[Blade of Agony] Before and after screens | p184

[Wiw]

I suggested having the option to toggle between models and sprites, and that was dismissed!

[RockstarRaccoon]

It's not the models, Wiiw, it's the resolution of their textures. I could barely make out the pixels on this one rock on the beach when I was playing C1M2 the other day, and that was sticking my face up against it at fullscreen resolution. There is a fine line between attention to detail and overdoing it.

Meanwhile, on Membrane, I got squeamish about a 512x512 terrain textures because we realized that the hires versions would be 2048x2048x32 (16MB per), and we figured having too many of those would make it hard to fit within our soft-limit of 1GB VRAM. (Imagine giving a mapper 20 of those to choose from: that's 1/3 a GB on terrain textures alone)

Geforce 1060 it's at 26-27 fps

Ryzen

The point is that we all have relatively recent graphics cards, and not shitty ones, and yet they can barely handle the sheer load of this behemoth.

(And interestingly enough Graf, that must be a pretty new, high end computer, seeing as this development machine I just bought uses a 1050ti, the previous model)

[Graf Zahl]

(And interestingly enough Graf, that must be a pretty new, high end computer, seeing as this development machine I just bought uses a 1050ti, the previous model)

My computer is 6 years old. I upgraded the graphics card two months ago. Before that I had a Geforce 550Ti, that mainly got replaced because it cannot do Vulkan. 6 years ago I spent a little more money to get a good CPU and that has paid off very well.

[Wiw]

It's not the models, Wiiw, it's the resolution of their textures. I could barely make out the pixels on this one rock on the beach when I was playing C1M2 the other day, and that was sticking my face up against it at fullscreen resolution. There is a fine line between attention to detail and overdoing it.

Would reducing the texture resolutions really make that much of a difference?

[Ozymandias81]

Let's stop this "is" or "isn't": it is all about models, a couple of skins, a couple of actors which probably could use tweaks (fog), shaders (grass and flowers use them FYI to give them a nice waving effect like wind through them) and the map which we'll investigate further.
Yet no one still said to me if you compare actual TITLEMAP and January one (check my previous comment) you notice significant improvements, at least all BoA team members noticed that.

[Graf Zahl]

Would reducing the texture resolutions really make that much of a difference?

No. This would be optimizing at the totally wrong place. Unless, of course this is run on some hardware with real memory access problems.

[RockstarRaccoon]

Ok, I think this probably feels like we're attacking, and that's not the intention, so I'm gonna go ahead and back out now. It's a good project guys,
great work so far, we're just saying it would benefit from framerate optimizations.

[Rachael]

Ok, I think this probably feels like we're attacking, and that's not the intention, so I'm gonna go ahead and back out now. It's a good project guys,
great work so far, we're just saying it would benefit from framerate optimizations.

If no one wanted to see the project succeed I doubt anything would've been said about it. We'd have been happy letting Tormentor live in a delusion that it's better than it really is.

By saying otherwise, I really believe from the bottom of my heart, that it's giving him an opportunity to see what other people see - to see the flaws that repel us away from the project - and to give him a chance to change it, so that one day we all might come back.

As I said, for me, it's really about the time and effort already put into this project - and just how much I would hate to see it go to waste because of some issues that couldn't be resolved when they needed to be.

I am not trying to verbally flog the team with my words. I just have an interest in actually being able to enjoy what I see as a real piece of art - that I can't because I really feel that the effort is being wasted in all the wrong places.

[Graf Zahl]

Yet no one still said to me if you compare actual TITLEMAP and January one (check my previous comment) you notice significant improvements, at least all BoA team members noticed that.

Yes, that map shows some major improvement. The reason for this is that the portal is badly defined. It contains sectors that have nothing to match in the upper part. And with the old version of the map this lead to buttloads of portal recursion because the main part of the map was in the place that would be visible through the portal, so it all got rendered again and again and again, even though it wasn't visible until the relocated version went off the screen. With your realignment you just moved the offending parts out of the way.

The general rule of thumb for portals is that the area of each sector that got the portal assigned must be fully covered by the linked part of the map.Otherwise the effect will be undefined, and what precisely happens depends on the nodes that will get built. It's still undefined in that optimized map so I cannot guarantee that it won't break later.

Do not count on sky sectors covering up your mapping mistakes, that's what essentially happened here.

[Ozymandias81]

Thank you very much Graf, that was exactly what I wanted to know since I was quite sure that it was the problem. Will investigate further then.

[doomjedi]

I'm having the most fps issues probably having only onboard generic GPU (and had so far minimum of 2 fps (!) on some Chapter 3 map areas which prevent me from even testing the maps).
So I don't claim for full objectivity, but fps is very important part of how I enjoy a mod personally, so alot of stuff said here resonate with me well. I don't care how beautiful the mod is if its fps sucks, and the team must IMHO listen well to the players in this thread. Mod is not a wall poster, it's meant to be played, and lag is difference between life and death in a shooter (and most details are missed anyway by the player in a well-paced shooter), and one can't aim only for high-end computers making retro-TC on a retro engine in retro community. There are consequences in using GZDoom, both good and bad, and bad ones can't be ignored.
As said, it's ok to push GZdoom engine limitations - but as long as you optimize them to be playable. Mere adding millions of stuff in - is not "pushing the engine", it's how much you can push in and yet optimize it to keep good fps - that's the achievement I can respect as "pushing the limits".
I think mappers would be surprised how little of those extra crafted detail players actually notice or care about. So whom are those for?

[Ozymandias81]

This is happening because of a recent internal discussion we had on BoA Team, even thanks to DoomJedi: 2d trees as an option for players.
It was something I had in my mind for aeons, but well didn't really get with it deeply since I should also "make photos" of 3d trees, cut them out and reuse them as sprites. If someone is willingly enough that would be really appreciated

[Wiw]

Finally!

[Tormentor667]

I think if you can replace this one portal with a traditional Boom sector to sector teleport somehow it'd make the map a lot more playable already. What's left after that needs to be done through engine-side adjustments, but I can outright say that the portal alone will be enough to prevent 60fps in the village, even if the engine side changes can eliminate 90% of the processing time.

Sorry for asking a dumb question, but what would you suggest instead? An elevator?

All effort is being focused on the visual aspects of it, but literally none is being made toward the things that really matter: what the player finds fun and engaging, as well as running with at least 30 FPS on an average system that GZDoom runs on. I think the team really needs to realize that it's okay to sacrifice visual quality for the other two things. It would do something that's taken such a monumental effort, as this clearly has, some real justice, and ensure that none of that effort goes to waste.

One of the real problems here is, that I have bought myself a very powerful system about 5 years ago - processor, gpu, all is still competable to systems from 2018 (and in most terms even better) so when I am working on my maps, I never have slowdowns, and I have no idea which kind of system I should "emulate" or use as benchmark. It's not about "I don't care", it's more about "I don't notice" Same fo the amount of trees, population that area with sprite variants might be the better route to go indeed.

Just removing the crossbeam bridges and replacing them with 3D floors makes it run at over 60 fps on my home system and 40 fps on my work Mac with an integrated Intel chipset.

I have never heard the term crossbeam-bridge before so what is it actually? How can I improve C3M3 by replacing these with 3d floors?

The saddest part is, last night I downloaded and played Tormentor's Refinery, and that map was even more intricate and detailed than any of the maps in this game, with detailed areas beneath complicated catwalk bridges, and a section where you fall through a floor to go through a giant worm thing which goes beneath another floor and pops out in the yellow key room, which is surrounded by detailed hallway. All of that looked amazing, and was done without any portal abuse, just a bunch of 3D floors, which, by the way, use a relatively small amount of resources, even if they are transparent or sloped.

For my understanding, 3d floors are more performance-intensive than portals - or did I get something wrong?

[gwHero]

For my understanding, 3d floors are more performance-intensive than portals - or did I get something wrong?

Absolutely not sure if I can speak in general here, but in my mod, I have replaced interactive line portals in the open field by 3d platforms, and the performance gain was very high.