To get it working:
- Make sure you have quicktime installed. If not get it from Apple
- Then you need to enable quicktime in Vegas. It's under Perferences->Deprecated Features
Deprecated, my ass :-0
I feel like I've been here before. Vegas refusing to play Mov files. OK Mov is an old format that Apple doesn't promote anymore. But I have about 6Gb of alpha blended explosions and blood splats in that format.
To get it working:
Deprecated, my ass :-0
I've been using Eevee pretty much since 2.8 came out. Turns out when I started using it, it made some assumptions that were wrong about how shaders needed to be set up. Yesterday I finally got irritated enough to figure out why my renders always looked a bit off.
I had 2 significant problems:
I was using DX normal maps when Blender uses OpenGL format. That's bad since they use different coordinate systems.
I was still feeding the Metallic and Roughness in using sRGB as the colour mode. I found out that actually they should be set to nonColor, just like the Normal Map.
So now I have to go back through all the materials I created in this project and fix them. Because I was lazy and didn't set this stuff up properly in the first place. There might be a lesson there somewhere....
3 weeks with an Ender 3 Pro, printing model and miniature related content. What did I learn?
Quality of the Prints
I tried the Ender with a .4 and a .2 nozzle. The .2 nozzle and a small layer height will give the best quality. However even at the highest levels layer lines are still visible. This basically means if you are making miniatures for display, it isn't good enough (in my opinion). If you just want models for a game and don't care too much about quality then may be its OK. Printing at high quality is very slow.
Another issue is glitches in the printer. Obviously the Ender3 pro is very cheap and we are trying for very high quality prints, so any glitches or slips are a bad thing. There are things you can do to minimize these like adding filters to the stepper motors, doing better calibration etc but there is a limit. I had small amounts of slippage in x and z directions on most of my prints. What this means is to get a perfect print you may have to print the model more than once.
The conclusion I came to was the extra quality provided by the finer nozzle wasn't worth the extra time involved in printing and that the printer is too slow.
Default Quality = 2 hours
High Quality = 6 hours
Before I got the printer I had done quite a bit of 3D printing using services like ShapeWays but I'd never had direct exposure to slicing. Cura is a free slicer and seems popular with the miniatures community. It generates the gcode (Tool path) for the 3d printer. I had assumed that the slicer would just "rasterize" each slice but actually the fill pattern is quite complex, aiming for best quality whilst having a hollow component. Cura is also where you add stuff like supports which are required to enable overhanging regions in the model. You can also orient a model so that the supports don't obscure details on the face or front of the model.
The biggest thing I learned about slicing was to trust the slicer. You can move up and down the layers and play the tool path in the preview option after slicing. Carefully checking the model Cura before printing saves hours of frustration with failed prints.
I used the fat dragon games profiles for the printer, they are popular and work well. Their models are also very nice if a bit simple, however they print without the use of supports and so are ideal for use with an FDM printer. Definitely worth checking out.
Pla is actually pretty nice to work with with sandpaper, knifes, file etc and takes paint well, but because of the limits of FDM you are going to need to do a lot of it. This means it's probably best not to add a lot of tiny surface detail in the 3d model. Use the printer to generate the gross shapes and then add surface details and texture the usual way with putty, etched brass etc. Similarly design the prints for ease of clean up. It's much easier to sand individual parts than a complete model in many cases.
What did I think of the Ender 3 Pro?
It didn't start well. Somebody told me they burn houses down. So a slow printer than can't be left unattended. Mine was completely unaligned out of the box and took significant rebuilding. The instructions are pretty much useless. There are no exploded diagrams anywhere that I could find. I had a magnetic bed, the plastic z-limit switch mount was designed for a glass bed which is thicker. So the z limit switch is in the wrong place. Based on info on reddit, I cut the locating tab off mine and mounted it myself in the right place. Oh and my print bad isn't flat which made bed leveling a challenge.
So if someone tells you putting an Ender together is like Ikea, they are talking nonsense. It took me 2 afternoons to get mine together.
Pretty much anytime you do anything with the printer you are going to have to level the bed. This is kind of a misnomer, what you are actually doing is making sure the bed is perpendicular to the print nozzle and is a good distance away from it. The bed leveling springs as supplied are crap (get some thicker ones) and "the thickness of a sheet of paper" isn't very helpful since that distance depends on the nozzle size in play among other things. Anway, bed leveling is a pain and takes about half an hour. It definitely helps if the printer is in a brightly lit location. It's hard to work on if you can't see.
I've run my printer for about 25 hours and I've had to tighten the x-axis belt tensioner. I've also discovered that my y-axis wheels are too tight and are eating themselves against the drive rail.
Changing filament isn't too hard. Changing nozzle takes about half and hour, plus you'll want to level the bed after.
What I'm saying is that you can expect to spend an hour fiddling with the printer for every 6 hours it spends printing... which isn't great. It isn't terrible, but it isn't great either.
I would say this printer is good for roughing out parts. It's not really up to doing wargame miniatures at commercial quality levels, but for models or terrain as long as you don't expect perfect surface texture quality and are prepared to put in some work on the finished article it's not bad if a little frustrating to use sometimes. Also as a prototyping machine to test prints before sending them out to Shapeways for higher quality printing its probably pretty good.
I had 2 days left with the 3D printer. I modeled up a little alien guy inspired by Babu Frik in the latest Star Wars, printed it on the Ender and then painted it. The wet look on the eyes is 5 minute epoxy and his 'tash is golden retriever hair I found on the carpet. Thanks Abby.
Still fighting with the fact that the printer is so slow and the quality isn't where it needs to be. I tried filling the very obvious layer lines with a mix of acrylic filler and paint. It sort of works, but its a pain.
Oh and before I forget...
I made a little RoJaws too. This was an experiment is parts breakdown. It seems to me that smaller pieces are better than trying to print as a whole statue.
I decided to model and 3D print a small piece of architecture and then make duplicates using a silicone rubber mould and urethane resin. The main reason advantage of resin casting over 3D printing is speed of duplication. This part took 5 hours to print at medium quality settings. Each cast part takes about 15 minutes.
I found a reference image of a holy well that looked ideal for a stand alone project. I modeled the building in blender. I modeled the door and doorway using basic low poly techniques and then used 3D sculpting to add texture and bricks to the model.
I then 3D printed the model on my Ender3 at medium quality.
I did some clean up on the print with sandpaper, files and a craft knife. This would have been a great time to add some extra damage/depth with a dremel or to add texture using textured paints and sand.
I made a simple mould box from balsa wood and then poured in Oomoo 30 from Smooth On. This is a soft silicone rubber moulding material. Because it has low viscosity, it’s easy for any trapped air to escape, thicker rubbers can result in trapped air bubbles which show up in the final cast. Oomoo 30 takes about 7 hours to harden before it can be used.
Next day I popped the 3D print out of the mould and I was ready to begin making castings using Smooth-Cast 300. This is a urethane resin that sets very quickly. Parts are ready to be pulled out of the mound in about 15 minutes. I made 2 test castings and was very happy to see that both came out very nicely with no air bubbles on the detailed side.
I assembled and painted one of the pieces to see how it looks. I’m quite pleased with how it came out.
If I did another project like this, I would make some of the gaps (for instance around the door frame) larger and I would bevel the edges. I think this would make them easier to print as well as making the mould stronger so it would last longer. I would make the mortar gaps on the walls deeper but I probably wouldn’t add much surface texture to the parts before printing. I think it would work better to add those to the 3d printing before making the mould with a dremel and other tools
You can see the printer slipped halfway through the print. I should have killed the print and done a fresh one with no errors.
Obviously a simple open faced mould like this only creates single sided parts, in the future I want to try a glove mould that will allow for more complex shapes.
The moulding and casting turned out to be very easy and none of the materials smelled bad which was a pleasant surprise.
Seems like a modular approach could create some quite complex buildings from just a few quickly cast parts.
This was fun!
Printed a Goblin from Fat Dragon Games using their custom profile for Cura and a .20 nozzle. I've deliberately painted it to emphasize the layer lines. At normal viewing distances of around a foot, the layer lines are not visible. I'm pleased with the performance of the printer although it's very slow at this quality level. This print took around 5 hours.
Fat Dragon Games has some excellent set up videos, some printer profiles that seem to work well with the latest version of Cura and their models are fun looking and seem well designed for printing.
It's important to realize this model is 32mm tall, so the photo is much larger than real life.
Like colour makes difference to me? I'll be painting everything I make anyway. My printer shipped with a small roll of white PLA. It seems to work great, but because it's white and slightly translucent it can be a bit hard to see, so I bought a giant drum of red PLA instead. It's slightly easier to evaluate the prints.
This does make me wonder though, why is PLA sold in such large rolls? For the small things I plan to print a single large drum is a lifetime supply. I would have rather bought a selection pack of different colours in shorter lengths. My current print is going to use 2g of material and takes ~4 hours :-0
Found myself in a situation where I needed a project for about a month so I picked up an Ender 3 Pro to have a play. I've seen some very impressive prints from these basic FDM printers, but I've also seen some terrible prints too....
This is Not a Setup Guide
I'm an idiot, don't listen to me. Apparently these things are as easy to build as IKEA. I found it a bit of a pain to put together.
Where are the instructions?
There are some instructions in the box, but most are on the micro sd card. There are also some test print files. However the test print models are pretty big, you might want to find something smaller to start with.
Get the base flat:
Take the base out of the box (the big bit shaped like an H) on a flat surface and check that all four feet are touching and that it doesn't rock. If it does, it's not aligned. You'll need to slacken off the bolts holding it together and get it straight.
Get the uprights parallel:
The next step would have you attach the 2 vertical extrusions to the base. However all the holes are slightly over sized so you can make life easy on yourself by attaching the top piece to the verticals on a flat surface to make a U shaped component. That way you know the verticals are pretty well aligned with each other. Attach them to the base, and then remove the top bar to finish assembly.
Adjust build plate:
The build plate moves back and forth on the y axis ( z is up and x is left to right). The plate runs on 4 wheels on the extrusion. If its wobbly, tighten the eccentric nut with the included wrench until it moves smoothly without wobbling.
Check for alignment:
With the printer assembled but powered off, it should be possible to smoothly move the parts along all 3 axes. Z will be tighter because of the worm drive, but the worm drive shouldn't be bent or under load. If the print head carriage is binding, your printer is bent. Undo the bolts and straighten it out. This is why I suggest starting with a known good flat base and getting the verticals perpendicular.
The original Ender came with a glass bed. The pro comes with a magnetic bed. The glass is much thicker than the mag bed which means the Z limit switch isn't in the right place. It's keyed to the frame with a little plastic tag. Reddit said cut it off so I did, then with the bed springs crushed down more than 50% I guesstimated a good spot for it. While we are talking about bed springs, the stock ones are a little weak, get some stiffer ones. They are an easy and cheap upgrade.
My unit came with the correct voltage PSU and English instructions but a French plug so go figure. Also since one of these machines once caught fire, I got a smoke alarm that lives by the printer
Assembling this thing was not at all straightforward or foolproof. Bring patience and your A game. It doesn't help that the entire unit is made out of black materials. If would be much easier to see what was going on if it were made in a lighter colour. It took me 2 afternoons to assemble because of the alignment issues.
Enough moaning, once I got it together it works fine :-)
It's important to finish things. If you abandon every project you start because it doesn't meet your idea of perfection you never learn the process of getting to the end. All projects go through an ugly stage, you just have to have the faith in yourself to push through, and that only comes from experience.
Fake News (Alien Edition) was inspired by the aliens in District 9 and some paintings by Chris Foss that featured Easter Island heads. I was hoping for a more "realistic" style but what I've ended up with looks a bit like Antz. Next project, one of the major goals will be to go for something more photo-realistic. That's been something I've been working on all year. Where to add detail, how to texture etc. Seems like this is something you just have to learn by doing, no short cuts...
I'm estimating that I have around 20..30 hours in this project, but it was spread over a couple of week. That means as I get better something like this should only take 2 or 3 days.
I've used Blender's sculpting tools a few times with limited success. This time I decided to sit down and watch some training videos rather than just guessing how to do stuff.
Found this series of videos by Grant Abbitt and I think they are pretty good. The use of constant detail, refining the shape and then cranking the detail up seems to be very logical and perhaps not what I was doing before. Also |Grant does a good job of showing you the finished product at the start of the video and his videos are clean and concise. Videos where you don't see the results until the end are a peeve of mine.
Time to make an Easter Island Head.
I'm currently working on an Alien outside new broadcast... Inspired by District 9 and Chris Foss's Easter Island head spaceship paintings. Here's an early shot processed to look like a Victorian photo. One of the important things here is even though the ships and head will be small on screen in the final product, I need to make them appear to be the same scale. If you imagine the ships to be about the size of a 737, then the head is huge, so I'll have to pay attention to that when adding the materials. This head looks like its about 4 inches tall.