Jump to content
phaeron

Altirra 2.30 released

Recommended Posts

I always have trouble finding rom images for the Atari - luckily there is the emulators.com site, and this zip file located HERE

 

Here are some more (if needed) in my collection.

System Roms.zip

Share this post


Link to post
Share on other sites

Hello,

 

First of all, congratulations, your emulator is awesome!

If i'm not doing something wrong, I just wanted to report that the game rom for Atari 5200 "Joust" doesn't work, I tried a lot of Atari 5200 roms, and just that one refuses to work, I tried with Altirra 2.3 and 2.4 test 14.

 

Thanks!

There are alot of bad 5200 ROM images out there. So you should try another one (like from atarimania.com) or something.

Edited by Shannon

Share this post


Link to post
Share on other sites

Is there any way to get rid of this moire effect?

 

post-21964-0-28424500-1376485265_thumb.png

 

This happens when the editor window size is below around 900 pixels wide (which it often has to be on my monitor because of space constraints when the debugger's open). I don't remember the earlier versions being like this, although I'm aware the filter modes have been augmented somewhat since then. The only way to get rid of the moire effect is to opt for square pixels in fixed multiples, which is razor sharp but obviously less flexible and not very realistic. Been experimenting with the filter modes some without success. Turning VBXE off changes the effect but doesn't banish it.

Share this post


Link to post
Share on other sites

Is there any way to get rid of this moire effect?

 

post-21964-0-28424500-1376485265_thumb.png

 

This happens when the editor window size is below around 900 pixels wide (which it often has to be on my monitor because of space constraints when the debugger's open). I don't remember the earlier versions being like this, although I'm aware the filter modes have been augmented somewhat since then. The only way to get rid of the moire effect is to opt for square pixels in fixed multiples, which is razor sharp but obviously less flexible and not very realistic. Been experimenting with the filter modes some without success. Turning VBXE off changes the effect but doesn't banish it.

 

Get a second monitor. ;)

Share this post


Link to post
Share on other sites

The moiré patterns are due to imperfect filtering of the image when it is upsampled. By default, Altirra uses a bilinear filter, which is fast and ubiquitous in graphics hardware, but is a pretty crappy filter by image processing standards. The bicubic filter gives better results, but you need 3D hardware acceleration available for it to work -- you may not have this if you are running Altirra in a virtual machine.

Share this post


Link to post
Share on other sites

The moiré patterns are due to imperfect filtering of the image when it is upsampled. By default, Altirra uses a bilinear filter, which is fast and ubiquitous in graphics hardware, but is a pretty crappy filter by image processing standards. The bicubic filter gives better results, but you need 3D hardware acceleration available for it to work -- you may not have this if you are running Altirra in a virtual machine.

 

Running on Win 7 with DirectX 11 graphics. Tried fiddling with the settings (using the bicubic filter), but didn't really make much difference. Guess I can live with it. ;)

Share this post


Link to post
Share on other sites

New test build with some more fixes:

 

http://www.virtualdub.org/beta/Altirra-2.40-test15.zip

http://www.virtualdub.org/beta/Altirra-2.40-test15-src.zip

 

This build fixes problems with the debugger sometimes switching to a Flyspeck 5 font, VDELAY issues with missiles, Ultimate1MB issue that let some games access the joystick when they shouldn't, and 65C816 mode doing read/read/write instead of read/write/write for read-modify-write insns in emulation mode.

 

Also new to this version is the ability to run the CPU in 65C816 mode accelerated, at either 3.58MHz (2x) or 7.14MHz (4x). This is very preliminary and WILL break a lot of games and demos, particularly ones that require precise cycle timing. It also runs a bit faster than it should right now because it allows access to hardware registers at fast speed instead of slowing down to 1.79MHz. However, it's useful for testing code that can use a faster CPU, such as a 65C816 OS. Besides productivity and BASIC programs, there are a few games that magically benefit from this -- Ballblazer runs great at 7.14MHz. I did have to put in a nasty hack to slow down the CPU to 1.79MHz for two cycles following a write to WSYNC, so I'd be interested in any experience people have with compatibility of this emulation vs. a real 65C816 accelerator.

Share this post


Link to post
Share on other sites

What were the circumstances that required the two cycle delay after a write to WSYNC? I don't do anything special for WSYNC but I know that you have to be very careful about READY and when you allow it to come active.

 

Bob

Share this post


Link to post
Share on other sites

There's a one machine cycle delay after a write to WSYNC before ANTIC halts the CPU. At 7MHz, the CPU core was running fast enough that it could get in a following STA instruction before WSYNC had taken effect. In hardware, this would probably require an implementation with a delayed write buffer since otherwise it would take two machine cycles for the 65C816 to get in a write to a hardware register.

Share this post


Link to post
Share on other sites

Yep, it sure does... whole bunch of instructions.

 

Looks like I should fix that!

 

I think I'll do my own ANTIC thing and pull READY after one CPU clock cycle. Put in a WSYNC latch that, when set, drops READY at the fall of the next CPU clock.

 

I execute all access to hardware at 1.79mhz. Same for any ROMs - carts or OS.

 

Got any more of these?

 

Bob

Share this post


Link to post
Share on other sites

Feature request ...

 

I'm a MIDI musician, and would like to be able to use Altirra as a MIDI module ... any way possible to add emulation of MidiMate or MidiMax hardware? The idea would be, to enable Altirra to connect to any installed MIDI hardware on the host PC and then be able to run programs written for MidiMate or MidiMax in emulation.

Share this post


Link to post
Share on other sites

Would it be possible to include SoundFont support for the MIDI emulation?

 

I noticed VLC has recently become able to play MID files internally using a specified SoundFont and it struck me then maybe Altirra could use something similar?

Share this post


Link to post
Share on other sites

Also new to this version is the ability to run the CPU in 65C816 mode accelerated, at either 3.58MHz (2x) or 7.14MHz (4x).

 

I ran a program to do some testing. The program runs mostly in the high RAM and is written in 65C816 code. Altirra set to PAL, 1 MB of high memory, my 65C816 OS. Execution time:

 

Real machine 1,77 MHz: 261 sec.

Altirra 1,79 MHz: 256 sec. (less or more OK)

Altirra 3,58 MHz: 128 sec. (as expected)

Altirra 7,14 MHz: 104 sec. (??? expected 64 sec.)

 

That last result is surprising for me, at a 4x clock you'd expect less or more 4x speed, but the improvement looks like 2.5x speed, i.e. with CPU running at 4.4 MHz.

Edited by drac030

Share this post


Link to post
Share on other sites

Antic DMA cycles currently lock out the CPU core for an entire machine cycle. I need to refine the logic for coordinating timing when ANTIC accesses fast RAM and when the CPU accesses the chip bus.

Share this post


Link to post
Share on other sites

Avery, by now you know I'm not a feature pest, in fact I'm one of the people who tell people not to keep requesting the same things over and over BUT I'm about to break my own rule but for a good reason.

 

When you are done with the hard core fixes etc and back to looking at your list of requests would you give a little thought to a graphical sprite and Char set dumper, the reason I ask is simple, there's a lot of talk about competitions and remakes so maybe if a set of simple tools was there to fast track the dumping of these two things it might just make programmers more likely to dig in to such projects.

 

I know people via the debugger can find char sets and sprite bases in a basic way but a little interface previewing the two items and selecting either blocks or individual sets would make it so much quicker for new and old devs to extract the data.

 

As always I merely ask that you give thought to this and do as you see fit, the simple point was to pop it in to your mind.

 

Purely a nice idea without any silly demands..

 

PS. I know people will say that dev's like to design their own stuff but as I said, I'm talking of remakes where you might want to pay homage to a game but redo the levels or simply for new devs to play around with their coding without the hassle of making char sets or sprites.

Edited by Mclaneinc

Share this post


Link to post
Share on other sites

Is there any way to get rid of this moire effect?

 

attachicon.gifCapture.PNG

 

This happens when the editor window size is below around 900 pixels wide (which it often has to be on my monitor because of space constraints when the debugger's open). I don't remember the earlier versions being like this, although I'm aware the filter modes have been augmented somewhat since then. The only way to get rid of the moire effect is to opt for square pixels in fixed multiples, which is razor sharp but obviously less flexible and not very realistic. Been experimenting with the filter modes some without success. Turning VBXE off changes the effect but doesn't banish it.

in the top-menu: View > Stretch Mode > Square Pixels (fixed multiples only)

Share this post


Link to post
Share on other sites

in the top-menu: View > Stretch Mode > Square Pixels (fixed multiples only)

The only way to get rid of the moire effect is to opt for square pixels in fixed multiples, which is razor sharp but obviously less flexible and not very realistic.

But yes - it does get rid of the moire pattern.

Share this post


Link to post
Share on other sites

I love accelerated 65C816 processor. Star Raiders uses 6502C processor but works fine with accelerated 65C816 processor. Star Raiders run much more smoothly with 2X or 4X accelerated mode, especially when enemy ship or asteroid "explode". All other games did not benefit anything I know so I did not test other games other than Star Raiders.

 

Atari Basic works fine on 65C816 processor and even accelerated, too.

Share this post


Link to post
Share on other sites

Sorry to bring this up again, but I wondered if anything had come to light about a permanent fix.

 

On my Windows 7 laptop I have had a recurring problem where Altirra forgets I'm using my 360 controller - it just stops taking input, usually mid-game. Someone posted a script which gets it working again, temporarily, but I find I have to run that a few times a day.

 

Well this week I bought a whole new machine, a desktop this time, also Windows 7. Installed Altirra and.... same thing. Four levels into Donkey Kong and the 360 controller fails again.

 

If there is another location I shoudl be asking this, just let me know, thanks!

Share this post


Link to post
Share on other sites

Is there an official place to make Altirra feature requests? Anyway I have one. When you do a frame capture it is based on a 320X192 pixel screen, and when forum members post these to show some graphics. The image when clicked on is barely larger than the thumb nail provided by the Atariage forum. What I would propose would be a submenu under the capture frame that let you select something like:

1X

2X

3X

4X

 

That would make it a lot easier to see the captured frames.

 

Thanks for your time.

 

Bob

Share this post


Link to post
Share on other sites

Got any more of these?

 

DLI timing in general looks like it's going to be a hassle. I tried changing the timing so that ANTIC DMA only takes one cycle on the fast bus -- like the way VBXE does it -- and ran into the problem that Ballblazer started breaking at 7MHz. The issue is that the DLI executes so much faster that it executes its writes to HSCROL too early and screws up playfield DMA, trashing the display list. I'm thinking that it might be necessary to force a few more slow cycles after WSYNC ends so that the write timing in horizontal blank is deterministic. This wastes a bunch of CPU cycles, but otherwise it'll be nearly impossible to do certain types of DLI changes safely at fast speed because there won't be a way to delay the required number of cycles with varying speed accelerators.

 

On my Windows 7 laptop I have had a recurring problem where Altirra forgets I'm using my 360 controller - it just stops taking input, usually mid-game. Someone posted a script which gets it working again, temporarily, but I find I have to run that a few times a day

 

I tried to reproduce this locally but couldn't do so -- the only possibility I can think of so far is that maybe my old 360 wireless controller acts differently (it is an early version I got at a conference). When this happens, does the input map still editor still see the button presses? Also, if you keep up the Game Controller control panel while playing, does it still see the inputs when the controller stops working in the emulator?

 

Share this post


Link to post
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.

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...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...