reasonably well this week, but trying to keep edge loops intact throughout the model significantly increased my vertex count. There are plenty of mistakes left to make, but I am getting better at linking assets during the creation but locally saving everything (assets, materials, images, etc.) packed in the final file. The important Lessons Learned this week:
- Subdivision Surface modifier is key for vehicles (see above)
- Without blueprints, or at least a 3-view drawing, it is easy for relative dimensions to 'adjust' during modeling. My 'need a human figure for reference' comment from last week still applies.
- Noise textures are great! I've started using them in almost every shader. Over the next couple of weeks, I expect to learn the lesson, 'Noise textures can be overused!' :)
The first several Models Of The Week (MOTW) will have an airport support equipment theme; this week's MOTW is a luggage conveyor.
Day 1: Spent the day finding reference photos and dimensions from the internet.
Update Day 2: Rough modeled supporting assets (wheels, tires, etc.) and the vehicle itself. I had vague memories of spinning curves to create objects. That must have been a Maya memory, because Blender will only spin a mesh.
Similar to the Model Of The Week (MOTW), the Image Of The Month (IOTM) is a self-inflicted challenge to improve my 3D skills. In the IOTM, the challenge is to create a 3D replica of a famous painting each month. Still in the Christmas spirit, I selected a painting by Robert Finale, Home For Christmas.
My intent is to provide weekly updates on the work in-progress.
One of my many challenges in 3D modeling is knowing when to stop. I started creating in 3D as a hobby, and endless tweaking wasn't a significant drawback. But when working with teams and/or with a customer, being predictable and staying on schedule is vital. To retrain this particular bad habits, I created several challenges for myself. Through this MOTW challenge, I will attempt to create a (admittedly simple) model each week; due by midnight Friday. In theory, I will choose a wide variety of models to create. In reality, the first few months appear to be leaning toward mechanical objects with which I am most comfortable. We'll see what the rest of the year brings.
I will post images of the results here. I've already started an informal pool to guess how many models I create within the specified schedule. Feel free to join the pool. The winner of the pool will likely win something terribly exciting, like a warm feeling of superiority. Comments and suggestions are always welcome.