4 ms·

The audio is the biggest memory blast for me. That low quality flash-style audio is just so nostalgic. Does anyone know what specifically causes that? Is that j

by jjice 2y ago

The audio is the biggest memory blast for me. That low quality flash-style audio is just so nostalgic. Does anyone know what specifically causes that? Is that just due to audio codecs at the time?

chefandy 2y ago
I don't know exactly how the kidpix sounds were programmed, but I'll bet flash was light-years beyond it. Flash used wav, QuickTime, real Media, and other audio files that were probably made using a DAW or real instruments like they are today, dropped into a wysiwyg Editor. They were usually highly compressed so they could be usable to people on slow dial-up connections which is where that crunchiness comes. In a lot of earlier applications with sound, such as games, the sounds were programmed in assembly using tone and noise generators. It was super fast and lightweight, but obviously one of the first older techniques to get left behind when it was feasible.
- chefandy 2y ago
  (random thought that popped into my head that might be interesting if you like this stuff-- there are lots of gui options for making tones and noise and I personally think it's super fun to explore these sounds.)
ssl-3 2y ago
AFAICT, Kid Pix was first released as public domain in 1989, and published a year or two later by Broderbund. We don't think much about it these days when buying and using a quarter-terabyte MicroSD card is rather passe, and when we can buy a complete USB sound card for ~$5 at Wal-Mart (we call it a "USB C headphone adapter" and it acts just like we'd expect a dumb adapter to act, but it's got the whole works integrated completely inside of the connector shell -- ADC, DAC, power amp, ...). Back then, audio was... problematic. It was unwieldy. At 16-bit stereo 44.1KHz ("CD quality"), a 5.25" double-density disk can hold less than two seconds of audio, and we didn't have the computational grunt for perceptive codecs like MP3. So we just used less of it: We sampled with fewer bits, and we did so less-often. A combination of low sample rate and primitive ADC implementations with not-so-good Nyquist filtering and other issues lead to both a chopped off top end, and a mess of sampling artifacts in an easily-audible range. A low sample depth lead to analog sounds being recorded as hot as possible -- often with deliberate clipping -- lest there be even fewer bits to contain the important parts of the sounds. Some of this could be mitigated with high-quality dynamic range compression, but good outboard compressors were expensive (and so were the 16-bit workflows that would allow that to be done digitally in software). It was a sea of tradeoffs. Further digital processing also had a bad effect on stuff (quantization errors, oh my!) -- somewhat akin to doing things like trying to scale the resolution of ASCII art. And even then, sometimes the audio was too big. So we'd truncate samples by just chopping off LSBs until the sounds were barely-useful even in a chonky fun drawing program (the original disksets from the SoundBlaster card came with "compression" software that did exactly this), just to make things a wee bit smaller because storage was relatively tiny. This was how we got the sounds of the early-ish days of PCM audio on home computers, and much of that carried on into the early flash days (due to bandwidth constraints, not so much storage; CD-ROMs had broadly fixed the storage problem for commercially-published software by then).

by jjice 2y ago

The audio is the biggest memory blast for me. That low quality flash-style audio is just so nostalgic. Does anyone know what specifically causes that? Is that just due to audio codecs at the time?

chefandy 2y ago

I don't know exactly how the kidpix sounds were programmed, but I'll bet flash was light-years beyond it. Flash used wav, QuickTime, real Media, and other audio files that were probably made using a DAW or real instruments like they are today, dropped into a wysiwyg Editor. They were usually highly compressed so they could be usable to people on slow dial-up connections which is where that crunchiness comes. In a lot of earlier applications with sound, such as games, the sounds were programmed in assembly using tone and noise generators. It was super fast and lightweight, but obviously one of the first older techniques to get left behind when it was feasible.

chefandy 2y ago
(random thought that popped into my head that might be interesting if you like this stuff-- there are lots of gui options for making tones and noise and I personally think it's super fun to explore these sounds.)

ssl-3 2y ago

AFAICT, Kid Pix was first released as public domain in 1989, and published a year or two later by Broderbund. We don't think much about it these days when buying and using a quarter-terabyte MicroSD card is rather passe, and when we can buy a complete USB sound card for ~$5 at Wal-Mart (we call it a "USB C headphone adapter" and it acts just like we'd expect a dumb adapter to act, but it's got the whole works integrated completely inside of the connector shell -- ADC, DAC, power amp, ...). Back then, audio was... problematic. It was unwieldy. At 16-bit stereo 44.1KHz ("CD quality"), a 5.25" double-density disk can hold less than two seconds of audio, and we didn't have the computational grunt for perceptive codecs like MP3. So we just used less of it: We sampled with fewer bits, and we did so less-often. A combination of low sample rate and primitive ADC implementations with not-so-good Nyquist filtering and other issues lead to both a chopped off top end, and a mess of sampling artifacts in an easily-audible range. A low sample depth lead to analog sounds being recorded as hot as possible -- often with deliberate clipping -- lest there be even fewer bits to contain the important parts of the sounds. Some of this could be mitigated with high-quality dynamic range compression, but good outboard compressors were expensive (and so were the 16-bit workflows that would allow that to be done digitally in software). It was a sea of tradeoffs. Further digital processing also had a bad effect on stuff (quantization errors, oh my!) -- somewhat akin to doing things like trying to scale the resolution of ASCII art. And even then, sometimes the audio was too big. So we'd truncate samples by just chopping off LSBs until the sounds were barely-useful even in a chonky fun drawing program (the original disksets from the SoundBlaster card came with "compression" software that did exactly this), just to make things a wee bit smaller because storage was relatively tiny. This was how we got the sounds of the early-ish days of PCM audio on home computers, and much of that carried on into the early flash days (due to bandwidth constraints, not so much storage; CD-ROMs had broadly fixed the storage problem for commercially-published software by then).