Beats of Fire

[khph_re]

nice Amiga check mark in that screenshot

[leathered]

Originally Posted by Bamiga2002

I think they're polishing da cock...

Well, I think said cock has been polished until it gleams =)

I'm getting the 2 player mode in – would in hindsight have been easier to incorporate from the beginning (probably) but progress coming along well. I want to get this in before finishing the AI – obviously it's going to suck if the goons just follow one of you around the screen!
Guy is fully playable – lacking a couple of moves and able to pull off a bogus pile driver that looks like Chun Li's headstomp. Turri has been getting on with mapping of the later stages.

As soon as I get the AI more finalised then I'll put something out there with whatever stages we've got. No promises, and as usual don't hold your breath...

I took some of Steve T's advice re. the palette – the 23 colour player/goon palette now gives us this:

[Shatterhand]

It looks amazing. Great job.

The ECS version won't run in a normal 68k 7mhz amiga, will it?

[s2325]

[ Show youtube player ]
so Amiga version was improved in almost everything

[Steve T]

Looks good, definitely more solid, less olive drab/sewer-y.

Maybe the orange wrestler is a bit radioactive looking... hard to say against that black background. email me the revised palette and I'll improve the bonus round some more.

Originally Posted by leathered

I took some of Steve T's advice re. the palette – the 23 colour player/goon palette now gives us this:

[Toni Galvez]

@leathered, thank you to code this wonderfull game, Amiga really needs a good Final Fight port.

I am a pixel artist, I work on videogame industry since 1999, I had Amiga since 1991 to 2001 (I had my Amiga 1200 with 060, 32MB... it was in the wardrove, my family throw to the trash ). Now I have Amiga 500(with aca500 and 1200 again).

I have a question, can I do with you a good Double Dragon port for Amiga adapting some of your Final Fight code?

[leathered]

@ Shatterhand - Well, ECS version was an experiment in getting the engine to run without AGA. It's a little wrong to say that it's planned – more to say that it's possible and that a demo exists. I'd really like to get around to that afterward, but it's hard to look forward to that with this to finish.

To dream for a minute though-

To get it working on a 7mhz machine – I'm really not sure – maybe with 16 colours?
Certainly possible (as Rich Aplin proved) – but again no promises. I used to put stuff together in Amos which required me to use 4-8 colours to get the speed I wanted (always 50 fps), but blitz definitely has the edge for speed. Guess I should learn to code properly =)
Bear in mind that Final Fight on ECS slows when there's more than a little going on in 16 colours – and that was by a pro coding in assembler...

No, reducing the colour would be a lot of work, but there's hope; the engine also works with a reduced amount of goons to fight – so we could just keep the colour and give you less goons on screen at once. Provided there are at least 2 it may still prove decent enough – they are queued and enter the screen as soon as there is a slot – so the action remains quite pacy.

The music loads into fast ram at present. The other problem for the ECS would be memory – you're gonna need a bigger boat if you have less than 2mb chip.
For the music – well there may be just enough memory left to squeeze a soundtrack in without fast, and questions like these need to be addressed at some stage.
I wasn't able yet to create something stable enough in blitz to facilitate proper allocation and clearing of memory for PT modules – the blitz commands (including newer libraries) don't seem to give enough for modules when you're tight on memory but the awesome Omed player library provided the solution for machines with fast mem. Listening to adrdesign's mods in this way is almost like having an mp3 of the arcade tracks.

So basically - it appears as though it wouldn't take much for me to create a version that will run, but I just can't vouch for the speed. If I can finish this then I'll just get onto it and we'll see


@ Tony Galvez – with me? Well I think I've got enough on my plate to be honest, but I'm not averse to others using the source of earlier demo at the top of this thread. It is not the easiest thing to read but I'll try to answer any questions.

The other thing is – it appears as though someone has beaten you to it terms of a modern conversion of Double Dragon for Amiga – http://ohmygibs.free.fr/ohmygibs/Blo...for_Amiga.html -so we'd be entering into an arms race of sorts.

But I'd be glad to help you do something Tony – whenever there is the chance and if I'm lucky enough to catch up with you later. Just I have to honour other commitments first.

@ s2325 Ahh the memories. But the music on real ST wasn't quite as bad as that, check the first part of this =) [ Show youtube player ]
I actually find it plays better than the Amiga version due to Haggar's immense pile driver spamming ability.

[dlfrsilver]

Yes ADRdesign mods from the arcade tracks i ripped are almost arcade perfect

[adrdesign]

Originally Posted by dlfrsilver

Yes ADRdesign mods from the arcade tracks i ripped are almost arcade perfect

Thank you! Just for the record, actually I didn't used your rips (all the channels were mixed), they worked to have a global idea of what to do. I explained in some PMs to turrican3 how where they made

Hi guys, fist of all I apologyze if my english is poor. I'll try to be clear

For the "Select player" tune (multichannel) I exported the original song to wav with bridge (player tool) unmarking the 2nd channel of the YM2151 synth (the other two are for drums and cymbal respectively, so unmarking them while exporting will allow to sample both channels).

after exporting the full song i've loaded it in wavosaur. I had the drums in the center and the synth sequence in the left channel.

In wavosaur i did a vocal removing to discard the center channel. Then chopped the longest note i founded for the synh bass and created the instrument. After maximizing and converting to 22050 8 bit I saved to wav (soundstudio is able to load 8bit wavs directly).

For the drums I reloaded the song and converted to mono selecting right channel. Maximized and resampled to 11025 8bit. Chopped and saved to wav.

As i can listed both channels isolated (synth and drums), it was easy to convert with that samples.

------


for the stage 1 song i sampled the synth riff loop and the drums with the same method and isolated them (one loop for synth, two different drum loops). mixed 100%drums and 66% synt riff i created both loops in channel 0.

the song is written by ear with fairlight and emax samples (dont know if "by ear" exists, i meant listening the song and trying notes on the tracker)

------

for the title song the process was longer but very interesing. 1st of all I searched a way to convert ROM tunes to MIDI, none for MAME games yet. I found a way to convert both GBA and SNES tunes directly from the rom (in useful software folder)

So i decided to work with the snes one.

Loaded the MIDI in Soundstudio and toggled 1-64ch mixing, selected 10 channels in mixing settings, converted each drum note to a sample (using Replace Notes function) and then the other instruments.

for the chords i exported the song with octamed muting the rest of the tracks. then loaded in octamed and chopped the sample in differet bits. Reconstructed the chord track and done. 4 tracks and 1:1 note converting. I didnt found a way to isolate the samples in the original song so I choosed from the fairlight collection

and that's all.

PS: after that I used other methods based on SNES music mixed with arcade samples (using SPC2IT) for some stages, and a new intro music with speech, based on the GBA port.

[dlfrsilver]

Very impressive, i listened to the mods, and they are 100% accurate to the YM2151 originals

[Toni Galvez]

Thank you, we will see in the future for a good Double Dragon conversion.

Please take this Final Fight picture.
Attachment 42697

I convert almost all the Final Fight graphics to 16 colors, they loose some detail, but I should use on a OCS/ECS version for sure. Maybe you can check and use them for another small and fast version. Then, we can have AGA and OCS/ECS version. In the picture, the dark grey is the color 0 for the sprites. OCS version may have a single scroll.

Have you maked the backgrounds on tilemap?

What do you think?

[dlfrsilver]

Toni, the attachment is invalid, can you replace it ?

[Shatterhand]

Regarding an ECS 7mhz version, I'd probably prefer less colors and more enemies on screen, as acurate as possible with tha Arcade version... but then I am not working on the game, so I can't ask for too much.

Slowdown is something very acceptable in a game like this IMO. Final Fight slowdowns, but it's not that bad IMO.

I would like to say the original Final Fight game for Amiga coded by R.Aplin wasn't a bad programmed game at all IMO. To me, it looks decent enough, it has big sprites moving at a decent speed, it has a good amount of frames of animation... maybe if it wasn't an Atari ST port it could had been a lot better, still I don't think it's awful programming...

... it's just awful GAME DESIGN. That's what really pisses me off about that game (and so many other Amiga games). It's clear the programmer was capable and the machine could handle it, is just that the programmer seemed to have no idea of how to do a good game (or maybe really lacked the time to polish it properly).

I sometimes feel a lot of those games weren't even tested properly. Today I played the Amiga port of Vigilante for the first time. How the heck could they screw that SO BADLY? And, again, I can't believe someone actually PLAYED that before the game was released and said "Ok, this is good enough to be released".

[Toni Galvez]

Final-Fight_16c_2.zip
Try this.

I forgot to say, I placed in the image the SNES/GBA graphics, then, they are smaller than the arcade ones, I think is better for the Amiga OCS/ECS.

If you like, I can arrange the background in tiles to make a map.

[dlfrsilver]

Toni, there is a problem however, by having all the sprites and background on the same picture, you decrease the amount of possible colors.

Coin-op games are using multiple palettes when computers only use unified palettes.
This has to be corrected when converting ^^.

[Toni Galvez]

I know that arcade games uses multiple palettes, I am making games from 1994, I converted all the graphics to 16 colors, I study well how to do that, I prefer to use some variated colors and give a good look to the game than using 3 shades gradients like original Amiga Final Fight.

I am sure of what I am doing, for a game like this, is the best we can do. Because all the characters share the colors with the backgrounds, all can have the 16 colors. There is no point on put 8 colors for the main characters and 8 colors for the background. If I remove the logo and other small details, will be the same colors.

@leathered, I can improve the detail in the backgrounds if the graphics will be used on a OCS version.

If you see, most of the arcade conversions on Amiga where in 16 colors and the graphics where so much ugly and the color palettes where washed colors most of the time.

Make a conversion of all the graphics yourself and show us

[leathered]

Originally Posted by Shatterhand

Regarding an ECS 7mhz version, I'd probably prefer less colors and more enemies on screen, as acurate as possible with tha Arcade version... but then I am not working on the game, so I can't ask for too much.

Slowdown is something very acceptable in a game like this IMO. Final Fight slowdowns, but it's not that bad IMO.

I would like to say the original Final Fight game for Amiga coded by R.Aplin wasn't a bad programmed game at all IMO. To me, it looks decent enough, it has big sprites moving at a decent speed, it has a good amount of frames of animation... maybe if it wasn't an Atari ST port it could had been a lot better, still I don't think it's awful programming...

... it's just awful GAME DESIGN. That's what really pisses me off about that game (and so many other Amiga games). It's clear the programmer was capable and the machine could handle it, is just that the programmer seemed to have no idea of how to do a good game (or maybe really lacked the time to polish it properly).

I sometimes feel a lot of those games weren't even tested properly. Today I played the Amiga port of Vigilante for the first time. How the heck could they screw that SO BADLY? And, again, I can't believe someone actually PLAYED that before the game was released and said "Ok, this is good enough to be released".

I am largely of the same opinion regarding the original conversion. I persevered with the differences and still find it a playable – but very different game. I think back then programmers just weren't given the time or resources to come up with the goods – it seems the remit was largely to get the graphics looking good for screenshots on the box. Double Dragon 2 was decent and looked the part but some moves missing that were present in DD1. For Final Fight that situation was even worse. These games weren't coded by large teams and probably without anyone playtesting, but I myself am bemused by this apparent 'regression' in very similar game design.

I'll put together the demo for ECS so you can make your own mind up about the speed – I would appreciate that as I don't have ECS machine to test. Bear in mind that you will need 2mb chip and roughtly 250kb fast – but I can cut the music for now if that's an issue.

You are right, if I was playing this on ECS I'd rather it was a close conversion rather than a poorly scaled down 'port' from another machine. And if it comes to it, all it needs is for someone to convert the gfx into fewer colours (actually 16 is a little inaccurate – we still have our great copper to help out) – let's say 4 bitplanes as opposed to 5.

Which brings us to...

Originally Posted by Toni Galvez

Attachment 42700
Try this.

I forgot to say, I placed in the image the SNES/GBA graphics, then, they are smaller than the arcade ones, I think is better for the Amiga OCS/ECS.

If you like, I can arrange the background in tiles to make a map.

Wow, firstly thanks for that!

I checked the floor tiles which appear to be 16*16 tile sets, but the total height of one section of the tiles was 195 – I'm guessing this just happened during the colour conversion.

This engine uses a height of 208 (13 tiles), of course it's no issue to change that. I also appreciate that using the GBA graphics should be a little smaller, and you would think so! But I also recall that the GBA adapted the SNES graphics which I am familiar with. Anyway whatever they did, the upshot is that the GBA gfx are no smaller than the arcade gfx which have been adapted.

What I ended up doing was a resize of roughly 90% from the arcade for the characters. The graphics for the stages are now taken directly from the arcade and re-coloured using a ruleset to 'emulate' the copper.

They get one 16*16 tile chopped (usually merged somewhere in the middle of the screen) leaving the playing area well alone (the floor tiles are always 100% ported). This gives us a mapped height of 208 pixels.

For the colour, there is a standard character palette of 23 colours (above) with another 9 colours for the background – which are copper 'optimised' for each stage. Obviously this will be different for 16 colours – but I would suggest starting at converting the character palette to 16 and going from there. Every stage can use a different palette!

The gfx you sent will be useful to keep for sure, and the results are surprisingly good.
I know you've got the bit between your teeth for a new project but if you feel like it – you can find plenty of work here FFAGA towers = )
I'll email you some of the stuff to look over.

[nobody]

Originally Posted by Toni Galvez

I know that arcade games uses multiple palettes, I am making games from 1994, I converted all the graphics to 16 colors, I study well how to do that, I prefer to use some variated colors and give a good look to the game than using 3 shades gradients like original Amiga Final Fight.

I am sure of what I am doing, for a game like this, is the best we can do. Because all the characters share the colors with the backgrounds, all can have the 16 colors. There is no point on put 8 colors for the main characters and 8 colors for the background. If I remove the logo and other small details, will be the same colors.

@leathered, I can improve the detail in the backgrounds if the graphics will be used on a OCS version.

If you see, most of the arcade conversions on Amiga where in 16 colors and the graphics where so much ugly and the color palettes where washed colors most of the time.

Make a conversion of all the graphics yourself and show us

Nice try but i feel it could be done way better with 16 colors. If i have time later i will make some graphics with 16 to prove this.

[Toni Galvez]

Originally Posted by nobody

Nice try but i feel it could be done way better with 16 colors. If i have time later i will make some graphics with 16 to prove this.

Yeah, of course, this conversion can be better, I can improve the backgrounds in case the graphics will be used for a Amiga 500 OCS/ECS.

I am only interested on a NON AGA version, sorry, I think the Amiga 500 deserves a good FINAL FIGHT and in 512k chip+512k slow/fast ram.

This is why I use 16 color graphics, because I want to have more characters in RAM, smoother movements, ingame music and SFX at once.

[nobody]

This is analyzing your palette, look how close some colors are, you are wasting at least 3-4 colors here, and you have only one green