Jump to content
IGNORED

Images generated by RastaConverter


Philsan

Recommended Posts

Interlace with tremendous sunburn....

Typical response. Care to point to any photo-realistic colour images on the A8 in 320*200?

 

Typical response.... comparing final stuff with work in progress...

 

Imagine this one in interlace....

post-2756-0-67824800-1343243378_thumb.png

 

emkay-demo100.xex

 

C64 needs interlace for more colours.

A8 needs interlace for more details.

 

Yes C64 requires interlace for more colors (although some additional mixes can be presented non-interlaced via pal blending)

Can the A8 display 320 hires pixels horizontally in interlace however? Not sure if one of the frames can be shifted 1 pixel hires increment. if it cant, it will always be x2 horizontal res maximum.

 

Can you provide me with the original of that picture you put. I just want to see how it looks non-interlaced on the c64 via my converter. :-) I am not trying to have a grudge match etc. C64 is great, so is Atari, each have their strengths and weaknesses.

 

Algorithm you are new here, but I know your work from csdb, but emkay is well known here for being full of shit and never delivering [anything other than underwhelming] 'amazing' examples when pushed and nothing but a dick head troll with zero talent and very little understanding of anything outside his blinkered flaming fanboy world of Atari 8bit....the most mighty machine in the whole world which apparently has ZERO compromises and is BETTER than a 68k Amiga amongst things in his tiny mind :D It's a shame because he seems to ruin every thread he posts in between 1% and 100%

 

Oswald gave up on his antics here longgggggggg ago :)

 

PS welcome here and thanks for all the great tools you've released for us mate.

Edited by macgoo
Link to comment
Share on other sites

Interlace with tremendous sunburn....

Typical response. Care to point to any photo-realistic colour images on the A8 in 320*200?

 

Typical response.... comparing final stuff with work in progress...

 

Imagine this one in interlace....

post-2756-0-67824800-1343243378_thumb.png

 

emkay-demo100.xex

 

C64 needs interlace for more colours.

A8 needs interlace for more details.

 

Yes C64 requires interlace for more colors (although some additional mixes can be presented non-interlaced via pal blending)

Can the A8 display 320 hires pixels horizontally in interlace however? Not sure if one of the frames can be shifted 1 pixel hires increment. if it cant, it will always be x2 horizontal res maximum.

 

Can you provide me with the original of that picture you put. I just want to see how it looks non-interlaced on the c64 via my converter. :-) I am not trying to have a grudge match etc. C64 is great, so is Atari, each have their strengths and weaknesses.

 

Algorithm you are new here, but I know your work from csdb, but emkay is well known here for being full of shit and never delivering [anything other than underwhelming] 'amazing' examples when pushed and nothing but a dick head troll with zero talent and very little understanding of anything outside his blinkered flaming fanboy world of Atari 8bit....the most mighty machine in the whole world which apparently has ZERO compromises and is BETTER than a 68k Amiga amongst things in his tiny mind :D It's a shame because he seems to ruin every thread he posts in between 1% and 100%

 

Oswald gave up on his antics here longgggggggg ago :)

 

PS welcome here and thanks for all the great tools you've released for us mate.

 

This was really unnecessary ...

Link to comment
Share on other sites

a magyar Atari user? I only know 3... the guys from HARD. ;) Svevasz.

 

Szevasz

a magyar Atari user? I only know 3... the guys from HARD. ;) Svevasz.

 

Szevasz too ;)

The HARD guys were not just simple Atari users, like me ... Though I haven't knew them personally nor met them.

Hi.I really like their work on A8.Anyone knows anything of them today?

Link to comment
Share on other sites

Algorithm you are new here, but I know your work from csdb, but emkay is well known here for being full of shit and never delivering [anything other than underwhelming] 'amazing' examples when pushed and nothing but a dick head troll with zero talent and very little understanding of anything outside his blinkered flaming fanboy world of Atari 8bit....the most mighty machine in the whole world which apparently has ZERO compromises and is BETTER than a 68k Amiga amongst things in his tiny mind :D It's a shame because he seems to ruin every thread he posts in between 1% and 100%

 

Oswald gave up on his antics here longgggggggg ago :)

 

PS welcome here and thanks for all the great tools you've released for us mate.

 

This was really unnecessary ...

 

Don't feed the troll. Such level of selfish faked up and unrealistic babble don't deserve any caution.

I wonder, if he has more recognition abilities in the real life.

Link to comment
Share on other sites

If something should be proven, it should be proven with solid facts, not made up with words. Credentials are via evidence and facts only and by work created, not by words. The example picture with the unrestricted color mixing of the atari's colors' is impossible on the real thing hence should not have even been posted. Either a snapshot or executable of the real image as displayed on the atari is all that is required.

 

Why can't people just discuss and share idea's instead of just arguing. One other person previously shared a link to interlaced images on the atari using this example pic in this thread which are true outputs.

Link to comment
Share on other sites

It would be nice to have a thread completely dedicated to the beautiful images generated by the awesome converter made by Jakub "ilmenit".

That tool really gave colors, once reserved to the classes, to the masses.

 

Original image (optional), converted image preview, Atari .xex file (filenames format: "Author's name_imagename" to keep Kaz atarionline.pl's format).

 

Posts without images are not allowed! ;)

Link to comment
Share on other sites

If something should be proven, it should be proven with solid facts, not made up with words. Credentials are via evidence and facts only and by work created, not by words. The example picture with the unrestricted color mixing of the atari's colors' is impossible on the real thing hence should not have even been posted. Either a snapshot or executable of the real image as displayed on the atari is all that is required.

 

Why can't people just discuss and share idea's instead of just arguing. One other person previously shared a link to interlaced images on the atari using this example pic in this thread which are true outputs.

 

1. Why can't people have a look at the thread's title, keeping the topic?

If you feel attacked by ... again something the A8 could do better... why not making a new thread?

 

2.You want to discuss and post that virtual picture, that has nothing to do with the thread, and isn't allthough a "true output".

 

3. We only can do proven argues, when a base is there... depending on the development tools.

 

So, here is a real screenshot.... not some dither added double picture(interlace)....

 

....

 

post-2756-0-22249500-1343457555_thumb.jpg

 

 

And, now , end offtopic. If you want to make a clear discussion, start a new topic.

Edited by emkay
  • Like 1
Link to comment
Share on other sites

Ok, a challenge for you all, and one which really pushes the limits of this program:

 

Original:

 

post-23798-0-65577700-1343620823_thumb.jpg

 

My pathetic attempt:

 

post-23798-0-10856800-1343620856_thumb.png synthpopalooza-rainbowrose.xex

 

I used knoll dithering, 10000 solutions, and gave up on it after a few days and 120M evaluations. Can anyone improve on this? This is a difficult one because of the sheer number of color changes per scanline.

  • Like 1
Link to comment
Share on other sites

Another bug (or a repeat of the same one I posted earlier). I was assisting A8ISA1 get a decent set of SMPTE colour bars. I fixed a blurry jpeg with artifacts in PhotoShop. I cut out all the bars, and filled them in with a single solid colour - taken from the middle of each bar. Every edge was crisply defined. I saved the image as a 24-bpp BMP at 320*240 resolution. Options for running Rasta Convertor 5.1 were - yuv colourspace, no dither, 10000 solutions, use only colours in picture, laoo palette. Ran to 12.7 million evals (an hour).

The destination PNG looks plerfect. The XEX screws up the bottom 3 vertical narrow stripes - they should be 3 shades of gray.

 

Source:smpte_320.bmp

RastaConvertor:

post-650-0-02223400-1343705290_thumb.pngpost-650-0-66503100-1343705288_thumb.pngpost-650-0-33667000-1343705289_thumb.png

post-650-0-55818300-1343705802_thumb.png(note the skinny blue & red bar - it should be all black)

output.xex

 

All files to run in "continue":

SMPTE.zip

output.xex

Link to comment
Share on other sites

 

The destination PNG looks plerfect.

 

Note the violett bar that had to be blue. And, as my 256 colour "warrior" screenshot shows, the red is too brownish. while the rest of the palette is ok.

 

Blue has a "brighter" state than white. Which is an error in the colour chosing.

Same with a straight blue, that gets turned into a violett.

And the old bug with the unnecessary different coloured line in the center of a unicoloured range.

 

 

post-2756-0-16123200-1343715473_thumb.png

 

emkay-zx450.xex

Edited by emkay
  • Like 2
Link to comment
Share on other sites

Another bug (or a repeat of the same one I posted earlier). I was assisting A8ISA1 get a decent set of SMPTE colour bars. I fixed a blurry jpeg with artifacts in PhotoShop. I cut out all the bars, and filled them in with a single solid colour - taken from the middle of each bar. Every edge was crisply defined. I saved the image as a 24-bpp BMP at 320*240 resolution. Options for running Rasta Convertor 5.1 were - yuv colourspace, no dither, 10000 solutions, use only colours in picture, laoo palette. Ran to 12.7 million evals (an hour).

The destination PNG looks plerfect. The XEX screws up the bottom 3 vertical narrow stripes - they should be 3 shades of gray.

 

Thanks for the bug report. I will take a look into that. Most probably this is the known bug with the imperfect emulation of ANTIC sprite repositioning in RC.

In this case the expected output is different then from the behaviour of the executable running on the real Atari (or Altirra). Well... only because of this bug I still use the beta versioning ;-)

I still did not have time to think up the best way to fix it not screwing the Phaeron optimizations.

  • Like 1
Link to comment
Share on other sites

Another bug (or a repeat of the same one I posted earlier). I was assisting A8ISA1 get a decent set of SMPTE colour bars. I fixed a blurry jpeg with artifacts in PhotoShop. I cut out all the bars, and filled them in with a single solid colour - taken from the middle of each bar. Every edge was crisply defined. I saved the image as a 24-bpp BMP at 320*240 resolution. Options for running Rasta Convertor 5.1 were - yuv colourspace, no dither, 10000 solutions, use only colours in picture, laoo palette. Ran to 12.7 million evals (an hour).

The destination PNG looks plerfect. The XEX screws up the bottom 3 vertical narrow stripes - they should be 3 shades of gray.

 

Thanks for the bug report. I will take a look into that. Most probably this is the known bug with the imperfect emulation of ANTIC sprite repositioning in RC.

In this case the expected output is different then from the behaviour of the executable running on the real Atari (or Altirra). Well... only because of this bug I still use the beta versioning ;-)

I still did not have time to think up the best way to fix it not screwing the Phaeron optimizations.

 

I don't see anything fancy going on in the raster program in that area, just a ton of NOPs. The error is also showing up both in Altirra 2.0 and 2.1 test so it's definitely not the bug that I found. The b3-based source I have seems to spot the problem and re-evaluate it. The problem might be that in the current versions there is some state that is not being tracked in the line cache key, which would cause the raster evaluator to use a cached solution that doesn't actually match.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...