Support NeoGAF

[JulianImp]

Do any of you guys know anything about creating speed videos of 3D modelling and/or level design? I'm thinking of recording some Unity/Unreal/Creation Kit work of mine, but I have no idea what kind of PC and software I need? Should I have a set plan before starting? Maybe not record something "live", but a recreation of some previous project of mine where mistakes will be much fewer? How do you do it?

I guess it depends on who you're asking, but personally I'm a lot more into live videos since they help you understand the creator's train of thought, and might even teach you something from their mistakes or the things they improve upon as they go. I'd only really consider watching scripted videos if they're about a very specific subject, since those are the times when you want to cut straight to the point (say, to see how someone creates a shader effect or a model in a particular aesthetic).

I tried streaming a couple nights ago and obviously got no watchers, but after I was done and checked the VoD I realized my microphone's volume was awfully low, so I'll have to look into that. I notice the stream's dropping video frames at times, but hopefully it's stable enough to be serviceable.

 

[Minamu]

I guess it depends on who you're asking, but personally I'm a lot more into live videos since they help you understand the creator's train of thought, and might even teach you something from their mistakes or the things they improve upon as they go. I'd only really consider watching scripted videos if they're about a very specific subject, since those are the times when you want to cut straight to the point (say, to see how someone creates a shader effect or a model in a particular aesthetic).

I tried streaming a couple nights ago and obviously got no watchers, but after I was done and checked the VoD I realized my microphone's volume was awfully low, so I'll have to look into that. I notice the stream's dropping video frames at times, but hopefully it's stable enough to be serviceable.

Yeah that's a good point. I wouldn't be doing it for views or revenue, it would be more to add another way of proving I can handle things like Unity and Unreal. That's why I mentioned mistakes, feels weird to record my failures and post on my portfolio or LinkedIn, you know? I've watched some guys on Youtube make complete levels via speed videos and it seems like a good way to show your skills, but it seems kinda orchestrated at the same time.

 

[JulianImp]

Yeah that's a good point. I wouldn't be doing it for views or revenue, it would be more to add another way of proving I can handle things like Unity and Unreal. That's why I mentioned mistakes, feels weird to record my failures and post on my portfolio or LinkedIn, you know? I've watched some guys on Youtube make complete levels via speed videos and it seems like a good way to show your skills, but it seems kinda orchestrated at the same time.

Yeah, it mostly depends on what you want to go for. Of course making mistakes or taking too long to do something might be bad if you want to show those videos as part of your portfolio, but if you want to polish your skills then it might be better to do live videos (or even stream as you work) to improve.

I don't know a thing about how portfolios work in modelling, but would speed videos actually help you more than just having your completed works as part of a demo reel? If anything, I think stuff such as adding the number of man-hours it took you to complete each work would be a lot more helpful than staged speed videos where you make something a second time.

 

[Ashodin]

Just sent over a request to Nintendo to work on the Switch. EXCITED

 

[Makai]

Just sent over a request to Nintendo to work on the Switch. EXCITED

How did you do that?

 

[JulianImp]

Had some trouble with Unity's built-in one-way 2D platforms, namely that the platforms would collide with the player and stop him from moving if you bumped into them from the sides. The proposed solution was changing the effector's surface arc to be 179 degrees or less (yay for hacks!), but that somehow screwed up with the collider-based grounded checks I was running for the player character, meaning I ended up adding a collider directly below the platform's and slightly wider than it, and made that a trigger zone that'd disable collision checks between the player and the one-way platform layers... which almost makes the whole platform effector system redundant, except for some cool corner-case interactions that prevent me from removing that component entirely.

...And I still have a corner-case interaction where jumping just high enough that you clear the trigger zone but low enough that you still hit the platform from the sides stops your horizontal momentum anyway. Hopefully it won't actually happen often enough in normal gameplay to become noticeable.

 

[mabec]

How did you do that?

https://developer.nintendo.com/ ?

 

[Blizzard]

Just sent over a request to Nintendo to work on the Switch. EXCITED

Last I heard you had to submit a video proposal/demo of a game to have a chance of getting approved as an indie developer. I'd love to apply but I don't have one ready yet. Did something change?

I'm already an approved Wii U developer if it makes any difference.

*edit* Yeah I just logged in again, and I can't see anything about the Switch on that site at all once logged in.

 

[JulianImp]

https://developer.nintendo.com/ ?

But the site only mentions 3DS and WiiU dev-ing so far... I guess talking about the approval process might go against some NDA, but it'd be nice to hear about it if it isn't the case.

 

[missile]

Thanks couldn't help myself and added some subtle lightning inside the fog

How you made it and what engine are you using anyways?



Meanwhile.
I did some further progress trying to come up with some proper pixelized
volume rendering. Well, I think I got something....

 

[Minamu]

Yeah, it mostly depends on what you want to go for. Of course making mistakes or taking too long to do something might be bad if you want to show those videos as part of your portfolio, but if you want to polish your skills then it might be better to do live videos (or even stream as you work) to improve.

I don't know a thing about how portfolios work in modelling, but would speed videos actually help you more than just having your completed works as part of a demo reel? If anything, I think stuff such as adding the number of man-hours it took you to complete each work would be a lot more helpful than staged speed videos where you make something a second time.

The main point a guy made that made me interested is that it visually proves I know how to use the tools, by actively showing me moving around the editor environment. But a finished product is probably just as good for such a purpose, though easier to "fake" I guess.

But all in all, I still don't know anything about the tools required or if my really old pc can handle recording Is it as simple as using fraps? The file sizes go up in several gigs really easy and there's a time limit too I believe.

 

[DemonNite]

How you made it and what engine are you using anyways?



Meanwhile.
I did some further progress trying to come up with some proper pixelized
volume rendering. Well, I think I got something....

Im using Unity but trying my best to make it not look like a Unity game lol

Heres a quick breakdown of the "Darkness" particle system:

https://gfycat.com/WelloffThirstyAtlanticspadefish

Basically, the base is a sphere shape particle system with 2 point lights in side. I personally think that getting a good looking fog/smoke particle system is making sure you have a good texture sheet animation to go with it.

In the end of the clip I dropped in a yellow point light for demonstration in how it looks in-game (from the ambient lights in the level). The particles that shoot out are cone shaped volumes.

 

[Timedog]

I'm fucking around with SFML and I really like it but holy shit it puts the L and R trigger on xbox controllers onto the same "axis". So the left trigger has values from -100 to 0, depending on how hard you hold it down, and the right trigger has values from 0 to 100. But since they're on the same "axis" if you hold both down you get a value of 0, since they cancel each other out.

This means that a very large portion of games can't be made with SFML if you want proper controller support. Any game where both triggers need to be used at the same time (Racers, FPS, TPS, etc). WTF.

 

[Popstar]

That's not SFML. That's just the way Microsoft's Xbox controller USB HID driver works. It's the same in any framework that uses Raw Input (or APIs written on top of Raw Input like DirectInput).

It's deliberately broken to encourage the use of Microsoft's proprietary stuff – which they can charge controller manufacturers a license fee for – instead of straight USB. Just Microsoft being assholes really.

 

[missile]

Im using Unity but trying my best to make it not look like a Unity game lol

Heres a quick breakdown of the "Darkness" particle system:

https://gfycat.com/WelloffThirstyAtlanticspadefish

Basically, the base is a sphere shape particle system with 2 point lights in side. I personally think that getting a good looking fog/smoke particle system is making sure you have a good texture sheet animation to go with it.

In the end of the clip I dropped in a yellow point light for demonstration in how it looks in-game (from the ambient lights in the level). The particles that shoot out are cone shaped volumes.

Pretty cool!

I will also try making a fog/smoke animation further down using some time
dependent density distributions which will behave like smoke or something.
Another cool thing would be to actually compute some rough 3d airflow and
transport a density along, but that's computationally somewhat demanding,
I guess. Yeah, lets keep it easy.

 

[Tain]

I'm fucking around with SFML and I really like it but holy shit it puts the L and R trigger on xbox controllers onto the same "axis". So the left trigger has values from -100 to 0, depending on how hard you hold it down, and the right trigger has values from 0 to 100. But since they're on the same "axis" if you hold both down you get a value of 0, since they cancel each other out.

This means that a very large portion of games can't be made with SFML if you want proper controller support. Any game where both triggers need to be used at the same time (Racers, FPS, TPS, etc). WTF.

Last time I messed with SFML I bypassed its input calls and just used XInput. It's really easy and splits out the triggers.

 

[DemonNite]

Pretty cool!

I will also try making a fog/smoke animation further down using some time
dependent density distributions which will behave like smoke or something.
Another cool thing would be to actually compute some rough 3d airflow and
transport a density along, but that's computationally somewhat demanding,
I guess. Yeah, lets keep it easy.

cant wait to see those results!

you're making an engine! you are already passed keeping it easy

 

[embalm]

I've been working on the Combat engine for my RPG and I let in a little scope creep. I got an idea for defensive stances and had to have it. So now if you have certain equipment equipped you'll be able to ready yourself against on coming attacks. Counter stance for example allows you to immediately strike back against any enemy that hits you.

There is something really satisfying about watching all these numbers pop up as monsters hit and miss.

I'm going to refactor my initiative list and finish up the character action menu, but I can see the light at the end of the tunnel now and it's a good feeling.

 

[missile]

cant wait to see those results!

you're making an engine! you are already passed keeping it easy

xD Oh yeah, thx for remembering me! Well, after month dealing with my engine
the phrase "keep it easy" turns into something like not to drain all the
cycles in a sudden. It's not so much about any one technique, but a quest on
how to implement it reasonable fast to be usable in a game. An 3d airflow
wouldn't be a big problem for me to implement, but anything meaningful (say
on a grid of say 128x64x64) would make my engine stop dead without any proper
optimization. But I will give it a try sooner than later!

 

[JulianImp]

The main point a guy made that made me interested is that it visually proves I know how to use the tools, by actively showing me moving around the editor environment. But a finished product is probably just as good for such a purpose, though easier to "fake" I guess.

But all in all, I still don't know anything about the tools required or if my really old pc can handle recording Is it as simple as using fraps? The file sizes go up in several gigs really easy and there's a time limit too I believe.

You can also use OBS (Opensource Broadcaster, I think), and if you're doing a sped-up timelapse then I guess you could also get away with having a pretty low capture framerate (maybe 15 fps for 4x speed, 10 for 6x or even 6 for 10x), since you'll be speeding the footage up anyway and that'd get you back to 60 fps.

 

[oxrock]

A buddy of mine commented that my original game logo was done in super low res and my axe's lack of detail bugging me, so I decided to work on my game's logo last night.

original:

updated:

I don't know why I've invested so much time working on this honestly, just needed a break from my regular grind I guess. Hopefully I can leave well enough alone and get back to working on my game again.

As for my game's progress, I'm getting close to the point where things can start being automated. (which is pretty important for a game like this) So I'm pretty excited about getting that going, even if just a little bit.

 

[Minamu]

You can also use OBS (Opensource Broadcaster, I think), and if you're doing a sped-up timelapse then I guess you could also get away with having a pretty low capture framerate (maybe 15 fps for 4x speed, 10 for 6x or even 6 for 10x), since you'll be speeding the footage up anyway and that'd get you back to 60 fps.

That sounds clever

 

[_Rob_]

Still trying to improve Clive 'N' Wrench's animation! After some feedback and time to reflect, it became apparent that they looked a little too graceful and not quite "jumpy" enough. As such, this version is a little snappier with some extra force and balance.

 

[Jimnymebob]

I'm tempted to dabble in making a 2D platformer with hand drawn art, opposed to sprites, and I'm wondering if there's a general consensus on which engine is the best for that?

I'm completely new, so I'd be starting from scratch with whatever. I have the Game Maker that was given away in the Humble Bundle a few months back, and then I guess there Unity and maybe UE4? I feel like I'm tempted to go GM, just because a lot of games I've played that used Unity suffered from some issues, particularly regarding frame rate- something that I obviously want to avoid in a 60fps 2D platformer.

 

[Pehesse]

I'm tempted to dabble in making a 2D platformer with hand drawn art, opposed to sprites, and I'm wondering if there's a general consensus on which engine is the best for that?

I'm completely new, so I'd be starting from scratch with whatever. I have the Game Maker that was given away in the Humble Bundle a few months back, and then I guess there Unity and maybe UE4? I feel like I'm tempted to go GM, just because a lot of games I've played that used Unity suffered from some issues, particularly regarding frame rate- something that I obviously want to avoid in a 60fps 2D platformer.

For a 2D hand drawn platformer, I still find Construct 2 to be a good choice (not C3!), as long as your main platform target is PC. If you have ambitions of a console release, then Game Maker might help, but you might have to use Unity or UE4 right off the bat.
My plan and advice would be to make the game for PC first using C2 or GM, then see if it warrants a console port - if it's your first game, the experience of releasing something is more valuable than targeting everything at once!

When doing a hand drawn game, your biggest problem will come from memory use, as hand drawn animation takes a huge toll no matter the engine (that is, if you do traditional hand drawn animation - if you use skeletal animation tools such as spine, then you can disregard most of the memory considerations here, and Spine is compatible with both GM and C2, I believe). Construct 2 recently removed their limit (at my request, heh :v), so I can vouch for that one - I don't know how far you can go in GMM with hand drawn stuff, it'll depend on the size of your sprites and amount of animation either way. (C3 currently can't handle large scale projects, so it's out)

For deciding between GM and C2, there are tutorials available for both to make platformers. I'd say C2 seems more natively built for platformers, and it also doesn't require any kind of coding (it's entirely visual scripting), whereas GM will require you to use GML, its scripting language. However, you already have a GM license, so there's that to consider! (the free trial of C2 will allow you to play around, but not build complex stuff)

So basically: do the tutorials, determine the engine you work best with, read up on how they handle memory in regards to what you aim to accomplish (ie: the details of the design of your game: layout sizes, number of objects/characters, animation, etc), and you'll be able to make an informed choice as to what will work best for you!


EDIT: I wasn't planning to just post this week's GIF at first since it's more animation, and you all know what it looks like by now, but since I've posted the message above anyway, might as well, or something? Anyway, Warrior combat moves, part 1 of too many, woo!

 

[timetokill]

Just to add to Pehesse's post, and to make your decision more complicated, Spine works with Unity too.

My main advice is: Do whatever is most comfortable for you. You can get 60fps out of any of those engines. Unity has the most power under the hood, but you'll also have to work a bit more for it.

 

[Hypereides]

Main hero sprite done. I'll probably have to add some additional frames for other states such as game over. If I don't manage to accomplish that prior to the exam, I'll continue working on it afterwards:

Also, level layouts.

 

[Pehesse]

Just to add to Pehesse's post, and to make your decision more complicated, Spine works with Unity too.

My main advice is: Do whatever is most comfortable for you. You can get 60fps out of any of those engines. Unity has the most power under the hood, but you'll also have to work a bit more for it.

...that, too, but then again, most things work with Unity :-D

Or another thought: I hear good things about Godot! Haven't tried it yet but it looks to be coming along, might be worth checking out? No idea about skeletal animation with that one, though. Does anyone use Godot here?

 

[Jimnymebob]

For a 2D hand drawn platformer, I still find Construct 2 to be a good choice (not C3!), as long as your main platform target is PC. If you have ambitions of a console release, then Game Maker might help, but you might have to use Unity or UE4 right off the bat.
My plan and advice would be to make the game for PC first using C2 or GM, then see if it warrants a console port - if it's your first game, the experience of releasing something is more valuable than targeting everything at once!

When doing a hand drawn game, your biggest problem will come from memory use, as hand drawn animation takes a huge toll no matter the engine (that is, if you do traditional hand drawn animation - if you use skeletal animation tools such as spine, then you can disregard most of the memory considerations here, and Spine is compatible with both GM and C2, I believe). Construct 2 recently removed their limit (at my request, heh :v), so I can vouch for that one - I don't know how far you can go in GMM with hand drawn stuff, it'll depend on the size of your sprites and amount of animation either way. (C3 currently can't handle large scale projects, so it's out)

For deciding between GM and C2, there are tutorials available for both to make platformers. I'd say C2 seems more natively built for platformers, and it also doesn't require any kind of coding (it's entirely visual scripting), whereas GM will require you to use GML, its scripting language. However, you already have a GM license, so there's that to consider! (the free trial of C2 will allow you to play around, but not build complex stuff)

So basically: do the tutorials, determine the engine you work best with, read up on how they handle memory in regards to what you aim to accomplish (ie: the details of the design of your game: layout sizes, number of objects/characters, animation, etc), and you'll be able to make an informed choice as to what will work best for you!


EDIT: I wasn't planning to just post this week's GIF at first since it's more animation, and you all know what it looks like by now, but since I've posted the message above anyway, might as well, or something? Anyway, Warrior combat moves, part 1 of too many, woo!

I'll start off by saying that animation looks fantastic.

Thanks for the detailed reply.
I definitely want to go fully hand drawn, as I'm personally not a fan of using skeleton based animation for 2D animation for the most part. For animation, I'd probably consider a lower amount of frames, but with exaggerated animations (think Guilty Gear Xrd).

An eventual console release would be incredible for me, so that's a big reason for GM or Unity over C2, but then the memory advantage of C2 seems invaluable too. I'll have to do some research about how GM handles that.

Game Maker language doesn't seem too bad from what I've seen so far (it seems quite logic based like how Timesplitters map makers were, so things like if key is pressed>action is performed), but obviously it's going to be more complex than that when it comes to fine tuning.

I like that in asking about narrowing down my selection of game engine, I've gained another option in Construct 2 ha ha! I'll definitely have a look at that though; it being visual based seems ideal for me. Is it possible to transfer projects between programmes without having to basically redo everything from scratch, or are they that different?

Just to add to Pehesse's post, and to make your decision more complicated, Spine works with Unity too.

My main advice is: Do whatever is most comfortable for you. You can get 60fps out of any of those engines. Unity has the most power under the hood, but you'll also have to work a bit more for it.

Thanks. Yeah, a solid 60fps is essential for me when it comes to platformers.

 

[Popstar]

@Makai It's time for a new thread.

If your new OP isn't ready it can be edited into the first post later.

Otherwise, I'll make the new thread, and noone wants that. GAF Indie Game Development |OT 3|: Advice on how to write your own engine

 

[timetokill]

@Makai It's time for a new thread.

If your new OP isn't ready it can be edited into the first post later.

Otherwise, I'll make the new thread, and noone wants that. GAF Indie Game Development |OT 3|: Advice on how to write your own engine

Agreed, it's taking way too long and we're well past the mark at this point

 

[nillapuddin]

question from a non-dev

I'm a big time fan of indie games, I try to support as much as I can and I appreciate all the work yall do, one of my favorite threads to lurk.

I currently don't have a Desktop PC and only have a SP3 so my gaming potential is limited, but I love playing on my Xbox. I was wondering, I was watching /build/ the other day and learned more about the Desktop Bridge for UWP Conversion process, I saw some random guy on reddit did it for Discord (just for himself) and praised the process.

Does this have actual possibility for converting Games to Win store apps and furthermore, Xbox accessible?

I'm not sure if the process is different for games specifically (have to through ID@Xbox?), anyways, hopefully someone can answer my question and if not, no worries.

Keep up the good work yall, look forward to lurking in OT3!

 

[Michael Armbrust]

Is there a good way to fix Unity's line renderer so it looks good at angles and with small segments?

Also anyone know how efficient is the line renderer when updating its points? I'm tempted to program my own mesh deformation script but I'm worried it'd be far too slow since I update points every frame.

 

[Makai]

@Makai It's time for a new thread.

If your new OP isn't ready it can be edited into the first post later.

Otherwise, I'll make the new thread, and noone wants that. GAF Indie Game Development |OT 3|: Advice on how to write your own engine

It'll be done today. Sorry for the long wait. Bad timing.

Is there a good way to fix Unity's line renderer so it looks good at angles and with small segments?

Also anyone know how efficient is the line renderer when updating its points? I'm tempted to program my own mesh deformation script but I'm worried it'd be far too slow since I update points every frame.

Just make your own. Line Renderer is inefficient anyway so you can probably beat that.

 

[Michael Armbrust]

Just make your own. Line Renderer is inefficient anyway so you can probably beat that.

Their new one is still bad? Guess I'll try my own. Thanks.

 

[Dynamite Shikoku]

Is there a good way to fix Unity's line renderer so it looks good at angles and with small segments?

Also anyone know how efficient is the line renderer when updating its points? I'm tempted to program my own mesh deformation script but I'm worried it'd be far too slow since I update points every frame.

What about vectrosity?

 

[DrNeroCF]

I'm tempted to dabble in making a 2D platformer with hand drawn art, opposed to sprites, and I'm wondering if there's a general consensus on which engine is the best for that?

I'm completely new, so I'd be starting from scratch with whatever. I have the Game Maker that was given away in the Humble Bundle a few months back, and then I guess there Unity and maybe UE4? I feel like I'm tempted to go GM, just because a lot of games I've played that used Unity suffered from some issues, particularly regarding frame rate- something that I obviously want to avoid in a 60fps 2D platformer.

Animate CC has a new plugin for Unity for exporting sprite sheets made in Animate / Flash, but I haven't messed with it. Flash is still king for animating and immediately being able to test the animations in gameplay without exporting / importing sprite sheets or such.

 

[Makai]

Their new one is still bad? Guess I'll try my own. Thanks.

Oh, I just remembered. Use this.

https://docs.unity3d.com/ScriptReference/GL.html

Editing vertices in a mesh wouldn't be that expensive except the way Unity handles that reallocates the whole vertex array every time you change any of them. You won't have that problem if you do a straight GL call. It's not very complicated - it's similar to raw OpenGL but if you're unfamiliar with that I'm sure there's Unity examples. Just do two triangles for each line instead of GL Lines, because that has variable support across GPUs.

 

[Michael Armbrust]

Oh, I just remembered. Use this.

https://docs.unity3d.com/ScriptReference/GL.html

Editing vertices in a mesh wouldn't be that expensive except the way Unity handles that reallocates the whole vertex array every time you change any of them. You won't have that problem if you do a straight GL call. It's not very complicated - it's similar to raw OpenGL but if you're unfamiliar with that I'm sure there's Unity examples. Just do two triangles for each line instead of GL Lines, because that has variable support across GPUs.

Awesome that should work perfectly!

What about vectrosity?

Thanks for the suggestion. I'll check that out more if GL doesn't work.

 

[Timedog]

What do you guys think of my new game I'm making? I think it's pretty cool, and also nice.

https://www.youtube.com/watch?v=a7h9OCssOlg (EXTREME epilepsy warning: flashing colors)

 

[MrHoot]

Heyo friendos, it's been a while !

Been working a lot on a bunch of different things. Most of all been really happy with the design and moves of my latest baddie right here, Isengrim (or Ysengrain in french)

Also it's nice to see you continue on your animations Pehesse !

 

[Popstar]

What do you guys think of my new game I'm making? I think it's pretty cool, and also nice.

https://www.youtube.com/watch?v=a7h9OCssOlg (EXTREME epilepsy warning: flashing colors)

Is that the Electronic Arts logo at the beginning? ◾️⬤▲

Congrats on the publisher.

 

[Timedog]

Is that the Electronic Arts logo at the beginning? ◾️⬤▲

Congrats on the publisher.

Thanks, it's been great working with the Madden team.

 

[DemonNite]

its got that charm of an old lucasarts adventure game

really impressive work there

thats really nice. I like that twitch the head does

 

[Pehesse]

Heyo friendos, it's been a while !

Been working a lot on a bunch of different things. Most of all been really happy with the design and moves of my latest baddie right here, Isengrim (or Ysengrain in french)

Oh boy, that's creepy! First thing I saw coming out of bed, and it was impressive :-D Nice job!

Also it's nice to see you continue on your animations Pehesse !

really impressive work there

..and thanks a lot, of course :-D

 

[Makai]

I had most of it done a couple weeks ago, but real life got in the way after that. There wasn't any rush, anyway - staff didn't mind a long delay.

http://www.neogaf.com/forum/showthread.php?p=238241304

Taking a risk and going unorthodox, but I think it's better for how we use this thread. Contributions encouraged.