Atari's Biggest Mistake

[Mindfield]

You know, that makes me think...I still have no intution, really, how like any sound or set of sounds can be sampled as a single WAVeform...

 

It's roughly analagous to a digital picture. Each pixel in a digital picture is really just a colour value -- in modern terms, a value between 0 and 16.7 million (or 24 bits). Taken individually, the pixels make no sense, but when strung together in a specific sequence, they make a picture. The higher the resolution, the more detailed the picture can be.

 

The same is true for audio, except replace "pixel" with "sample" and change the interpreter from a video card and monitor to a sound card and speakers. Each sample in a CD-quality WAV file has a value of 0 to 65,535 (or 16 bits) and is played back at a rate of 44,100 (44.1KHz) samples per second (analagous to screen resolution). The higher the sample rate, the more clearly and accurately the sound can be reproduced. Most music is mastered at either 48KHz or 96KHz and then downsampled to 44.1KHz for CD playback. (It sounds wierd, but sampling at higher rates than they're intended to be played back at and then downsampling for the final product produces much cleaner results -- much like scanning an image at 600DPI and reducing the size later for posting on the web -- you get a more detailed picture than if you'd scanned it at the size you wanted it at in the first place)

[kisrael]

You know, that makes me think...I still have no intution, really, how like any sound or set of sounds can be sampled as a single WAVeform...

It's roughly analagous to a digital picture. Each pixel in a digital picture is really just a colour value -- in modern terms, a value between 0 and 16.7 million (or 24 bits). Taken individually, the pixels make no sense, but when strung together in a specific sequence, they make a picture. The higher the resolution, the more detailed the picture can be.

Thanks for the explanation...that much matched my understanding I think. What I have more trouble is the next step up....like, I see where a cluster of pixels resolves into a small piece of a picture, but it still defies my intuition that sound does the same thing...

 

I mean, audio-synth is pretty difficult. Thinking back to my C=64 programming book days, I leearned all about sawtooth and other waveforms, so I see how to generate computer-sounding tones. But how a linear series of those "sound pixels", momentary point samples, can then be resolved into a big symphony...it's amazing. (I know what a single pixel looks like, but could a single "point of sound" ever sound like anything , or does it always depend on context of the next and previous points...I think it's very dependent. a visual pixel has RGB values and a brightness and what not, but a sound pixel is just a single scalar measurement...only in a series with its brothers does it pick up so many amazing things like polyphonicism, reverb, pitch, etc....)

 

I mean, on the one hand I know it has to work because of my mental model of an eardrum....I can get a feel that my eardrum takes whatever I'm listening to and vibrates a single waveform to it...I guess the amazing part is how skilled our brains are at breaking up that single waveform into seperate distinct parts. (What is that mathematically, something to do with a "Fourier transform"?) And that idea that any sound hitting one ear is a single waveform is how I get how Edison's wax cylinder can record sound...just make a mechanism for "writing down" that waveform, and then for moving a speaker in the shape of the waveform...

 

Going back to the image analogy, sound is weird because it's so linear. I mean you take a picture of some stuff, and, unless the objects are semi-translucent, each pixel maps to a single surface of whatever is being photographed (I know that's an oversimiplification). But Sound...well, it's like all sound is totally translucent, and unless one sound source is VERY LOUD you can always hear the other sounds at the same time.

[Shannon]

Well see using your (pardon the pun) analogy one can easily connect picture and sound.

 

Just think of each pixel as having a different color and when those colors are combined they form a picture. A pixel by itself is just a pixel. Sound works the same way. Each recorded value, 8-bit/16-bit, represents the volume of the sound at any given moment. These volumes when combined formed the sound.

 

It is hard to visualize something you hear. But if you've ever looked at a visual representation of a sound sample you can get a pretty good idea.

[mos6507]

Memory might be escaping me, but I don't recall running any software on my 65/xe that didn't look like a game that was on the 800. Sure we had some XE carts, but It really wasn't a leap foward. It's not hard to tell the difference between a Commodore 64 game and anything realeased for the Atari's. The graphics were far superior on the 64 and we never had anything to compete with it.

 

That's a matter of opinion. The C=64 may have supported 320x200 but that day-glo fixed 16-color palette drives me absolutely nuts compared to the more artistic gradations possible on the Atari 8.

[mos6507]

Given how much I gripe about the PS2 not having 4 built in controller ports, I guess I should kvetch about Atari ditching the concept on all its 8bit lines...

 

And they really didn't have to, had they stacked another PIA onto the design to make room for the banking bits.

[mos6507]

Last Question

C) Can the Atari (Regardless of how hard it is too pull off) produce the (SAME or BETTER) graphics that we see on the C64?

 

Even a 2600 can generate certain kinds of graphics the C64 can't. The gradients in Enduro come to mind.

 

The thing is, during the early 80s a lot of companies developed software for one target platform and then hastily ported it to the "lesser" platforms.

 

Depending on the perceived market, they would spend more or less time refining the port. Since there were quite a few platforms at that time, it made sense to use tools to do things like convert title graphics over.

 

Doing a DLI or a GED-like mid-scanline color register swap for more colors per screen is very hard to automate, even worse with 80s toolsets. So I think that's the reason so many title screens look worse on the Atari 8-bit.

 

Games that were specficially targetted for the A8 as the primary platform, like Archon or Ballblazer, look better. I think Ghostbusters also looks great on the A8, which is no surprise since David Crane did that one.

 

When you write for a lowest-common-denominator pseudoplatform then you are going to be less likely to do DLIs and other tricks to up color counts on the A8.

[mos6507]

This just makes me wonder if the tactic of losing money on the hardware and making money on the software was such a good idea to begin with.  Being that creating a new console is so cost consuming to the point where companies start using similar technology.

 

CPU is only part of the mix, though. Sony is trying to up the ante with the PS3. Everyone wants to know if they are actually on to something new or not with the cell architecture.

[Mindfield]

With regard to the audio vs. video, it's difficult to conceptualize the comparison because while they're analagous, they reach different and mutually exclusive senses. To further confuse things, it is actually possible to "see" sound by way of a sound editor or oscilloscope -- and most sound engineers are used to being able to see and even to some degree identify sound visually. (For example, it's easy to sample a a drum loop and separate it into its component parts by sight if you know what bass drums, snares and hats look like in waveform) No such analog exists for images -- it's purely vision-oriented.

 

Put simply, the prameters of a waveform basically define the type of sound made. The distance between waves, the height of each peak and the depth of each valley, the shape of each wave -- be it left/right saw, triangle, square, round, or any combination thereof -- all, in rapid succession, define the sound we hear.

[mos6507]

I think I read once that he got royally pissed off at Nintendo for something(at a trade show) and lost out at being the sole distributor for Nintendo. Or something along those lines

 

People keep looking at this through the virtue of hindsight.

 

Nintendo only became something because of the vacuum left by Atari's collapse. Atari would have collapsed before the NES would even have a chance to come out, and depending on the business deal, Atari's troubles might thrown the NES into some kind of legal no-man's-land and left the Sega Master System in its wake for all we know.

 

I see no reason why Atari would have wanted the NES based on what they saw. By that time Atari had never put its label on someone else's console. GCC created the 7800 on Atari's behalf, but it was still created to Atari's own specifications.

 

I know if Atari had released the NES as its own I would not have recognized it as an Atari any more than I did the ST. It didn't follow the same architectural style of the 2600, 5200, and to a lesser extent the 7800. I would have been wondering why Atari had stooped so low, the company that practically invented videogames had felt the need to release someone ELSE'S console as its own?? A travesty. And forcing you to use your right thumb for the fire button and your left thumb on a joypad?

 

I reject the notion that the only reason games rebounded was because games like Mario came out.

 

Pitfall II showed that this kind of game was inevitable as memory prices went down.

 

Had Atari held onto the Amiga it would have WIPED THE FLOOR with the NES 8-bit as the Amiga was a good 5 years ahead of anything in the pipeline.

 

So it's mutually exclusive. If Atari hadn't gone off track, Nintendo would not have taken over and it would never have been necessary for Atari to do anything but TROUNCE the NES with its own engineering prowess.

[kevin242]

Had Atari held onto the Amiga it would have WIPED THE FLOOR with the NES 8-bit as the Amiga was a good 5 years ahead of anything in the pipeline.

 

So it's mutually exclusive. If Atari hadn't gone off track' date=' Nintendo would not have taken over and it would never have been necessary for Atari to do anything but TROUNCE the NES with its own engineering prowess.[/quote']

 

I think you are forgetting Super Mario Brothers and the fact a console

based on the Amiga chipset would have been too expensive. Atari prob

would of done much better with the Amiga tech in their computers, but

I dont think it would of played a role in videogames prob until the next

generation.. and by then the Amiga chipset was as not as far ahead

(compared to SNES/Genesis).

[www.atarimania.com]

I think Ghostbusters also looks great on the A8, which is no surprise since David Crane did that one.

Ghostbusters looks OK on the A8 but it's less detailed than the C64 version. Still, the game doesn't exactly push either system very far. The A8 version was ported by Glyn Anderson.

 

--

Atari Frog

http://www.atarimania.com

[kisrael]

Put simply, the prameters of a waveform basically define the type of sound made.  The distance between waves, the height of each peak and the depth of each valley, the shape of each wave -- be it left/right saw, triangle, square, round, or any combination thereof -- all, in rapid succession, define the sound we hear.

I know. I think I'm just amazed at how good we are at picking out seperate voices from a cacophany. (And there's other crazy stuff like some skills we have at tuning in certain sounds, like it's easier to hear specific conversations in a crowded room...)