 Previous Messages |
Posted By
IstvanV
on 2017-10-21
10:50:19
|  Re: ep128emu is updated!
Reading in every raster line is simpler, the NICK chip has smaller internal memory than TED, it (re-)reads a 16 bytes long line parameter block in the horizontal blanking period of each line, but does not remember attributes from the previous lines. More can be read in this document: http://enterprise.iko.hu/technical/ET1-3_Nick_Chip_Programmers_Guide.pdf
The Exorcist port by Geco is not complete yet, but there is already a beta version, currently only as an ep128emu snapshot file: https://enterpriseforever.com/konvertalas/exorcist/?action=dlattach;attach=19247
|
|
Posted By
MMS
on 2017-10-16
10:34:55
 | Re: ep128emu is updated!
BTW one question: why the chip needs to read memory in every rasterline?
I mean it is nice for raster tricks, but in normal case the video chip should read out the color only at every 8th rastertime, and should store it till for the next 7 rasterlines without change. Otr maybe misunderstood someting?
BTW it raises the question, if a FLI PETSCII is possible too? Imagine an inverse SPACE character with 8 shades of grey or a Extended backgound color mode with different background color (extra to ink color) in each attribute line  Anyhow I suppose these are too complex for manual drawing purposes and the PETSCII character set it too limited as it is now (missing some shapes to be 100% complete).
|
|
Posted By
MMS
on 2017-10-16
10:24:58
| Re: ep128emu is updated!
The EP Scramble looks pretty decent game.
And fortunately both EP and C16 versions using much less intensive pink and red, than the original arcade version
|
|
Posted By
IstvanV
on 2017-10-16
08:36:50
| Re: ep128emu is updated!
The lack of color attributes is because the video chip does not have a feature similar to bad lines in TED and VIC-II. In each raster line of a character, 2 bytes are read by the NICK chip from video memory, and 1 is reserved for the Z80. The text modes use one of the two bytes as the character code, and the other is read from the character set. A palette of 8 colors and other parameters of the current line are read during horizontal blanking.
There is a feature similar to ECM that uses the most significant bit(s) of the character code to select the upper or lower 4 colors of the palette in 4-color character mode, or one of the 4 pairs in 2-color mode, but this would not be enough to convert Exorcist without losing colors.
By the way, the Skramble port is already complete, there is a description (in Hungarian) and a few screenshots here: http://www.ep128.hu/Ep_Games/Leiras/Skramble.htm
|
|
Posted By
MMS
on 2017-10-15
16:40:12
| Re: ep128emu is updated!
Yeah, I did not realize it was Istvan, otherwise no questions would be raised
Istvan, thanks for the explanation, so really there are wait states.
Actually, it is a kind of crazy not adding color atributes to the character screen. I think the best would be two bytes, one for the character and one byte forEACH character as a background color. You can imagine, how good a PETSCII gfx could be with such a mode? (OK, with a little more advanced character set)
(I think a similar thing killed the SAM Coupe too, despite it's 6MHz CPU, it was only a little bit faster than a 3.5MHz 48K Spectrum. Only 12.000 sold)
|
|
Posted By
IstvanV
on 2017-10-15
10:34:46
| Re: ep128emu is updated!
On the Enterprise, access to the last 64K of RAM is slower because it is shared with the video chip, but the rest of the memory can be used without wait states. The EP64 does not have any additional faster RAM, so the code has to run in video memory. Without clock stretching or wait states, the Z80 is maybe 2-3 times slower than the 6502 at the same clock frequency, it depends on the code, some routines can be optimized relatively better on one CPU or the other. Exorcist on EP64 is only about as fast as it would be on a 2.32 MHz Z80 in the ideal case.
The Enterprise port also needs to do more work since it runs in the equivalent of hi-res graphics mode instead of text. The NICK chip does support character modes, but they are limited by the lack of color memory (separate color for each character).
|
|
Posted By
siz
on 2017-10-15
08:53:12
| Re: ep128emu is updated!
That was IstvanV who said that the C16 is faster than the EP64. And I believe him. If someone knows for sure that must be him.
I think 4MHz Z80 is about the speed of the 1MHz 6502 architecture due to newer internals of the 6502 family. I'm not sure where I read this though.
|
|
Posted By
MMS
on 2017-10-15
16:30:09
| Re: ep128emu is updated!
Interesting, that they state that the EP64 is slower than the EP128, although only differ from each other in the size of the memory. (if you play on EP64, then the sprite routine is slower).
The most interesting part, that someone stated on the forum, that the 4MHz Z80 based EP64 is slower than the C16.  .
I know there were a lot of objective measurements, and the 4.0MHz Z80 is pretty fast. Maybe due to the cooprocessors (Nick, Dave), there are "bad lines" on the EP too?
|
|
Posted By
gerliczer
on 2017-10-14
04:24:52
| Re: ep128emu is updated!
Let me carry on with the /OFF
Someone converted, or maybe it is better to say ported or rewritten, Exorcist for the Enterprise computer.
/ON
|
|
Posted By
MMS
on 2017-07-07
14:02:50
| Re: ep128emu is updated!
Hi Istvan,
Sound like a nice challenge! I am sure I will check it when ready, I am sure to had a workaround!
Surely the sound need improvement, as with C16 version I always got headache after 20 minutes. ;-D
OFF
My best friend in secondary school was an E128 fan, while I was dedicated to Plus/4. we had a lot of discussion on the two computers, and I do remember he frequently highlighted the 256 colors on screen, but I am confident he never mentioned the reduced screen resolution 
|
|
Posted By
IstvanV
on 2017-07-05
15:23:19
| Re: ep128emu is updated!
The NICK chip has both advantages and disadvantages compared to TED, but in the particular case of character mode this game, it is mostly disadvantages: there is no hardware horizontal smooth scrolling, and no separate color byte for each character. Thus, the graphics had to be limited to 4 colors, and all characters are low resolution. On the other hand, the start address of the screen can be anywhere, and not just at 2K boundaries, this allows for faster scrolling without having to move large amounts of data. Smooth scrolling was achieved by using multiple character sets, and translating pairs of the original 40 characters to 174 characters with a table. The remaining 82 characters are used by sprites.
By the way, the port is mostly complete, it still needs some tweaking and better sounds, but it is playable.
|
|
Posted By
MMS
on 2017-06-25
16:10:31
| Re: ep128emu is updated!
/OFF
you surprised me. I though the E128 video IC ("Nick") is much more advanced than what the Plus/4 has.
/ON
|
|
Posted By
gerliczer
on 2017-06-25
08:12:03
| Re: ep128emu is updated!
/OFF
IstvanV is working on porting Skramble to Enterprise. As the video IC lacks the capability to do horizontal fine scrolling, he's using an interesting technique. The scrolling is emulated via multiple character sets.
/ON
|
|
Posted By
IstvanV
on 2017-06-14
15:27:47
| Re: ep128emu is updated!
New binary packages of 1.2.11 beta are now available at GitHub: https://github.com/istvan-v/plus4emu/releases
|
|
Posted By
bubis
on 2017-06-10
05:04:14
| Re: ep128emu is updated!
Great stuff!
|
|
Posted By
IstvanV
on 2017-06-09
07:29:43
| Re: ep128emu is updated!
There is a fix to this issue in the GitHub source code now, but it still needs to be tested if it is correct. A few other fixes and changes since version 1.2.10 are listed in the NEWS file (https://github.com/istvan-v/plus4emu/blob/master/NEWS).
|
|
Posted By
BSZ
on 2017-05-27
16:17:11
| Re: ep128emu is updated!
I run the inv_test.zip in real hardware, and I show this results:
$00/$D0/$C2: http://bsz.amigaspirit.hu/pics/plus4-videotest4-00D0C2.png
$00/$D1/$C6: http://bsz.amigaspirit.hu/pics/plus4-videotest5-00D1C6.png
(The original version is captured, the name is "...test3-00D1C4", but the forum motor not enabled 3 links... )
The 00D1C4 version is 24.9-25.0 (×2 of course) FPS, the second, 00D0C2 version is 25.0 (×2) FPS, the third 00D1C6 is 24.9-25.0 (×2) FPS. The pictures captured again the BT878 card, the external VGA converter working buggy, if the raster lines number not standard. (!= 312 lines.)
|
|
Posted By
gerliczer
on 2017-05-27
14:53:19
| Re: ep128emu is updated!
Hi BSZ,
The second picture looks very similar to the one IstvanV inserted into his reply in the issue tracker. But I'm more curious about the inv_test.zip he attached to his last post. I downloaded and run it in plus4emu but the result is a border in that strange pink or whatever colour it is and a BREAK.
|
|
Posted By
BSZ
on 2017-05-27
14:31:18
| Re: ep128emu is updated!
This picture is digitized by the external CBVS/S-Video to VGA converter, based on SAA7113? Photo from the monitor:
http://bsz.amigaspirit.hu/pics/plus4-videotest2.jpg
The interesting "double line" is not shown... The digital processing of this CVBS signal not correct.
|
|
|
|
|
|
Posted By
Gaia
on 2016-12-29
14:32:07
| Re: ep128emu is updated!
It's interpolating the neighbouring 50 Hz frames at refresh barriers on the expense of more washed out frames and slightly increased latency.
It's very cool, though these days I prefer setting my monitor refresh rate to either 50 or 100 Hz 
|
|
Posted By
Mad
on 2016-12-28
16:53:41
| Re: ep128emu is updated!
The "resample to monitor refresh rate" option is pure gold, btw.. So the scrollers don't even jitter anymore.. Would be interessting how this is done.
|
|
Posted By
gerliczer
on 2016-12-28
15:57:26
| Re: ep128emu is updated!
Me neither. But I did assume that it should be accessed without jumping through hoops, so I took the time and tried to read everything on the screen.
|
|
Posted By
Luca
on 2016-12-28
15:29:04
 | Re: ep128emu is updated!
Oh wow, I don't know how basically GitHub works  Got it, thanks.
|
|
Posted By
gerliczer
on 2016-12-28
09:48:46
| Re: ep128emu is updated!
What is the problem? There is a link called releases above that magenta or what is that colour line. That leads to the binary releases. The installer folder is part of the source tree.
|
|
Posted By
Luca
on 2016-12-28
09:37:16
| Re: ep128emu is updated!
Ok, dunno how to use this installer, mmm... 
|
|
Posted By
MMS
on 2016-12-21
22:40:07
| Re: ep128emu is updated!
WOW, that is fantastic! I am really happy to see this post!
Plus4emu is still my favourite emulator, mainly because of the faithful CRT TV emulation and lot of possible video settings.
With them I can reproduce well my childhood's Orion Viking television screen .
Thank you, Istvan! ;-)
|
|
|
|
Posted By
Csabo
on 2016-12-21
14:31:03
| Re: ep128emu is updated!
Any ideas on what's new? This should be front-page news.
|
|
Posted By
gerliczer
on 2016-12-21
13:11:58
| Re: ep128emu is updated!
IstvanV updated the plus4emu to 1.2.10 beta a couple of days ago. This is a pleasant surprise after so many years of inactivity.
|
|
|
|
Posted By
SVS
on 2016-12-11
14:24:31
 | Re: ep128emu is updated!
I did try to write him an email a pair of years ago, but... no answer :-/
|
|
Posted By
gerliczer
on 2016-12-11
14:15:00
| Re: ep128emu is updated!
I wouldn't raise my hopes too high. IstvanV frequents their forum regularly but here he's been last seen in July 2011.
|
|
Posted By
Litwr
on 2016-12-11
12:44:40
| ep128emu is updated!
https://sourceforge.net/projects/ep128emu/files/ep128emu2/
It is the first updates since 2011. So it is possible to see plus4emu updated too... I had made a bug and suggested features list...
|
|
