I'm kind of surprised that the board has to wait two minutes before even starting up the warm-up-the-bubble-memory routine but it can play synthesized voice samples beforehand.
The text is misleading. The warm-up actually happens during the first part of the countdown. By the time the screen comes on, it's already at temperature and reading the game. Bubble memory is slow to access, because bubble motion is limited by pesky magnetic effects, and in some way, the inductance of the drive coils.
@@evil-wombat Is that period when the numbers appear not giving it a bit of time for the temperature to stabilize then starting read/verify of the program at 13 hence the momentary lag in the music?
@@soundspark I believe the actual heating process happens even before the music starts. While the warmup is happening, a synthesized voice slowly counts down from 50, but this usually stops long before the voice countdown reaches zero. When the bubble memory reaches the right temperature, the second-stage bootloader is loaded into RAM- I believe this contains the code for the on-screen countdown and the music (though I may be wrong).
@SuperSmashDolls according to wikipedia, "Konami used their new G400 BIOS for this project, and modified it a bit. The main CPU was a Motorola 68000 at 10 MHz. There was a separate Zilog Z80 for sound control, which drove two AY-3-8910s, a custom Konami SCC (K005289), and a VLM5030 speech synthesizer." I guess that's why there could be music/speech/some graphics while it was warming up.
@@Lethaltail It's probably not attempting to read the bubble memory until that point where the ROM says it's been warming up long enough to operate. Past that point it may be verifying checksums prior to passing control on to the game program.
I suppose it could have been used like many block-based systems of today, where it isn't entirely random, but tiny block sections of it can be randomly read?
Blast from the future! Most of these use a major/minor loop architecture, which makes random seeking reasonable. The way it works is data is stored in "sectors" (which are confusingly called "files" in the early specs). You can sequentially seek to the sector you want, and each "step" of the seek operation (in theory) takes as much time as it would take to read out one bit. Once you've seeked to the sector you want, you use the transfer gate (or in later designs, the block replicator) to sequentially read out all the bits of that sector. So if you're reading out one big thing, it's really not bad. There are also some papers out there for how to allocate a large block of data across multiple sectors, to minimize the seek time between them.
Kind of, but the Bubble Memory in the case of the konami system stored far more than high scores; the game itself is on the bubble system cartridge while the ROM chip is present to provide somewhat of a BIOS during this warm-up sequence.
It's not too far-fetched to compare bubble memory to "tape on a chip". The game is stored on this on-chip tape, and just like loading ZX81 and C64 games from tape into RAM, this konami system had to load the game from the bubble memory into RAM to be able to play it. On the other hand, game tapes didn't need to be pre-heated to 30°C to be able to be played back without the tape tearing or jamming, but that's basically what happens in bubble memory if it's not at the right temperature. As far as I know, current knownledge is that the code that plays the "morning music" is actually in the first block of the bubble memory, so again very much like tape-based home computer games that loaded a small piece of code that draws the title screen and continued loading of the actual game in background.
2:17
Thirty-- SHUT UP!!
Yeah, that was rude.
最初この起動曲を聴いたときに、モーツァルトをDX7で演奏したカンジのメロディに鳥肌が立ちそうになったのを
20数年経った今でも思い出します。
Thanks 4 Your Uploadings!
ノイズまでもが尊く愛おしいバブルシステム(*^^*)
VLM5030のカウントダウンからモーニングミュージックに切り替わる時のブツブツ音が鳴るまでドキドキものですよね。
うちのはカウントダウンの繰り返しです・・・
泡のように記録が消えゆく磁気バブルメモリ…。(ﾟｰÅ) ﾎﾛﾘ
ウチの婆さんもバブルシステムです🥺
I'm kind of surprised that the board has to wait two minutes before even starting up the warm-up-the-bubble-memory routine but it can play synthesized voice samples beforehand.
The text is misleading. The warm-up actually happens during the first part of the countdown. By the time the screen comes on, it's already at temperature and reading the game. Bubble memory is slow to access, because bubble motion is limited by pesky magnetic effects, and in some way, the inductance of the drive coils.
@@evil-wombat Is that period when the numbers appear not giving it a bit of time for the temperature to stabilize then starting read/verify of the program at 13 hence the momentary lag in the music?
@@soundspark I believe the actual heating process happens even before the music starts. While the warmup is happening, a synthesized voice slowly counts down from 50, but this usually stops long before the voice countdown reaches zero. When the bubble memory reaches the right temperature, the second-stage bootloader is loaded into RAM- I believe this contains the code for the on-screen countdown and the music (though I may be wrong).
@@evil-wombat What happens if for whatever reason the bubble memory doesn't reach temperature when the voice reaches zero?
@@soundspark that's a good question. I've never owned one of these, so I don't actually know.
Presented by konami
Getting ready
Fun piece of arcade history.
@SuperSmashDolls according to wikipedia, "Konami used their new G400 BIOS for this project, and modified it a bit. The main CPU was a Motorola 68000 at 10 MHz. There was a separate Zilog Z80 for sound control, which drove two AY-3-8910s, a custom Konami SCC (K005289), and a VLM5030 speech synthesizer."
I guess that's why there could be music/speech/some graphics while it was warming up.
You can hear a pulse at 2:17, then a faster pulse when the video is about to play the morning tune...
2:17
You can see the text get drawn on the screen as the music begins and the screen goes blue.
You can also hear the system lag as it reaches 13.
Perhaps that's when the program starts loading.
@@soundspark would be nice to see a top-down view of ram at that moment (or the whole time)
@@Lethaltail It's probably not attempting to read the bubble memory until that point where the ROM says it's been warming up long enough to operate. Past that point it may be verifying checksums prior to passing control on to the game program.
Here's an interesting project for the 8-Bit Guy/Keys
モーニングミュージックも三丁目の夕日もモーツァルトのアイネ・クライネ・ナハトムジークが元なんやなあ（´ω｀）
ずらすと合うのか?
GETTING READY 48 GETTING READY 47
I suppose it could have been used like many block-based systems of today, where it isn't entirely random, but tiny block sections of it can be randomly read?
Blast from the future! Most of these use a major/minor loop architecture, which makes random seeking reasonable. The way it works is data is stored in "sectors" (which are confusingly called "files" in the early specs). You can sequentially seek to the sector you want, and each "step" of the seek operation (in theory) takes as much time as it would take to read out one bit. Once you've seeked to the sector you want, you use the transfer gate (or in later designs, the block replicator) to sequentially read out all the bits of that sector. So if you're reading out one big thing, it's really not bad. There are also some papers out there for how to allocate a large block of data across multiple sectors, to minimize the seek time between them.
Kind of, but the Bubble Memory in the case of the konami system stored far more than high scores; the game itself is on the bubble system cartridge while the ROM chip is present to provide somewhat of a BIOS during this warm-up sequence.
Perhaps what you are thinking of is an IPL, assuming the ROM no longer provides any services once the game starts.
It's not too far-fetched to compare bubble memory to "tape on a chip". The game is stored on this on-chip tape, and just like loading ZX81 and C64 games from tape into RAM, this konami system had to load the game from the bubble memory into RAM to be able to play it. On the other hand, game tapes didn't need to be pre-heated to 30°C to be able to be played back without the tape tearing or jamming, but that's basically what happens in bubble memory if it's not at the right temperature.
As far as I know, current knownledge is that the code that plays the "morning music" is actually in the first block of the bubble memory, so again very much like tape-based home computer games that loaded a small piece of code that draws the title screen and continued loading of the actual game in background.
*PrEsenTeD bBY*
*KoNaMee*
*GeEtiNg ReeAady*