5-color mode with almost no restrictions

[+MrFish]

um I thought that pot of goulash was 4 colors....

 

It's 5 colors; maybe you can't see the gray in between the green and the beige -- without zooming.

 

It's not in 160-mode, wide pixels, though -- unless you want to stretch it, which wouldn't look right.

 

Otherwise, you could get it to look something like this -- with the colors shown above.

 

 

And even then, it's 5 colors with almost no restrictions; so you'll inevitably have some clash somewhere...

Edited December 6, 2017 by MrFish

[Sheddy]

Well spotted on the aspect ratio! It was such a quick and dirty conversion I just remember thinking after resizing horizontally to 192 (widescreen), what an coincidence that vertical was 240, the absolute limit without interlace trickery, and called it a day

Edited December 6, 2017 by Sheddy

[+MrFish]

It was such a quick and dirty conversion

 

Yeah, I assumed as such.

 

 

I just remember thinking after resizing horizontally to 192 (widescreen), what an coincidence that vertical was 240, the absolute limit without interlace trickery, and called it a day

 

I was thinking the same thing when I converted it too.

[fox]

Shift data may be a problem. The playfield shift register already runs at maximum speed in ANTIC 4 so doubling the fetch rate won't speed up the shifter; this means that it will only get a chance to shift out four bits before it is reloaded again. Tests I did years ago suggested that when this happens the new byte is merged into the shift register similarly to when players are reloaded. Since the playfield shifts out bits MSB first, this will cause bit 7 from the name to merge with the character data. I don't remember offhand if the effective result is an AND or an OR.

Take a look at the 4-pixel wide right-pointing triangles:

Is inverse video applied on the shift register output or the load value?

[phaeron]

Take a look at the 4-pixel wide right-pointing triangles:

80col.png

Is inverse video applied on the shift register output or the load value?

 

Shift register output. Bits 7 and 6 are captured from the character name byte and processed in parallel to the shifter with CHACTL and the current mode to handle character attributes, and then applied to the shifted data in the ANx bus output logic. There's no pre-shift processing as the input from the data bus or line buffer goes straight into the shifter. See ijor's reconstructed ANTIC schematic for details.

[fox]

The shape of the triangle is not helpful in determining if it's an OR or a full load. But with the inverse logic happening afterwards, it's not an AND.

[phaeron]

Should be an OR, with the post-inversion and looking at the earlier characters in that row. That's also the way I have it implemented in Altirra when emulating this bug. It's hard to tell from the schematic since bitwise AND and OR are duals of each other with inverted data and it's difficult to track polarity through 7 layers of inverters.

[fox]

As expected, non-bad lines are affected by ANTIC's 48 bytes of RAM:

[ijor]

If I understand correctly what you are doing, reloading the PF shift register before it finishes shifting, then yes, the result is an OR.

 

It is indeed difficult to track the chain of inversions, and I wasn't always fully coherent labeling the polarity of the signals. But the internal data bus (that I labelled Din) is positive, and there are two inversions before the wide NOR that combines the values. So the values are positive at that NOR. Also the input to the shift register at the far left side is a logic one with a single inversion before the NOR, hinting again that the polarity is positive at that node.

[Synthpopalooza]

Could this method be used for a 4x8 80 column mode in Graphics 0?

[phaeron]

Could this method be used for a 4x8 80 column mode in Graphics 0?

 

No, ANTIC only has 63 bytes of line buffer. There isn't enough bandwidth on the memory bus to fetch both character names and character data for 80 columns, and not enough on-chip storage to buffer the names. But 63 columns, if you're willing to put up with a blank line between each row? Maybe.

 

Edit: Derp, 48 bytes of line buffer, not 63.

Edited January 9, 2018 by phaeron