English Amiga Board


Go Back   English Amiga Board > Coders > Coders. Asm / Hardware

 
 
Thread Tools
Old 17 November 2016, 16:18   #141
DanScott
Lemon. / Core Design
 
DanScott's Avatar
 
Join Date: Mar 2016
Location: Tier 5
Posts: 1,209
16 colours for background and sprites (enemies)

16 colours for the main character HW sprites... his armour changes colour as it powers up, so can change the HW sprite palette for this "effect"

Overlay score panel can be done as multiplexed hardware sprites too, as the main character (IIRC) always stays central in the screen.

I think a very good conversion of Black Tiger can be done. I will look at putting something together as a test
DanScott is offline  
Old 17 November 2016, 16:35   #142
sandruzzo
Registered User
 
Join Date: Feb 2011
Location: Italy/Rome
Posts: 2,281
Quote:
Originally Posted by DanScott View Post
16 colours for background and sprites (enemies)

16 colours for the main character HW sprites... his armour changes colour as it powers up, so can change the HW sprite palette for this "effect"

Overlay score panel can be done as multiplexed hardware sprites too, as the main character (IIRC) always stays central in the screen.

I think a very good conversion of Black Tiger can be done. I will look at putting something together as a test

Keep In mind what weìre doing for rygar
sandruzzo is offline  
Old 17 November 2016, 21:33   #143
DanScott
Lemon. / Core Design
 
DanScott's Avatar
 
Join Date: Mar 2016
Location: Tier 5
Posts: 1,209
Yes, i know full well what you're doing for Rygar this is a different game, different style (slightly)... and a different approach for Amiga conversion

Keep going with Rygar.. it is looking good so far.. but needs a lot of work still
DanScott is offline  
Old 17 November 2016, 21:34   #144
saimon69
J.M.D - Bedroom Musician
 
Join Date: Apr 2014
Location: los angeles,ca
Posts: 3,516
Quote:
Originally Posted by DanScott View Post
16 colours for background and sprites (enemies)

16 colours for the main character HW sprites... his armour changes colour as it powers up, so can change the HW sprite palette for this "effect"

Overlay score panel can be done as multiplexed hardware sprites too, as the main character (IIRC) always stays central in the screen.

I think a very good conversion of Black Tiger can be done. I will look at putting something together as a test
I could help in doing the in-game music; got specialized in two channel music lately, and i can bet most of the in-game tracks can be stepped down to that so to have the remaining two fir SFX
saimon69 is offline  
Old 17 November 2016, 21:43   #145
DanScott
Lemon. / Core Design
 
DanScott's Avatar
 
Join Date: Mar 2016
Location: Tier 5
Posts: 1,209
Quote:
Originally Posted by saimon69 View Post
I could help in doing the in-game music; got specialized in two channel music lately, and i can bet most of the in-game tracks can be stepped down to that so to have the remaining two fir SFX
thumbs up.. the music is quite simple in Black Tiger

I have so much experience with this coin-op

and as an 17 year old, was so dissapointed with the conversion... who knew at that time, I would go on to program some well known Amiga titles
DanScott is offline  
Old 18 November 2016, 00:48   #146
sandruzzo
Registered User
 
Join Date: Feb 2011
Location: Italy/Rome
Posts: 2,281
Quote:
Originally Posted by DanScott View Post
Yes, i know full well what you're doing for Rygar this is a different game, different style (slightly)... and a different approach for Amiga conversion

Keep going with Rygar.. it is looking good so far.. but needs a lot of work still
I know rygar is different, what I meant, is that we can do a great conversion too here

Rygar is very complex as game play, japanese are very tough
sandruzzo is offline  
Old 18 November 2016, 20:35   #147
sandruzzo
Registered User
 
Join Date: Feb 2011
Location: Italy/Rome
Posts: 2,281
In order to speed up things, you could use 2 big interliaved planes:

1-3-5 and 2-4-6. With this layout, you can blit objects that require less colors, but you can still have the whole 6 or 5 planes for the comples ones.
sandruzzo is offline  
Old 19 November 2016, 09:03   #148
Lazycow
Registered User
 
Lazycow's Avatar
 
Join Date: Oct 2014
Location: Germany
Posts: 195
@sandruzzo: Nice idea, didn't came up with that, yet. But there's a drawback: If you blit 3-plane blitter objects - e.g. for shots or explosions or something, then the background has to be empty. If the object is placed over backgrounds, the colors will be off...

So I guess it would only work in very specific setups.
Lazycow is offline  
Old 19 November 2016, 10:25   #149
sandruzzo
Registered User
 
Join Date: Feb 2011
Location: Italy/Rome
Posts: 2,281
Quote:
Originally Posted by Lazycow View Post
@sandruzzo: Nice idea, didn't came up with that, yet. But there's a drawback: If you blit 3-plane blitter objects - e.g. for shots or explosions or something, then the background has to be empty. If the object is placed over backgrounds, the colors will be off...

So I guess it would only work in very specific setups.
the second blit, will take less cycles since you don't need the whole coockie-cut
sandruzzo is offline  
Old 19 November 2016, 12:50   #150
Lazycow
Registered User
 
Lazycow's Avatar
 
Join Date: Oct 2014
Location: Germany
Posts: 195
what do you mean with 2nd blit?
1. blit shape with mask (do you mean this with cookie-cut?)
2. blit background over shape (restore background)
Both blits would be faster with 3 planes. But it only works if the other planes are empty. (or do I miss something?)
Lazycow is offline  
Old 19 November 2016, 12:55   #151
sandruzzo
Registered User
 
Join Date: Feb 2011
Location: Italy/Rome
Posts: 2,281
Quote:
Originally Posted by Lazycow View Post
what do you mean with 2nd blit?
1. blit shape with mask (do you mean this with cookie-cut?)
2. blit background over shape (restore background)
Both blits would be faster with 3 planes. But it only works if the other planes are empty. (or do I miss something?)
1. blit shape with mask (do you mean this with cookie-cut?) first tree plane
2. make an hole into background (faster than cookie-cut) last tree plane
3...other operations
sandruzzo is offline  
Old 19 November 2016, 13:03   #152
Lazycow
Registered User
 
Lazycow's Avatar
 
Join Date: Oct 2014
Location: Germany
Posts: 195
Quote:
Originally Posted by sandruzzo View Post
2. make an hole into background (faster than cookie-cut) last tree plane
Ahh...
Ok, you would need a 2nd blit for each object, but still, that could be a little bit faster, indeed. Nice idea!
Lazycow is offline  
Old 19 November 2016, 13:06   #153
sandruzzo
Registered User
 
Join Date: Feb 2011
Location: Italy/Rome
Posts: 2,281
Quote:
Originally Posted by Lazycow View Post
Ahh...
Ok, you would need a 2nd blit for each object, but still, that could be a little bit faster, indeed. Nice idea!
Yes. so if we need to have small object with less colors, we can use this simple trick. We have 2 possible layout 8 colors plus 8 colors! Not bad!
sandruzzo is offline  
Old 05 July 2020, 19:37   #154
mcgeezer
Registered User
 
Join Date: Oct 2017
Location: Sunderland, England
Posts: 2,702
There was a thread recently on Facebook relating to Xenon 2 which discussed the opinions of how bad and good people liked the game. After delving into the code of the game a little Galahad gave me a link to this thread.

The Amiga version of the game uses the Blitter extensively to so you can't realistically call it a direct ST port as they did go to some effort on the Amiga version.

But the reason for this post is around the speed and technique used on the scrolling, the programmers decided to blit absolutely everything (even the star field according Galahad but I haven't seen this in the code).

The first 2 bitplanes in the game handle the background in the game, they then seem to blit the entire foreground. It looks like they don't make any use of the sprites.

So my question is, did they make the best use of the Amiga hardware? Could they have improved the game speed wise?

My initial thoughs were that they could have used Half Brite mode given the gradients of the palette but that doesn't seem possible from my calculations.

They didn't make any use of the hardware sprites though which would certainly improved things in places.

What else could they have done?

Geezer
mcgeezer is offline  
Old 05 July 2020, 20:21   #155
Hewitson
Registered User
 
Hewitson's Avatar
 
Join Date: Feb 2007
Location: Melbourne, Australia
Age: 41
Posts: 3,772
From a programming perspective I'm not qualified to say what they could have done better.

As far as I'm concerned there is absolutely no excuse for the framerate. Games like Hybris and Apidya look much nicer and run far more smoothly.
Hewitson is online now  
Old 05 July 2020, 21:31   #156
mcgeezer
Registered User
 
Join Date: Oct 2017
Location: Sunderland, England
Posts: 2,702
Quote:
Originally Posted by Hewitson View Post

As far as I'm concerned there is absolutely no excuse for the framerate. Games like Hybris and Apidya look much nicer and run far more smoothly.
Hybris and Apidya don’t use parallax scrolling.

But this is the whole point of exploring the xenon 2 paradox, could it have been done better with the constraints they had? (Better as in frame rate)
mcgeezer is offline  
Old 05 July 2020, 21:56   #157
Galahad/FLT
Going nowhere
 
Galahad/FLT's Avatar
 
Join Date: Oct 2001
Location: United Kingdom
Age: 50
Posts: 8,986
Quote:
Originally Posted by Hewitson View Post
From a programming perspective I'm not qualified to say what they could have done better.

As far as I'm concerned there is absolutely no excuse for the framerate. Games like Hybris and Apidya look much nicer and run far more smoothly.
Both also don't have remotely as many things going onscreen either.

When Xenon 2 gets hectic, it gets VERY hectic.
Galahad/FLT is offline  
Old 05 July 2020, 23:53   #158
roondar
Registered User
 
Join Date: Jul 2015
Location: The Netherlands
Posts: 3,408
Quote:
Originally Posted by mcgeezer View Post
Hybris and Apidya don’t use parallax scrolling.

But this is the whole point of exploring the xenon 2 paradox, could it have been done better with the constraints they had? (Better as in frame rate)
I've thought about it for a bit as I like puzzling about this kinda stuff

Setting aside the number of objects thrown around on screen (as Galahad points out, it can get very hectic which means 50Hz is going to be challenging regardless of any other consideration), the main issue is the dual layer screen. The method used here is not efficient at all and doesn't suit the Amiga's hardware well. However, it's not so easy to just replace it with a way better suited for the Amiga as the game's design choices aren't really "compatible" with the way dual layer screens are normally done on the Amiga.

Normally you'd either use Dual Playfield mode or a Sprite layer. Neither are a good fit here: Dual Playfield mode doesn't have enough foreground colours and a Sprite layer can only display 128 unique pixels per scanline (which is not enough for the graphics used on the background layer).

Now the above two compromises will definitely help increase the frame rate, but will degrade the looks.

So what's left?

Well, I've thought up one potential way to improve things compared to the current "redraw the entire FG" option. What you could do is change the game to run in 5BPL as follows:
  • Put one of the two background layer bitplanes in bitplane 5
  • Combine the other background layer and the 4th foreground bitplane in bitplane 4 (you can probably do this with a masking blit, I'm pretty sure you won't need full cookie-cut)
  • Bitplanes 1-3 are as normal
Since the scrolling is slow in Xenon, it's probably possible to do the masking over several frames instead of needing to do it every frame, which will further help (though this needs an extra buffer to prevent issues with Bob drawing).

To further improve speed, use hardware sprites for as much as possible (an easy way would be to try to draw as many of the player bullets as possible using sprites).

I'm pretty certain that a method like above will run faster than what Xenon 2 uses at the moment. I'm not so certain it will make the game run at 50Hz though.
roondar is offline  
Old 06 July 2020, 07:10   #159
Tigerskunk
Inviyya Dude!
 
Tigerskunk's Avatar
 
Join Date: Sep 2016
Location: Amiga Island
Posts: 2,770
A steady 25 fps would be a huge improvement already, imo.
Tigerskunk is offline  
Old 06 July 2020, 09:17   #160
mcgeezer
Registered User
 
Join Date: Oct 2017
Location: Sunderland, England
Posts: 2,702
Quote:
Originally Posted by roondar View Post
Normally you'd either use Dual Playfield mode or a Sprite layer. Neither are a good fit here: Dual Playfield mode doesn't have enough foreground colours and a Sprite layer can only display 128 unique pixels per scanline (which is not enough for the graphics used on the background layer).
Would you get any gains by multiplexing the background layer with sprites (similar to Risky Woods?) I know RW points to the same graphics data per sprite but due to only needing 4 colours does it not make it possible?

Pretty sure you covered this in one of your cool videos about a free-form sprite layer.
mcgeezer is offline  
 


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
Critical Mass by Parallax, what machine will it run on? Leffmann support.Demos 16 11 August 2021 19:41
Lionheart Parallax question trydowave Retrogaming General Discussion 19 03 February 2020 08:24
Parallax scrolling layer with sprites possible? phx Coders. Asm / Hardware 21 12 July 2015 18:15
Parallax scrolling meant nothing to me until... killergorilla Amiga scene 26 12 February 2006 16:40
Parallax scrolling in DPaint (tutorial) Stein Retrogaming General Discussion 2 17 January 2006 22:18

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +2. The time now is 17:01.

Top

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.
Page generated in 0.10981 seconds with 16 queries