ZX Spectrum birthday


Those videos ARE the 128K's music (sames the same as what you can download here: http://www.worldofsp....cgi?id=0002408). There's no way you could get that sort of music in-game with 1-bit software audio. It sounds even cleaner than the demo Craig posted. I tried to find a video of the original but didn't have a lot of luck finding anything that definitively showed it.


There's a top 10 for 48K music on Youtube.. they don't sound too good. And the description on the first one implies none of them are during gameplay.
 
Last edited by a moderator:
Not very robust things I imagine.That coloured stripe is quite iconic though.

They were actually pretty robust - the plastic body was a softer ABS, quite thick and not brittle, you'll not see too many with cracks, they either survive or were abused to death (Daley Thompson's decathlon, I'm looking at you and your spawn). The physical weak points were the DC input socket - not really weak, but because there was no power switch it got used way too much as a 'reset', since most games didn't offer you a 'quit' option so you could embark on the process of loading a new game and boiling the kettle.


The Keyboard membrane - would have survived longer had games like Daley Thompson's olympic button mashing and their ilk not come along in such numbers. Replacing them was a doddle thankfully, and they can still be found today.


They used to get packed in sports or shopping bags and hauled about a lot too (The Spectrums, not the membranes!).

They were partially made of metal so they were pretty solid, although it was possible to pry them apart as I think they were glued together.

The metal template was attached with a kind of double-sided tape at the edges, but that was it - the plastic case was screwed together, no glue. Even on the rubber feet, which were also held on (well, almost - they did fall off in the cold) with a sticky backing. One of the hazards of computer clubs was having your spectrum stripped for spare parts while your back was turned.


Anyone curious about spectrum music capabilities should also check out 'Zombie Zombie', the follow-up to Ant Attack, it boasted 'multi channel audio!' - though I seem to recall it was only during the menus, not in-game. Would have spoiled the atmosphere.
Actually I have played daley thompsons decathlon on a spectrum. It was an absurd button mashing thing with stick figures.This was one of the speccys i saw that was covered in sellotape.It was in a friends house in the 1980's.his brothers used to mainly play daley thompsons decathlon on it. I think there was javelin throwing and a high jump vault in it etc.This could be why it was so battered up.If memory serves some of the keys were no longer responsive on the keyboard either.
 
This was one of the speccys i saw that was covered in sellotape.It was in a friends house in the 1980's.his brothers used to mainly play daley thompsons decathlon on it.(..) This could be why it was so battered up.If memory serves some of the keys were no longer responsive on the keyboard either.

That would account for the tape. When you needed to replace the membrane you had to yank the metal template off first - ideally you bought a replacement one with the membrane and it would then have fresh gummy tape on it and sit firmly back over the rubber mat. This being the Speccy though, the additional expense and almost certain need to repeat the process at some point (rather than give up insane twitchy finger games) meant most people put the old, now slightly upwardly bent, template back on. That's where the tape would come in. Then come off, then get more piled on top again. Aesthetics mattered little in the pursuit of high scores.
 
What about the 6502 do you think makes it less conducive to rendering audio in real-time? It's not like it couldn't do interrupts. It was unusual to see both software generated audio and gameplay at the same time, but the same goes with Z80 based consoles like Sega Master System, where voice samples would usually happen during pauses in gameplay. Do games on the original 16/48K Spectrums really often have music that plays continuously while the game plays? Because w/o timer interrupts that means everything you code has to be done in tiny chunks that don't run for more than a pretty small number of cycles. Even at a really low sample rate like 1KHz that only gives you 3500 CPU cycles between when you have to update that sound bit, which would amount to a few hundred cycles.. that's got to be an absolute nightmare to program a whole game around. Not only because you only have that many cycles but because you have to keep track of how many cycles everything is using so you know when to update the bit.


I can see why the 128K model added that AY-3 chip.. which was pretty bottom of the barrel in terms of sound quality in 1987 but still a huge step up.

I think the 6502's in available hardware at the time were massively slow in clock terms. The Z80 in the Spectrum was at 3.5mhz and the 6510 (6502 variant) in the C64 was at 0.989mhz. Even the "fast" beeb's 6502 was only 2mhz.
 
Last edited by a moderator:
I think the 6502's in available hardware at the time were massively slow in clock terms. The Z80 in the Spectrum was at 3.5mhz and the 6510 (6502 variant) in the C64 was at 0.989mhz. Even the "fast" beeb's 6502 was only 2mhz.

That's the old MHz myth for you. 6502s were lower clock speed but instructions took less cycles on average. They weren't the same instructions - 6502 programming was generally more restrictive and cumbersome, at least IMO - but it was still generally capable of performing a lot more at the same clock speed. I'd probably take a 2MHz 6502 over a 3.5MHz z80 in most cases. I wouldn't favor a 1MHz 6502 over a 3.5MHz z80 with everything else equal but I'd definitely pick a C64 with everything else it has going for it, including less need to spend CPU time on graphics.
 
I think the 6502's in available hardware at the time were massively slow in clock terms. The Z80 in the Spectrum was at 3.5mhz and the 6510 (6502 variant) in the C64 was at 0.989mhz. Even the "fast" beeb's 6502 was only 2mhz.

That's the old MHz myth for you. 6502s were lower clock speed but instructions took less cycles on average. They weren't the same instructions - 6502 programming was generally more restrictive and cumbersome, at least IMO - but it was still generally capable of performing a lot more at the same clock speed. I'd probably take a 2MHz 6502 over a 3.5MHz z80 in most cases. I wouldn't favor a 1MHz 6502 over a 3.5MHz z80 with everything else equal but I'd definitely pick a C64 with everything else it has going for it, including less need to spend CPU time on graphics.

IMHO the Z80a found in the Speccy and AmstradCPC was far superior to the 6502. LDIR - used for moving blocks of memory around, index registers, cp wasn't restricted to nz or nc - one branch instruction could mean 'less than', 'greater than', I could go on.


The thing the 6502 had was hardware designers who realised the limitations of the CPU and built things like the VIC and SID chips.


The Daddy, ahead of it's time was the Sinclair QL.


Still love the 68000.


Sinclair positioned it as a business computer. Looking back it's easy to see how ridiculous a decision that was.


Still...
 
Last edited by a moderator:
It was pretty sad that Sinclair was obsessed with making a 'serious business computer' as the QL could have been a real powerhouse for gaming, well ahead of its time at a great price.
 
Back
Top