Blood 3 tech idea
#31 Posted 21 March 2017 - 01:23 AM
#32 Posted 21 March 2017 - 08:18 AM
BloodGuy, on 20 March 2017 - 11:02 PM, said:
In the 90s we all had old 4:3 screens and, just as fuegerstef explained, and you could not really see the single pixels like on todays displays.That's the effect i want to create with this.
I will prepare my demo for download then you can see what I mean =D
Well to be frank, you failed very badly . Upscaling is not the answer, all it does is create a blurred mess. Upscaling also does not give the retro look you were trying to achieve. Since you are obviously not going to take my advice and study the composition of the original game, at least search google for a CRT pixel shader you can incorporate into your engine.
#33 Posted 21 March 2017 - 09:27 AM
icecoldduke, on 21 March 2017 - 08:18 AM, said:
Hahahah. after playing some time in my game shows me you are right. the upscaling is bullsh*t.
In Austria we call this a touch in the toilet.
Deleted that feature
#34 Posted 20 May 2017 - 10:56 AM
Extended my experiment and now work on a sector-based Build-Engine-Replica with it's own editor and script-language. And I tried a 3rd-person-gameplay which works very fine. The game-character you see is again a 3D->2.5D realtime conversion.
The editor with the map you see below
lol, maybe I make a pixeled DarkSouls
#36 Posted 12 August 2017 - 06:22 AM
It doesn´t matter if you can´t create gameplay, you can focus on your feature and gameplay experts could take care of it.
#37 Posted 12 August 2017 - 10:09 AM
Gambini, on 12 August 2017 - 06:22 AM, said:
Talk about putting the cart before the horse.
#39 Posted 12 August 2017 - 01:22 PM
Hendricks266, on 12 August 2017 - 10:09 AM, said:
Yeah because all engine demos are full of awesome gameplay.
#40 Posted 12 August 2017 - 01:42 PM
The idea of the tech is not bad at all, I could see a great indie game using it to create a unique art style but when it comes to Blood it's misguided. Back in the day you did not really see low-res sprites, you saw highly detailed enemies, the technology itself was not an artistic choice, it was a limitation the developers had to find clever ways to get around.
This post has been edited by Zaxx: 12 August 2017 - 02:37 PM
#41 Posted 13 August 2017 - 07:43 AM
Zaxx, on 12 August 2017 - 01:42 PM, said:
The idea of the tech is not bad at all, I could see a great indie game using it to create a unique art style but when it comes to Blood it's misguided. Back in the day you did not really see low-res sprites, you saw highly detailed enemies, the technology itself was not an artistic choice, it was a limitation the developers had to find clever ways to get around.
Indeed. When I saw Duke Nukem 3D in the 90s for the first time... I thought "Damn, that looks so real."
Hell, I remember the first time I saw Doom, went into a pawnshop that was running a version of the game with the demo loops going, and it was E1M2, the area with the two sets of stairs that go up to where the keycard is. I remember seeing the player traversing that area, seeing the sky outside the window and being blown away by just how vivid and real it looked.
#42 Posted 13 August 2017 - 07:44 AM
Gambini, on 12 August 2017 - 01:22 PM, said:
A single shader is a visual gimmick, not an engine.
#43 Posted 13 August 2017 - 11:56 AM
A budding indie developer wanting to make the next Duke Nukem 3D (more likely the next Witchaven but lets be optimistic) could license his 3D-to-sprite shader... or they can just prerender sprites, which gives more consistent results. if they wanted more than 8 directions they could easily render that as well, disk space is hardly an issue nowadays (the art files for Blood with Plasma Pak take up 24.8 MB, for Duke 3D Atomic Edition (once extracted) they are 19.4 MB).
As for the idea to render at a lower res then upscale to the screen's resolution... that is actually already being done in modern games to improve framerates on low performing systems
#44 Posted 13 August 2017 - 01:35 PM
Hendricks266, on 13 August 2017 - 07:44 AM, said:
So it could be implemented on Eduke32. I would take care of adding gameplay to it, since a shader like that would come VERY handy for my mod (next to the shadetables to voxels and model skins i´m still waiting ).
#45 Posted 13 August 2017 - 02:03 PM
Gambini, on 13 August 2017 - 01:35 PM, said:
Never. This shader is somewhere between poseury and degeneracy. It's completely misplaced as an attempt to obtain a """retro""" aesthetic. Its only strength is in its technical achievement.
Gambini, on 13 August 2017 - 01:35 PM, said:
I've received the same request from someone else as well, in addition to my own needs, so rest assured it is in the queue.
#46 Posted 13 August 2017 - 10:41 PM
Of course, the difference of those games to this one are practicality vs pretension. Blood wasn't about "pixels".
This post has been edited by leilei: 13 August 2017 - 10:57 PM
#47 Posted 14 August 2017 - 02:41 AM
#48 Posted 14 August 2017 - 02:28 PM
But hey, Galileo was almost burned alive for thinking the way he did...
#49 Posted 14 August 2017 - 02:59 PM
Pixels ≠ Retro.
Pixels ≠ Retro.
Pixels ≠ Retro.
#50 Posted 14 August 2017 - 03:32 PM
Daedolon, on 14 August 2017 - 02:41 AM, said:
It refreshes after a certain distance. And while it's not really a shader, it's still the same principle of rendering a model to a texture. Trespasser's messy hybrid of a software/hardware renderer probably helps make that possible....
This post has been edited by leilei: 14 August 2017 - 03:35 PM
#51 Posted 14 August 2017 - 07:48 PM
Hendricks266, on 14 August 2017 - 02:59 PM, said:
Pixels ≠ Retro.
Pixels ≠ Retro.
Pixels ≠ Retro.
I did read the article, but i don´t agree with it.
I have a different vision of what´s retro and for such, this shader would be of great benefit. For that matter, everything we do prenteds to be retro too, since polymost and polymer are both polygonal engines which downgrade to run two and half dimentional maps. Retro is in short terms about the feel, nobody wants to input DOS commands to run a game, nor wait for a diskette to load for five minutes to then realize there´s an error in the filesystem. You choose what takes you to it, and for me an enemy that moves smoothly, can be looked from up or down but still retains that pixelated print when you get close to it is indeed retro and in a good way.
#52 Posted 14 August 2017 - 10:35 PM
Gambini, on 14 August 2017 - 07:48 PM, said:
This ... it is about the feel of the gameplay, not the looks of pixels.
This post has been edited by fuegerstef: 14 August 2017 - 10:36 PM
#53 Posted 15 August 2017 - 02:02 AM
fuegerstef, on 14 August 2017 - 10:35 PM, said:
Not just feel of the gameplay, but also feel of the presentation. Where pixels have a part, as well as chiptune music, synth/waveform sound effects and so on.
#54 Posted 15 August 2017 - 03:29 AM
Gambini, on 14 August 2017 - 07:48 PM, said:
No, they remove the technical limitations of the software renderer. You're arguing that because they don't have boundless scope, they are somehow on the same level as this shader.
Gambini, on 14 August 2017 - 07:48 PM, said:
No one is arguing this.
Gambini, on 14 August 2017 - 07:48 PM, said:
This meets my definition of poseury. You're latching on to a technical limitation as representative of the art form.
#55 Posted 15 August 2017 - 01:51 PM
Modelling an enemy to then take 8 side lo-res pictures of it and then convert them to 8bits= retro
Modelling an enemy and doing the 8bit lo-res convertion on realtime= poseury
This post has been edited by Gambini: 15 August 2017 - 01:51 PM
#56 Posted 15 August 2017 - 01:59 PM
#57 Posted 15 August 2017 - 02:09 PM
Now people downgrade models from modern games to sprites many times. It´s not like your personal taste dictates the meaning of what´s retro or what´s not.
I know for certain it´s not pouseury or whatever, because i envisioned with having something like this a few times. And it was for the sake of creating something that feels like old games did yet took advantage of new software and hardware.
#58 Posted 02 April 2018 - 11:03 PM
The way I see it, the writer is too focused on problems that are personal in nature, and ends up writing an article that is full of contradictions and circular logic to justify his renouncing from pixel art.
For instance, when comparing the Zelda image to Bubsy's, he acknowledges that the first one is superior art, even though the second may have better technological features, but then he summons this excuse:
"Problems arise when it comes time to convey to a non-artist what constitutes quality art [...] However, it is easy to explain that the second image has a higher level of technology. To the average person, I’m sure it’s self-evident. Some may even be so taken with the spectacle of added color and resolution that they might think Bubsy has the better artwork! I could write you an entire book on why that is absolutely not the case, but that’s the thing – it’s not the audience’s responsibility to read that book. It’s my responsibility deliver them quality in a language they understand".
This is a fallacy. It is hard to convey anything to someone who is not versed into that anything, this applies to art and to technology. Saying the second image has more advanced technology because of a higher resolution is as easy, or as hard, as saying the Zelda image has better art because we can clearly distinguish the elements on the image. Secondly, as with anything, researching into a subject will give you better judgement on it. The audience and the IGN reviewer are not excused from judging graphics poorly because they refused to study graphics in the first place. Why should their ignorance be the artist's fault?
Like many said, the author seems fixated on people 'getting it' - his art -, and he'll be working for the lowest common denominator if that is the case. However he tries to dress it up with phony wording, it is what it is. The passionate guy is not making pixel art anymore because an ignoramus said something he didn't like. That's not what I call passion.
It comes off as if the game he worked on bombed and because of that he had to find something to blame. Pixel art is a niche, just like dungeon synth and chip tunes, and you can't expect them to sell as much when they don't have as large of a public as pop music. I would be fine with the article if it came clean and said he's not doing pixel art anymore because it doesn't sell as much as 3D realism, but what he does is pandering, plain and simple.
I could go on about the article but I would be repeating most of the comments that disagree with his views. In the case of Blood, I agree with Gambini and disagree with Hendricks, art style is important, of course, but it is not all. Was Blood made the best they could within their limits? Yes. Would Monolith have made the game graphically better (from a technical perspective) if they could? Sure. But Blood was cutting edge stuff 20 years ago, today you'll need to limit yourself to a pixelized look to make the art look like Blood's. Fitting 100.000 polygons models in it will look bad.
It surprises me that Hendricks is saying the shader 'pretends to be retro'. Is Ion Maiden pousery as well? It certainly tries to look like the old build games. Why are the visuals limited to sprites and 2.5D maps instead of going full 3D? Why even use such a limited engine? Why latch it to technical limitations? That's an odd position to defend when your own game limits itself.
The shader can be tweaked to sucessfully reduce visual complexity to what the game has while being a lot less work than making a billion sprites. It does a good job in emulating voxels, while being seemingly easier to work with, especially for animated stuff.
#59 Posted 03 April 2018 - 12:54 PM
Lune, on 02 April 2018 - 11:03 PM, said:
You have provided no evidence for this assertion. Blood's art style is in no way dependent on pixelated graphics. Phredreeke's upscale pack completely destroys your claim.
Imagine if instead of photographing Kevin Kilstrom's clay models of the enemies and converting them to low-resolution low-bit depth image assets, a 3D scanning technology was used to convert them to 100,000 polygon models complete with color- and shade-accurate texture mapping. Blood would look better than ever, and just as coherent.
Lune, on 02 April 2018 - 11:03 PM, said:
If your 3D polygonal models, texture maps, and any other realtime processing done on the scene can already achieve an accurate and convincing "retro" aesthetic, then a shader that pixelates your models is unnecessary. If you use one anyway because you think pixels = retro, you are a poseur.
Lune, on 02 April 2018 - 11:03 PM, said:
Because it's what we know, and what we have been taking care of for the past 15 years. We are not trying to make a "retro game" in itself. Plenty of low-effort Wolf 3D clones on Steam have that territory covered. We are trying to make a good game using the skills we have cultivated. The use of old technology has been drummed up by our marketing team to great effect. Dusk and Amid Evil are achieving retroness just fine despite using Unity and UE4.
Lune, on 02 April 2018 - 11:03 PM, said:
"Reducing visual complexity" does not require pixelation.
Lune, on 02 April 2018 - 11:03 PM, said:
Pixels are not the same as voxels. If you want voxels, then convert your models to voxels, or write a shader that quantizes polygons and converts them to cubes (or XY-gridlocked billboarded rectangles if you hate yourself). "Emulating voxels" with this shader would be poseury because they're not the same.