Jump to content

Kismet

Members
  • Content Count

    494
  • Joined

  • Last visited

Community Reputation

193 Excellent

About Kismet

  • Rank
    Moonsweeper

Profile Information

  • Gender
    Not Telling
  1. This seems to be the only direct evidence. Suffice it to say, I don't think the CIC actually ever interfaced with the ROM itself, I'm more inclined to say it has to be some kind of intentional or accidental manufacturing design that might have only been discovered by people using game genies. Fun fact. I wanted a game genie when I was a kid, and it turns out that there are a whole series of game genies that didn't work with my SMB3 cart, which meant we went to more than one store to find one version of the game genie that worked with it. So my early experience with this suggests that there are different versions of the NES game genie. It's likely that the developers of the game genie might even have known about this, since they probably didn't want to be responsible for melting games (yet the GG itself warps the cartridge slot.) Anyway, the evidence suggests: a) Either a chip on the cart is dying, resulting in this effect when the CIC is disabled b) The ROM chip itself has a design flaw that results in more current being pulled if the CIC isn't pulling any current. At any rate, I probably wouldn't risk the cart, though I feel the NT Mini's greatest value is when it's jailbroken so that legitimate vintage carts can be copied to the device, so maybe just do that.
  2. This sounds like perhaps the ROM chip itself was failing for an unknown reason (consider the age of NES and SNES era hardware, and this isn't an unreasonable assumption), one of the redditors mentioned they had blue PCB capcom carts fail as well, and another mentioned the CIC lockout being disabled could cause it. It would be kind of amusing if capcom had invented a kill switch to kill the cart if played without a CIC with the assumption that it's an illegal copy, but that seems like too much effort. Capcom did make suicide tricks exist on their arcade boards, but for a cart? Unheard of. In my opinion it's most likely the EPROM itself and maybe Capcom used non-5V EPROM's or had a mixture of them. Without someone going out there and finding a few more of those exact carts and trying them on the same Super NT, we probably will never know. I'd assume that the firmware voltage switch could prevent any further overheating.
  3. Not the overscan, that I expected. Like the moving platforms in 1-4. I'll have to check some other stuff, but it just kinda jumped out at me that I don't remember it that way.
  4. I'd rather not. Even if seeing square, clean, pixels was not the developers intended way of seeing the game, adding scanlines, blurring or barrel warping is not what everyone remembers. Heck I played the NT Mini Noir today (88$ import fees, good god) with my vintage SMB3 cart, and two things didn't seem right: 1. The audio sounds... worse? I have the vintage NES here as well, but maybe I'm just remembering the audio having more bass because when I played it in 1991 or so it was always sent through the stereo with the 12" woofers. 2. There's a lot of visual artifacting that I don't recall being a thing on the vintage console either, let alone 1991. I don't have that many real carts laying around to test at the moment.
  5. I just got my shipping notification, almost exactly 10 months after it had been paid for. I think Analogue should probably move closer to the original Zimba 3000 proposed earlier on in the thread, a console that can do everything with just a cartridge adapter, any controller can be used with a controller adapter/usb. I've bought three consoles so far, and my only nitpick in the build quality so far has been that the Super NT's loses gripping power to the cartridge, which I kinda solved by sticking cardboard down the side of it to firm it up. Clearly the cartridge slot is a weak point, and having something that can be replaced would extend the life of the console as well. I kinda hope they just build the Z3K with the pocket as the base point. "Here's a SoC with a FPGA that you can program", like after the recent Nintendo crackdown on emulation with the Super Smash Bros tournament, it seems like the path forward here is figuring out how to make two to four consoles in two to four different locations act like one console. I see two ways for that to happen without resorting to software emulation (and subsequent crackdown on the inevitable pirated games) : a) FPGA consoles that the controller I/O can be redirected over the internet, and the video stream can be compressed by a hardware ASIC and sent. b) a dedicated hardware device that you can plug your controllers into on one end (leave streaming to twitch/youtube,) and on the other end into the remote console. Think of this as "friends codes" without a server. Plug in the ethernet and the second,third,fourth player cables to the console, and on the player side plug your controller into it. Configure it with the built in web-server. The interesting thing, is that since the SD resolution consoles (everything from the NES to the N64/PS2) is rather small, a lower-complexity codec with lower latency could be used, or the user could switch it to a lossless codec (eg ZMBV) and record it to the SD card, or stream it remotely. At anyrate, it's one of those questions that will remain unanswered without a change in copyright law, is how to get people in remote locations to be able to play the same copy of the game for multiplayer reasons. The actual laws of physics prevents being able do this in any practical form.
  6. That's not how it works. When eBay ends a listing, the listing fees are credited back to the person who listed it. If the listing already ended due to a winning bidder/buyer than it's a dispute between the buyer and seller, and if paid by paypal, falls under Paypal's policy. https://www.paypal.com/us/brc/article/presale-policy-and-reserves If they paid the seller by some other means contact the payment provider to say no item was received. If the seller tries it a third time, they are banned from eBay for life. Do not bid on pre-orders, because you may just end up out the money, and do not list pre-orders because you're committing fraud.
  7. I never got an email about the pocket, which is disappointing. I'm still waiting for the NT Mini Noir to be shipped, so it's probably not the best idea to have two expensive projects in limbo.
  8. Clone has the same connotation as "counterfeit", so any kind of Cloned hardware will be seen as inferior, and thus that's why you wouldn't call it that, even if it's technically true, because it's also technically true of inferior counterfeits based on cheap mass-produced parts. For the FPGA consoles, the most logical and safest thing to call them is "Modern (8/16)-bit game console, plays all licensed games like you remember them" As far as amount of error precision one would accept, it's acceptable, in most cases, for minor audio or video glitches that are a direct result of a full digital signal. As a weird example, when I was setting up the HDMI capture (and the capture device wasn't picking up the 1080p signal) I noticed that there's some faint noise in the black area of about 1/3 from left side of the screen that was only picked up because I accidentally did "flood fill"in the paint program's still image of the dark area of it. It's completely unnoticeable otherwise, and if the FPGA is emulating it, it's reason enough to believe that there's a reason for it. The background was flood-filled with RGB:255,0,255 to make it readily apparent. I'm not sure if other games do it, but it was discovered entirely by accident. As far as scanlines go, I know some people want them, but I don't because to me, all the scanlines and scalers look worse than integer scale, and I'd rather have the image aspect-ratio correct on a high resolution screen, and have the result window-boxed than try to make it look like a CRT in any shape. In software emulators, typically the scaler/shader is applied after the game output, and the consequence is that it's trying to do a one-size-fits-all at the cost of 1-2 additional frames of latency on an already laggy render pipeline. You just won't get around that unless you can jettison the operating system, and short of something like retropie building it's own close-to-metal OS for it, a lot of these ARM SoC's are no better than what you can run on a desktop OS.
  9. So after some fiddling around, I found that if I stuck a piece of cardboard (eg cereal box cardboard) down the back of the cartridge slot, with the SD2SNES cart in it makes the sd2snes just snug enough to not wobble. This is a US model SD2SNES. The same issue was had with the Japanese Super Gameboy 2. However, I also found that there is a strange "alarm" sound from Bishojo Senshi Sailor Moon Another Story ( 5F776488E366B6D6B44F6E8A01D9953D ) that happens on the real cart, and the SD2SNES, during both "press start" screens while it plays Moonlight Densetsu. It should sound like below. But sounds like this: https://filebin.net/xssoeh4zoxegxnzr I'm playing through the rest of the game listening for anything else that sounds bad, but haven't encountered anything else yet.
  10. Not to start complaining but I have two issues come up with the SuperNT 1. Firmware 4.9, 1080p (50 and 60) no longer works through the Micomsoft SC-512N1 (FPGA HDMI PCIe 2.0 capture card), it used to. And the strange thing is, the capture card is showing a sync lock at 1080p. 2. Mechanically the cartridge slot is allowing the SD2SNES and Super Gameboy to wobble and it's making it fail. My guess the problem with 1 is that something has changed. Capturing from the card just shows "no signal", meanwhile the monitor attached still works. Unless this is a driver issue with the card, I'm not sure why 720p would work and not 1080p. The current driver is 1.1.0.188.5, 2020.02.03. The problem with 2, is something I mentioned way back at the beginning when I got a defective Super Gameboy, it doesn't fit in the cartridge slot quite right, allowing the edge connector to wobble, easily. This also happens with the SD2SNES, so even moving the SuperNT causes it to lockup.
  11. If anything, the FPGA would have to be the same generation, but that doesn't mean it has to be the same model. So it's probably still a Cyclone V 5CEBA2U15C8N. A2 and A4 have the same pins, and so does A5 with the U BGA. Vs the Super NT 5CEBA4F23C8N. So presumably if they were able to source the exact same FPGA they might have just used the same FPGA again, but if not, there would be some kind of minor redesign to fit a FPGA that is a slightly different model with the same number of pins. With that said, unless they needed to order a certain number of FPGA's to get a bulk discount, it would seem a bit weird to stick a model with more LE in it since the per unit cost still have to be more for a better chip.
  12. There are bugs that happen if the hooks are enabled on a SD2SNES in some games, which also happen on a real SNES, for example Seiken Densetsu 3 will flip out a lot. I haven't run into bugs in FF3 using the real cart, but I haven't sat there and played it through.
  13. The problem I see is that OSS falls into two very broad software camps: GNU-GPL-Stallman, where anything and everything not GPL is evil. This is what some of the MiSTer people sound like. BSD-PublicDomain-Preservation, where the core issue is making the software available perpetually without barriers(eg legacy hardware, operating systems, system exclusives, etc) The world benefits regardless of which camp produces software. The GPL crowd however actively villainize anyone who doesn’t use the GPLv3 license because their goal is to have the source code stay public, regardless of security, investment or other incentives. That tends to work out good for preservation of game engines and software tools, but is antithesis to creative content like game assets and scenario code/script since it means anyone can recompile it, making it impossible to retain control over compatibility and data mining. On the other side, where software and hardware are allowed to have a mix of open (BSD/MIT/PD) and closed code, means there is less effort to get a project to market. But at the same time there are bad actors out there that actively pillage the public domain to sell you things you can get for free if you simply knew how to compile it. Where MiSTer and Analogue come in is that the latter is building a platform of a known quality and lower barrier to entry than even the rubbish counterfeit game consoles and copier devices have ever been able to do. It may feel like sour grapes, but unless people want to pay around $1000 to get the experience of a 30 year old hardware it’s unlikely that FPGAs will come down in price short of Intel deciding to build its own FPGA retail platform.
  14. I'm Assuming that the second FPGA is basically "the DAC" if it were compared to the Super NT or Mega SG. So it might be invoked when plugged into the USB-c dock, but not in the same way, since I'd assume that a usb-c dock that Analogue makes would be an actual USB-C dock that happens to support game controllers and the "DAC" analog output. Typical USB-C docks used in the enterprise are either powered or unpowered, and have 2 Displayport (or HDMI/VGA) outputs, 5 USB 3.1 ports, Realtek Audio and Ethernet controllers, analog audio output/input and some have sd-card slots. Don't be fooled though, there are THREE kinds of docks. "Displayport" docks, "Displaylink (software GPU)", and Thunderbolt. Thunderbolt docks allow 2 monitors at 4K, standard USB-C DP docks only support two 1080p monitors even if they have 6 ports on them.
  15. Seems to me the logical thing would be to have the "HDMI" part of the DAC report itself to the SNT/MSG/Future-console-here via the EDID and have the FPGA devices output the native output straight to the DAC device. If connected to anything else, then just report the capability based on the NTSC/PAL switch. You shouldn't need an OSSC/XRGBMini with the DAC. But people do strange things in the name of wanting a consistent experience.
×
×
  • Create New...