English Amiga Board

English Amiga Board (https://eab.abime.net/index.php)
-   Coders. General (https://eab.abime.net/forumdisplay.php?f=37)
-   -   NeoGeo Metal Slug conversion to Amiga 500 - madman task ;-) (https://eab.abime.net/showthread.php?t=89396)

s2325 13 November 2017 22:08

Quote:

Originally Posted by BippyM (Post 1199263)
That does look quite good.

even better on CRT display

Amiga1992 13 November 2017 22:38

Quote:

Originally Posted by Galahad/FLT (Post 1199298)
Saw that Amstrad Beast yesterday, pretty amazing what they achieved, shows that the version released could have been so much better.

Well but people all the time think that a tech demo = a game and that's totally not true.

On Pouet,Overflow clearly explains it, and people should pay attention to this kind of explanation:
Quote:

Originally Posted by Overflow @ Pouet.net
This is a demo - not a POC for a game. A game with such quality (on sound and parallax scrollings) is not possible.

Part of the demo is cheating.
Think about it: a small one-screen intro... on a 512KB cardridge!?
As the best example, all played notes from the original MOD were generated and saved to rom: 368KB!
More than 1/3 of cputime during the show is spent by copying samples from rom to ram to be played.
Others cheats are: 52KB (out of 64KB) ram dedicated to screen (incl. 16KB just for the small wall scroll at bottom),
generated code in rom for balloons and moon, generated code for delta-packing animations in sprite,...
512KB were full quickly!

Part of the demo is not cheating.
From the nfo: promote Amstrad GX4000 as one of the most powerful 8bit platforms.
That's it: GX4000 is just too powerful! ;)
There's no cheat there, I'm using standard features from the hardware:
* dma-AY on 3 channels
* 4096 colors - palette: 16 for background + 15 for sprites
* colorful sprites - which are zoom-able (watch those trees)
* 1-pixel H/V scroll
* splitscreens & programmable scanline interruptions
From the nfo again: that's an easy job to show that GX4000 is not the crap most of people were lauhging at.


Galahad/FLT 13 November 2017 23:03

Quote:

Originally Posted by Akira (Post 1199317)
Well but people all the time think that a tech demo = a game and that's totally not true.

On Pouet,Overflow clearly explains it, and people should pay attention to this kind of explanation:

Ahem. I said that what was originally released could have been so much better, I didn't EVER say it could have been exactly like that demo, but it shows that with a bit of effort, what was released on Amstrad could have been better.

Better now? ;)

Amiga1992 13 November 2017 23:07

Quote:

Originally Posted by Galahad/FLT (Post 1199327)
I didn't EVER say it could have been exactly like that demo

I know! And I wasn't talking about you at all. But yeah thanks for the clarification, otherwise people who don't know better might start saying that "Galahad of Fairlight said it is possible!" :P:crazy

Galahad/FLT 13 November 2017 23:28

Quote:

Originally Posted by Akira (Post 1199332)
I know! And I wasn't talking about you at all. But yeah thanks for the clarification, otherwise people who don't know better might start saying that "Galahad of Fairlight said it is possible!" :P:crazy

Having seen modern efforts, I think the Amstrad CPC464 got REALLY and properly short changed during its lifetime, more so than the Amiga and its ST ports.

Amigajay 13 November 2017 23:30

It’s still a good demo I won’t take that away, but if you compare the Amstrad 6128 plus disk version that came out in Sep 1990 was £329 compared to an A500 for £399 I know which one we all would pick, so whilst it’s still an 8bit machine, a 1990 computer shouldn’t be shocking us tbh

robinsonb5 14 November 2017 00:02

Imitating Shadow of the Beast on 8-bit machines seems to be the in thing at the moment. I stumbled across this yesterday:
https://www.youtube.com/watch?v=VHmkvWwNq8s
On a BBC micro? Really? (Admittedly with a newly-released ULA that massively improves its colour capabilities)

(Now all we need is someone to port it to the Archimedes! Yes, I know, I'm going to hell for that!)

As for the Amiga's limit for something like Metal Slug - has anyone actually calculated / measured theoretical and real-world maximum numbers of pixels / BOBs the Blitter can shift in a single frame, for ECS 8+8 Dual Playfield and likewise for AGA with fastest fmode?

AnimaInCorpore 14 November 2017 07:37

Quote:

Originally Posted by britelite (Post 1199245)
Kind of, demos are more fun to make compared to games. But mainly because you don't have to worry about boring parts like game mechanics, level design and so on.

Yeah. Just imagine all the arcade ports for the Amiga if the implementation of the gameplay wouldn't have been so boring...

britelite 14 November 2017 07:57

Quote:

Originally Posted by AnimaInCorpore (Post 1199397)
Yeah. Just imagine all the arcade ports for the Amiga if the implementation of the gameplay wouldn't have been so boring...

To be honest, I don't think most developers would go for arcade ports anyway, but rather their own games ;)

modrobert 14 November 2017 09:09

Quote:

Originally Posted by Master484 (Post 1199249)
Here is a quick mock up screenshot on how Metal Slug might look on the A500, using the assests that Trachu zoned:

http://eab.abime.net/attachment.php?...1&d=1510597394

Although I changed the colors from 16 to the 8 + 8 dual playfield format, with pretty good results.

Also, the player could be a sprite, so in reality it could have it's own 16 color palette.

So theoretically the graphics themselves might look pretty nice on the A500.

But a 2 MB machine is needed to store all the GFX, so A1200 and 16 + 16 Dual Playfields sounds better.

But of course who says that AGA should always use a 16+16 color playfield? It can also use 8+8, which would allow more frames, more speed, less loading and so on.

I haven't tested how fast A1200 is in 8 colors, but should be super fast.

So maybe 8+8 on AGA would be the best choice? But you decide, no way I'm going to do a crazy project like this. :great

That looks great, impressive work with the colors.

I remember reading somewhere that AGA (A1200) also introduced some improvement regarding hardware sprites (can't find the info atm). Still, nowhere near NeoGeo sprite capabilities, so I guess any kind of custom "sprite engine" in assembler to port the original NeoGeo code is out of the question?

Master484 14 November 2017 11:37

Quote:

I remember reading somewhere that AGA (A1200) also introduced some improvement regarding hardware sprites (can't find the info atm). Still, nowhere near NeoGeo sprite capabilities, so I guess any kind of custom "sprite engine" in assembler to port the original NeoGeo code is out of the question?
That's right, on AGA sprites maximum horizontal size was increased to 64 pixels. The older OCS / ECS sprites could only be 16 pixels wide. But both on AGA and OCS we still have only 8 sprite channels.

But the Neo Geo has 900 sprites, and everything that moves in the screen, including the backgrounds, are sprites. So in all likelyhood it's totally impossible to make an Amiga sprite engine that would be capable of directly replicating all Neo Geo sprites by just multiplexing the 8 Amiga sprites. :)

---

I made an updated memory plan for A500 8 + 8 Dual Playfield:

50 kb = two 288*224 8 color "game area" bitmaps (front playfield)
125 kb = five 288*224 8 color level "background" bitmaps (back playfield)
80 kb = music and sound (although the demo most likely has none)
30 kb = reserve
220 kb = graphics

Game code goes to Slow RAM.

With this plan, in a "real game" there would need a loading break only every 5 screens.

One tank frame is 2,6 kb in 8 colors. So we can have 84 "tank frames"...This is equal to five standard 320*256 Amiga screens full of graphics.

So 220 kb should be able to hold a good amount of frames for the following: Player, enemy soldier, two tank types, helicopter and some bullets and explosions.

So RAM wise a short 5 screen infinitely looping demo level would be 100% possible on the 512K A500, and speed wise I think it would be 17 FPS + player and scroll at 50 FPS.

dlfrsilver 14 November 2017 12:02

Quote:

Originally Posted by Galahad/FLT (Post 1199337)
Having seen modern efforts, I think the Amstrad CPC464 got REALLY and properly short changed during its lifetime, more so than the Amiga and its ST ports.

Well you nailed pointed it correctly. the CPC is really able of excellent results, if the programmer has the will to use it correctly.

And yes, if we consider the Amiga was plagued by shoddy ST ports, the CPC was plagued with half-bastardized Speccy and C64 ports, that left most of us questioning.

Some famous people like Dave Perry of Aladdin's Fame was one of the best CPC coders, and he did good, colored, playable games.

But he is not the only one, Andrew Deakin at Ocean, John Brandwood or James Higgins at Ocean were good CPC coders too.

Rafaelle Cecco the Brit-french coder of Stormlord I & II also showed clearly the CPC could really do excellent games.

but nowadays, the CPC demos guys made either demos and also games showing the CPC could go higher than the other 8 bits computers.

A coin-op conversion is prepared for Ghosts n goblins, on the CPC plus. It's a faithful replicate of the arcade machine, minoring the resolution.

zero 14 November 2017 12:31

I think the CPC just didn't sell enough units to get the attention it needed to really shine. If you compare it to how much effort went into the C64 and how far games on that system came during it's lifetime... But by the time the CPC464 came out it was just too late to get that critical mass of coders working on it.

nobody 14 November 2017 13:37

The CPC and CPC+ are different things. The CPC+ has sprites, scrolling and 4096 palette and can do 16+16 colors. The original CPC has nothing, is just a Spectrum with a CGA graphics card.

Megol 14 November 2017 13:48

Quote:

Originally Posted by dlfrsilver (Post 1199442)
Well you nailed pointed it correctly. the CPC is really able of excellent results, if the programmer has the will to use it correctly.

And yes, if we consider the Amiga was plagued by shoddy ST ports, the CPC was plagued with half-bastardized Speccy and C64 ports, that left most of us questioning.

Some famous people like Dave Perry of Aladdin's Fame was one of the best CPC coders, and he did good, colored, playable games.

But he is not the only one, Andrew Deakin at Ocean, John Brandwood or James Higgins at Ocean were good CPC coders too.

Rafaelle Cecco the Brit-french coder of Stormlord I & II also showed clearly the CPC could really do excellent games.

but nowadays, the CPC demos guys made either demos and also games showing the CPC could go higher than the other 8 bits computers.

A coin-op conversion is prepared for Ghosts n goblins, on the CPC plus. It's a faithful replicate of the arcade machine, minoring the resolution.

It's easier doing hardware targeting hacks in a demo than in a game. Remember that the CPC had 4(?) different CRTC chips which require different code to do scrolling etc. IIRC some effects couldn't even be done on all chips, and it's acceptable to make a demo that doesn't run on a percentage of machines while for a game it isn't.

Never actually coded for CPC so may be wrong. Looked at it once thinking about porting scrolling techniques for a completely different platform.

And the CPC+ is almost a 16 bit class machine. Not comparable.

nobody 14 November 2017 13:58

If I remember well the spectrum coders were the best at the time so the CPC owners were lucky they got some excellent colored ports like target renegade, chase hq, RoboCop, Batman, operation wolf, nebulus etc. That was about the best a CPC can do.

dlfrsilver 14 November 2017 14:06

Quote:

Originally Posted by zero (Post 1199445)
I think the CPC just didn't sell enough units to get the attention it needed to really shine. If you compare it to how much effort went into the C64 and how far games on that system came during it's lifetime... But by the time the CPC464 came out it was just too late to get that critical mass of coders working on it.

It's not comparable, the C64 was sold as a toy in USA in big hypermarket, plus the usual IT shops.

It had a very large scope. The CPC problem was else : Amstrad suffered from a bad hardware reputation in UK. And the developers prefered to dev first for C64 and Spectrum.

Problem : the CPC requires developments from "stronger" computers that itself, like the Amiga or the Atari ST. The result are plain bad from C64 and spectrum. Many examples showed it throughout its life.

dlfrsilver 14 November 2017 14:07

Quote:

Originally Posted by nobody (Post 1199464)
The CPC and CPC+ are different things. The CPC+ has sprites, scrolling and 4096 palette and can do 16+16 colors. The original CPC has nothing, is just a Spectrum with a CGA graphics card.

the CPC can do hardware scrolling with the CRTC. The speccy cannot :)

nobody 14 November 2017 14:11

Quote:

Originally Posted by dlfrsilver (Post 1199479)
the CPC can do hardware scrolling with the CRTC. The speccy cannot :)

This is not hardware scrolling, it's character scrolling. We have discussed that before. Example game: Wonderboy 1. Didn't helped much because C64 version is miles ahead but CPC is still playable.

dlfrsilver 14 November 2017 14:31

Quote:

Originally Posted by nobody (Post 1199481)
This is not hardware scrolling, it's character scrolling. We have discussed that before. Example game: Wonderboy 1. Didn't helped much because C64 version is miles ahead but CPC is still playable.

nope, it's not character based, it's byte per byte based :

http://www.cpcwiki.eu/index.php/Prog...using_the_CRTC

Wonderboy CPC is exactly what i was saying above. That's your typical C64 port on CPC.

the graphics are grainy and awful, never mind the colors used, and the hum... scrolling ?

this is a very crap use the CRTC.


All times are GMT +2. The time now is 13:00.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.

Page generated in 0.07961 seconds with 11 queries