Like what? I didn't know perforce worked with unity, huh. We've been using SourceTree and Svn Tortoise together with Bitbucket and while it hasn't been perfectly smooth, some files do cause trouble, nothing has really hindered any real progress. Not sure why we switched but it's a lot easier to commit stuff now. Haven't run into any merge conflicts yet, I have no idea if that part works as intended :/
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
GAF Indie Game Development Thread 2: High Res Work for Low Res Pay
- Thread starter Jacksinthe
- Start date
- Status
LordRaptor
Member
the big one was Scenes (ie maps), which level designers needed to greybox, artists needed to set dress, sound guys needed to add environment triggers and AI guys needed to navmesh (because the level designers and artists never bothered to) but admittedly this was in older versions of unity so they might have improved that now
I haven't tried collaborate even though my team applied for it and we got in x_x;
Does it allow for more than one people to edit an scene at the same time? That's the only "bottleneck" I've found using Unity and Git.
On the Perforce comments, I think it's kind of the same, but it adds up that if someone has checked out an asset no one can edit that asset on its PC, you have to wait until the person who has it checked it out unchecks it, either by reverting the asset or submitting it. Also unity metadatas are a fucking nightmare
Does it allow for more than one people to edit an scene at the same time? That's the only "bottleneck" I've found using Unity and Git.
On the Perforce comments, I think it's kind of the same, but it adds up that if someone has checked out an asset no one can edit that asset on its PC, you have to wait until the person who has it checked it out unchecks it, either by reverting the asset or submitting it. Also unity metadatas are a fucking nightmare
GroundCombo
Member
Changing all the text components in our UI to use the now-free TextMesh Pro is teaching me a lot of lessons about components, prefabs, manual labor and pain. It's worth it though; TMP is so much better than the stock text that it hurts. And it supports some features that we absolutely will need in the future, so there's that.
I've been able to push scenes with no problems with SourceTree at least. But since we use both ProGrids and ProBuilder, I think each person needs to have the same addons installed in their Unity for the maps to load correctly. Beyond that, I don't think multiple people will have any issues with modern Unity.
I have no idea, but I doubt it. There's no checkout system like in UE4 and/or Perforce as far as I know, so I don't know how merging of scenes work, if at all.I haven't tried collaborate even though my team applied for it and we got in x_x;
Does it allow for more than one people to edit an scene at the same time? That's the only "bottleneck" I've found using Unity and Git.
On the Perforce comments, I think it's kind of the same, but it adds up that if someone has checked out an asset no one can edit that asset on its PC, you have to wait until the person who has it checked it out unchecks it, either by reverting the asset or submitting it. Also unity metadatas are a fucking nightmare
VisceralBowl
Member
I used RGB 1.000, 0.766, 0.336 but I also thought the gold color looked a bit off. It could be the scene's lighting.Thx for making that up! Hmm ... it looks a bit dull. The Fresnel reflection
also looks quite weak towards grazing.
Here is my current attempt for gold;
I use the specular color for gold, i.e. linear RGB (1.0, 0.71, 0.29), and
Schlick for the Fresnel reflection.
It should also be noted that the specular color for gold only holds when the
scene is illuminating with D65 light, i.e. RGB(1,1,1) with the monitor set to
sRGB. Also, if the RGB color of the light is different, the given specular
color for gold won't hold any longer. In that case the specular color needs to
be adapted via some chromatic adaptation. Anyhow, for games nobody will
notice, I guess. And for non-games you won't even use RGB colors.
I think I can still further improve on the result in modelling the grazing
reflection a bit differently. Will see.
timetokill
Banned
I poked around and liked what I saw, but afaik it still can't export to anything yet (including .exe) so I'm sticking to Studio 1 for a while.
Hi. I'm planning on doing a hand drawn top down 2D game. I've been thinking about the animations and my conclusion is that my best solution is doing a basic 3D model, animate and draw over it. What is an easy modeling and animating tool (free if possible).
I'm also open to suggestions in how I should approach te this. Thanks in advance.
I'm also open to suggestions in how I should approach te this. Thanks in advance.
LordRaptor
Member
Blender is free, has a vast number of available tutorials, has a built in animation system with things like inverse kinematics, and will let you export individual animation frames as images - but its UX is very contrarian, and you'll probably want to delve into its settings and add a lot of the newer additions that are not enabled by default (like making left mouse button select things, and turning radial menus on so not everything is via memorised keyboard macros).
Alternatively, you could look into some of the 2D animation software out there like Spine or Spriter (and Unity just bought out a similar system) which will let you assemble "paper dolls" from 2D images and attach a skeleton to them for thing like inverse kinematics and mesh deformation
Was there something said about on what this specular color was based on?
Either it's wrong or it's based on a different illuminant (perhaps D50, will
have to check that). For gold, I can now computed the specular color myself
leading to 1.0, 0.71, 0.29 under D65 lighting, which is also the value given
in Real-Time Rendering (3rd Edition), like:
Gold 1.00,0.71,0.29
Silver 0.95,0.93,0.88
Copper 0.95,0.64,0.54
Iron 0.56,0.57,0.58
Aluminum 0.91,0.92,0.92
(but they, too, aren't saying on what these colors are based on
) What could mess up the colors in your rendering may be the result of the
HDR/tone-mapper pulling the colors towards gray when fitting the HDR
colors/intensities into the limits of the sRGB gamut. If the highlight is
below, say, 1.5 everything should be fine though. You can check by turning
the tone-mapper off.
VisceralBowl
Member
Substance designer also gives the same reflectance value. No idea how Epic/Allego got their values.Was there something said about on how this specular color was based on? Either
it's wrong or it's based on a different illuminant (perhaps D50, will have to
check that). For gold, I can now computed the specular color myself leading
to 1.0, 0.71, 0.29 under D65 lighting, which is also the value given in
Real-Time Rendering (3rd Edition), like:
Gold 1.00,0.71,0.29
Silver 0.95,0.93,0.88
Copper 0.95,0.64,0.54
Iron 0.56,0.57,0.58
Aluminum 0.91,0.92,0.92
(but they, too, aren't saying on what these colors are based on
What could mess up the colors in your rendering may be the result of the
HDR/tone-mapper pulling the colors towards gray when fitting the HDR
colors/intensities into the limits of the sRGB gamut. If the highlight is
below, say, 1.5 everything should be fine though. You can check by turning
the tone-mapper off.
I disabled the default tonemapper for this pic. The commands for UE 4.15+ have changed now that they're using ACES by default. I'm no longer sure how to change the tonemappers or view the raw output as not all of the old commands are working.
Still working on Light Fairytale.
Here's a few more work-in-progress shots from the same location as last time.
Love the blurred background look
BatavianBunny
Neo Member
I did press that "publish" button in Steam finally.
Now if only VR gamers would like arcade games instead of wave shooters ^^
phew
Wooo! Good luck man!
So we know that there are at least some inconsistencies among them.
I ripped an ACES mappers (global one, independent of other pixels) from the
net and implemented it straight. Indeed, it takes a bit of the color
saturation away making the picture look less vibrant.
Make Diesel Power Great Again lol xDI did press that "publish" button in Steam finally.
Now if only VR gamers would like arcade games instead of wave shooters ^^
phew
Btw, what's the thing below the car, heated exhaust?
LordRaptor
Member
Allegorithmic do have a PDF explaining their methodology which also cites additional references, so they're basing it on something certainly.
There is a heat distortion effect at the exhaust pipes but the cubes beneath the car mainly visualize the player where the ground contact is. Kind of a hover ground effect ripping the floor appart
LordRaptor
Member
I like it, its certainly a nicer solution than drop shadows which in any light source other than directly overhead are misleading as to player position
It appears to me that the difference for gold is perhaps just a difference in
gamut mapping (gold falls outside the sRGB gamut), but I don't know what sort
of mapper was used in the book Real-Time Rendering to get their value for gold.
Whatever, I think the value for gold can be improved by improving the
interpolation scheme for the daylight (D65) values and also by using a more
improved gamut mapper which doesn't just scale by the largest value.
Perhaps it would be cool to write an entire daylight simulator to compute
all Dxx illuminants for computing the specular color for gold etc. with
respect to different states of the day. Or why not a Planck black body radiator
such that we can compute the specular colors under, say, tungsten lighting?
Would be interesting to see how gold renders under warm lighting say CCT 2400K
for example. With the black body radiator it would also be possible to render
the specular colors under say skylight, i.e. around CCT 10000K. Well, now that
I talked about it, I think I'm going to program such a simulator next week and
will try to render some of the metals under some warm/cold lighting.
LordKasual
Banned
For anyone interested: Lab Zero is currently looking for a 2D animator for contract work on Indivisible. All info here: https://labzerogames.com/jobs-2d-animation-contractors/
I thought I'd give it a try myself, mainly to see if I could make it, but I fear this is outside my skills. I'm trying to get the application done remotely quickly, as I feel it's an integral part of what they're looking for, and I'm wondering how to balance the cleanliness of the animation. Here's my current attempt for the Ajna anim:
The right hand motion is bothering me, though I feel a bit constrained because of the imposed keyframe's hand position (could've just done another motion, you might say, but I didn't get any other looking right).
I'm mainly wondering if I should spend time cleaning up the rough torso/hand lines more, so they're in line with the head/legs, or if that's time wasted as they're specifically mentioning that cleaners will rework the animation and that some roughness is acceptable.
So, thoughts? Should I rework the torso/hands to be cleaner? Should I completely rethink the right hand motion because it's too off? Should I just give this up and go back to working on Pacha?
There's still a whole other motion to do, which will take me at least another day, and I fear taking longer than 2 days for the overall application would be overkill, but I have no prior experience with contract animation, so I don't actually know - thoughts on that appreciated as well!
(also, good luck to any other who'd apply!)
Dude that looks awesome. I guess you're talking about how her hand jerks during the kick?
I think its fine, but you could use that to your advantage if you maybe added a single "impact" frame during the kick. Like where her weight keeps moving into the kick but her leg stops. It would make the attack look alot more crunchy / powerful. But I feel like that kind of thing is usually artificially added with hitstop...depends on the game, IDK.
But I think you did a damn fine job. Your animations definitely remind of me Lab Zero's more fluid style.
Edit:
Oh i'm late, sorry lol
Essential Mathematics for Games and Interactive Applications
(Damn did they make the third edition cover ugly)
This is a new series that is pretty good:
Foundations of Game Engine Development: Mathematics
https://www.amazon.com/dp/0985811749/?tag=neogaf0e-20
The author has written various other mathematics books before that are also published on Amazon. This new series he is doing is self published.
Foundations of Game Engine Development: Mathematics
https://www.amazon.com/dp/0985811749/?tag=neogaf0e-20
The author has written various other mathematics books before that are also published on Amazon. This new series he is doing is self published.
Hey all. I'm trying to put a small team together to work on a Gone Home style game.
Basically, I just really need a project right now that I can focus on and I want to make a game that requires less from my weak points (code, animation, art, audio) so I don't give up on it. I also feel like having a team would help me stick with it as well.
Is anyone looking for a project to join, or know of where I can find a team?
Basically, I just really need a project right now that I can focus on and I want to make a game that requires less from my weak points (code, animation, art, audio) so I don't give up on it. I also feel like having a team would help me stick with it as well.
Is anyone looking for a project to join, or know of where I can find a team?
Actually, I've heard that Unity has another new project system in beta, saw a link on fb the other day. Could be something!
LordRaptor
Member
Unity 5 final comes out at the end of this month (along with EOL support for any legacy 'bought outright' pro owners), so I wouldn't be surprised if they need something to upsell the new rebranded Unity 2017
ThatStupidLion
Member
For anyone interested: Lab Zero is currently looking for a 2D animator for contract work on Indivisible. All info here: https://labzerogames.com/jobs-2d-animation-contractors/
I thought I'd give it a try myself, mainly to see if I could make it, but I fear this is outside my skills. I'm trying to get the application done remotely quickly, as I feel it's an integral part of what they're looking for, and I'm wondering how to balance the cleanliness of the animation. Here's my current attempt for the Ajna anim:
The right hand motion is bothering me, though I feel a bit constrained because of the imposed keyframe's hand position (could've just done another motion, you might say, but I didn't get any other looking right).
I'm mainly wondering if I should spend time cleaning up the rough torso/hand lines more, so they're in line with the head/legs, or if that's time wasted as they're specifically mentioning that cleaners will rework the animation and that some roughness is acceptable.
So, thoughts? Should I rework the torso/hands to be cleaner? Should I completely rethink the right hand motion because it's too off? Should I just give this up and go back to working on Pacha?
There's still a whole other motion to do, which will take me at least another day, and I fear taking longer than 2 days for the overall application would be overkill, but I have no prior experience with contract animation, so I don't actually know - thoughts on that appreciated as well!
(also, good luck to any other who'd apply!)
Good luck keep us posted! I dont know what the restrictions are for the animation tests, but this is looking good! I would say the shoulder pop is more jarring to me than the hand. I know its a few days late but good luck! Also, dont know if you tried this already but it might help to keep the left foot/leg more planted rather than keep rotating through the whole action, looks really good though! Love seeing your anims,
Thank you both
I'll keep your advice in mind for next time! I honestly don't think this attempt will go through, but I actually don't mind, since I have Pacha to work on anyway. On that note, progress will be slightly slower for a little while, as I've just thrown off my back pretty badly (again) - let that be a lesson to anyone who doesn't take proper care of theirs. It'll catch up with you when you least expect it!BatavianBunny
Neo Member
Hey guys... don't know if its ok to to post on here for Kickstarter or Greenlight? I'm so wracked with excitement and nervousness!
Anyway... if we can, please just delete this.. if not, give it a look please!
Our Kickstarter page https://www.kickstarter.com/projects/criticalforge/forged-of-blood
Our Steam Greenlight Page: http://steamcommunity.com/sharedfiles/filedetails/?id=877620687
Our album: http://imgur.com/gallery/MwkSG
Anyway... if we can, please just delete this.. if not, give it a look please!
Our Kickstarter page https://www.kickstarter.com/projects/criticalforge/forged-of-blood
Our Steam Greenlight Page: http://steamcommunity.com/sharedfiles/filedetails/?id=877620687
Our album: http://imgur.com/gallery/MwkSG
Here is a first implementation of a pseudo-spectral renderer. xD
I implemented the black body radiator's SPD (taken from Wiki) to generate
myself some spectral light sources. The D ones will take a bit more time, for
I want to go with the SPD vector-basis approach for the D illuminats because
I think this is a pretty cool approach for generating many other spectral
sources instead of using only tabulated values fixing the source. Anyhow.
So I quickly tried to compute some new specular colors for gold unter some
given blackbody temperatures like;
2856K (a tungsten light bulb, illuminant A)
9500K (skylight)
(for reference I used tabulated D65)
2856K
specular color gold: (1.000, 0.369, 0.022), linear, RGB, scaled (R:1.76)
D65 (tabulated, linear interpolated)
specular color gold: (1.000, 0.766, 0.336), linear, RGB, scaled (R:1.05)
9500K
specular color gold: (0.911, 0.785, 0.498), linear, RGB
The 2856K case is pretty tough, for the specular color is way out of sRGB,
much more than the D65 case. But it's interesting to note that the color for
9500K lies in sRGB and represents as such the best reproduction. But even if
the colors are out of gamut, it is still much better than just using the
specular color for D65 in all other cases. For, using the specular color from
D65 in the 2856K and 9500K case would make the reflection much more yellowish.
Only the specular color for gold was spectrally corrected in here. The renderer
is still full RGB. And all diplayed RGB values are linear. To match the
blackbody color in RGB the in-engine light source and background RGB color
were computed from CCT (2856K, 9500K).
Anyhow. That's my first attempt. Bugs are all mine!
I implemented the black body radiator's SPD (taken from Wiki) to generate
myself some spectral light sources. The D ones will take a bit more time, for
I want to go with the SPD vector-basis approach for the D illuminats because
I think this is a pretty cool approach for generating many other spectral
sources instead of using only tabulated values fixing the source. Anyhow.
So I quickly tried to compute some new specular colors for gold unter some
given blackbody temperatures like;
2856K (a tungsten light bulb, illuminant A)
9500K (skylight)
(for reference I used tabulated D65)
2856K
specular color gold: (1.000, 0.369, 0.022), linear, RGB, scaled (R:1.76)
D65 (tabulated, linear interpolated)
specular color gold: (1.000, 0.766, 0.336), linear, RGB, scaled (R:1.05)
9500K
specular color gold: (0.911, 0.785, 0.498), linear, RGB
The 2856K case is pretty tough, for the specular color is way out of sRGB,
much more than the D65 case. But it's interesting to note that the color for
9500K lies in sRGB and represents as such the best reproduction. But even if
the colors are out of gamut, it is still much better than just using the
specular color for D65 in all other cases. For, using the specular color from
D65 in the 2856K and 9500K case would make the reflection much more yellowish.
Only the specular color for gold was spectrally corrected in here. The renderer
is still full RGB. And all diplayed RGB values are linear. To match the
blackbody color in RGB the in-engine light source and background RGB color
were computed from CCT (2856K, 9500K).
Anyhow. That's my first attempt. Bugs are all mine!
What are you after?
Thought I'd put this here.
I work in PR (we've done some games, got another decent sized gaming project launching in July) but I want to expand my reach a little.
If anyone wants any help (pro bono of course) with stuff like press releases, dealing with media inquiries, pulling together lists of journalists to send out to etc then please do drop me a PM.
I work in PR (we've done some games, got another decent sized gaming project launching in July) but I want to expand my reach a little.
If anyone wants any help (pro bono of course) with stuff like press releases, dealing with media inquiries, pulling together lists of journalists to send out to etc then please do drop me a PM.
Thanks a lot!
VisceralBowl
Member
@ missile
I did some research this morning and tweaked the scene's lighting intensities to be more physically correct. In addition I also baked the light which fixed a few problems I noticed with previous dynamic-only setup. While the scene no longer matches yours the values are more correct, the reflections are a little more accurate (UE's reflection probes only capture light when using lightmass) and most noticeably the extra light has slightly shifted the colors slightly.
The images use the same fixed exposure settings. I swapped the black fog for an atmospheric fog as the black fog was causing weirdness with the bakes.
I did some research this morning and tweaked the scene's lighting intensities to be more physically correct. In addition I also baked the light which fixed a few problems I noticed with previous dynamic-only setup. While the scene no longer matches yours the values are more correct, the reflections are a little more accurate (UE's reflection probes only capture light when using lightmass) and most noticeably the extra light has slightly shifted the colors slightly.
The images use the same fixed exposure settings. I swapped the black fog for an atmospheric fog as the black fog was causing weirdness with the bakes.
I did press that "publish" button in Steam finally.
Now if only VR gamers would like arcade games instead of wave shooters ^^
phew
Looks nice! VR arcade games are where it's at, but I'm biased.
Just_myles
Member
Man you are always in experimentation mode. So interesting the things you do.
I uploaded a very early alpha demo of Beastmancer in IndieDB, in case someone wants to try it and give me some feedback (this version is for Win64 systems): http://www.indiedb.com/games/beastmancer/downloads/beastmancer-alpha-demo-win64
NipplesAndToes23
Member
I might try going back into Unity one last time.
Looks better. But there is an old saying; a reflection is only as good as the@ missile
I did some research this morning and tweaked the scene's lighting intensities to be more physically correct. In addition I also baked the light which fixed a few problems I noticed with previous dynamic-only setup. While the scene no longer matches yours the values are more correct, the reflections are a little more accurate (UE's reflection probes only capture light when using lightmass) and most noticeably the extra light has slightly shifted the colors slightly.
The images use the same fixed exposure settings. I swapped the black fog for an atmospheric fog as the black fog was causing weirdness with the bakes.
...
environment it reflects. So, yeah, it looks better. But I'm still not
convinced about the gold reflection. There seems to be some inconsistencies
and the question is where in Unreal (assuming my one is correct xD) the
problem occurs. I still think it's some form of post-processing which offsets
the color of the reflection.
I think we should render are more easy setup without any post-processing to
see whether we are able to reproduce the specular color at all.
Here is my new setup;
backround color: (1.0, 1.0, 1.0)
ground plane: size 2x2, color: (1.0,1.0,1.0) and (0.5,0.5,0.5), diffuse 0.5
sphere: center (0.0, 0.25, 0.0), radius 0.25, diffuse 0.0, specular color
gold (1.0,0.766,0.336), roughness small (make the spot small at least)
light: position (0.0, 2.0, -1.0), attenuation off
camera: position (1.0, 0.8, 1.0), lookat (center of sphere)
HDR: off
gamma: off
The only thing left on, on my end, is the the RGB clipper clipping off RGB
values (hard) to 1.0, which is the case for the highlight in my example, but
isn't important for our case unless the highlight has a long "tail" covering
the sphere to a greater extend changing colors. Hence, keeping it small won't
have any impact for the measurement to come.
Given this setup the center pixel has a viewing angle of 0, i.e. dot(V,N)=1
(make the triangles small enough). Under these conditions any Fresnel
implementation should reproduce the specular color, i.e. the F0 reflectance of
the material, in our case: (1.0,0.766,0.336), or RGB(255,195,86) (truncated).
The color of the center pixel in my rendering is RGB(255,195,86), verifying
the computation.
Whatever the lighting, post-processing etc., if this specular color can't be
reproduced at normal incidence, the colors of the reflection will be off no
matter how cool the environment. Hence, if you can reproduce this specular
color, anything else is just some fancy addon to direct the representation of
the scene.
I hope to come out of experimentation some day to actually ship a product. xD
But won't happen as fast, for I'm still after some special styles/looks I
envision. One of such things is stylized reflection at grazing producing cool
looks. Here is a shoot I did for the sphere as of late manipulation the
reflectance at grazing. The background and the sphere won't do this effect
justice, it will have a much greater impact on low-poly graphics. Cool thing
is, the colors of the highlight will also take on the color of the changed
reflectance, sort of simulating some fancy metallics.
VisceralBowl
Member
I did more research and it may in fact be altered. Check your PMs!
Indeed, it's altered. I'm wondering what's the rational behind it.
Edit:
I always had this feeling that there was something wrong with Unreal's
graphics when looking at peoples work using it. It always looks a bit dull.
I don't know if you're referring to Unreal Engine 3 (UDK) or Unreal Engine 4 specifically, but it can produce a massive variety of outputs. If people just use default settings then there can be some similarities, sure.
If you don't mind me asking, why do you go back and forth in the thread asking people to check things in Photoshop and UE4 instead of instrumenting them yourself? It would probably be more efficient to do that, especially since UE4 has full source access, even for the editor.
Don't know which version people use, but from what I've seen I got this
impression from.
Well, yeah, I don't have PS nor UEx. However, the reason is that > 2 eyes see
more than just 2. And there are people in here pretty confident using said
programs able to reproduce similar stuff quickly. I also find it interesting
how the results different in the end when done by different people.
Fair enough. If you're ever curious about say, source code values or formulas used to calculate shader output or whatever, the UE4 source is free to access (and improve, since I think some people submit changes that make it into later versions).
Made this a little bit back but I recently updated the second to last sprite for what I think I will use as my main character. I had some help finalizing the design but in the end it sort of delve into bickering so I had to make the final decision myself. I think it looks nice. The 2nd Gen 3 sprite is the WIP battle outfit.
But yeah, I think it is good and time to move on to other matters.
- Status