Bug or Feature?

Hi

Just having some fun with my dev build of chipsounds trying out my new Pokey oscillator code. Here we have the eternal programmer question...

Is THIS SOUND a bug or a FEATURE?

Safe to say, chipsounds will contain some extra "note quite accurate" fake-chips in there.

3 Brothers

What do you do when you want to lower your stress level? Solder of course :)

I wanted to free my breadboards and "stabilise" my test suite against hardware, so i spent a few hours doing those two new boards.

Pokey Trials

Here's a photo of my current breadboard for my Pokey. Just missing a clock divider hooked to my 3.57Mhz crystal to get the desired 1.79Mhz. (doing alright with a 2.0 MHz one at the moment for diagnostics).

Pokey is odd, really. and a challenge to integrate in my current Synthesizer code. All bit patterns (or distortions) vary across frequencies and even sometimes retriggering. This seems to be due to the various Pokey internal sub clocks not keeping synch at all time (frequency divider vs 5bit poly vs 4bit poly vs 9/17bit poly).. to make matters worse, there is also various tweaks to the frequency divider (2 channel mode vs 4 channel mode, etc).

Heres a VERY HARSH sound example of 4 different triggerings of a 9bit noise pattern. warning., this is GRITTY! (the last one is more musical, which is interresting)

My goal is to find the best solution to give the user the widest palette of variations and randomness, representative of someone actually programming a song on a a800 (contemporary tracker writers ares still going mad with those sync problems...)

Small update

Sync Buzzer is DONE! ...
Good thing is that new oscillator code opens the door to tons of other cool new sounds... but you will just have to wait until the demos are out!!!

More AY-3-8910 and YM2149F hacking

One of the last oscillator code bits I need to do is to replicate the Envelope buzzer (and Sync Buzzer) techniques used on those chips.

So again, like with all facets of this project, i just HAD to get my hands dirty!
I used a bread board to experiment (left), and then made a protoboard(right) that proves a bit less flaky than the original. (and well i needed my breadboards for some other stuff)

All my boards now follow a simple setup where I can talk to any 8 bit chips using a serial line (74hc595) and a few chip select lines. I can use either the an Atmel/Arduino/MIDIBoxCore as an interface between the 8bit chip and the PC running Bidule.

It goes without saying that also made a special MIDI encoding for every chip I have on external boards, so that microcontroller code modifications to drive each new chip was minimal (there are limits to my masochistic tendencies)

In short: I can - direct in Bidule- just "type in" the register changes that i need for the chip and then feed the audio back in bidule for analysis and recording... many GB worth of that in fact.

I can also, directly in C++, code MIDI to "Chip register language" transforms to actually get some live musical results out of them.

More boards and pictures soon... Maybe my 1U RACK'ed NES is next :)

One by One ... SN76489AN

The SN76489AN is a very basic chip, probably second to the VIC-I. Even though the AY-3-8910 is a tad bit more complex (the envelopes), in many places you will read both the SN and the AY described as "PSG"s or programmable sound generator.

• Basic 50% duty cycle square
• "1/15" duty "cyclic noise, (rarely used)
  
• 32767 bit long Noise pattern. (different on chip variations)

I was lucky to discover that the 32k-1 bit pattern is different on the SN94624N (used in some TI/99's), and Chip sounds includes both patterns exactly replicated.

That chip was used in way too many computers, consoles and Arcade machines to list, probably due to its simplicity and price... It was however included as a discreet chip on my first video game console, the ColecoVision.

Just a little site to collect my findings and describe my experiments in developing this product.
Already 3 years of work have passed since I started this and first time I actually write about it.
The reason is, well one is always afraid of doing some work for nothing, especially when the concept is pretty bold... to make the Chip Tune VSTi to cover them all.

NAMM 2009 is in two days.
Still things to sort out.... just in the demo.

So its all blips and basic square wavs right? Afraid not!