Friday, November 1, 2013
Sentinel Armor Texture - All those little details
There are a lot of details on this armor. I didn't notice most of them because the Sentinel Armor is almost pure black, but after finding the grey version of the armor, just wow. Blood Dragon wasn't like this. Lines, plates, and layers everywhere. Sure, there were a lot of shapes, but they were easy to trace and make vectors out of. Sentinel is comprised almost exclusively of curved, more organic details.
Right now, I have a lot of line segments outlining the shapes. They'll be combined later into the final designs on the armor.
Labels:
3d modeling,
Dragon Age: Origins,
Photoshop,
Sentinel Armor
Sunday, October 13, 2013
Sentinel Armor Progress 6
I've started laying down the base colors for the Sentinel armor. Now, you may notice how both armor models aren't pitch black. That's because I'm using a diffuse texture from a palette swap armor model to make it easier to trace all the filigree.
Like before, nearly everything at this stage is a vector using blending styles The filigree is an exception. The traces are vector lines, but I stroke them onto a layer with the vector mask disabled.
Labels:
3d modeling,
Dragon Age: Origins,
Photoshop,
Sentinel Armor
Sunday, October 6, 2013
Sentinel Armor Progress 5
UV unwrapping complete. Well, kind of. As you can see, I still need to take all the pieces and shove them inside that black, square outline. I think I'll make it double height and aim for a 1024x2048 texture. That way, I don't have to scale the UVs down so much and can avoid having to make a bigger texture to compensate.
------------------------------------
The MyBookLive is alive again. Swapped harddrives, ran an unbricking script to "install" the new drive, and copied files back over. Now, I wait.
------------------------------------
I was playing around with blog themes and tried Dynamic Views. It looks pretty, but runs slow and is barely customizable. You can't choose one view and remove the menu items for the others; it's all or nothing unless you edit the theme's source.
Extra widgets do not show, color choices seem to be irrelevant, you can't remove existing widgets, FireFox became unresponsive half a dozen times when I looked at customization previews, and it adds extra spaces in posts which then messes with formatting.
Tuesday, October 1, 2013
MyBookLive dying: Confirmed.
According to the SMART status, the drive is dying. The read error metric is at 1, which is apparently a bad thing. The drive can still work, but it looks to be all downhill from here. I don't expect any support from WD after tearing the thing apart, so I'm adding "another hard drive" to my list of things to buy.
----------------
I'm almost done unwrapping the Sentinel armor. Just one elbow pad left.
I still don't know what kind of delivery mechanism I'm going to use. I think a quest would be appropriate or even a new dungeon. Placing items on NPCs has turned out to be a bad idea. They wander off, get killed and deleted, fall through the ground, don't die, etc.
The messenger system worked well for the Bassrath-Kata weapons. However, players had to use the console to get all the weapons because I set it up so they would only get one item after finishing the Dragonborn quest. Not a great way to do things. Instead of the messenger system, I think I'll create a quest NPC and let him handle giving stuff away.
----------------
I'm almost done unwrapping the Sentinel armor. Just one elbow pad left.
I still don't know what kind of delivery mechanism I'm going to use. I think a quest would be appropriate or even a new dungeon. Placing items on NPCs has turned out to be a bad idea. They wander off, get killed and deleted, fall through the ground, don't die, etc.
The messenger system worked well for the Bassrath-Kata weapons. However, players had to use the console to get all the weapons because I set it up so they would only get one item after finishing the Dragonborn quest. Not a great way to do things. Instead of the messenger system, I think I'll create a quest NPC and let him handle giving stuff away.
Labels:
Dragon Age: Origins,
Sentinel Armor,
Skyrim,
troubleshooting
Monday, September 23, 2013
Sentinel Armor Progress 4
I'm unwrapping the armor starting with the helmet and the hands. Up next: Boots.
I tried unwrapping in 3ds Max 2012. It was...interesting. The problem was how radically different the UI was compared to 2011. I had to guess that each icon meant and hope that the developers and I agreed on that meaning.
The Peel function is interesting. I divide most of the geometry using shading groups. So, Peeling splits the UV map according to those groups, saving me from doing it manually. Neat trick.
Now, I'm trying to decide how I want to make the mod. The armor I have right now, minus the helmet, contains geometry common to nearly all the massive armor in Dragon Age: Origins. So, a different armor is just a retexture (sometimes, just a palette swap). I can release what I have now in a small mod and make more later or I could model the other unique pieces and release the whole thing at once. I probably won't come to a decision until after I complete the Sentinel armor. I need to see how big one armor set will be to determine whether or not the Workshop will support more.
I tried unwrapping in 3ds Max 2012. It was...interesting. The problem was how radically different the UI was compared to 2011. I had to guess that each icon meant and hope that the developers and I agreed on that meaning.
The Peel function is interesting. I divide most of the geometry using shading groups. So, Peeling splits the UV map according to those groups, saving me from doing it manually. Neat trick.
Now, I'm trying to decide how I want to make the mod. The armor I have right now, minus the helmet, contains geometry common to nearly all the massive armor in Dragon Age: Origins. So, a different armor is just a retexture (sometimes, just a palette swap). I can release what I have now in a small mod and make more later or I could model the other unique pieces and release the whole thing at once. I probably won't come to a decision until after I complete the Sentinel armor. I need to see how big one armor set will be to determine whether or not the Workshop will support more.
Labels:
3d modeling,
Dragon Age: Origins,
Sentinel Armor
Sunday, September 8, 2013
Sentinel Armor Progress 3
Finished basic shape of the legs, arms, and torso. Now I need to move everything into place. The pauldrons, tassets, and whatnot were based on the shape of the original so I have to fiddle with them to make them fit over the new shell (The shoulder gear needs to drop 5 units and move back 1 or 2).
Once everything is fitted, merged, and tweaked, I'll start adding more details and refining the shapes (boots are too rounded). Then, it's off to UV land!
---------
After breaking the tabs off of my MyBookLive and disassembling it to get the harddrive, I found out that the backup I was trying to copy was corrupt. It copied over fine locally, but when I opened it, the filesystem was gone. File recovery software only recovered the temporary internet files (seriously?!). I resorted to pulling the original drive and rummaging through file remnants to get data back. Folder paths were shot, but the files were mostly recoverable.
I really don't know what burned the backup. It hasn't done that to me before and there were no errors during the process. Next time, I'll backup to a USB harddrive and keep it there instead.
Friday, September 6, 2013
MyBookLive being difficult
I made a backup of a computer a while ago just in case something went wrong. And sure enough, something went wrong and that backup is now needed. There's one small hiccup though: I made the backup to a MyBookLive.
Why is this a problem? I can't access the file and rummage around in it or get it off the device. Copy/Paste via Explorer (Win7), robocopy, ftp, and Copy/Paste via Konqueror in two versions of Linux all fail. I'd try a wireless connection to halve the transmission speed just in case the device cannot sustain long periods of transfer, but a 20 hour transfer time is too long (not to mention keeping a machine on for that long is going to create a lot of heat). I've moved the device and plugged it directly into the router, but that didn't help.
My first attempt to copy/paste went smoothly until about 90% completion, at which point the MyBookLive lost its net connection. No ping, no login, no access whatsoever. Over 3 days, every attempt after that has failed sooner and sooner with the same symptoms. Now it fails at 0.2% completion. I've updated the firmware twice (there was an intermediate firmware update before the latest could be installed). There are no errors, no bad packets, just an abrupt "No longer accessible" message and I can no longer communicate with the device for a few minutes.
I'm trying OSX now and while it hasn't failed yet, the traffic has stopped several times at less than 6% completion. I gave file splitting a shot with 7zip, but it reported an ever increasing time remaining and I stopped it after it quoted 41 hours. The odd thing was that while 7zip was creating nice 500MB chunks, the connection never dropped. It's like the device requires a give-and-take that is more significant than ACKs.
As much as I don't want to, it looks like I'm going to have to disassemble the drive completely (WD makes it hard to get to the drive itself without going through all the plastic bits and circuit boards), plug it directly into my PC, move the file off, and then put it back together. It's not like there's a warranty on it anymore (it expired years ago).
Odd Note: OSX reports the file as 10GB LARGER than what Windows or Linux report.
Update:
OSX just failed to copy it via Finder and the drive was not accessible for a few minutes, again. Disassembly it is.
Why is this a problem? I can't access the file and rummage around in it or get it off the device. Copy/Paste via Explorer (Win7), robocopy, ftp, and Copy/Paste via Konqueror in two versions of Linux all fail. I'd try a wireless connection to halve the transmission speed just in case the device cannot sustain long periods of transfer, but a 20 hour transfer time is too long (not to mention keeping a machine on for that long is going to create a lot of heat). I've moved the device and plugged it directly into the router, but that didn't help.
My first attempt to copy/paste went smoothly until about 90% completion, at which point the MyBookLive lost its net connection. No ping, no login, no access whatsoever. Over 3 days, every attempt after that has failed sooner and sooner with the same symptoms. Now it fails at 0.2% completion. I've updated the firmware twice (there was an intermediate firmware update before the latest could be installed). There are no errors, no bad packets, just an abrupt "No longer accessible" message and I can no longer communicate with the device for a few minutes.
I'm trying OSX now and while it hasn't failed yet, the traffic has stopped several times at less than 6% completion. I gave file splitting a shot with 7zip, but it reported an ever increasing time remaining and I stopped it after it quoted 41 hours. The odd thing was that while 7zip was creating nice 500MB chunks, the connection never dropped. It's like the device requires a give-and-take that is more significant than ACKs.
As much as I don't want to, it looks like I'm going to have to disassemble the drive completely (WD makes it hard to get to the drive itself without going through all the plastic bits and circuit boards), plug it directly into my PC, move the file off, and then put it back together. It's not like there's a warranty on it anymore (it expired years ago).
Odd Note: OSX reports the file as 10GB LARGER than what Windows or Linux report.
Update:
OSX just failed to copy it via Finder and the drive was not accessible for a few minutes, again. Disassembly it is.
Thursday, September 5, 2013
Plain old text post
I've been working on the torso of the armor. Instead of turning the Skyrim female body into quads like I did for BDA, I scaled and translated the original armor, roughly molded a cylinder around the torso, then adjusted it to fit around the Skyrim female body. It's less work, but also lets me keep the polycount down until I want to bump it up.
------------
I'm getting into Unity now. I'm not too thrilled with the brush tools because they're quite limited compared to Hammer and UDK. There's also an acceleration to the 3d view that really slows things down. Sure, I can pan quickly, but I prefer flying through the map (like in Hammer, UDK, CrySDK, and 3ds Max).
Right now, I'm looking at the tools from SixBySeven. They seem to add what I consider missing functionality to the editor.
------------
I'm getting into Unity now. I'm not too thrilled with the brush tools because they're quite limited compared to Hammer and UDK. There's also an acceleration to the 3d view that really slows things down. Sure, I can pan quickly, but I prefer flying through the map (like in Hammer, UDK, CrySDK, and 3ds Max).
Right now, I'm looking at the tools from SixBySeven. They seem to add what I consider missing functionality to the editor.
Wednesday, August 14, 2013
Sentinel Armor Progress 2
Got most of the armor around the hips done. Now, it's just the big pieces: chest, arms, and legs.
A lot of the pieces are carbon copies of others. For example, the blue tassets are the same piece copied over and over. While there are quite a few pieces to this armor, the copies will make it a lot easier to texture. I might do sets instead of having all 6 of them use the same texture just because I don't want it to be repetitive. Also, since there's no unique bits across the mirror, I can unwrap half of the chest and not worry about seams like I did on BDA (not a big issue, but still).
Labels:
3d modeling,
Dragon Age: Origins,
Sentinel Armor
Monday, August 12, 2013
Sentinel Armor Progress
As always, female body first because it's my character's gender (eliminates save swapping and summoning the character generation menu for the early stages). Sentinel Armor doesn't conform to the bosom, so I should be able to scale the armor to the male body instead of remaking it like I did for Blood Dragon Armor.
The gloves are almost identical to the Blood Dragon gloves. I took the base mesh, fiddled with exporting/importing their skin envelopes, and removed all but the hands. From there, I removed the knuckle guards and used the trace-spline-rotate method for creating the wrist guards. Massive Armors in Dragon Age: Origins are asymmetrical for the hands and pauldrons, so no quick mirroring for me. This is going to be interesting to unwrap later on.
---------------------------------------------
Since I haven't heard from the artist in a month and he isn't responding to me, I'm going to go ahead and post an image of the helmet I mentioned in my last post. Just note: Full credit to him for the initial sketch. I made the base model shown below.
The back of the head isn't complete yet. I'm trying to figure out what I want to do with it. There's this ongoing theme of protrusions that look like receptacles for tubes, but not all frames have them. Given how this particular one is close to Excalibur's, I think I'll draw from that. Then there's the organic wisps, swirls, bumps, and such that comprise Warframe's art style. I have no idea how I'm going to do that, but it should be informative in either case (I might actually try to use Mudbox for more than painting UV seams away!).
I have no idea how I'm going to unwrap this. Everything is currently using Max's default color selection. It is also MeshSmoothed a couple times.
The gloves are almost identical to the Blood Dragon gloves. I took the base mesh, fiddled with exporting/importing their skin envelopes, and removed all but the hands. From there, I removed the knuckle guards and used the trace-spline-rotate method for creating the wrist guards. Massive Armors in Dragon Age: Origins are asymmetrical for the hands and pauldrons, so no quick mirroring for me. This is going to be interesting to unwrap later on.
---------------------------------------------
Since I haven't heard from the artist in a month and he isn't responding to me, I'm going to go ahead and post an image of the helmet I mentioned in my last post. Just note: Full credit to him for the initial sketch. I made the base model shown below.
The back of the head isn't complete yet. I'm trying to figure out what I want to do with it. There's this ongoing theme of protrusions that look like receptacles for tubes, but not all frames have them. Given how this particular one is close to Excalibur's, I think I'll draw from that. Then there's the organic wisps, swirls, bumps, and such that comprise Warframe's art style. I have no idea how I'm going to do that, but it should be informative in either case (I might actually try to use Mudbox for more than painting UV seams away!).
I have no idea how I'm going to unwrap this. Everything is currently using Max's default color selection. It is also MeshSmoothed a couple times.
Labels:
3d modeling,
Dragon Age: Origins,
Sentinel Armor,
Warframe
Monday, July 22, 2013
The Long Silence
I've been busy and distracted for a while. I put more time into playing Warframe, getting A+ Certified, and working on side projects.
As I said in my previous post, Warframe is getting really grindy for new players now. Since I've been playing for a while, I've already completed a majority of the game before scaling difficulty, lower drop rates, and additional "BUY PLATINUM" neon signs. I'm a bit upset at how I'm not nearly as efficient at leveling gear as I was before the scaling difficulty change and how this also messed up item farming (lower drop rates+ harder to solo areas = more risk, less reward). Then again, I have more "but this was how it used to be" comparisons now.
In other news, I am now A+ Certified and have 3 years to either take a newer version of the test, completed a higher level cert, or enroll in Continuing Education to keep my status from expiring. All those options cost money and I'm still unemployed. Bah. And yet, I still have access to a computer with a decent net connection through which I can play games. What? Needless to say, I'd be screwed on my own in my current state.
On the Warframe forums, there was a fan sketch of a Prime Warframe's helmet, so I did the sensible thing and made it in 3ds Max. I've since submitted it to the sketch's author and am waiting to see what kind of organic details he places on it. I should probably inquire as to its progress...
Some friends and I are planning a game about a year out and our artist drew a couple sketches of armor. As above, I've been making it in 3ds Max and trying to fill in the details where the sketches are just gradients or solid color. It's also front view only, so I'm improvising on the back. No images to show here until the game gets further along.
EDIT:
This is apparently my 100th post. I guess I'll have to make some media (read: Continue Sentinel Armor) for the next one.
As I said in my previous post, Warframe is getting really grindy for new players now. Since I've been playing for a while, I've already completed a majority of the game before scaling difficulty, lower drop rates, and additional "BUY PLATINUM" neon signs. I'm a bit upset at how I'm not nearly as efficient at leveling gear as I was before the scaling difficulty change and how this also messed up item farming (lower drop rates+ harder to solo areas = more risk, less reward). Then again, I have more "but this was how it used to be" comparisons now.
In other news, I am now A+ Certified and have 3 years to either take a newer version of the test, completed a higher level cert, or enroll in Continuing Education to keep my status from expiring. All those options cost money and I'm still unemployed. Bah. And yet, I still have access to a computer with a decent net connection through which I can play games. What? Needless to say, I'd be screwed on my own in my current state.
On the Warframe forums, there was a fan sketch of a Prime Warframe's helmet, so I did the sensible thing and made it in 3ds Max. I've since submitted it to the sketch's author and am waiting to see what kind of organic details he places on it. I should probably inquire as to its progress...
Some friends and I are planning a game about a year out and our artist drew a couple sketches of armor. As above, I've been making it in 3ds Max and trying to fill in the details where the sketches are just gradients or solid color. It's also front view only, so I'm improvising on the back. No images to show here until the game gets further along.
EDIT:
This is apparently my 100th post. I guess I'll have to make some media (read: Continue Sentinel Armor) for the next one.
Friday, June 14, 2013
Sentinel Helmet - Model Progress
I've finished the helmet's shape for now. I was considering chamfering all the edges on the wings to produce the smooth shapes from the normal map, but no one's going to be that close unless they've blown the model up several times or imported into a 3d package. So, I made some really close polygons resembling the shapes and assigned all the hard edges to the same smoothing group.
There is a sun shape on the forehead where the red glow dot is supposed to be, but I'm not sure if I want to make that geometry or not.
This is really the only unique thing about the Sentinel Armor, besides the texture. It seems like everything else is exactly the same for all massive armors in Origins/Awakening.
-------------------------
I came across a thread on the Warframe forums with a sketch for a certain frames's helmet redesign and, well, decided to give it a go. The sketch is front view only, so I'm using a side view from the wiki to get a head size.
Labels:
3d modeling,
Dragon Age: Origins,
Sentinel Armor,
Warframe
Monday, June 10, 2013
Starting the Sentinel Armor Helmet
I took some time to blast through Warframe and level a lot of gear, but it's time I got back into making things. So, I've targeted the Sentinel Armor set from Dragon Age: Origins - Awakening.
--------------------------------------------------------
Come to think of it, I haven't heard a POST beep in a long, long time. Errors, sure, but POST? I don't think they do that anymore.
--------------------------------------------------------
Come to think of it, I haven't heard a POST beep in a long, long time. Errors, sure, but POST? I don't think they do that anymore.
Saturday, June 1, 2013
Going for A+ Certification
After studying for over 6 months, I've decided it's now or never for getting A+ Certified. There's always more and more stuff to learn, but if I keep waiting, I'm never going to be "ready". That, and since 2011, all future A+ Certs last 3 years instead of for life. So I'll have to recertify or pay a monthly fee and demonstrate my "Continuing Education" to Comptia.
I know my way around most PCs, but I tend to get stuck in my assumptions. An example I keep using goes like this:
Question: A client calls and says their headphones stopped working this morning, but were fine yesterday. After going through numerous solutions, you cannot fix the problem. What do you do?
2 goes against best practices as outlined in the A+ book I'm reading. Also, I don't do that. Additionally, it isn't good for getting return clients.
1 is something I'd do off the bat while running the client through solutions I could think up.
3 looks like the best answer here. I can't fix the problem which means it could be a) something I've overlooked or b) this is a symptom of a much larger problem, not the problem itself. If escalation will solve the problem when I couldn't, I'll escalate it.
So, I chose 3. The correct answer was 1. Why? The test assumes once you get the call, you run through everything you can think of AND THEN you research the problem on the Internet. Everything is in steps. There's probably more to it, but I'm no psychologist.
I know my way around most PCs, but I tend to get stuck in my assumptions. An example I keep using goes like this:
Question: A client calls and says their headphones stopped working this morning, but were fine yesterday. After going through numerous solutions, you cannot fix the problem. What do you do?
- Research the problem on the Internet
- Scold the client for being an idiot
- Escalate the issue
- Recommend the client buy a new set of headphones
2 goes against best practices as outlined in the A+ book I'm reading. Also, I don't do that. Additionally, it isn't good for getting return clients.
1 is something I'd do off the bat while running the client through solutions I could think up.
3 looks like the best answer here. I can't fix the problem which means it could be a) something I've overlooked or b) this is a symptom of a much larger problem, not the problem itself. If escalation will solve the problem when I couldn't, I'll escalate it.
So, I chose 3. The correct answer was 1. Why? The test assumes once you get the call, you run through everything you can think of AND THEN you research the problem on the Internet. Everything is in steps. There's probably more to it, but I'm no psychologist.
Thursday, May 30, 2013
Warframe
I've been playing quite a bit of Warframe lately. I picked it up during the Global Game Jam earlier this year and, well, I've been having quite a bit of fun. I have a couple more missions on Pluto and have never been to the Orokin Void, but the grind is starting to set in. I've endured long grinds before (RF Online Shield Miller @ level 48), but every update seems to tack on more and more grinding.
Understand though, as of this writing, the game is in open beta so anything can change. I'm just not a fan of having to grind for items with a low drop chance against enemies that are simply bullet sponges. Well, it wouldn't be better if the enemies weren't bullet sponges, but still. Then you have the Dojo from Update 8 which added 20 weapons, a decent chunk of which take an obscene amount of time and resources to make.
Take Ignis, the flamethrower, for example (stats taken 5-30-13). To get the opportunity to make the weapon you need a chem lab:
Detonite Ampule 500
Plastids 10,000
Ferrite 20,000
Polymer Bundle 25,000
Credits 500,000
Research Time 3 days
Once you can craft it, then you need
Detonite Injector 2
Nano Spores 5,000
Ferrite 5,000
Forma 1
Credits 30,000
Build Time 24h
Now, this wouldn't be that much of an issue except for the following (no hard numbers here):
Detonite Ampule - 1 per drop. Uncommon
Plastids - 10-30 per drop. Uncommon
Ferrite - 50-100 per drop. Common
Polymer Bundle - 100-400 per drop. Common.
Nano Spores - 100-200 per drop. Common.
Detonite Injector - Crafted via purchasable blueprints. 12 hour build time after Chem Lab is built.
Forma - Randomly awarded from Void missions, Tier 3 login rewards, purchased with real money, or awarded as a blueprint for an Alert mission. 24 hour build time.
So, to get all the resources, you need to invest a huge amount of time. Now, the Chem Lab costs could be offset if you join a clan, but since I made my own clan, I can't leave (didn't know that beforehand) and there's no one I really play with to invite. This leaves me footing the entire bill if I decide to go for it. Since you have to build hallways, reactors, and other support rooms just to be allowed to build a Chem Lab, it effectively raises the difficulty of gaining those resources by what feels like an order of magnitude.
But I digress, since I have a whole arsenal of weapons to build, let alone use, I can at least keep the grinding interesting. And, before anyone goes off, Warframe is a Free-To-Play game, meaning they have to make money somehow. I understand it (well, not like an Economics professor level of understanding). The usual money crunch is when new gear is released in an update. Those new items are not available to non-paying folk for a few weeks; after which they'll start appearing as drops or Alert missions if the RNG gods smile on you.
One more thing, I have a strong...dislike towards Lech Kril. Stupid, camera flipping, drop hoarding, bullet sponge of a boss.
Understand though, as of this writing, the game is in open beta so anything can change. I'm just not a fan of having to grind for items with a low drop chance against enemies that are simply bullet sponges. Well, it wouldn't be better if the enemies weren't bullet sponges, but still. Then you have the Dojo from Update 8 which added 20 weapons, a decent chunk of which take an obscene amount of time and resources to make.
Take Ignis, the flamethrower, for example (stats taken 5-30-13). To get the opportunity to make the weapon you need a chem lab:
Detonite Ampule 500
Plastids 10,000
Ferrite 20,000
Polymer Bundle 25,000
Credits 500,000
Research Time 3 days
Once you can craft it, then you need
Detonite Injector 2
Nano Spores 5,000
Ferrite 5,000
Forma 1
Credits 30,000
Build Time 24h
Now, this wouldn't be that much of an issue except for the following (no hard numbers here):
Detonite Ampule - 1 per drop. Uncommon
Plastids - 10-30 per drop. Uncommon
Ferrite - 50-100 per drop. Common
Polymer Bundle - 100-400 per drop. Common.
Nano Spores - 100-200 per drop. Common.
Detonite Injector - Crafted via purchasable blueprints. 12 hour build time after Chem Lab is built.
Forma - Randomly awarded from Void missions, Tier 3 login rewards, purchased with real money, or awarded as a blueprint for an Alert mission. 24 hour build time.
So, to get all the resources, you need to invest a huge amount of time. Now, the Chem Lab costs could be offset if you join a clan, but since I made my own clan, I can't leave (didn't know that beforehand) and there's no one I really play with to invite. This leaves me footing the entire bill if I decide to go for it. Since you have to build hallways, reactors, and other support rooms just to be allowed to build a Chem Lab, it effectively raises the difficulty of gaining those resources by what feels like an order of magnitude.
But I digress, since I have a whole arsenal of weapons to build, let alone use, I can at least keep the grinding interesting. And, before anyone goes off, Warframe is a Free-To-Play game, meaning they have to make money somehow. I understand it (well, not like an Economics professor level of understanding). The usual money crunch is when new gear is released in an update. Those new items are not available to non-paying folk for a few weeks; after which they'll start appearing as drops or Alert missions if the RNG gods smile on you.
One more thing, I have a strong...dislike towards Lech Kril. Stupid, camera flipping, drop hoarding, bullet sponge of a boss.
Saturday, April 27, 2013
Trian's Maul Geometry
I forgot to post this while I was making Trian's Maul. Just a quick screenshot of the old model (left) and all the geometry I put on the new model (right).
EDIT:
Due to someone claiming I've stolen textures from Dragon Age 2 (curiously, not Dragon Age: Origins), here is a cropped screenshot (no need to show the tools window or title) comparing the Bassrath-Kata greatsword textures. On the left, mine. On the right, DA2's.
Labels:
3d modeling,
Dragon Age 2,
Dragon Age: Origins,
Photoshop
Tuesday, April 16, 2013
Questionable job hunting
Got a response from a job application.
They asked me for contact info and a good time to call. I plastered said info all over my coverletter, resume, and email application. *Flag 1* (okay fine, I omitted contact times just to see what would happen)
Their website doesn't exist either. *Flag 2*
Well, the server responding to the site works (whois reports GoDaddy). It completes the 3way handshake and ACKs the HTTP request, but it doesn't send the page and the connection stays alive. *Flag 3*
I'm not quite sure if this company exists or not. It was a craigslist posting (some companies don't like posting who they are), they're responding from the craigslist email *Flag 4*, and the person who responded has 3 phone numbers in 3 different area codes. *Flag 5*
Because this is a QA position and "why not?", I responded with an overview of what Wireshark captured.
They asked me for contact info and a good time to call. I plastered said info all over my coverletter, resume, and email application. *Flag 1* (okay fine, I omitted contact times just to see what would happen)
Their website doesn't exist either. *Flag 2*
Well, the server responding to the site works (whois reports GoDaddy). It completes the 3way handshake and ACKs the HTTP request, but it doesn't send the page and the connection stays alive. *Flag 3*
I'm not quite sure if this company exists or not. It was a craigslist posting (some companies don't like posting who they are), they're responding from the craigslist email *Flag 4*, and the person who responded has 3 phone numbers in 3 different area codes. *Flag 5*
Because this is a QA position and "why not?", I responded with an overview of what Wireshark captured.
Sunday, March 10, 2013
BDA - Updated...or not.
Found a solution to the vertex count problem: http://forums.nexusmods.com/index.php?/topic/521561-3ds-max-nif-importer-exporter/page-8#entry5243549
Editing normals above the Skin Modifier but below the BSDismember Modifier fixed it. I wonder if this'll fix the viewangle clipping on the female body...
The male version was all ready to be deployed until I saw the upload window.
3MB over. That's one texture too big. Looks like I'll have to halve the textures for Steam after all. Which isn't a bad thing given the limited upload space modders have on the Workshop. I blanked mine out for some reason. *shrugs*
On the plus side, I now feel it's ready to release as a Beta on the Nexus.
EDIT:
Bwahahahaha! Found a glow texture that is bigger than it should be. Lowering its resolution will put me just under the upload limit.
Editing normals above the Skin Modifier but below the BSDismember Modifier fixed it. I wonder if this'll fix the viewangle clipping on the female body...
The male version was all ready to be deployed until I saw the upload window.
3MB over. That's one texture too big. Looks like I'll have to halve the textures for Steam after all. Which isn't a bad thing given the limited upload space modders have on the Workshop. I blanked mine out for some reason. *shrugs*
On the plus side, I now feel it's ready to release as a Beta on the Nexus.
EDIT:
Bwahahahaha! Found a glow texture that is bigger than it should be. Lowering its resolution will put me just under the upload limit.
Saturday, March 9, 2013
BDA - Nif exporter unrelenting
I didn't preserve the _0 models in the same scene as the _1 models, so I had to go back one version to the _0 models I initially exported. Since 3ds Max thinks preserving skin weights is stupid, I couldn't import/export/merge one scene's models into the other. Merging resulted in either the existing or imported models losing skin data or, if I merged the skeleton, the skin data was preserved for what imported, but broken for everything else because the "new" bones replaced the "old" bones (They're the same bones!). "Save" and "Load" for the skin modifier did nothing whatsoever. Well, "Save" did indeed save the envelopes, but "Load" didn't load them; presumably because of the same new/old bones thing.
Oh, and Skin Wrapping 1 over 0 didn't work.
Anyways, I buckled down and remade the _1 torso models. I started by welding all vertices within .1 units of each other on the _0 (the only vertices that close were between the glowing panels on the chest) and followed up with scaling it all out, again. Upon export, I expected everything to go well, like on the female armor. By merely writing that last sentence, you already know that didn't occur. Still off by 4.
Here's the weird bit. If I do not weld, the vertex count is 3214. If I weld, the vertex count is 3319.
I tried exporting as .obj and importing that into nifskope, but the _1 has 7 extra vertices. Exporting and importing a .3ds kept the same vertex count, but I do not have a Skin Instance or Dismember modifier, so the imported mesh is only good for a static model (and probably not even that).
*sigh*
EDIT:
I've tried putting everything into one smoothing group using the Smooth and the Smoothing Groups section of Editable Poly. While the poly count did go down, the _0 model has 3 more vertices than the _1. This would have been a suboptimal solution given that parts of the armor do require different smoothing groups to look correct in game.
Oh, and Skin Wrapping 1 over 0 didn't work.
Anyways, I buckled down and remade the _1 torso models. I started by welding all vertices within .1 units of each other on the _0 (the only vertices that close were between the glowing panels on the chest) and followed up with scaling it all out, again. Upon export, I expected everything to go well, like on the female armor. By merely writing that last sentence, you already know that didn't occur. Still off by 4.
Here's the weird bit. If I do not weld, the vertex count is 3214. If I weld, the vertex count is 3319.
I tried exporting as .obj and importing that into nifskope, but the _1 has 7 extra vertices. Exporting and importing a .3ds kept the same vertex count, but I do not have a Skin Instance or Dismember modifier, so the imported mesh is only good for a static model (and probably not even that).
*sigh*
EDIT:
I've tried putting everything into one smoothing group using the Smooth and the Smoothing Groups section of Editable Poly. While the poly count did go down, the _0 model has 3 more vertices than the _1. This would have been a suboptimal solution given that parts of the armor do require different smoothing groups to look correct in game.
Labels:
3d modeling,
Blood Dragon Armor,
not working
Friday, March 8, 2013
BDA - Nif exporter strikes again
Well, it's doing it again. Same vertex order, same number of vertices, same number of faces, but the exported nifs have a different vertex count. But not only are the vertex counts off, the polygon/triangle counts are off. I didn't even touch the skin modifier (the fiddling of which caused the problem last time).
This time, it's the new chest and arms that are missing pieces. Everything else is scaled up and tweaked from the female body (you could probably weight slide between them as is).
When I compare the welded vertex counts at a .1 threshold, the _0 model ends up with more vertices in the arms than the _1. At .01, neither loses vertices. I think I'll try skin wrapping _1 over _0 next.
Labels:
3d modeling,
Blood Dragon Armor,
not working
Thursday, March 7, 2013
BDA - Male Progress
The _0 model is pretty much done. I just have a couple skin tweaks left. The problem is going to be "growing" it to the _1 model. As, well, they are drastically different physically. It goes from the scrawny Skyrim Courier to Brick from Borderlands; quite a big change.
The female armor was kind of easy. The _1 was just that, a bigger woman. Same overall shape; more accented curves. Male_1, not so much. Instead of scaling the entire part up, it looks like I'll have to choose a couple polygon rings and scale them with Soft Selection on. I suppose this is why you maintain good edge loops.
Tuesday, March 5, 2013
BDA - Male Texture
I've got the major details in for the chest, but the arm bands are still unadorned. The dragon doesn't wrap around across the back seam, so I don't have to patch it like I did on the female torso. To avoid reinventing the wheel and save time, I think I'll copy the detail layers from the female torso textures.
Saturday, March 2, 2013
BDA - Male armor
Well, there it is. I built the torso out of Skyrim's male_0 body instead of trying to preserve the UVs of the female torso. It would have been nice to keep the UVs, but I didn't want to mess up the masculine look by trying to de-feminize the armor. I'll have to create another texture for the male armor, however, I can make the texture only cover the new geometry. So, it'll be one more texture, but smaller than a complete remake.
This could have worked too. It saves the existing UVs and looks masculine, kind of.
Saturday, February 23, 2013
Blood Dragon Armor - Adding light
The second beta version is released with the new textures and skin weights. It looks a lot better now.
Riding on the high of the release, I delved into the Creation Kit to figure out how to add a light to the armor. Why? It glows, so shouldn't that glow have some effect on the world?
It took a while, but I figured it out and, in the process, found my next video tutorial.
I had wanted a spotlight with a low fov to simulate the light coming from the visor, but I couldn't rotate the light at all. Everything I tried just wouldn't work and I did not understand the Quaternion rotation the Interpolator node had. I did change the rotation value by entering random numbers, but that had no visible effect on the light's orientation. I guess I'm stuck with an omnidirectional light.
Riding on the high of the release, I delved into the Creation Kit to figure out how to add a light to the armor. Why? It glows, so shouldn't that glow have some effect on the world?
It took a while, but I figured it out and, in the process, found my next video tutorial.
I had wanted a spotlight with a low fov to simulate the light coming from the visor, but I couldn't rotate the light at all. Everything I tried just wouldn't work and I did not understand the Quaternion rotation the Interpolator node had. I did change the rotation value by entering random numbers, but that had no visible effect on the light's orientation. I guess I'm stuck with an omnidirectional light.
Friday, February 22, 2013
Nif exporter being a pain
I made a few changes to the Blood Dragon Armor's skin weights to address issues I had with the initial weighting. Now, the upperarms don't raise several polygons on the breasts and the stomach plate isn't awkwardly deforming. But there's a hitch: The _0 model is losing vertices upon export. These same changes, when applied to the _1 model don't change the vertices on export.
So, what's going on? Initially, I thought it was because I moved the belt away from the main body to reduce selection scope. Not doing that, however, did not prevent the vertex disappearance. Next, I thought the breastplate was to blame. Once I shifted the weights around and exported, the vertices disappeared upon export.
The vertex count did not change in 3ds Max and, considering the vertex count changes upon export into .nif for whatever proprietary reason, I assumed the exporter was to blame. Now, the _1 did not lose vertices, which led me to believe the issue was spacing. The _1 model's vertices are too spread out to be hit by the welding.
To test this, I tried changing the weld parameter's precision when
exporting the _1 mesh. Values ranged from blank spaces to 0 to 0.0 all
the way to 0.000000. No change. When applied to the _0 mesh, though,
every single value had no effect on the lost vertices.
The only solution I can think of to prevent this is to weld the vertices of every element. This would require making the _1 model for a third time. Why? The welding would change both the vertex count and vertex numbering with no guarantee the same exact operations will produce the same count and numbering on the _1 model. I think.
Edit:
Welded every single vertex within .01 units on both models. Still getting the 2 missing vertices.
Edit:
After deleting elements until the discrepancy stopped, I've found the pauldrons are where the vertices go missing. Now I get to either a) leave them slightly messed up or b) separate them into a third component of the armor and fix their weights.
So, what's going on? Initially, I thought it was because I moved the belt away from the main body to reduce selection scope. Not doing that, however, did not prevent the vertex disappearance. Next, I thought the breastplate was to blame. Once I shifted the weights around and exported, the vertices disappeared upon export.
The vertex count did not change in 3ds Max and, considering the vertex count changes upon export into .nif for whatever proprietary reason, I assumed the exporter was to blame. Now, the _1 did not lose vertices, which led me to believe the issue was spacing. The _1 model's vertices are too spread out to be hit by the welding.
_0 model: 10740 vertices |
_1 model: 10740 vertices |
Both exported. _0 missing 2 vertices |
Edit:
Welded every single vertex within .01 units on both models. Still getting the 2 missing vertices.
Edit:
After deleting elements until the discrepancy stopped, I've found the pauldrons are where the vertices go missing. Now I get to either a) leave them slightly messed up or b) separate them into a third component of the armor and fix their weights.
Thursday, February 21, 2013
BDA Textures
I've been focusing heavily on the Blood Dragon Armor's textures for the past few days. Mainly the torso. In the beta release, I completely forgot the knee pad details (text, dots, and triangles) and the texture itself looks terrible in daylight.
The armor is really washed out and the details aren't that visible. Sure, it looks awesome at night, but looking good half the time isn't ideal.
If you look back at the development shots I posted of the armor, you'll probably notice the slew of smoothing groups I created for each component. Those groups produced the "details" that distinguished one piece from another. An example being the plating around the visor on the helmet.
I set these groups up to help me see what the model would look like before I started texturing. I thought they would be enough to convey the same message when placed in game. That's not to say they don't help, but small lighting changes can't always compete with the normal and specular maps which produce much stronger effects.
Now, I have the original textures for reference, but they aren't that great. I mean, they look good in game, but when you look at them directly, they're very noisy with very faint scratches that are obscured by said noise. I look at these textures and go, "Well now what?"
Since I generate the normal and specular maps from the diffuse and rarely make major modifications, my initial texture was flawed from the get-go. While I do have separate groups for each of the texture types, I kept them somewhat similar. Increased contrast here, toned down details there, but nothing dramatic.
To solve my texture problem, I've decided to go to the root of my texture and start again. I've added a lot of shading/hightlights/scratches/grunge, increased the visibility of details (what used to be 20% opacity is now 50%), increased the strength of the noise, and cleaned up the dragon's jagged appearance. The specular map has contrast increased twice (100 followed by 30 or so) and a higher brightness to compensate. The normal map now covers every piece of armor instead of un-modelled details.
The armor is really washed out and the details aren't that visible. Sure, it looks awesome at night, but looking good half the time isn't ideal.
If you look back at the development shots I posted of the armor, you'll probably notice the slew of smoothing groups I created for each component. Those groups produced the "details" that distinguished one piece from another. An example being the plating around the visor on the helmet.
I set these groups up to help me see what the model would look like before I started texturing. I thought they would be enough to convey the same message when placed in game. That's not to say they don't help, but small lighting changes can't always compete with the normal and specular maps which produce much stronger effects.
Now, I have the original textures for reference, but they aren't that great. I mean, they look good in game, but when you look at them directly, they're very noisy with very faint scratches that are obscured by said noise. I look at these textures and go, "Well now what?"
Since I generate the normal and specular maps from the diffuse and rarely make major modifications, my initial texture was flawed from the get-go. While I do have separate groups for each of the texture types, I kept them somewhat similar. Increased contrast here, toned down details there, but nothing dramatic.
To solve my texture problem, I've decided to go to the root of my texture and start again. I've added a lot of shading/hightlights/scratches/grunge, increased the visibility of details (what used to be 20% opacity is now 50%), increased the strength of the noise, and cleaned up the dragon's jagged appearance. The specular map has contrast increased twice (100 followed by 30 or so) and a higher brightness to compensate. The normal map now covers every piece of armor instead of un-modelled details.
New diffuse vs old diffuse |
That's a bit better |
Friday, February 15, 2013
BDA Beta release on Skyrim Workshop
http://steamcommunity.com/sharedfiles/filedetails/?id=127060429
So far, there haven't been any new issues reported. Then again, it's only been a couple days since release.
Given the silence, I'm off to try and make the armor fit the _1 body better. The clipping may never go away, but I want to reduce it as much as possible.
So far, there haven't been any new issues reported. Then again, it's only been a couple days since release.
Given the silence, I'm off to try and make the armor fit the _1 body better. The clipping may never go away, but I want to reduce it as much as possible.
Wednesday, February 13, 2013
BDA - Almost there
Got the weight scaling sorted. Kind of. The glowing bits had their skin weights reset for some reason. Also, it turns out order matters when it comes to named nodes in the nif. So, when exporting, you need to select multiple pieces in the same order for _0 as you do _1 and vice versa.
I'm tweaking the textures now
I think I'm going to tweak the skin weights some more. The side plates clipping into the thighs bugs me. There's also the annoyance of the _1 proportions making it nigh impossible for the pauldrons and gardbraces to not clip through each other. There's a happy medium around here somewhere.
Edit:
907MB for the source textures...yikes. The torso's source is 500MB.
When exported to dds, the textures come to 61.8MB. This is what I get for using 2048x2048 and 2048x4096 textures. I do believe a half-res version will be required.
I'm tweaking the textures now
The glow should probably be brighter and more saturated |
I think I'm going to tweak the skin weights some more. The side plates clipping into the thighs bugs me. There's also the annoyance of the _1 proportions making it nigh impossible for the pauldrons and gardbraces to not clip through each other. There's a happy medium around here somewhere.
Edit:
907MB for the source textures...yikes. The torso's source is 500MB.
When exported to dds, the textures come to 61.8MB. This is what I get for using 2048x2048 and 2048x4096 textures. I do believe a half-res version will be required.
Sunday, February 10, 2013
BDA - Putting it all together
I finished the preliminary textures for all the armor pieces (diffuse, specular, normal, and glow). However, they're a bit too clean and don't quite look right. I was going to work on dirtying them up, but I ran into yet another issue with the armor.
I made a backup of the .nif while I was trying to get rid of that viewangle clipping issue (still not fixed). So, aside from a name change and a defrag, the file hasn't changed at all since that first screenshot. No amount of fiddling with the first "working" .nif makes it light up normally (despite what http://wiki.tesnexus.com/index.php/Skyrim_common_mod_issues tells me).
I tried exporting to a new file and got this:
The glowing polygons all move with the body, but the main torso won't.The skin modifiers work in 3ds Max, but when I export, only the glowing polygons work. Exporting just the main torso doesn't work at all. It'll only show the t-pose. Even after fixing up the body partitions, vertex color flags, UV set number (4097), Unknown flag (8 -> 0 and vice versa), Has Normals flag, using existing LightingShaderProperties, and Updating Tangents, nothing works.
I can import the nif into 3ds Max and it'll deform. If I re-export it from there, the model moves with animations, but is permanently black. No manner of tweaking fixes it.
I've already installed and reinstalled different versions of the NIF exporter and, from the initial viewangle clipping problem, even tried a completely different machine and version of 3ds Max.
I'm pretty sure both are deforming |
The Bassrath-Kata Bow, the Duster, and now, the Blood Dragon Armor set. None have gone smoothly.
The Bow resolved itself with a bit of fuss, the Duster still has random polygons shoot off into oblivion, and the Blood Dragon Armor won't let me export the torso properly. The boots, helmet, and gloves exported like a charm, but the torso? No no no no, can't have that now, can we?
Edit:
My export settings:
Second Edit:
The T-pose issue: SLSF1_Skinned became unchecked. "..."
Still no idea what's going on with the viewangle clipping or why the Duster randomly spazzes.
Labels:
3d modeling,
Blood Dragon Armor,
not working,
Photoshop,
Skyrim
Wednesday, February 6, 2013
BDA Texture Progress 2
I've added the tarnish and scratches to the torso's diffuse texture. I'm working on the normal and specular maps now. Up next, the helmet, gloves, and boots.
I've reached a decision on the armor's skinning woes: I'm going to release it, broken. I can't find a solution to the problem, haven't found the same problem online, and SoMuchMorpher only produces the "You must specify at least one bone" error whose only solution so far has been "Lol. After I looked at it, I figured out what was wrong. :D" as a YouTube comment from a year ago.
Reasoning: I'm going to crowd-source finding a solution to this problem. I haven't made enough noise on my own, so having however many people downloading, using, and pointing out the problem should help. If it works, I'll know why the problem occurred in the first place. If it doesn't, at least I'll have released another armor mod.
Labels:
3d modeling,
Blood Dragon Armor,
Photoshop,
Skyrim
Thursday, January 31, 2013
BDA Texture Progress
I was hoping for a complete image of the Bioware Dragon, but they are either too elusive for me or non-existent outside of the studio. So, I found a couple images that covered most of the dragon and used the Pen tool in Photoshop to draw the rest in. From there, I erased most of the splatters that covered the image and pasted the finished dragon on the torso textures twice and matched them up along the seam (It required some skewing and masking). The dragon and its placement isn't perfect, but I'm happy with it.
I've stripped the armor's scene of everything but the new armor mesh. I'm thinking there's something wrong or missing in the hidden assets, so if I delete everything but what I made and import an existing Skyrim armor with the skeleton at the same time, I won't have the view-angle clipping. Should it fail, I may have to find someone who is better versed in 3ds Max and ask if they could do it for me.
Sunday, January 20, 2013
Checking things off...or not
I finished the sticker and poster a couple days ago.
My Portal 2 map and Backstock work haven't had any progress lately. My desktop has failed to go to hybrid sleep twice (fade to black followed by a hang) and both times Steam has decided it needs to "update" Portal 2. "Update" meaning download all 6.6GB of content and completely replace all files.
I've tried moving all existing Portal 2 content to the download directory, but I didn't change the timestamps and a casual glance did not reveal a progress log. So, Steam proceeded to redownload everything.
Checking game cache integrity doesn't help either. It just says "1 file failed to validate" and continues downloading everything. Subsequent checks freeze and make Steam unusable ("Hey, Hey wait. Validation window is active, can't let you do that, Dave."), while it continues to download everything. I haven't checked after it finishes downloading. Mainly because I'm afraid it'll start downloading everything again.
Setting "Do not automatically update this game" once the download starts doesn't do anything (makes sense, but only if files are overwritten; they aren't until everything is downloaded). I've set it again and disabled cloud support to see if this download frenzy starts again.
I can start the Authoring Tools while the download is going, but once the download stops and file moving begins, things will get hairy (locked resources). It doesn't matter anyways because the game won't start. Well, it will in that it loads the intro video (completely ignoring the global -novid parameter set in Steam) and then crashes because it can't find the background image (materials/console/portal2_product2_widescreen.vtf). This crash isn't new as the Source Filmmaker couldn't handle this error when it came out, either (I haven't started it since).
You'd think Valve could, I don't know, have a default color to display should the background not exist. Maybe the error I see is the start and the game simply goes into catastrophic failure so fast, it can't show all the other errors (given the error box spam I've seen since Windows 98 SE, I doubt this is the case), but that doesn't explain why the game doesn't die until I acknowledge the missing background (suggesting there are no other errors to show).
Also, Steam is ignoring game activity. It used to stop downloading if I so much as started the Portal 2 Authoring tools. Now, it keeps downloading regardless of what I start (also confirmed with a friend's machine). Steam just sucks up all available bandwidth until it is sated.
Steam is still intermittently starting games other than what I double-clicked on. "Double click on Borderlands 2? Well, I'd better start Blade Symphony." I use the alphabetical list and don't recall it choosing a game after my choice alphabetically, just before.
My Portal 2 map and Backstock work haven't had any progress lately. My desktop has failed to go to hybrid sleep twice (fade to black followed by a hang) and both times Steam has decided it needs to "update" Portal 2. "Update" meaning download all 6.6GB of content and completely replace all files.
I've tried moving all existing Portal 2 content to the download directory, but I didn't change the timestamps and a casual glance did not reveal a progress log. So, Steam proceeded to redownload everything.
Checking game cache integrity doesn't help either. It just says "1 file failed to validate" and continues downloading everything. Subsequent checks freeze and make Steam unusable ("Hey, Hey wait. Validation window is active, can't let you do that, Dave."), while it continues to download everything. I haven't checked after it finishes downloading. Mainly because I'm afraid it'll start downloading everything again.
Setting "Do not automatically update this game" once the download starts doesn't do anything (makes sense, but only if files are overwritten; they aren't until everything is downloaded). I've set it again and disabled cloud support to see if this download frenzy starts again.
I can start the Authoring Tools while the download is going, but once the download stops and file moving begins, things will get hairy (locked resources). It doesn't matter anyways because the game won't start. Well, it will in that it loads the intro video (completely ignoring the global -novid parameter set in Steam) and then crashes because it can't find the background image (materials/console/portal2_product2_widescreen.vtf). This crash isn't new as the Source Filmmaker couldn't handle this error when it came out, either (I haven't started it since).
You'd think Valve could, I don't know, have a default color to display should the background not exist. Maybe the error I see is the start and the game simply goes into catastrophic failure so fast, it can't show all the other errors (given the error box spam I've seen since Windows 98 SE, I doubt this is the case), but that doesn't explain why the game doesn't die until I acknowledge the missing background (suggesting there are no other errors to show).
Also, Steam is ignoring game activity. It used to stop downloading if I so much as started the Portal 2 Authoring tools. Now, it keeps downloading regardless of what I start (also confirmed with a friend's machine). Steam just sucks up all available bandwidth until it is sated.
Steam is still intermittently starting games other than what I double-clicked on. "Double click on Borderlands 2? Well, I'd better start Blade Symphony." I use the alphabetical list and don't recall it choosing a game after my choice alphabetically, just before.
Wednesday, January 16, 2013
Getting back to it
Took a nice break from developing anything and shot down enough Hyperion Loaders to sink a freighter or two. Now, I feel refreshed and ready to tackle my ToDo list:
- Make poster and stickers for a local business
- Make new Portal 2 map
- Detail Backstock maps
- Recreate more Dragon Age series weapons to Skyrim
- Work on BDA texture (if it's going to fail, it'd better look awesome)
Saturday, January 12, 2013
BDA - Breaktime
Four days and I've gotten nowhere with the clipping issue. I've asked for help on the Nexus forums, but no one has replied and my thread was quickly shoved to the 4th or 5th page. I can't find anything online either. It's all "I forgot to select the armor for the BSDismember modifier. I'm an idiot. Lol!"
I can make it shift around randomly, but I cannot make it go away. I cannot use Skin Wrap on vanilla armor as it makes the issue appear, I am not going to recreate the entire armor from scratch in the hopes that some cosmic alignment will make it work, and I'm losing patience with weighting each vertex by hand.
I need a break from it. Probably a good long one.
Friday, January 11, 2013
BDA Progress 9 - Halted
Well, it seems no matter what I do, short of recreating the armor from scratch in a brand new file, the selective clipping persists.
I've tried the following:
So, my question is this: Why must every model I create that involves vertex weighting go wrong? The Bassrath-Kata bow refused to work until I repeated the same steps for the umpteenth time, the Leather Duster occasionally shoots a vertex off into the distance, the Blood Dragon Armor port has vertices bunching up regardless of where I weight them, and this Blood Dragon Armor remake changes z-order for no reason other than "screw you, author".
EDIT:
New development: If I export the belt and tassets, Nifskope refuses to render them or show to XYZ axis markers until I delete all BSDismemberSkinInstances in the file.
I've tried the following:
- Using SkinWrap off of female_0 body with no modification
- Recreating BSDismemberModifier
- Having multiple sections in BSDismemberModifier
- Splitting armor up into multiple meshes
- Using dummy textures
- Using no textures (Nifskope only)
- Adding additional geometry to make backfacing elements match frontfacing (No simplified geometry on the back of elements; perfectly shaped)
- Changing the distance of affected elements from the main body of armor by a couple units (any more and it is floating)
- Manually matching weights for layered elements
- Copy/pasting LightingShaders from vanilla Skyrim armor
- Copy/pasting AlphaProperties from vanilla Skyrim armor (specifically, the Heavy Iron Armor with pauldrons)
- Not using Alpha Properties
- Switching between Editable Poly and Editable Mesh (with Skin and Dismember recreation)
- Detaching backfaces as separate elements
- Exporting to BSFadeNode (Skyrim crashed, but Nifskope still displayed the issue)
- Updating tangents in Nifskope
- Deleting the Skin Wrap Modifier and creating a new one
- Separating most of the elements into their own objects (only made more objects act up)
- Used a completely different computer with 3ds Max 2012 x64 to export
- Removing and reinstalling a different version of the nif exporter
- Using 3ds Max 2011 x86
- Running 3ds Max 2011 as administrator
So, my question is this: Why must every model I create that involves vertex weighting go wrong? The Bassrath-Kata bow refused to work until I repeated the same steps for the umpteenth time, the Leather Duster occasionally shoots a vertex off into the distance, the Blood Dragon Armor port has vertices bunching up regardless of where I weight them, and this Blood Dragon Armor remake changes z-order for no reason other than "screw you, author".
EDIT:
New development: If I export the belt and tassets, Nifskope refuses to render them or show to XYZ axis markers until I delete all BSDismemberSkinInstances in the file.
Tuesday, January 8, 2013
BDA Progress 8 - Export to Skyrim
Not a lot of activity since the New Year. I've been busy with various stuff around the house and what work I've done on the armor hasn't been noteworthy (unless you want to hear how I set the skin weights on vertices 10,001 and 5078).
-------------------------------
I'm doing animation tests in Skyrim (see my last post for why). So far, there have been a few issues with the armor.
The first issue is Skyrim clipping the armor when the pieces clipping with each other aren't even intersecting. The first image shows the back of the armor. Look at the distance between the plate and the tasset. Now, look at the second image. The rear tasset is clipping through the plate.
This only occurs if you zoom out from the armor. If you zoom in 1 or 2 notches away from going into first person, the clipping goes away. I've tried separating the elements further, but that doesn't help. I'll try setting skin weights to see if that'll fix it (the tasset's top isn't as influenced by the pelvis as the plate) and, should it not, adding an alpha property.
The second issue is pauldron position when in combat mode. They rise too far up and expose the shoulders too much. I raised them in 3ds Max to better clear the breastplate, but it looks like I'll have to drop them down a bit.
The third issue is also represented in the above 2 screenshots. The torso has no specular map. Cutting out the glowing pieces on the breastplate skews the UV map. I'll have to make the cuts and adjust those parts of the UV, but I might be stuck with more distortion/skewing than desired.
Once these issues are ironed out, I'll start on the female_1 body to get weight sliding working. I've also just realized I'm going to need 2-3 additional helmets per gender because of the different head shapes for the other races. Oh boy...
-------------------------------
I'm doing animation tests in Skyrim (see my last post for why). So far, there have been a few issues with the armor.
The first issue is Skyrim clipping the armor when the pieces clipping with each other aren't even intersecting. The first image shows the back of the armor. Look at the distance between the plate and the tasset. Now, look at the second image. The rear tasset is clipping through the plate.
This only occurs if you zoom out from the armor. If you zoom in 1 or 2 notches away from going into first person, the clipping goes away. I've tried separating the elements further, but that doesn't help. I'll try setting skin weights to see if that'll fix it (the tasset's top isn't as influenced by the pelvis as the plate) and, should it not, adding an alpha property.
The second issue is pauldron position when in combat mode. They rise too far up and expose the shoulders too much. I raised them in 3ds Max to better clear the breastplate, but it looks like I'll have to drop them down a bit.
The third issue is also represented in the above 2 screenshots. The torso has no specular map. Cutting out the glowing pieces on the breastplate skews the UV map. I'll have to make the cuts and adjust those parts of the UV, but I might be stuck with more distortion/skewing than desired.
Once these issues are ironed out, I'll start on the female_1 body to get weight sliding working. I've also just realized I'm going to need 2-3 additional helmets per gender because of the different head shapes for the other races. Oh boy...
Wednesday, January 2, 2013
BDA Progress 7 - Weighting
Skin-wrapping the armor with the female_0 body produced a good base to start from. I'm going through each piece and assigning weights manually now.
The most annoying part is, as usual, the front and rear tassets (anything that covers both legs, actually). Both thigh bones' influences abruptly stop in the middle of each tasset (look at the front tasset, on the right of the image). This means when the legs move, each side of the tassets passes through the other (they flip inside out) and stretch only in the middle (instead of gradually across the entire surface).
3ds Max 2011 still doesn't show skin deformations properly. And, as always, it is only in quadrants 2 and 3 on the XY-grid; well, it is always worse there. Quadrants 1 and 4 only show slight jiggling. The thing is, this jiggling/undesired deformation only occurs in Max, never in game. The jiggling comes regardless of how many bones are influencing. So, I can have only a single bone influencing a single vertex but, once I move the bone, the vertex's position fluctuates wildly.
I'm probably going to need a 75/25 thigh/calf split on the poleyns (knees). 50/50 causes them to change size and, as you can see in the jump, separate from the wrap that is supposed to hold them in place.
The most annoying part is, as usual, the front and rear tassets (anything that covers both legs, actually). Both thigh bones' influences abruptly stop in the middle of each tasset (look at the front tasset, on the right of the image). This means when the legs move, each side of the tassets passes through the other (they flip inside out) and stretch only in the middle (instead of gradually across the entire surface).
3ds Max 2011 still doesn't show skin deformations properly. And, as always, it is only in quadrants 2 and 3 on the XY-grid; well, it is always worse there. Quadrants 1 and 4 only show slight jiggling. The thing is, this jiggling/undesired deformation only occurs in Max, never in game. The jiggling comes regardless of how many bones are influencing. So, I can have only a single bone influencing a single vertex but, once I move the bone, the vertex's position fluctuates wildly.
Skin weighting - Nailed it. |
Subscribe to:
Posts (Atom)