"7-planes" feature and more..

[Toni Wilen]

btw, B2 map rotator is very corrupted if run on OCS Agnus A500. (tested on real A500, looks same as OCS WinUAE configuration except that little copper timing error which can also be seen in ECS mode)

[coyote]

Originally Posted by Toni Wilen

btw, B2 map rotator is very corrupted if run on OCS Agnus A500. (tested on real A500, looks same as OCS WinUAE configuration except that little copper timing error which can also be seen in ECS mode)

Yes. I am aware of that.
I have tested it on A500 with ocs agnus but the guy who let me test it really didn't like me so I simply didn't have the courage to ask him to lend me his amiga to try to find the workaround. Unfortunately, I haven't known many people with amigas around me in those days. And those that had it usually had ecs agnus. One even had full ecs a500. At the demo party the guys were nice enough to let me try the demo on the official machine before the compo so I was sure it would work. And later I didn't care to make a new release... Shame on me...

[Toni Wilen]

Originally Posted by coyote

Yes. I am aware of that.

I'll just noted it here before someone says this is some kind of emulation bug

About B2 flag. Does it do some preprocessing or is everything is done in that longish level 1 interrupt routine that has lots of blits?

[coyote]

Originally Posted by Toni Wilen

About B2 flag. Does it do some preprocessing or is everything is done in that longish level 1 interrupt routine that has lots of blits?

There are some predrawn 16 pix high strips (the shades I guess), and some hor and vertical strips to cut off the edges and to make seams, but I should really look into source to answer what I am doing in the middle checkered part. I remember that I've also drawn two checkered strips, but how exactly I am using them I don't know.
I'll get back to you tomorrow.

[Toni Wilen]

B2 Flag: there is no bug, need to switch cycle-exact before map zoomer ends.

When zoomer ends, there is blit still active that writes to same area as flag pattern. If it ends too late (=non-ce mode) = it overwrites part of flag pattern.

[andreas]

Well, coyote & Toni...

I feared to write some nonsense, hence I had to wait a bit longer...

But Toni indeed mentions extra halfbrite (EHB) mode in the beta log!

Sooo...coyote could it be that your demo which uses the 7-bitplane trick cannot work on a real Amiga 1000 then?
At least the early models - I have one of those, far away from my home atm - were NOT able to do EHB. (iirc)

[Toni Wilen]

Originally Posted by andreas

Well, coyote & Toni...

I feared to write some nonsense, hence I had to wait a bit longer...

But Toni indeed mentions extra halfbrite (EHB) mode in the beta log!

Sooo...coyote could it be that your demo which uses the 7-bitplane trick cannot work on a real Amiga 1000 then?
At least the early models - I have one of those, far away from my home atm - were NOT able to do EHB. (iirc)

Technically it has nothing to do with EHB. Denise confuses 7 planes (which does not exists) with EHB mode which is 6 planes, max supported in OCS/ECS.

Demo only uses 5 planes, 6th plane is blank = no EHB effects used.

But there is still possibility that A1000 does not support this hack, with or without EHB capable Denise because this hack expects side-effect both from Agnus (7 planes is identified as 4 plane mode DMA cycle diagram) and Denise (7 planes = 6 plane mode enabled)

A1000 Agnus isn't exactly same internally as Fat Agnus versions. (highly unlikely there are internal bitplane function changes but better test it..)

[andreas]

I agree that an A1000 test (real one only!) will definitely be a good nice-to-have.

Originally Posted by Toni Wilen

Technically it has nothing to do with EHB. Denise confuses 7 planes (which does not exists) with EHB mode which is 6 planes, max supported in OCS/ECS.

Demo only uses 5 planes, 6th plane is blank = no EHB effects used.

Well I was just trying to establish a correlation between your information in the beta log!

- OCS/ECS "7-planes" mode didn't enable EHB mode

This reads as: Now - after your fix - it will enable EHB mode. But youve just stated, that - ... eh?

[Toni Wilen]

Originally Posted by andreas

I agree that an A1000 test (real one only!) will definitely be a good nice-to-have.



Well I was just trying to establish a correlation between your information in the beta log!


This reads as: Now - after your fix - it will enable EHB mode. But youve just stated, that - ... eh?

You got what you asked

Demo does not use EHB effects = enables EHB mode without using EHB. Clear enough?

[andreas]

Yup, thanks and sorry for the fuss
Demo enables EHB mode just for the fun of it? That's just as if I add 5 variables in a function of my code because I could need them "someday"

To a more serious note, Toni, I do not like that too much yet. We should definitely retrieve another demo which uses "7-bitplane" trick and which (unlike coyote's) does use EHB in combination, just to check that everything works correctly.

The 7bpl trick can be used to speedup chunkytoplanar conversion up to 17%. The c2p screenmode called "blitterscreen" use scrollregisters
with a mask in the 5th bitplane to do a c2p merge for free.

More info at the ada.unterground.net forum
.
Maybe there is a similar 9bpl trick in the AGA chipset.

set bplcon0 to $1210 (9bpl) and fill BPLdat5-8($118,$11a,$11c,$11e) with masks.

If this use 4bpl dma it will be a 12bit chunkytoplanar revolution.
.
A 12bit/24bit c2p use the last 2 bitplanes as a static mask. in Ham mode. The Mask represent wich color to change at each pixel (RGBB)

For a 1x1 resolution the screen must be set to superhires. (4 pixels form one lowres pixel) The problem is that a 6bpl superhires screen take alot of DMA. A 4bpl Superhires screen is as fast as a 4bpl lowres screen.(with 64 bit Fetch). If this trick works. Fast 1x1 truecolor can be done!

I tested the 9bpl trick in winuae, but it doesn't seem to work. If anyone could test this on a real AGA amiga it would be great.

[Toni Wilen]

Originally Posted by Sp_

If anyone could test this on a real AGA amiga it would be great.

I will test this today or tomorrow.

[Toni Wilen]

Does not work in AGA

I tried plane values from 9 to 15. Only got blank screen (same as zero planes)

Same result with or without Indivision AGA.

Having separate "Agnus" and "Denise" BPLCON0 would have been great, and even better would have been bitplane on/off bitmask

[StingRay]

Originally Posted by andreas

I agree that an A1000 test (real one only!) will definitely be a good nice-to-have.

I've just tested it on my A1000 (thanks for giving me a reason to turn on that old lady again ) with Brainwalker's intro and it worked! Will test with the Lazy Bones demo when I'm in the mood to transfer it from the PC. =)

[coyote]

Originally Posted by andreas

Yup, thanks and sorry for the fuss
Demo enables EHB mode just for the fun of it? That's just as if I add 5 variables in a function of my code because I could need them "someday"

When you set 7 bitplanes the screen is in halfbrite mode. This is not something I did for fun. As Toni said - when you set 7 bpl Denise treat is as 6 bpl which is halfbrite. This is simply how it behaves. The difference is that only 4 dma channels are active (which does not slow down copper) while on the other hand all 6 bpldat registers are used. You then put some data in bpldat5 and bpldat6 with processor or copper and there you go. Since I did not see any use of halfbrite at the time I simply filled the bpldat6 register with zeros to avoid some possible residual pattern. In bpldat5 register I put a pattern which in that particular intro enables me to access the upper 16 colors of the color table without sprites (which is the usual method I suppose). With this I got the access to all 16 upper colors (sprites give you access to only 15 colors because one color is transparent), and more importantly, copper didn't get slower because only 4 bpl dma is activated. So I can use copper to change the color as fast as it can - 8 lores pix.

Originally Posted by andreas

To a more serious note, Toni, I do not like that too much yet. We should definitely retrieve another demo which uses "7-bitplane" trick and which (unlike coyote's) does use EHB in combination, just to check that everything works correctly.

What do you mean by "everything works correctly"?
If you mean "is it working on A1000" then Toni answered it - someone should test it. Ehb will of course not work, but the 4dma 5bpl mode might, but then again it might not.
If you mean "is the ehb working on A500" then yes. You can fill the bpldat6 with a pattern. I didn't need it so I filled it with zero.

[StingRay]

Originally Posted by coyote

If you mean "is it working on A1000" then Toni answered it - someone should test it.

I am pretty sure it will work, I tested it with this intro by Brainwalker which also uses the "7BPL feature" and it worked without any problems on my A1000 so I don't see why yours shouldn't work. Will test today evening I think. =)

Just curious, how did you find this trick? By coincidence (i.e. accidentally writing a wrong value to BPLCON0)?

[coyote]

Originally Posted by Sp_

A 12bit/24bit c2p use the last 2 bitplanes as a static mask. in Ham mode. The Mask represent wich color to change at each pixel (RGBB)

Can't you do on aga similar to what can be done on ocs/ecs?

In other words: turn on the HAM bit but turn on only 4 bitplanes. This will initiate the HAM mode in which 4 dma channels will be used for bpldat1-4 while bpldat5 and bpldat6 will be used as a pattern. Simply fill them once and that's it.

If I remember it correctly I did something similar in B2 for the morphing globe routine. I used HAM with 5 bpls activated and filled the bpldat6 with a pattern. This way I saved some dma accesses. I used HAM to fill the horizontal lines.

I think HAM with 4 bpls will also work. Maybe HAM works even with fewer bpls, but this gives you nothing speed wise because 4 channel dma is always dedicated for screen. (Toni feel free to correct me if I am wrong...)

[coyote]

Originally Posted by StingRay

I am pretty sure it will work, I tested it with this intro by Brainwalker which also uses the "7BPL feature" and it worked without any problems on my A1000 so I don't see why yours shouldn't work. Will test today evening I think. =)

So I was not the only one that found this?!?
I must check it!

Originally Posted by StingRay

Just curious, how did you find this trick? By coincidence (i.e. accidentally writing a wrong value to BPLCON0)?

No.
I intentionally put 7 in bplcon0 and tested the behavior.
Here is my original announcement which was somewhat later put on the internet by a friend (Kristijan Zibreg) because at that time I didn't have the internet.
http://groups.google.co.uk/group/com...0073c7a1342b56

[andreas]

What do you mean by "everything works correctly"?
If you mean "is the ehb working on A500" then yes. You can fill the bpldat6 with a pattern. I didn't need it so I filled it with zero.

Exactly, that's what I meant! Test functionality with an actual pattern unlike in cleared state. But ... if we test on A500, the only reasonable test would regard EHB enabled by the 7-bitplane trick only. I know that standard-established EHB mode works on A500, so there's no point in (re)testing that.
However, that special way by which EHB is (automagically) invoked e. g. in your demo (using the aforementioned software trick), ought to undergo a test IMHO.

Originally Posted by StingRay

I've just tested it on my A1000 (thanks for giving me a reason to turn on that old lady again )

It's been a pleasure for me.

[coyote]

StingRay,

I just watched the Brainwalker's intro but can't say that I understand...
Do you by any chance know for what exactly the 7bpl trick is used in that intro?