Space Invaders - Why does it skip a beat?

[SlowCoder]

Throughout the game, the invaders march at a beat ... except when there are exactly 4 of them on the screen. Why do they skip a beat when there are 4? Is this a technical thing, like timing, or is it an intentional thing?

[Nukey Shay]

Routine LF437:

 

Table LFFC8 holds the "# invaders left" comparison values. The Y index is increased in value as the comparison values are checked, one-by-one. When there are only 4 imvaders left, the Y index is equal to 5. That index is then used to read from table LFFD4. Those values are used as a framecounter comparison. Because the 5th value ($0B...i.e. %1011) in that table is not symmetrical as far as the bitpattern is concerned, the invaders are delayed an uneven number of frames.

 

So the answer is a little bit of both, really. Technical because of how the AND opcode functions with non-symetrical bitpatterns...by choice because the invaders would be moving at the previous or next speed otherwise.

[+wood_jl]

Routine LF437:

 

Table LFFC8 holds the "# invaders left" comparison values. The Y index is increased in value as the comparison values are checked, one-by-one. When there are only 4 imvaders left, the Y index is equal to 5. That index is then used to read from table LFFD4. Those values are used as a framecounter comparison. Because the 5th value ($0B...i.e. %1011) in that table is not symmetrical as far as the bitpattern is concerned, the invaders are delayed an uneven number of frames.

 

So the answer is a little bit of both, really. Technical because of how the AND opcode functions with non-symetrical bitpatterns...by choice because the invaders would be moving at the previous or next speed otherwise.

 

I'm shamefully too ignorant to understand most of that, but I sort of put together a layman's crude interpretation of what is clearly beyond my intellect, and even then, I'm still impressed. In short, damn, that's a specific and in-depth answer.

[Nukey Shay]

Suffice it to say that there is not an "even" delay that falls between those used for 5+ and 3 invaders. There could have been, if the routine at LF437 was not so simplistic. Toy around with the 8 values in each table (LFFC8, LFFD4) if you want.

Edited July 28, 2011 by Nukey Shay

[mtshark7]

I love Nukey’s technical answers! Sadly enough I do understand MOST of his technical answers with me being a CE

 

 

I always thought it was a timing issue myself, and it's actually a good thing cos when they are like that is when I make my move and destroy the last 4 w/o getting hit and moving on to the next level.

[+Random Terrain]

I love Nukey’s technical answers! Sadly enough I do understand MOST of his technical answers with me being a CE

What is a CE? Cephalopod Enthusiast? Cabalistic Engraver? Coquettish Ecdysiast?

[mtshark7]

Computer Engineer

 

I love Nukey’s technical answers! Sadly enough I do understand MOST of his technical answers with me being a CE

What is a CE? Cephalopod Enthusiast? Cabalistic Engraver? Coquettish Ecdysiast?

[SlowCoder]

I love Nukey's technical answers! Sadly enough I do understand MOST of his technical answers with me being a CE

What is a CE? Cephalopod Enthusiast? Cabalistic Engraver? Coquettish Ecdysiast?

 

[SlowCoder]

I love Nukey's technical answers! Sadly enough I do understand MOST of his technical answers with me being a CE

 

 

I always thought it was a timing issue myself, and it's actually a good thing cos when they are like that is when I make my move and destroy the last 4 w/o getting hit and moving on to the next level.

 

I figured it was timing also. Having been a programmer since I was about 10, I understood it too. I just don't have the time or gumption any more to try to figure it out for myself.

[ComputerSpaceFan]

Slightly off topic, but can any of you technical wizards explain to me why coin-op Space Invaders has its own weird timing issue? When you're down to the last invader, he moves right to left faster than he moves left to right. I always assumed it was something in the design of the clock cycles because it even happens in MAME and other emulated versions of the game.

 

Any ideas?

[Nukey Shay]

That is how the actual arcade game handles it (so MAME is emulating it correctly, at least). Why it's there...who knows? Somebody at MAMEworld ought to know.

[Rybags]

It's deliberate, to disorient the player.

 

There were lots of little strategies like that in old games - even the terrain colour change in Scramble levels was a subtle way of putting the player off.

[Nukey Shay]

I would guess that it's a carry status glitch feature (uncorrected...as you say, to disorient the player) That would easily explain why it happens in one direction and not the other.