do FPGAs count as emulation?
Thalidomide Vintage Ad Shirt $22.14 |
UFOs Are A Psyop Shirt $21.68 |
Thalidomide Vintage Ad Shirt $22.14 |
do FPGAs count as emulation?
Thalidomide Vintage Ad Shirt $22.14 |
UFOs Are A Psyop Shirt $21.68 |
Thalidomide Vintage Ad Shirt $22.14 |
They're the funniest scam in retro gaming right now. Even more than the mini consoles.
When used as a console itself I agree, but for use in flashcarts and odes, they’re pretty neat.
How are mini consoles a scam? They are literally marketed as souvenirs
that's not how they are perceived by the buyers.
zoomers know nothing about emulation so they think fpga and mini consoles is the same experience as the real deal.
downloading a pc emulator achieves better for free.
>no way bro why emulate when you could emulate the way I like to emulate?
i don't emulate, especially when nowadays one can buy a console for cheap and add an sd card/ hard drive and download all games ever released for it.
real hardware, 100% compatible.
Yes. Why the frick would I spend $300 on an emulator when I can just download a different emulator for free?
>open source hardware and software you can build yourself from commodity parts
>scam
>being sold the electronic equivalent of snake oil and believing your pile of commodity parts is magically somehow more accurate emulation than its software equivalents
>not a scam
lol imagine believing this
Exactly my point. Imagine believing your pile of parts is magic. But there are plenty of people who do.
>magic
It easily connects to a CRT, doesn't have any additional input delay over original hardware and is easy to use. I'm not sure 'magic' is but it's been pretty great so far, don't really understand what issue you have with it
>don't really understand what issue you have with it
The issue I have with it mostly comes from literal children trying to tell me that it's magic. I thought I was pretty clear on this. A bunch of kids, bordering on cultlike behavior, are trying to convince me that a handful of MAME drivers rewritten in verilog are somehow superior to software emulation by orders of magnitude. A percentage of these kids will even go as far as trying to convince me that rewritten MAME drivers are somehow not emulation at all, because the driver is running on a magic box made out of magic parts.
Then they try to tell me that their magic box is "100% cycle accurate" and is "completely indistinguishable from a real [insert console here]." Ignoring the fact that they have no fricking idea what "cycle accurate" actually means, I'm left wondering which console it is they're recreating with 100% accuracy? Which hardware iteration of the original NES is the "correct" one? Which revision of the SNES is the "real" one? It wouldn't matter even if they could answer this question, as they'd still just be emulating whatever version they decided was the "right" one.
But no, magic box can transform into any retro console you want with 100% accuracy, no emulation required.
dude it's easy to connect to a CRT and feels console authentic. for someone who likes the feel of real hardware on a crt, getting a mister saves you an insane amount of money compared to buying the consoles. hell a neo geo and flashcart alone will run you more than a mister
I prefer it to software emulation, and I say that as a retroarchgay who knows how to setup auto frame delay and runahead. Who cares if it's still emulation?
>A percentage of these kids will even go as far as trying to convince me that rewritten MAME drivers are somehow not emulation at all
It isn't if its a hardware implementation though. Quality of the reproduction is irrelevant. There are plenty of official and unofficial system-on-a-chip consoles out there that are inaccurate. The Genesis and NES have dozens. But if its a custom ASIC and not a Pi or something, we don't call it an emulator. We call it a clone. FPGA is a hardware implementation and so its a clone, not an emulator. If we're going to call it all emulation then the word loses all meaning since then even the Genesis 3 or SNES Jr. would be emulation. You could technically get around this by specifying software emulator and hardware emulator but that's stupid and needlessly verbose. When people talk emulation they specifically mean a piece of software that emulates a piece of hardware. Once you start using FPGA you're building a hardware clone. The quality of the implementation in either case is irrelevant. There are some emulators that are more accurate than some clones.
>It isn't if its a hardware implementation though.
But it's not. It's MAME drivers rewritten in Verilog.
Verilog is a language to design circuits, not to code something that runs on a CPU.
You can technically do a 1:1 exact copy of the original hardware on a transistor level or you can do similar hardware that accomplishes the same thing studding emulators and the little documentation of the original chip.
Shit like "X FPGA is just X emulator written in Verilog" is the most moronic shit you can say.
>Shit like "X FPGA is just X emulator written in Verilog" is the most moronic shit you can say.
Except that it's completely accurate.
It's irrelevant what the source is once you're using HDL. Accuracy isn't what's important here. If I take ZSNES and convert it to verilog, test it on an FPGA, and then send my verilog to a factory to produce a thousand ASICs then I just built a very shitty SNES clone, not an SNES emulator.
>Accuracy isn't what's important here.
One of you finally admits it.
You're the one who doesn't understand what you're talking about. MiSTer isn't transistor accurate but that doesn't mean it's an emulator. It's a clone. Whether it's better or worse than software emulation is beside the point. At some level it's semantics but the semantics is the topic of conversation. FPGA is simply not emulation.
To clarify, there's a very good argument that MiSTer cores are overpromising and underdelivering. MiSTer skates by on the presumption of hyper accuracy while not actually being more accurate in practice than a lot of emulators. The MiSTer community tends to knowingly gloss over this or outright quash discussion of it. But that doesn't mean MiSTer is emulation. It's still a hardware clone.
finally after ages of wasted bandwidth /vr/ comes to a concise and correct answer. I'm going to quote this on every FPGA thread for the remainder of my days.
Yeah I feel like I've been oversold on the accuracy of the project based on what is possible rather than what is currently in the cores.
I guess you kinda have to hype people to get initial adoption to make development worthwhile.
Not that upset about it tbh, it's a pretty good little machine for playing on CRTs
A lot of it is because the FPGA option is so new the push right now is just to get shit running. I bet that once Saturn is out and all the biggest arcade architectures are implemented we'll start seeing a lot more focus on refining the cores to reach the accuracy that's been promised. Right now all the contributors are spread out just getting the games to work in the first place and people don't want to wait for perfection before they can actually use the thing.
As long as you're describing a circuit, using a hardware description language, which is what Verilog is, then you've implemented it in hardware. Emulation is colloquially understood to be simulating hardware with software. With FPGA you're simulating hardware with hardware. You're effectively laying silicon. You're replicating the process by which the original chips were made to begin with. It's "emulating" in the dictionary sense that it means "to imitate" but in this context calling FPGA emulation is just plain confusing since it obscures the fundamental difference in what it means to be a hardware implementation.
>You're effectively laying silicon.
Stopped reading right here. There is no amount of drivel that makes this not rewritten MAME drivers. Sorry.
>It's MAME drivers
>Source: My ass
Pic related, for example. Way to let the thread know that you're completely clueless about any of this.
>it uses some of the same ROMs as MAME, so it totally must be the same drivers "rewritten in Verilog"!
HAHAHA what a fricking moron. Do you even know what MAME ROMs actually are?
>i got blown the frick out and am now desperately trying to move goalposts
This is going to be hilarious.
So you don't know what a ROM or a driver is. Got it, moron.
That page doesn't even mention MAME drivers.
Not needlessly verbose. Accurate. Which is ultimately the focus of this discussion, to be accurate about the terminology and to dispel stupid marketing ploys. FPGA is hardware emulation. There's nothing verbose about using two accurate words instead of one single word that fits whatever narrative you're accompanying here. Let it go, man. FPGA is hardware emulation. it is STILL emulation.
The problem is that midwits conflate the two as equivalent, that FPGA is equivalent to running software.
no, the problem is people spent $500 on a dumb box because they don't know anything about technology and now try to justify it to everyone with their insane claims of it being completely accurate because muh fpga
What if I'm interested in FPGA and enjoy tinkering with it? We're not all bing bing wahoo consoomers.
>because they don't know anything about technology
The only reason why someone would be interested in FPGA is because they understand the value in the technology, so if you fail to see it then you're probably the one who doesn't know jack shit
You're pretty hung up on this mister thing, despite it only costing $250ish. But OP is talking about FPGA in general, for instance an FPGA SID replacement. I know there have been ARM on a package replacements for SID but FPGA is clearly better in this instance.
Exactly. We know that FPGA is emulating hardware on a dictionary term basis, but when you say FPGA is emulation, dumb people will just assume what you said. The term "emulation" carries a stigma.
You’re way too passionate about something you’re incorrect about, friendo. I think you should put this energy into a more active hobby, not a passive one like game consumption. you’d feel a lot better.
thanks,
Richard
I don't have MiSTer friend, but everything you're saying can be said about the "pile of parts" your software emulators run on too.
There's really no difference between an FPGA and having an ASIC manufactured to the same spec. That development is typically done on FPGA before being put in "real" silicon anyway.
It's more that FPGAs look kinda janky and they need a separate little daughterboard with voltage translators since they can't run off 5V. It works but would be more elegant and seamless if the original chip could be recreated exactly.
>FPGAs look kinda janky
The MiSTer certainly does, can't argue with you there.
>But if its a custom ASIC and not a Pi or something, we don't call it an emulator. We call it a clone.
This is probably the most succinctly I've seen that put. A $5 NOAC is still a hardware clone after all.
>snake oil
but it actually does what it says: it lets you play retro video games. yes there are other solutions, sometimes with devices you already own (like a pc), and yes there is some hyperbole surrounding it, but it does what it's supposed to and it does it pretty well. if you're enthusiastic about retro games, then any form of preservation should be exciting to you. and there are many examples in which an fpga is absolutely necessary (special chips in snes flash cart for example).
i can't imagine how shitty and boring your life is to be this much of an anti-mister autist.
>how to say you're some third worlder emulating without saying it
>do FPGAs count as emulation?
no
fpga has been around for decades, apeman. before that: asic and other technologies.
it's neither.
>>I love that people do this, but at the same time, that board is orders of magnitude more powerful computationally than the device it's fixing.
back in the 1980s we didn't have access to dirt cheap fpga so we can recreate custom chips using such powerful systems. frick going back to that dire period of time where everything was controlled by chip manufacturers and suppliers. now we can develop our own custom chips without needing millions of dollars.
>fpga has been around for decades
But it's never been commercialized like this, you dumb frick
>low-key hobby tinkerers
>commercialized
Interesting how gays are seething so hard about a community trying to keep old platforms alive.
The core project is alright. But it's a big stretch to call the people around this "low-key" anymore. Everyone knows about mister now, it's shown up on big søytubers and all.
And honestly, it's not very different to software emulation. I can happily use stuff like switchres. Latency-wise, well many WRs are done on those still, I can't imagine it being much of a real problem.
The biggest turnoff for me is how the most powerful system supported by FPGA emulation is PS1. I want to see GameCube, that would be far more interesting, mister or not.
>I want to see GameCube, that would be far more interesting, mister or not.
Gen 6+ consoles won't benefit from FPGA. In fact, a lot of those consoles are best played with software emulation where models can be rendered in 4K with anti-aliasing, lighting patches, etc.
PS2 definitely would benefit, you're not fixing PCSX2's fricked up float handling without impractical FPU LLE.
What consoles do you say would benefit from an FPGA? 4th gen consoles which already have highly accurate emulators?
>The biggest turnoff for me is how the most powerful system supported by FPGA emulation is PS1
Current FPGAs support about 300,000 transistors which is just enough to replicate the PS1's chipset.
Current FPGAs support tens of billions of transistors champ. But they cost more than your daddy makes in a year.
Friend you are wrong.
Dolphin has less lag than a real GameCube. From 6th gen and on you really don't benefit that much from FPGA.
>Dolphin has less lag than a real GameCube
How?
Through the sheer force of cope and mental gymnastics
https://mobile.twitter.com/kadano/status/780762623990267904?lang=en
morons.
>60ms lag on official console on a CRT
calling bullshit on that one
That's a known fact with multiple tests out there, these games do not run close to the metal at all. So you provide evidence to the contrary or shut the frick up.
sure, got any tests that aren't melee?
no.
Then how do you know it’s not just specific to the one game?
Melee is the fastest responding GC game, most GC games don't run at 60 fps
Yah I’m definitely calling bullshit on that one
Alright, have a nice day
>Dolphin has less lag than a real GameCube
not how it works, compulsive lying idiot.
It’s literally emulation homosexual. Cope harder
>It’s literally emulation homosexual
false. it's not emulating anything. it's a clone of hardware. it's not emulating anything, it's as close to the real chips you can get.
>It's a clone of the hardware.
That's the textbook definition of emulation lmao.
>Chip that runs code to *EMULATE* the functionality of hardware is not emulator
Get outta here anon.
An FPGA is not a "chip that runs code."
It's an IC you literally program to act on certain way. How's that not EMULATION if you use it to EMULATE a video game system?
The FPGA is not running a software program though, it is loading a configuration which describes a layout of circuits and logic which remains static until the state is cleared.
>chinks literally steal the design docs/specs of chips
>chip is literally made with the same process as the original
>still never is 1:1 identical with the original
>one moron can make a fpga and it is magically perfect because "FPGAS ARENT EMULATION"
ok moron
to be fair Commodore couldn't even make SIDs sound consistent from batch to batch
That's not unusual either, Megadrive sound was different across revs too. For what it's worth as well it wasn't unusual for early systems to have general purpose logic devices driving their integration, effectively precursors to FPGA. The c64 uses a PLA to control its bank switching and the ZX Spectrum uses a ULA to implement display and system glue. The main difference is these are mask "programmed".
In case of Mega Drive that was different board revisions due to changes in the audio mixing circuit. The SID however had filters that came out slightly different in different batches of chips (this ceased to be the case on 8580 chips which all sound identical)
8580 also fixed the volume bug which allowed samples to be played on the older SID. Regardless there have been multiple revisions of chips in consoles, they are cost-reduced and combined, notably the SNES was single-chipped. As this didn't happen in the 70s like the 6502, they weren't hand taped from logic designs they would have been described at a high level with logic to be implemented in VLSI, highly likely that a tool like Verilog was used in this process. FPGA is just an uncommitted version of this with onboard prebaked primitives like RAM.
>8580 also fixed the volume bug which allowed samples to be played on the older SID
So much for Ghostbusters, alas.
There was a simple mod which would could fix that, just bridge two pins with a resistor.
Reportedly the 8580 was closer to the designers' original intentions but many people prefer the 6581. it's like the Star Wars DVDs where the creator's intentions aren't necessarily what the fans want.
Not sure what that has to do with FPGA configuration, but it's clear you have a marginal comprehension of the situation so it's good that you have decided to not be part of the FPGA community because you are a lamer.
>The FPGA is not running a software program though, it is loading a configuration which describes a layout of circuits and logic which remains static until the state is cleared.
Doesn't matter, since it's still EMULATING a video game system's functionality. Doesn't matter if the emulation happens on FPGA or running on your computer through software means, it's still EMULATION.
No it's not emulating. If I have the literal schematics of a logical device and implement that exactly in logic then it is effectively identical. Now some FPGA core implementations may be black box, but some like the NEOGEO graphics are literal transcriptions of the silicon. You could take that Verilog or VHDL and create an ASIC which is a literal exact copy of the cloned hardware. You could drop in an FPGA on a package which is also a literal exact copy.
Face it, all things being equal FPGA is better than software emulation and is just one step away from an ASIC.
>all things being equal FPGA is better than software emulation and is just one step away from an ASIC.
Sure, there are inherent good points in FPGA EMULATION (and it's honestly speaking great to see those implementations). But, since it's still basically an IC (no matter if it's reprogrammable or not) acting as the system it's EMULATING, I'd say it is still EMULATION. The term EMULATOR conveys both hardware and software level implementations.
Cope.
>EMULATION
What's with all the caps? You sound like an Englishman on holiday asking for directions.
Very bizarre cope. How much do you have invested into fpga emulation machines?
A fraction of what I make every month. I wish it was more expensive so morons like you would have no chance in hell of buying one.
Sauce?
https://en.m.wikipedia.org/wiki/Field-programmable_gate_array
>my source is other moronic zoomers
>here's a link to a wikipedia page i didn't even read and couldn't understand if i did
lol
>a wikipedia page i didn't even read and couldn't understand if i did
Talking about yourself there?
>n-n-no u
I accept your concession
>midwit chimes in
theyre good for stuff like the gaemboy afaik theres no software emulation device for those that can do link cable with the original consoles, i dont get why chinks haven't done it though
Yes, it's hardware emulation.
It's software emulation.
Fundamentally false, anon. There are arguments to be made, but that isn't one of them.
HDLs are still a type of computer language.
That's like saying schematics are software
This is a good analogy. Yes, you can use a computer to draw up schematics but those schematics are blueprints for a physical object. People tend to think Verilog is just another programming language like C++ or Python when it's actually more like CAD.
No. It's exactly like saying one type of computer language is a type of computer language. Unless you're designing the masks for your part by hand you're using a computer language. If your part needs to load a program (ie an FPGA bitstream) in order to do what you need then you're using a computer language.
Such as?
Who fabricated your ASICs, what were you quoted?
Oh. Obviously not going to give details of suppliers and financials.
>trust me dude i made tens ok Ks in fpgas but its top secret tech so if i told u i would have to kill u 😉
>If your part needs to load a program (ie an FPGA bitstream) in order to do what you need then you're using a computer language.
That's irrelevant because synthesizing and FPGA and manufacturing an ASIC uses the exact same HDL. The FPGA has a software layer to actually perform the synthesis but the FPGA isn't "running" the code like an OS would run a program. You use HDL to design a chip, test it on an FPGA because you want to see it perform, and then when you're satisfied you send the HDL to manufacture and lay silicon. Calling it a "computer language" is like calling CAD software a "computer language." Yes, you're using a computer but you're not writing software. You're writing a schematic.
>do FPGAs count as emulation?
Yeah, but it's possible for them to be un-detectable or frame-perfect emulation. They don't have to be, and some of them really aren't frame-perfect/undetectable/whatever, but they're almost all better than software only emulation.
Now to look up this fpgasid you posted...
From the site on it:
Introduction
The chip MOS6581 also known as the Sound Interface Device (SID) of the Commodore 64 computer is known for it's characteristic sound. Unfortunately, many of these devices produced between 1982 and 1993 reach the end of their life span these days. It is getting harder to get a replacement part. Thus, we need a proper substitute.
And this is what the FPGASID project is all about!
It has been started to overcome with this situation and reaches even higher - it aims for a better SID than any SID has ever been before...
The Mission
Create a pin compatible drop-in replacement for the MOS6581 device as well as for its successor, the MOS8580.
Hu. I guess if you need such a device, I don't know if it really would be emulation. Emulating a single chip, in a pin-compatible-manner, and nothing else. Interesting.
I love that people do this, but at the same time, that board is orders of magnitude more powerful computationally than the device it's fixing.
>Unfortunately, many of these devices produced between 1982 and 1993 reach the end of their life span these days
Which is an asspull. An IC doesn't have a set "lifespan" but it can be damaged from a variety of factors mostly related to poor storage or user stupidity. This is certainly a problem with SIDs as they've often been killed from ESD. While replacements are certainly a good idea, mere use doesn't wear a chip out.
>mere use doesn't wear a chip out
Assuming there aren't cooling or power issues with the chip, true.
Fun fact: the GPUs used in high-end clusters do have a limited lifespan, because of the high temps they run at. They are expected to die as they slowly cook themselves to death over the clusters useful lifetime. 100% wasn't the case for the chips in a C64 though.
Commodore's chips definitely got hot though and heat sinking would be beneficial to them. The VIC-II was known to be quite an oven. Now there was a guy who was working on an FPGA VIC-II but it's only a prototype atm but SIDs are more in need of a modern replacement solution because the filters get damaged from ESD very easily.
ICd do degrade over time, do a google search before saying such things.
It's not just thermal expansion, actual usage can wear gates down.
Plus ICs *DO* have a set lifespan, on these miniscule levels, over time they degrade and can short out or gates just stop working.
Claiming otherwise is bullshit. But we're talking about 100+ years here.
>Collectors on suicide watch.
100 years means the chip has to have 876,000 hours on it before it wears out. I don't think you'll be alive for or ever use it enough to worry about that.
I'm not talking about ON, just standing in storage. Degradation to the materials happens even when you're not using it. Plus other things like solder whiskers etc.
even if that was true there is no IC not even the oldest ones from the 70s that are anywhere close to 100 years old. i mean 100 years ago was 1922. i doubt you'll have to worry about that in your lifetime. more like your grandson's problem.
I'm not gonna be alive in 2098 to see if any 100 year old N64s are still working, so...meh.
>mere use doesn't wear a chip out.
Thermal expansion can wear out the solder bonds inside the chip.
>Thermal expansion can wear out the solder bonds inside the chip
Chips that run hot like many of Commodore's ICs can definitely have this issue which is why heat sinking is recommended. If its peak operating temperature is more than 20 degrees above room temperature it's too hot and needs a sink.
The C64's MOS chips are notoriously crappy and fail all the time.
It's partly due to poor board design and complete lack of thermal design by Commodore. the chips actually NEED a heatsink, but they were sold without one.
It's planned obsolescence at it's finest.
Commodore were really behind the times technologically; the MOS fab in New Hope was never once updated and right until its closure in 1992 was still using the same archaic equipment that had been in there since the Ford Administration. All of the 6xxx MOS chips used an ancient NMOS process that throws off a lot of heat. Intel and Motorola had moved past that and adopted the more modern HMOS process (basically improved NMOS using a smaller chip die that doesn't guzzle as much current and has better thermal characteristics) by 1978 but Jack Tramiel was insistent on the idea of "If it's not broken, don't fix it."
>Motorola had moved past that and adopted the more modern HMOS process
>Motorola more modern
Oh god. When Moto was more modern than you, you knew you were in trouble. That place was piled kindling waiting for a spark to burn it down for decades before it really died. I temped in their factories over the summer when I was young, and knew a bunch of their engineers. They beat the hope out of every one of them who had good ideas. Awful place from top to bottom.
they hit gold once with the 68000 and that was apparently a fluke they were never able to replicate
And that was Motorola's downfall. They got complacent off fat 68000 profits and let Intel overtake them.
companies that strike gold once and get too lazy for it tend to pay the price. Ford almost died that way when Henry didn't want to let the antiquated Model T go. also Commodore basically succumbed to the C64/Amiga 500 cash cow.
They introduced HMOS in 1982 and took three years and scads of dead PLA and TED chips before they figured it out.
that's pretty fricking bad
CMOS was available since 1974 but not used much in that time due to the higher cost. the RCA 1802 was the first and only CMOS microprocessor for some time and it did have quite a bit better thermal characteristics and less power usage than its contemporaries.
They got better as time went along. Later C64 board revisions were more robust and the C128 was surprisingly bulletproof considering what a tangled maze the PCB is.
I'm actually more interested in it for the sound chip angle myself
would be really cool to have a makeshift SIDStation
Its odd how most FPGA owners don't use CRTs. Like why go the extra mile for authenticity and accuracy if you're playing on a shitty monitor?
I know emugays tend to not use CRTs either but they usually don't put on a whole show about being faithful
>Its odd how most FPGA owners don't use CRTs
???
The IO Board to use analog out is the most popular thing people put on a misterfpga, after the ram board. And people only put it on to get the analog video.
Funny thing is that the analog video is actually sharper with greater color depth when converted from the HDMI's Direct Video Mode compared to just using the direct VGA output from the optional Analog IO Board. Either way, the MiSTer's analog video and audio are sharper than all original hardware, even with RGB and audio bypass mods.
Modern TVs have reached parity in input lag with CRTs, while gaming monitors even surpass them. So it depends on your priorities.
>Modern TVs have reached parity in input lag with CRTs, while gaming monitors even surpass them. So it depends on your priorities.
Delusional
It’s basically true and there’s next to no advantage to a CRT except for nostalgia.
The CRT worship on this board is sometimes creepy.
That's not fully true though, CRTs are still the best way to display retro games. They look beautiful in them, and they're cheap, while a low lag TV can be expensive.
You're basically moronic when you have no idea how CRTs and fixed pixel displays work. There are a plethora of advantages with CRTs that include outputting different resolutions natively, the ability to output resolutions that other displays cannot, unmatched motion clarity, zero display latency, excellent color reproduction, contrast and black levels, etc.
This is correct. There's plus and cons to both CRTs and modern LCD tvs. At least input lag isn't aconcern anymore. So like I said earlier it's a matter of priorities.
As an example, I like big screens, don't have the space for a big CRT and low input lag is the most important factor for me when playing games. So a good OLED makes sense. If I had the space I'd have a frickhuge CRT for sure for all the other benefits.
fricking sucks how you have to mess with all kinds of menus and hidden settings to calibrate HDTVs. Out of the box they have overblown contrast ratios and motion blur shit for normies, and you have to turn game mode on. It gets me paranoid and I check if my settings are right everytime I boot a game. At least CRTs give peace of mind that you're getting optimal results just by turning it on.
The amount of coping CRT gays have on this board to justify their CRT ownership is unbelievable. I can't think of anything on any other board, even Ganker, that compares
>while gaming monitors even surpass them
no digital display will ever be faster than the direct electrical connection from video chip to electron gun, it's physically impossible. all digital displays have to buffer.
They already do though.
they literally don't though
They do. CRTs take 16ms to draw a full image. It's 8ms at the middle (where gameplay happens) and 16ms at the end. So CRTs have 8ms of input lag, but these monitors literally take 1ms to draw the entire image.
It looks like you're assuming that light travels instantly to the screen, but it's not how it works. There's a circuit inside the screen that is responsible for getting the color information from an electric signal and drawing it line by line. The Sync signal is responsible for telling the screen where the colors go and where a line begins and ends. You can see this happening with a high FPS camera.
uhh but what about the very first scanline? checkmate.
>It's 8ms at the middle (where gameplay happens) and 16ms at the end. So CRTs have 8ms of input lag,
that's a moronic ass statement buddy
>these monitors literally take 1ms to draw the entire image
no they don't, LCD's update from top to bottom too, and you get the added fun of digital buffering and pixel response times
>It's 8ms at the middle (where gameplay happens) and 16ms at the end.
>Comparing 360hz monitors to 60hz crts.
Anon unless you're actually pumping the full refresh rate these are gonna draw/input at the same rate as the CRT + 1-2ms processing (imperceptible)
For Input lag they're functionally the same for /vr/ stuff. But it comes down to how the image is just drawn/displayed differently. Until there are high refresh rate displays have some sort of black frame insertion these games will just feel different on a crt, not to mention things like color reproduction, ghosting, black levels, etc.
Tho I'm a CRTgay so take my word with a grain of salt
Yes
Yeah, its just better because gate arrays are parallel but they have the worst downsides like the illusion of being opensource
It depends. They can be accurate if you recreate a chip at the circuit level. For example Curt Vendel obtained the original schematic diagrams for the Atari POKEY/ANTIC however it should still be stressed that those aren't definitive either because a chip may have modifications made to it when production begins, or changes along the way that the diagrams don't reflect. In that case reverse engineering is more accurate.
that's nice however 80% of the retro ICs people are likely to care about have their schematics in Japan somewhere and are consequently not accessible to baka gaijin (as opposed to Atari docs which were in California). even if you had them you couldn't read them.
Over extended periods, everything can eventually be reverse engineered. Technology is advancing fast and newer methods are more and more accessible for cheaper to the everyman for software and hardware reverse engineering.
Why do you think most advances have been made in the last few years even though this has been a thing for decades?
http://sam.zeloof.xyz/category/semiconductor/
What if we'll eventually be able to make new SIDs in our basement that will work like the original not an FPGA?
Probably.
We already have tiny DIP packages FPGA/CPLD based replacement chips that you only need to program once.
As anon said there's really no difference on a digital side, but for analog like sound, the process itself can have a impact on the sound, SID partially owns it's sound to also it's NMOS process.
Indeed anon, our ears being analog complicates things too.
or well the 6581 does because the filters varied between batches of chips. the 8580 sounds more like the designers' original intention but it's more like the Star Wars DVDs where the creator's original intention is not necessarily what the fans preferred.
>We already have tiny DIP packages FPGA/CPLD based replacement chips that you only need to program once.
But they're still 2.5V and need a board with level translators. There used to be 5V FPGAs but those disappeared years ago.
True, but honestly drop-in FPGA solutions for individual ICs is actually kind of silly. There aren't too many situations where a system's failure rate is going to be meaningfully attributed to a specific chip. And if the goal is to keep the hardware as original as possible you run into a Ship of Theseus problem pretty quickly.
>And if the goal is to keep the hardware as original as possible you run into a Ship of Theseus problem pretty quickly
If you repaint a vintage car and put new tires and gasoline in it does it count as "original"?
>Ship of Theseus problem pretty quickly.
I like how it was proposed in John Dies At The End:
>There aren't too many situations where a system's failure rate is going to be meaningfully attributed to a specific chip
in case of Commodore hardware it usually is
>you must quest to the mystical land of nippon to obtain the sacred chip scrolls written in the ancient language no foreigner could ever hope to comprehend
Calm down a little, weeb. It's not that serious.
I doubt they still exist anyway. Although (to use one example) Ricoh might still have the SNES chipset schematics in a file cabinet somewhere they probably don't and Japanese companies are notorious for not holding onto anything once it's not needed anymore.
Of all companies, Nintendo is actually the most likely to still have that stuff around. Meanwhile Sega of Japan probably burned the Genesis ones to spite the baka gaijin.
It's more likely they burned the Saturn schematics as I'm sure they consider the Saturn damnatio memoriae.
Saturn was successful in japan. Go ask SoA about the AMAZING Saturn launch and excellent marketing decisions. Those are damnatio memoriae.
saturn was not successful in japan
Nintendo has stuff laying around because they're a super early adopter of what's now known as cloud storage but all the way back in the late 90's. Exactly how organized and well inventoried all that data is well that's another matter entirely.
Meanwhile companies like Square Soft kept everything on floppies and workstation hard discs which is why they lost so much shit over the years. There's frick all remaining from the pre-merger times for them.
My guess is that Nintendo having been a toy company for a century means it was used to physical storage compared to most other game publishers which were typically founded as software companies.
>projecting weeblet can't into moon or electronics
Many such cases
Nothing ironic about it. Pieces of paper don't make someone skilled and knowledgeable. Anyone who's ever interviewed a candidate with multiple useless degrees can tell you that. What is ironic is that the TIA was perfectly cloned 40 years ago.
Fair, was more a figure of speech than a literal statement, but you are correct. And I too found that odd considering the 1000s of South American and Taiwanese 2600 clones floating around in the 80s, and those were single chip designs too. You can still buy deadstock on eBay, but I don't feel like spending $150 on one.
The old SoCs were great. You can still find them cheap. Even on ebay, for a fraction of that.
Yeah it certainly seems so, I've made a few saved searches for them. Though the RAMBO units are all like $150 for some reason.
thanks anon, agreed entirely
Ironically, when Curt had to design the 2600-on-a-chip ASIC for the Flashback 2, his ended up with major incompatibilities despite having Atari's schematics to work from. Hard to say whether it was incompetence or something changing over time like you said. But, the cart slot hack was a bonus feature and it is (obviously) compatible with all the games they built into it. Video quality is also dramatically better than on a real composite modded 2600, which is the main reason I rather like mine.
The TIA had no less than ten different revisions and the 2600 Jr had a modernized version made with a newer process that broke compatibility with a few games most notably Kool Aid Man.
The Jr. has a ton of revisions too, including the mythical "one-chip" 2600. Don't get me wrong, I recognize the difficulty of the task, Curt was just a bit of a nut-case on top of it all.
No? They're equivalent to emulation in terms of reproduction quality, but that's orthogonal to the type of technology being used.
Yes, you're not even actually playing the game unless you're using original hardware, with RF on a 20''< CRT, and your parents are fighting in the background.
>digital FPGA
>for an analog synth
lol
There are programmable analog ones but they're somewhat uncommon.
Sure but that's not one
The analog ones are called FPAAs btw
/thread
for me it's 3
>do FPGAs count as emulation?
When they're being used for emulation, yes. What else would you call it? If you put one in your video scaler for purely signal processing purposes then that's not emulation. Stupid question tbqhwyf
Of course. No amount of seething or coping can ever change that fact.
Technically yes, but if you're doing 1:1 hardware emulation on a transistor level, there's really no difference, you're just using a more complex chip instead of fabricating a chip yourself.
EITLIM5 ?
It's hipster emulation for pretentious onions consumers.
depends
If it's just doing a chip then no, however if it's doing multiple functions then yes
https://atariage.com/forums/topic/153133-how-long-do-you-think-a-machine-will-last/#comments
This also quotes the 100 year figure but again it's hard to say because no IC is 100 years old or even close to it.
here's a record from 1913 being played. also VWestLife is a good guy, he's involved in the retro computer community and posts on VCFED.
as well as a Victrola. i dunno if your DMG Gameboy will still be working when it's 119 years old but that's for your grandkids to find out.
You know those single chip plug-n-play Genesis consoles with like three games on them? An FPGA console is just a variant of those except an FPGA can be updated.
And the reason they need to be updated is that they're far far from perfect. Once a perfect FPGA core for a specific console exists it no longer makes sense to keep it on an FPGA since a single chip clone would be so so much cheaper to manufacture in volume.
>Once a perfect FPGA core for a specific console exists it no longer makes sense to keep it on an FPGA since a single chip clone would be so so much cheaper to manufacture in volume.
You need production runs of tens of thousands before that makes sense.
It does make sense for the Famicom, Genesis, Game Boy Color, and SNES apparently, as we have decent mass-produced SOACs of all of them.
Well, clearly famiclones and megadrive clones were manufactured in large volumes. Either way, it remains a barrier to any community effort like MiSTer. A dozen or so developers writing cores for a few thousand users is viable with FPGAs, you need a lot more than that to make the jump to ASICs.
Oh I don't disagree, certainly.
>You need production runs of tens of thousands before that makes sense.
Based on what factors? A reddit and a youtube?
I've done runs as small as 5K. To say it "made sense" would be a ridiculously comical understatement. It was less than 5% the cost of using programmable logic.
Details?
>I've done runs as small as 5K
There used to be specialty fabs who would do custom IC runs on demand if you needed only a small quantity but I'm pretty sure they disappeared once cheap FPGAs became commonplace. Of course even if they still existed they couldn't recreate an SID in its original form as they didn't use MOS's antediluvian 1970s fabricating tech.
you could make a new SID, but it would have to be made with modern low power CMOS tech in which case there's no real difference between that and SwinSID. well maybe you'd have an easier time getting analog filters in there but it would still not behave exactly like an original 6581 chip (even the 8580 was somewhat different owing to a newer process).
It's even easier to do much smaller runs today. It's mainly a question of cost. But none of this will help you make a SID. Commodore couldn't even make them consistent across batches of the same version, let alone versions. Even if you could make a SID to spec, it almost certainly wouldn't sound like the SID in your C64.
This is why your employer doesn't even trust you with the secret mac sauce recipe
So you don't have a clue how these things work, on any level. Got it.
>Even if you could make a SID to spec, it almost certainly wouldn't sound like the SID in your C64.
of course you couldn't. even the 8580 SID was rather different due to the more modern process used.
And this isn't even just a SID thing. That's just what gets all the attention. Many old sound chips have noticeable differences.
Mister seems cool, but for the cost of one it would be more prudent to buy a Ryzen mini PC and turn that into an emulation machine and hook it to your TV like a Mister. You can run Retroarch on it and get the input lag low with certain settings and get essentially a zero latency set up with the right configs and hardware.
A Ryzen emulation machine can be repurposed for something else if you ever need to. A Mister won't be good for much else except circuits you design for it. If it becomes unsupported you're on your own or you have a useless electronic.
I have a dedicated emulation machine, though it's not a new mini PC. It's a 10 year old desktop that I have outputting RGB to my CRT. If it ever dies, I'll get some sort of mini PC to replace it.
>Mister seems cool, but for the cost of one it would be more prudent to buy a Ryzen mini PC
If you're in the US, and know where to get the DE-10 nano, the cheapest Ryzen mini PC is still more expensive. Don't buy from scalpers senpai. Ever.
I think Mister's greatest feature is not relying on an OS and a bloated frontend like Retroarch that crashes all the time, but that doesn't push me over the edge to buy one at current prices.
It'd be interesting to know whether conventional PCs could run some sort of stripped down RTOS to minimize latency on inputs, video and audio for emulation.
Why not? And this already exists, there's several OSs that do this including Lakka.
>Retroarch
absolute cancer
i guess you could call it an analogue emulator. As I understand, it's not replicating the code like an emulator does but rather replicates the signal
It's replicating the digital circuits.
idk what that means
It can be programmed to have the same circuit configuration as the original chip.
I don't understand, Why would they not be emulation ??!?
For me, it's just a very convenient emulation box that does a lot of consoles very very well, has an active community of devs, nice usability and QoL features, connects easily to a CRT and is much smoother to use on a modern display compared to an rasp pi. Shame about the price though! I love mine, truly don't care about all the FPGA aspects, it's just the perfect living room device for retro games.
Do they have customizable chips/boards that can change themselves to emulate the hardware and wave generation of real sound chips inside old consoles yet?
>That's because you and that mouse in your pocket were in your dads ballsack.
I'm willing to bet it was an optical mouse.
Simulation, as FPGAs attempt to model the internal structure of a chip.
it depends
>the nervous samegayging begins
keks are incoming
Yes, it reproduces the function of a different computer
Check out my Super Nintendo emulator, guys.
ah, yes the SNES jr. reprogrammable FPGA I love playing C64 games on it
>posts stupid as frick misunderstanding multiple times
>gets called out
>desperate coping and projection begins
>still can't prove he has even a basic understanding of what he's screeching about
>here comes the paranoia, right on schedule
Dude, just take the L and stop embarrassing yourself.
>still trying to run damage control because he doesn't have a clue how his snake oil box actually works
keks were had
ITT: the assembly language LARPer schizo strikes again
>t. the assembly language LARPer schizo
Yes. That being said, who cares, just play some damn video games.
FPGAs are theoretically nice for a modern replacement for custom chips in retro gear, but not perfect since they're made with modern manufacturing methods not 1985 manufacturing methods so results may vary.
If SIDs had continued to be made to the present, they would still end up that way. I mean you can get new Z80s and 6502s, they still make those for embedded designs but using modern low power processes.
That is true and those chips have differences; notably the absence of undocumented 6502 opcodes.
I believe they were present in all NMOS core 6502s but not CMOS ones.
Yeah, it's a side-effect of the manufacturing process as
implied.
Would there be a way to get the same effect of the old chips out of an FPGA implementation by changing the circuit? If it's possible I could imagine a scenario where you get to pick between accurate transistor logic with less accurate results or inaccurate transistor logic but accurate results.
If it's something like undocumented opcodes, sure. You could just incorporate that into your new design.
the FPGAs you get now are of course modern 2.5V CMOS chips so while they can replicate the basic functions of, say, an SID, the chip will still behave differently and not have the same glitches/exploits. however if SID had continued to be produced it would eventually end up being made with modern chip processes just as those new Z80s you can get for embedded systems are.
whos that under 4chin
the larpy kid in the mario allstars thread who was autistically arguing that no one realized the bricks were weird until some romhack came out a few years ago
Ganker
This thread is as good as any for my blog post
I never had strong opinions on hardware vs software emulation. I bought a Super NT, because I wanted to use original controllers and cartridges. That just makes playing more fun to me. Also not having the option to use save states, makes me engage with the games on a deeper level.
I then thought it would be fun to play some games on the go and at the same time peek a bit into the Genesis library.
After I read and watched many reviews I bought an Anbernic handheld. I was told that those devices emulate SNES with ease. Up to PS1 is well emulated, but Gamecube and Dreamcast are more miss than hit. Imagine my suprise when SNES games are not at all smooth sailing. I experienced many additional slowdowns and small sound glitches, that it is just not enjoyable to me.
So in the end I learned software emulation is fine only if you have a powerful PC that can do run ahead and all that jazz, otherwise it is a trash solution.
Thanks for reading
Support for the nt mini noir has pretty much dried up. Upgraded firmware to the latest introduced more bugs and the 6.6 jailbreak is based on the latest firmware and has those bugs. Analogue has become a disaster that can’t even do basic support for their own product with correctable firmware updates. It’s been a year since the last official update.
Damn that sucks. What kind of bugs are we talking?
Bouncing menus in various nes titles that show the health ammo counts on the bottom of a screen. Graphical issues galore in about 20-30 games some of which are high profile games. (Gi joe - rygar, crystalis)
The jailbreak 6.6 can’t dump dragon warrior 3/4 because copynes won’t work with their file size. So, the save dumping feature is kinda pointless if the 2 highest profile rpg can’t be backed up at all.
The GitHub for the jailbreak is just a ton of open cases with no results. Analogue Twitter gets you no answers on their own firmware issues. It’s real pathetic that the nt mini noir is their most expensive product with zero support now. They moved on to the analogue pocket and good luck if you bought that pos. The promised accuracy is NOT there.
ask matei
Mike would kick a casual in the nuts for suggesting that emulation is acceptable to play games.
Anything that isn't oem hardware is emulation.
the only correct answer.
I think when it's used as a single chip replacement it's not
>I think
yes, because all of these posts are nothing but opinions
>1+1=racist
checkmate wokeist
This. THIS.
THIS RIGHT FRICKING HERE YOU GAY homosexual FRICKS!!!!!!!!!!!!
HE'S RIGHT!!! SO FRICKING RIGHT!!!
OMG I'M CRYING!!!
THANK YOU SO MUCH GODDDD!!!!
HAHAHAHAHAHAHAHAHA!!!!!!!
YOU STUPID ATHEIST FPGA NASA WORSHIPING homosexualS!!!!
I HATE YOUUUUUU ALLLLLLLL!!!!!!!
DIE DIE DIE DIEEEEEEE!!!!!!!!
HAHAHAHAHAHAHAHA!!!!!!!
How about a 1 chip snes? or PS1 ASIC cpu on PS2 slims? Those are OEM
Are those modernized 6502s WDC sells for embedded use "emulation" because they don't behave exactly like the 6502 in a VIC-20?
Emulationgays in full cope mode. Face it FPGA is better than software running on a PC, and a RPI is not even in the running it's a joke.
arrest yourself shillgay and go to bed
Yes but the only difference is they cost money. They're an absolute scam.
The software is free you idiot