8Luminances VS 16Luminances

[José Pereira]

Maybe this seems a little bit strange but I always thought that the 8 more shadows of Gray/White we have on GTIA aren't following the same others in the 128Pallete:

 

Top you have Normal: 0,2,2,4,4,6,6,8,8,10,10,12,12,14,14

Bottom GTIA luminan: 0,1,2,3,4,5,6,7,8,09,10,11,12,13,14,15

 

The GTIIA specific: 1,3,5,7,9,11,13,15 seems totally out of the Normal increment/decrement of lighetning luminances.

They seem something totally gray different from the others.

And what I always thought is that this luminances could be very usefull on some Gfxs. that the normal Black/6Grays/1White don't aren't good.

 

Will try on the other colours luminances to see if they are like this...

 

Here's the '.g2f' and '.xex' of this: 8 and 16 shades.zip

(The top normal are from left to Right:

Backgr.- Black

And then: PF0,PF1,PF2,PF3,P0;P1,P2 (Players in double width Mode)

[Rybags]

They aren't. Either you've drawn the lower part the wrong way around, or the utility you've used has some sort of bug.

 

ed - I see the problem.

 

You should shift the bottom part to the left by 2 blocks.

 

Luma 0 (top) should be paired up with Luma 1 (bottom) at it's right part, you've got Luma 1 paired up in the right hand part of Luma 2 (top).

Edited October 19, 2011 by Rybags

[José Pereira]

I noticed this for the first time in all that GTIA Demos at YouTube and this night no sleep and want to give it a try.

What I have here is on Atari800WinPlus with Laoo pallete (people says that laoo was the one that fits best)

 

I also read somewhere that the luminances don't increment/decrement the same nº sequence...

But here, even more than this is that they seem another Gray, not the tradintional Gray we are used to on A8)

[Rybags]

The luminences don't have a linear progression, but that's the case with most computers.

 

ie, luma 1 and 2 are very different but luma 14 and 15 are very similar.

 

But luma value 2, 4, 6 etc. in GTIA mode 9 is the same as 2, 4, 6 etc. from any other mode, and the odd ones only available in mode 9 fit neatly in between.

 

Bottom line - you got your picture wrong.

[José Pereira]

The luminences don't have a linear progression, but that's the case with most computers.

 

ie, luma 1 and 2 are very different but luma 14 and 15 are very similar.

 

But luma value 2, 4, 6 etc. in GTIA mode 9 is the same as 2, 4, 6 etc. from any other mode, and the odd ones only available in mode 9 fit neatly in between.

 

Bottom line - you got your picture wrong.

 

You are not understanding me.

I got it o.k.

The bottom Line are the same as the above ones and between them are the specific 'odd ones' from GTIA Mode.

 

 

I was saying that I know that the sequence would not be the same, but what I was trying to say is that the Grays only GTIA seems another colour, another Gray.

They don't seem other Luminances of the traditional Gray of the A8 160x Resolution.

 

 

Here there are all A8 colours and correspondant luminances:

And here all the colours and luminances in 1File=1colour/luminances (all the 16colours '.g2f' and 'xex'):

8 and 16 shades.zip

In case that anyone want to see and analyze in a real Machine as I am using atari800WinPlus emulator with Laoo pallete...

Edited October 19, 2011 by José Pereira

[José Pereira]

From what I am seeing here it seems that instead of having 16colours and 16luminances GTIA Mode and the same 16colours but only 8luminances on 160x resolution, it seems that you've actually would have 32colours.

 

 

It seems that the odd values on GTIA Mode aren't that colour but these luminances are indeed a different colour .

Likle for example:

-> 160x Resolution colour0 you'll have luminances 0,2,4,6,8,10,12,14

I would called colour0A with their 8luminances

-> GTIA: you'll have these but also 1,3,5,7,9,11,13,15

I would called them colour0B with the 'only 8 GTIA luminances'

 

 

Seems that the problem isn't on they don't follow a Mathematical luminance sequence.

Seems that the 'only GTIA' are from a different colour than the ones of the 160x resolution.

Edited October 19, 2011 by José Pereira

[Rybags]

The alignment between the "16 luma" and "8 luma" bars is all wrong.

 

Luma 1 (thin bar) should be aligned with the right hand of Luma 0 (thick bar)

[miker]

31 luminances?

 

http://atari.fandal.cz/detail.php?files_id=740

[Rybags]

Now you'll really break his brain!

[miker]

[José Pereira]

I see that we can allign also like this:

But I still think the same.

 

I am trying to say something, I don't know if it's English or not.

But probably it isn't...

 

Forget it, it isn't usefull anyway...

I'll be back when something usefull I have to show.

 

It seems that lately I am not very well here...

And no, he isn't 'breaking my brain'.

I am not all this stupid.

 

 

I know that It would be best not erase this from my desktop:

(Why don't you go playing with something else?)

Edited October 19, 2011 by José Pereira

[Rybags]

There ya go, you lined it up right.

 

 

No need to have a period about it.

[José Pereira]

I am not stupid nor a clown!

But I still see things different...

[candle]

Gary, could You try to setup GTIA mode change bug and see how many luminances you'll get? I just wonder if it will reset to CTIA mode, or stay somewhere between

[emkay]

The colours ARE defined by the hardware, as "S-Video"

GTIA produces 8 different brightness steps... using the palette, 16 brightness step get reached by the bitmap decoding.

The colours are the results of the 2nd part of the video "mixer" in the Atari. It puts simply the 2nd signal of GTIA (always available 16 colour steps) together.

A Difference of the resulting colours is to find in the displaying device.

On PAL monitors it mixes colours every 2 scanlines. NTSC machines put artefacting in.

TFT (TN) Panels , and depending on the viewing degree, colours and brightness can change heavily.

[Rybags]

Gary, could You try to setup GTIA mode change bug and see how many luminances you'll get? I just wonder if it will reset to CTIA mode, or stay somewhere between

 

Not sure... GTIA mode switching, the thought is that whatever GTIA mode you subsequently go to will work as intended. Just hires mode doesn't because the latch controlling hires won't go back to the setting for hires (according to ijor's interpretation).

 

I guess another look at the decap/schematics might be in order, but there's only a 7-bit palette. But maybe there might be some trick with mode change and existing pixel data where we could access the "odd" lumas in other modes by getting GTIA "stuck" on the lowest bit.

 

But it wouldn't be totally useful if it did work, you'd have it forced on all colours for the remainder of the scanline.

[emkay]

The grey are 100% grey, because they simply were build on the luma line. Every variation there is a problem of the displaying device.

Another point here is the RGB range of the colours in the emulator. To reach the mostly correct colour, the R.G.B. values may vary and result in slightly different colours. But that doesn't happen on the real machine.

The variations of different colours let the eyes produce additional variations "optical illusions".

[emkay]

And,to bring it together.... The also slight variations of the RGB range adds with the colour variation of the display.

You know, you can have 100 TFT Panels, and see a different picture on every....

TN Panels is the worst, the industry made, besides the Plasma... they both cannot handle colours right.

Analog CRTs can, and the new OLED displays do well (but OLEDs were still in development)

That's why professional picture studios have always a CRT in their stash

Or, you have to pay 1000s of $ € .... whatever, to have Colour -DSP "raped" TFT, showing colours almost correct, or to have it usefull for the business at least.

Edited October 19, 2011 by emkay