Inviyya - OCS Shoot Em Up

[TjLaZer]

Just wanted to say that the game is very nice, and good job. Keep up the good work.

[Tigerskunk]

Originally Posted by TjLaZer

Just wanted to say that the game is very nice, and good job. Keep up the good work.

Thanks man...

[sparhawk]

Originally Posted by Steril707

The real problem here is, that with NTSC you can blit a lot less per frame.
And games are made with PAL in mind, so you can either go for slowdowns, 25 fps or a lot less objects on screen.

So that's the reason. I was a bit surprised, because I thought, it should be faster on NTSC as you have less screen to move, and also with the copper, you don't need to handle those extra lines.


But it makes sense that the timing changes as well, as a screen is drawn in a shorter time. Quite interesting take on this, I hadn't considered before.

[hooverphonique]

But is a true NTSC machine not exactly as fast (per frame) as a PAL machine?

If you switch a PAL machine to NTSC, the base clock of the system doesn't change, but a TRUE NTSC machine has higher base clock, AFAIK.

[Predseda]

7.09 vs 7.16 MHz IIRC

[ross]

7159090Hz NTSC
7093790HZ PAL

To make an exact count we have to consider 2 whole NTSC frames because of the alternating length lines (for PAL it is simple because they are constant).

In NTSC there are 263 lines with 228 DMA slots and 263 with 227 slots.
In PAL there are 626 with 227 slots.

So for NTSC 119665 and for PAL 142102. A 18.75% difference.
The processor that is 1% faster is of little use.

And it is a difference regarding the DMA usable slots for the display of a video frame, the ones that are used for sprites, copper and blitter, that is, those used by games.
For those wondering why the processor is slightly slower in PAL it is simply to synchronize exactly on the 227 DMA slots per PAL frequency line.

That's all.

[roondar]

Originally Posted by hooverphonique

But is a true NTSC machine not exactly as fast (per scanline) as a PAL machine?

If you switch a PAL machine to NTSC, the base clock of the system doesn't change, but a TRUE NTSC machine has higher base clock, AFAIK.

Not quite. As Predseda pointed out earlier, it all has to do with the video standards. Apart from 50 vs 60 Hz (and a different colour carrier), there's not actually that much of a difference between PAL and NTSC.

The important part is that both run at about 15Khz horizontal sync and both have a bandwidth of about 3.5MHz.

So, if you design a computer like the Amiga and use the video system as the base for your clock, both PAL and NTSC systems will be roughly equally as fast (as both will be clocked at a multiple of about 3.5MHz).

The difference will be that PAL frames last longer, so your system is capable of pushing more pixels per frame over in PAL land. But it really is more or less the same speed as in NTSC land. Why then do games appear faster in NTSC? That is due to the vertical refresh difference. Which means frames in NTSC countries happen more quickly. However, both NTSC and PAL frames are actually roughly the same bandwidth. So both NTSC and PAL will process approximately the same number of pixels in a second. Just divided differently across time.

Maybe this helps clear it up?
Feels a bit too technical and off-topic to me though, so perhaps I'll drop the subject now.

Edit: I misread that, you are actually correct that they are the roughly same speed. My post just tried to explaine why. Sorry for any confusion.
Edit 2: ok, so I might need to take reading classes (and perhaps glasses) . Ignore my earlier edit: NTSC and PAL Amiga's are only "equal" in speed on a scanline basis, not a frame basis...

[malko]

Originally Posted by ross

7159090Hz NTSC
7093790HZ PAL [...]

Thanks ross. Really interesting

[hooverphonique]

Yes, of course the clock difference (1%) is way too small to make up for it. And with the same number of dma slots per line, it wouldn't even matter. Silly me

[ross]

Originally Posted by hooverphonique

Yes, of course the clock difference (1%) is way too small to make up for it. And with the same number of dma slots per line, it wouldn't even matter

And of course the situation is even worse for NTSC games in a real situation because this 18.75% is valid only if all the uses of the DMA slots are reduced proportionally.
But this is practically never true because there are equal uses for many resources (bitplanes DMA, sprites, copper ..).
The only thing that can vary with few problems are the blitter objects (or eliminate some features, such as multiplexed sprites).

What figures are we talking about? Let's do some calculations in the case of Inviyya.
Bitplanes DMA: 20*4*192 = 15360 slots
Sprites DMA: 16*192 = 3072 slots
Sprites multiplexing (copper): 48*106 = 5088 slots
Copper generic (setup, splits, colors, waits, sync..): 10*200 = ~2000 slots
Then there are the refresh slots, the audio slots and the unusable: ~2000 slots.
We approach ~27520*2 out of ~142000, so we can consider to have almost 60% of the slots available that can be shared between the blitter and the CPU.

We can estimate that in very crowded situations the blitter takes 75% of the available slots and the CPU the remainig 25%. This means, for PAL, ~32000 slots per frame.
On NTSC the CPU usage is very similar since the same engine is used.
So 25% has to go up 18.75% and the CPU takes 30% of the frame.

Then for NTSC I will have ~(119000-55000)/100*70/2 = ~22400 slots for the blitter.

Here, to all intents and purposes we can actually put 30% less objects on screen than the maximum possible on PAL
(EDIT: otherwise we have a frame skip, ie slowdown).
So even if you are led to think that 20% of objects less is sufficient (difference between 50fps and 60fps), it is usually not.

[note that there are many approximations and assumptions, so the values may vary quite a bit ]

[MarkBogaard]

Originally Posted by Steril707

The game has been tested on a config like that, so i don't have any idea what is different on your setup here, i am sorry... :/

There are just so many different factors that can play into why a game plays or not on all these different setups. It's kind of horrible, compared to when you are coding to something with a unified setup like old consoles or the C64, for instance..
In the end I can only guarantee vanilla machines, with the notion that the game seems to work right now on most standard setups ->usually<-.

No worries, I have my childhood A500 for playing games and here it works and looks awesome!

[Tigerskunk]

Originally Posted by MarkBogaard

No worries, I have my childhood A500 for playing games and here it works and looks awesome!

Nice to hear that, my friend..
Check out the updated version which I will release soon. Will be a good bit more difficult, and have some bugs fixed.

[nikosidis]

Any new demo this weekend

Very exited to try it.

[Tigerskunk]

Just need to fix one thing, all the other stuff is finished...

[HonsterGamer]

Loved the demo always been a shoot em up fan since hybris was released

[Tigerskunk]

Here we go again..

Version 2 of the demo can be downloaded here:

https://tigerskunk.itch.io/inviyya


As always, I am extremely interested in hearing your feedback regarding bugs, playabilty, etc.

Have fun, hope you guys like the additional challenge..

[roondar]

Been looking forward to this. Gonna give it a spin right now

[Tigerskunk]

And I have to say LOTS of thanks to our ross here, for having a deep look on some stuff that wasn't running so well in my game so far, finding and removing quite a few bugs, and optimizing the living shit out of the game.

I am super impressed by his work...

[roondar]

Ok, I played a bit of it now (will continue after this post as it's good fun ).

First impressions: I feel it's much improved over the last demo, the difficulty is a good bit higher and that's good in this case. I also like the new weapon power up system.

I do have some minor things you may or may not want to do something with. Nothing major, but I hope to be of some small assistance with some constructive criticism

The first is that I noted that the first robot that walks in after you enter the base proper sometimes doesn't spawn (edit: not the one in the entrance tunnel, the first one after that). It usually does, but I had it not spawn at least once. The second is that the enemies firing at you from behind sometimes seem to do so with perhaps slightly too little time for you to react* (this could simply be skill though, I mention it because the first time it happened I didn't realise why I died).

Last is more of a question: the "wave" weapon seems to be less powerful than the single bullets. I'm not sure it actually is, but the bigger enemies took longer to kill with it. Is the idea that it's better suited for multiple targets? Or am I missing something?

Overall I'm impressed, it's quite nice and plays very well

*) I just noted they sometimes fire when already off-screen. Perhaps this is what caught me off-guard?

[thairacerjp]

hope we could have a whdload version for test on my CD32 tf330, but as I saw in demo on youtube, it look very smart, Polish , like the music and the style, I will give my impression when I could play lol ! thanks Steril707