Atari Video Music on the 2600?

[madscijr]

A long time ago I got me an Atari Video Music (HAD to have one after seeing it in "Over the Edge"!) - it worked pretty well, but took up a lot of space and I wasn't going to be able to justify maintaining ANOTHER ancient device, and ended up selling the thing in a time of need. 

 

From time to time I wonder how we might make an add-on for the Atari VCS (serving as inputs from your stereo) to turn it into an honest-to-goodness Video Music console? 

 

Many moons ago, some friends threw a party and hooked up a line out from the stereo to an old cathode ray TV to make a primative music visualizer like this and this, affectionately named "herbert". 

 

I wonder how an audio signal might be converted to some kind of paddle input on the Atari?

We might have the left and right audio channels being read as paddles 1-2, while paddles 3-4 provide knobs for input, and a bunch of digital inputs from paddle buttons 1-4, the difficulty, b/w - color, game select, and reset switches, to match the controls of the video music. (The gain controls would probably just be physical potentiometers built into the audio input adapter.)

 

How could this be done? Thoughts? 

 

Edited November 17, 2020 by madscijr

[R.Cade]

That's a neat idea, and shouldn't be too hard if you have some electronics experience.

 

[Mr SQL]

Here's the manual for the Audio Spectrum Analyzer cartridge for the Radio Shack Color Computer:

https://colorcomputerarchive.com/repo/Documents/Manuals/Applications/Audio Spectrum Analyzer (Tandy).pdf

 

The Color Computer had an audio input which the VCS lacks however the SuperCharger cartridge has an audio input jack, I wonder if the AD converter in it could be utilized to write a video music program. If so you could use asm or SuperCharger BASIC.

 

[+SvOlli]

I don't think there is any chance for a simple reason: speed. The audio signal was fed directly into the AVM chip to be processed instantly, while it can take several frames to get a reading from the potentiometers. The magic about the AVM is that it's pure analogue technique, and for this, if you want a reproduction close you should rather go for current FPGA hardware instead of the 2600.

 

I also had to have one a couple of years ago to make this:

The backstory for this is here: https://xayax.net/trsi/atari_video_music-the_demo/

[Mr SQL]

17 hours ago, SvOlli said:

I don't think there is any chance for a simple reason: speed. The audio signal was fed directly into the AVM chip to be processed instantly, while it can take several frames to get a reading from the potentiometers.

Awesome Video Music Machine Demo, and explanation of the choreography!  

 

Interesting idea to use the potentiometers from the paddles, I was wondering how accessible the AD converter in the SuperCharger is because the Audio Spectrum Analyzer cart had utilized the AD converter in the CoCo to sample the audio spectrum.

 

I don't know what type of AD converter the SuperCharger is using but it it is pretty maxed out cycle wise pulling data at a high baud rate.

 

The rate is adjustable so perhaps the audio spectrum could still be sampled at a lower rate with greater accuracy than the paddle pots could do?

 

[madatariscientistjr]

Just wondering if anyone has made or seen any developments along the lines of a homebrew Atari Video Music?

 

Someone mentioned an issue with speed and cycles - I think the sample rate wouldn't have to be that high, right? Most music is what, 120 bpm, 160 bpm max? Which comes to 2-3 beats per second. We're just trying to move an image to the beat of the music.

 

I'm envisioning something with minimal hardware you can order inexpensively from a Digikey or even Amazon, and using the existing paddle inputs and not a supercharger (which most people don't have access to).

 

If it's a lost cause or done more easily and inexpensively with a Raspberry Pi or Arduino then so be it, but it would be a very neat thing to see on an Atari VCS! ?

 

Edited March 22, 2022 by madatariscientistjr

[Mr SQL]

On 3/22/2022 at 3:11 PM, madatariscientistjr said:

Just wondering if anyone has made or seen any developments along the lines of a homebrew Atari Video Music?

 

Someone mentioned an issue with speed and cycles - I think the sample rate wouldn't have to be that high, right? Most music is what, 120 bpm, 160 bpm max? Which comes to 2-3 beats per second. We're just trying to move an image to the beat of the music.

 

I'm envisioning something with minimal hardware you can order inexpensively from a Digikey or even Amazon, and using the existing paddle inputs and not a supercharger (which most people don't have access to).

 

If it's a lost cause or done more easily and inexpensively with a Raspberry Pi or Arduino then so be it, but it would be a very neat thing to see on an Atari VCS! ?

 

That's an interesting idea.

 

 I wonder if you put a Dynamic Microphone in place of the 100 Ohm resister in the Paddle.

 

 

 

 

[alex_79]

On 3/22/2022 at 8:11 PM, madatariscientistjr said:

...and using the existing paddle inputs and not a supercharger (which most people don't have access to).

The paddle inputs are digital (apart for the ability to "ground" them when writing to a TIA register, they're identical to the inputs for the joystick buttons). They can just read a logic "1" or "0" depending if the voltage applied to the pins is above or below a threshold value.

 

There's a capacitor connected externally to those pins. That's supposed to be charged through a resistor (the paddle potentiometer) connected between the pin and a fixed +5V voltage. The game code measures the time it takes for the cap to charge above the threshold level. As already said, this can take up to a few frames depending on the resistance (that is the position of the pot).

 

An audio signal is a varying voltage, and you need to sample its instant value for this application.

The SuperCharger input is digital too: the SC tapes contain a series of pulses (single cycle of a sine wave) with fixed amplitude (volume), which results in the digital input to switch between logic 0 and 1 accordingly.

I don't see this very useful to sample music for this application either.

I think you'd need an external ADC (analog to digital converter) connected to any input other than the paddle ones, as the capacitors are of no use for this and I think they would just complicate things instead.

 

 

-------------------------------

Anyway, here's an experiment for the SuperCharger that I did just for fun...

Note that it doesn't even try to replicate the output of the AVM. It just displays... something based on the audio input, but given the limitations mentioned above (and the fact that is just a quick and dirty test that I put together - it's just about 150 bytes), most of the time it doesn't really seem to follow the rhythm... ?

scVM.wav
The TV type switch is used to select the video output (COLOR=60Hz, BW=50Hz)

 

WARNING: it displays quickly flashing images! Seizure risk for people with photosensitive epilepsy!

NOTE: I highly suggest to build a proper adapter when connecting the Supercharger to a stereo source, as the mono plug will otherwise short the right channel and not all devices are designed to handle that situation.