kevtris
-
Content Count
859 -
Joined
-
Last visited
-
Days Won
20
Posts posted by kevtris
-
-
Before Kevtris releases any more cores for it, I hope he fixes audio issues in popular games. Just play Streets of Rage 2 and before you've successfully chosen your character, you'll hear some weird stuff. Especially on the character select screen. No amount of tweaking has fixed that for me yet.
I already fixed that one a week or so ago, but I have not pushed the update yet. I fixed a few other audio issues as well, such as the continuous tone on Phantasy Star IV
-
6
-
-
As far as I can tell, here's two ways used to detect it in the games. The first is checking to see if the extra RAM exists. The other way is to write to some of the AY registers, and see if you can read back the values you wrote. Adding the AY should fix detection on everything else.
-
Then why does my other display fill the entire screen and not on this TV for 720p without any interpolation?
These are the settings I use on 720p. There is no shimmering and interpolation is not enabled. My other display fills the screen but my other TV does not. The Nt Mini doesn't have this issue along with it's cores. I want to use 720p because I want to see everything without interpolation.
Maybe somebody else can explain or tell me if they're having issues like I'm having.
It's because the nt mini has 240 lines of picture information. 240*3 = 720. The msg and snt only output 224 lines (well most of the time, and for NTSC msg always). This means a 3x scale will give you 672 lines like mentioned above, thus you will get some letterboxing. You will have to use vertical interpolation if you wish to exactly fill the screen, or you could chop some rows of pixels off your monitor, I guess. (not responsible for loss of warranty, sanity or marriage).
There's simply no way to exactly fill the screen with an integer scale vertically for this reason. It's a mathematical impossibility.
-
6
-
-
I think I found out how they are prototyping things and why it's taking so long...
They are using Tortilla-Board technology.
-
2
-
-
Well they show a thing with a cart in it, but no footage of it running or anything.
-
7
-
-
Too bad the dithering solution has some negative side effects, but I'm happy it's there just the same even if it isn't something that we can enable and just forget about while it does its magic. I bet the worst case examples like Virtua Racing benefit much more than they're harmed by the feature.
Are SG-1000 games on a Genesis Everdrive running with the correct SG-1000 color palette here, or is it like it is on a Genesis 1/2 with incorrect colors? I'm assuming it's the latter.
You get proper TMS9918a palette on SG-1000 games.
-
3
-
-
well it's even worse than that, the only thing they even kind of implied to have running on that board was a spinning cube. everything else was 'illustration purposes only'
They showed last year in august their dev board (complete with photoshipped pics to strip identifying information) running centipede. Of course we don't see anything but it running centipede- no OS or anything, so I suspect it was just windows like I did. The pictures do appear genuine- I set this board up just like they had, and got it to run. Compare the pictures in this article with pictures of my board. They are nearly identical. I still am not sure why that one guy in the last picture has his hand down the back of his pants...
-
7
-
-
I present to everyone... the Poly Ataco Mega Box VCS! It has a 64 bit AMD CPU! and umm, electrons, and chips and things!
First, the M.2 SSD is installed. It has so much POWER that it sticks out past the end of the PCB, and it needs to be held down with a special retention device.
Well crap. It has so much drively power that it wasn't bootable... Never fear, the SSD of doom will fix it right up! SATA to the rescue
After the drive situation is sorted, I proceeded to install Windows 10, earm, Linux on it...
Loading taco.exe....
Success! We have tacos!
Oh yeah, it runs Millipede too.
Seriously though, it is no joke. I did indeed get this AMD dev board to install windows onto the SATA drive via a USB stick, and loaded MAME onto it and ran Millipede. It took around 3-4 hours total, with about 1.5 hours spent futzing around with EDID hacking to fix my monitor's broken EDID so that I could get the entire image to display. (known problem with this old LCD). This shows that Atari did basically no work last year getting that board to run outside of installing an OS and an emulator. Proving that anyone can make an Ataribox prototype in their own home!
One thing, I didn't manage to get into the BIOS so I could not make the M.2 bootable. This is a dev board so it might not be possible, or there's some weird way to do it. If anyone knows, I'd like to hear. I don't even know what BIOS it runs- it doesn't say on boot.
-
28
-
-
Did we reach 500 pages yet? Woot. yes, yes we did!
-
9
-
-
Looks a lot like the Dev board the VCS team used. Does it use the Bristol Ridge chip? I see the board is dated 2014, and the BR chip came out in 2017. Is this why the Dev team blurred out the date of the board because it would be revealed the board is 3 years older than they claimed? I'm sure this was discussed when the first pictures of the board came out, but that seems like ages ago.
As far as I can tell, it's a newer version of the board they have. If you check their board pictures that they photoshopped you can barely see the '14" part of 2014 on it, right where it is on mine. My board was made in 2015 some time. Going by the sticker on the bottom, my board has this CPU on it:
The AMD FX-8800P is the top-tier Carrizo APU of 2015 for mainstream laptops. It is a SoC containing two Excavator CPU modules (with 4 integer and 2 FP units counted as 4 compute cores), a Radeon R7 graphics card with 8 compute cores (512 shaders), a DDR3-2133 (35 Watt cTDP) memory controller and a video engine (including UVD 6 with hardware HEVC / H.265 decoding).
The performances of the CPU and GPU cores depend on the configurable TDP. This cTDP can range from 12 to 35 Watts depending on the cooling solution and size of the laptop. Thus, clock speeds of the compute cores will vary. E.g., the GPU can score between 1992 points (15 Watts) to 2753 points (+38% with 35 Watts cTDP) in 3DMark 11 (P).
If you look carefully at their pictures, there's a "35W" written in sharpie on it, so they are using a 35W tdp CPU, most likely. The CPU on my board is also 35W tdp. I suspect their board is slightly older, because it uses some through hole U shaped current sense resistors (these are the silver things that look like small binder clips), while my board has smaller, SMD ones. Now that I have this board, I can tell what they have plugged into it, and where.
On one of the pictures there's an SSD or 2.5" harddrive laying next to the PCB on the bubble wrap. This is because that's where a board mounted SATA connector lives. You can see it peeking out from the edge of the board in my picture.
We have no way to know what CPU they have mounted to that dev board, but it's soldered in. I think but haven't checked, that the pinout of many of the CPUs is the same, so you can perform an upgrade to your product by replacing the CPU and little else.
I have not gotten the board to boot proper, because I am waiting on some RAM. It takes standard laptop style DIMMs, and there's two slots for same on the bottom. When I power it up (3.2A, 10V on the battery connector. yes... it has a complete charging circuit and battery connector!) and press the power button (on the left side, near the top) it gets stuck in a boot loop, with various numbers showing on the POST display (the 7-segs at the bottom left). A green LED flashes each loop, and a red LED lights continuously. There's another green LED that shows standby power is on.
Hopefully I can get it to boot windows or something and run tempest 4K on it... proving anyone can make an Ataribox prototype!
Would that dev board in its default layout fit inside an Atari 5200 shell? Just thinking out loud for ways the VCS project could advance faster by changing the enclosure.
Yeah... I just need a clear jaguar case now to make it complete...
-
15
-
-
I will just leave this here for now...
-
8
-
-
Ikari mentioned that the save state feature with the sd2snes pro would not work on the Super NT, don't know exactly why.
My question is, are the hardware on the Super NT the problem, or the FPGA implementation that makes this impossible. Suppose it has something to do with how the APU works, that the flashcard have no idea of what music is playing, it makes this this "separately".
So the sd2snes needs to emulate the APU? Please correct me.
Remember I tried an experimental build with the sd2snes with save states, it "worked" but somewhat buggy. Audio did not save the state but that is a known issue that can't be done in the original sd2snes.
It is going to rely on overdriving the SNES bus and force its values onto the data bus while the SPC is trying to also drive the bus. It's basically a bus conflict between the sd2snes and the SPC that the sd2snes probably will always win. I am not sure if it will work properly on every SNES, but it might if it can drive the bus harder than the SPC can on every rev of SNES. The reason it won't work on the snt is because the SPC signals are mainly internal and the FPGA will listen to its internal SPC data before external data.
The sd2snes will put the SNES' SPC "to sleep" and then all audio will be done on the cartridge only, feeding the audio data into the cartridge audio input. Even though the internal SPC is put into an endless loop, it will still be outputting data to the bus when the SPC registers are read, since that's done in the hardware itself. The sd2snes will also output data at the same time, causing the bus conflict that it will win usually when games read the SPC registers.
This also means the audio quality will be down to the cartridge now, since the internal SPC is no longer being used.
-
1
-
-
I had a third type of game gear I have not heard of being reported; It checked the copyright string like the TMSS one did, but didn't have the blue and white screen saying it was licensed or manufactured by sega. I only discovered this trying to run my own ROMs on it. My original launch system played it just fine, but the other GG didn't play it, and just black screened. Turns out it was checking for the TMR SEGA string.
-
I knew you would say this so of course last night before making my post I re-downloaded plenty of NSFs and tested them again. I used JoshW's archive linked from here: http://www.vgmpf.com/Wiki/index.php?title=NSF -> http://nsf.joshw.info/
With so many of them working in players on the PC yet failing on the NT Mini, I don't know what to say. They could, in theory, all be dirty rips. I'll check one or two of them out in a debugger, but could you at least look into your NSF player for a bug causing crashes / glitched music?
You might remember many pages back that I posted oscilloscope traces of a failed FDS load on the NT Mini, and you said that perhaps you needed to add a stronger filter to the IRQ line in your FPGA core for the next release?
I don't really remember much, I'm working 7 days a week, 10-12 hours a day for the last 3 years so I can't remember what I had for supper the day before. Sorry 'bout that. I am hoping to do some fixes to the ntm/snt this year, but I have to get through my current obligations before I can do that.
-
Hi, Kevtris. I know you're nonstop busy, but I still hold out hope that you can someday fix the NT Mini core bugs, like the FDS IRQ not working on both my RAM adaptors on the Mini (meaning no loading of FDS disks), and the SMS FM sustain emulation being somewhat broken...
But did you know that the NSF player core has several bugs? The most disappointing one for me was in FDS NSFs.
From my quick test of Konami & Nintendo NSFs, the following FDS ones simply crash/freeze the player: Exciting Boxing, Exciting Soccer, Green Beret, Famicom Golf JP Course, Zelda II (Link no Bouken), Time Twist, Doki Doki Panic.
Some FDS NSFs will play, but will crash / play obvious erroneous notes starting from a certain song number. These include: Almana no Kiseki (crashes from song #9); Dracula II (#2); Exciting Baseball (#4); Gyruss FDS (#2); Meikyuu Jiin Dababa (#3); Famicom Golf US Course (#C); 3-D Hot Rally (#5)
It depends on what FDS NSF rips you are using. If they are dirty rips, it won't run them. There can't be CLI or writes to the PPU in there. A lot of NSFs used to write to the PPU and turn NMIs on and things like that. As for IRQs from the port, maybe there's a hardware issue with the IRQ pin. I used the FDS RAM adapters (I have three different ones and three drives and all 6 worked) many times. I tried the FDS units on 20 different boards, as well and it worked on all of them.
-
The asm for all of the plugins is available in the copynes zip file on my website. I used tasm (telemark assembler) to assemble them. It's nothing too special or hard to make a plugin. Hacking the existing CNROM plugin would probably be the best solution. You could just force it to dump 16 banks all the time and then use a hex editor to trim it down if need be.
-
2
-
-
The dev on this replied back to me and said it's CNROM, Mapper 3, yet it still doesn't dump right via NT Mini.
I guess it could be wasting 96K of the 128K PRG ROM, and only using 32K of it. The CHR might be similar- it could be using all 128K of it for CNROM duty- this would be 16 banks of 8K each. My CNROM dumper plugin only goes up to 32K though I think it is. So this is why you have missing graphics and such.
-
I don't think my MHROM dumps more than 64K of PRG/CHR, and that has 128K. So that's probably part of it. It could also be some kind of copy of the colordreams mapper (#11) - that has 8 bits nominally but you can do 4 since there's less ROM. I don't see any info/data on how the jumpers on the board work, but it can't be difficult. There's only a single '161, and nothing else, so it cannot be mapper 30.
There's quite a few famicom-only '161 mapper boards that exist, but I am not sure right now what the mapper numbers are at the moment. A few minutes with a multimeter though should clear it up. Just trace the data lines from the PRG ROM to the '161, then the '161 to the upper address lines of the two ROMs and that will tell how it's wired and thus what mapper it is.
-
2
-
-
Yes, I can see the games. They just don't load properly.
Because there's ambiguity in the 2600 mappers in use, some games require a special file extension to work. Otherwise it will just choose the most common mapper for that particular file size. The "atari 2600 release notes.txt" t tells you which extensions to use for the various mappers.
-
1
-
-
Dang man. I hope you're getting well compensated. They are amazing systems but wouldn't be what they are without your work AND you are still working that many hours on it?
well they ain't gonna finish themselves. :-)
-
4
-
-
It looks like a mess on the NT Mini. Not sure why. It doesn't run well on the N8 Everdrive with the latest v1.6 firmware. I can't even run the v2.0 firmware, so I can't test that.
This is on the NT Mini:
Edit: Got the v.20 rc7 (latest release for the N8ED. It's worse than the NT Mini itself lol)
Yeah something happened to my MMC5 implementation when I ported it from the original FPGA NES prototype, to the nt mini. It used to work perfect for all the koei games, but after the port they had various issues. I am guessing the same thing is wrong with Sim City. When I get time I will try to fix it, but have been super busy doing the usual 7 days a week, 12 hour a day grind on the msg. Yep, I worked on xmas day all day too. No breaks or vacations for me.
-
4
-
-
Will there be another firmware branch... or is it a if(serialnumber>X){fancyboot(} ?
nope, it's the same. It's just firmware x.x or greater.
-
Maybe Kevtris can answer that... otherwise, I'll ask Analogue. I bought it to collect it for sure, but I still might be using it.
no, the Ghostly unit has unique boot screens, you can only see them on that model.
-
1
-
-
the Super NT and upcoming Mega SG both run on a Cyclone V, do they not? How does the included FPGA compare pricewise?
So exactly how does Analogue, RetroUSB get around using HDMI on their devices? Last I checked, anyone can go to Mouser or Digikey and buy HDMI connectors in bulk.
Maybe getting the HDCP license is only to use the HDMI logo (which isn't featured on the AVS or Super NT AFAIK) or to generate and/or display protected video streams requires the ten grand per year?
Or if the FPGA already has the HDMI driver built into it, or some other IC chip within the console has a built in encoder, than the chip manufacturer has already paid the HDMI license.
I really don't know if the license can apply to a circuit on the device and not the device itself. But there is no general protection against using HDMI ports for other uses which do not meet the standard.
One company did a kickstarter for a universal controller converter, to allow to plug everything into a PC. The pentagon shaped dongle had a USB on one side and four HDMI ports that plugged into custom controller adapters.
cough cough. something something check the adopter list :-)
-
1
-
Like us on Facebook
Follow us on Twitter
AtariAge on Instagram
Watch us on YouTube
Chat on Discord
New Atari Console that Ataribox?
in Atari 2600
Posted
I don't think I could handle it. I only work on hardware that gets released.