Could PoKey do this

[Rybags]

No... the preferred way would be a digital capture oscilloscope. But I've only got a CRO. Next best would be capture by PC card at a high sample rate.

[Foebane]

https://brokenbytes.blogspot.com/2018/03/a-48khz-digital-music-player-for.html

 

 

FAKE!

[Sheddy]

Yep, if it was linear, the maximum volume would be

15 + 15 + 15 + 15

(as Rybags said, 4bit + 4bit + 4bit + 4bit)

but in reality it is even less than 60

Jim Slide may approach 6 bit, not 7 as I had put

[+CharlieChaplin]

Afaik, Pokey cannot playback samples / digis / MODs / etc. with 8Bit resolution (only 4Bit or with tricks up to 6Bits). Thats why they invented the Covox enhancement in the 90s. It is a simple thing to add 8Bit digi playback to the A8. There were single Covox devices (selfmade ones and some commercially sold by Lotharek) and several enhancements that included a Covox (e.g. Simple-Stereo by Candle which is still available at Lotharek's shop, IDE+ version S, RAM-320XE, etc.).

Covox_Info.zip

Edited March 31, 2018 by CharlieChaplin

[emkay]

Yep, if it was linear, the maximum volume would be

15 + 15 + 15 + 15

(as Rybags said, 4bit + 4bit + 4bit + 4bit)

but in reality it is even less than 60

Jim Slide may approach 6 bit, not 7 as I had put

Never checked it. But it could be interesting to have a look at why the plain volume does just add , while the generators add in a non linear way.

[Sheddy]

Nope, plain volume is non linear too. Volume never gets close to 60. Same as generators.

 

I need an idiot guide to pulse width modulation. How can average volume of pulse over same time be more finely controlled than fastest 4 bit direct volume writes? I didn't understand Phaerons code when I had a quick look. Obviously I'm missing something fundamental

[Rybags]

With PWM the granularity is a single cycle within whatever the carrier frequency is - generally you want the carrier frequency to be as high as possible but of course since our machine is a bit slow there's limitations.

So, if e.g. you were to use 2 scanlines for PWM your carrier frequency would be just under 8 KHz and you'd have 228 graduations which is almost equivalent to 8-bit sound quality.

 

But of course hitting the exact cycle isn't always possible so there'd be a bit of loss of precision there but the main problem is that we get an unfiltered and undesirable 8 KHz carrier noise.

 

The whole philosophy of PWM is that you're providing "full voltage" for a fraction of the overall timeframe (=duty cycle). It's used in the real world for all sorts of things including audio playback, and many cheap DC to AC inverters use it (generally at a v-high frequency of 20-40 KHz).

[Sheddy]

thanks Rybags. Hmm, I'll still have to think that through a bit, along with what Xuel wrote earlier.