Jump to content
IGNORED

FPGA Based Videogame System


kevtris

Interest in an FPGA Videogame System  

682 members have voted

  1. 1. I would pay....

  2. 2. I Would Like Support for...

  3. 3. Games Should Run From...

    • SD Card / USB Memory Sticks
    • Original Cartridges
    • Hopes and Dreams
  4. 4. The Video Inteface Should be...


  • Please sign in to vote in this poll.

Recommended Posts

People claim an Everdrive has never fried a system, but that is untrue because one fried the PPU on my NES. I bought one for a friend and was setting it up and it destroyed it immediately. I won't put another one of those things (at least the NES one) in my systems.

 

QFT.

 

I know of 2 dead consoles caused by mismatched voltages. When using old electronics, it's best to match everything and keep component stresses low.

Link to comment
Share on other sites

Yes, it was legit, I purchased it from SAG. They don't know what caused it and paid for a replacement as well as the fee to have my NES shipped out to my modder to fix, so at least they took care of me. That said, I don't trust them. I ended up setting it up for my buddy using my Mini, but that'll be the last time I touch one.

 

Edit: for what it's worth, I was just looking back at my emails with them and whatever happened fried the Everdrive as well. So I guess I got a bad unit, but frying my childhood top loader that I just had RGB modded was too traumatic to bother trying again :woozy:

Good customer service for SAG.

 

I'm curious, who modded your childhood top loader for RGB?

Link to comment
Share on other sites

I'll start this post with what to me is the big ellefant in the room. Why does analog advertise these 8bitdo pads as lag free when they add about a frame of lag?

 

Tl;Dr: 8bit do controllers are NOT perfect replicas of the originals. Its easier to get diagonals on them, wanted or unwanted, and they feel different to play on. They are still overall good precise controllers that work as intended. (with a frame of lag)

 

Sorry dude. I lost you. I am talking about the d-pad of the 8bit do controllers. I have no idea how your post relates to that at all. Now you're talking about "It’s a bigger issue for high-level players of certain kinds of games where accuracy is key." I hope you don't mean I am not skilled enough to perceive a difference. I have half a dozen Arcade sticks, and about 30 different controllers from Genesis originals, Saturn originals, fightpads of all sorts, older hori fighting pads, moderns fightpads etc...

 

I did clearly say, it's easier to get diagonals with the 8bitdo controller, but it isn't that bad. Now again, why it's a bigger issue to some people and not to other I'm not sure, but I can play fine with the 8bitdo controller and get the inputs I want. MY 6 years old daughter can also play yoshi's island with this controller, I have been watching her play the game, and she's doing just fine. In yoshi's island if you press diagonal down you'lll crouch. If you press diagonal up yoshi will throw his tongue up instead of forward. She doesn't have an issue with that at all. She eats fowrard as intended and doesn't croutch when she doesn't want to.

 

I'll say this again and for the last time. It is my personal experience with the 8bitdo controllers, that it is much easier to get diagonals on them. I do get an unwanted diagonal now and then if I'm not careful, but overall the controller is precise and works as intended. I can see some people would have more issues with it depending on their grip, how they like applying pressure on the d-pad etc... I also ask myself wether there are discreppancies between controllers.

 

Of course, there is one last thing that needs to be said: When you want to play a game well, you need to get a pad and get used to it. You need to practice on that pad till you master it. If what you want is get really good, alternating between the 8bitdo controllers and SNES controllers is a bad idea. They feel very different, and switching between them will most likely throw you off your game.

Where did 8bitdo ever claim that they were lag-free? I’ve seen some clueless users say that but anyone who knows how BT works knows that isn’t true. I wouldn’t measure it in “frames” though. [emoji6]

 

While certain games and certain play styles are fine, I disagree that you can simply practice and get used to it for any game/level of play. Again, most people would never notice the issue on, say, an SNES ASCII Pad but it is simply impossible to play something like Tetris Attack at that speed when pressing right five times in a fraction of a second can end up leaving you a row or two above/below where it should be despite never coming near the up or down button. It’s no different with this controller.

 

You would have to slow down to accommodate the D-Pad which will drastically impact your play options (and, thus, scoring). If you are pushing the limits of how fast you can possibly press a direction but you have to slow down to keep it from sending directions you didn’t press then it has already impacted your ability to play a game as well as you were capable of playing it. In this case it makes the game impossible to play at the level we’re easily capable of. I’m nowhere near as good as Ichinisan but even I can’t play with those D-pads. I never pretended it was normal. Heck, his ability to max out the score in under 10 minutes shows that he’s playing beyond what was reasonably expected. Even when they made the score 10x higher in Pokémon Puzzle League and later versions he maxed it out right away (all 9s).

 

There is no “getting used to it” that could enable a high-level Tetris Attack / Panel de Pon player to play with a controller like NES Classic Edition, 8bitdo SN30-SF, SN/SF30, SNES ASCII Pad, etc. There is only slowing down and that is not an option when “playing well” means going fast. The controller makes or breaks this, and in this case it breaks it. I’m not saying that the controller is bad (I LOVE it) and even I insisted that this is not terrible for other people, even pointing out that Nintendo has made controllers with the same flaw. You really don’t have to tell me that it’s OK for most people (casual or otherwise). I agree and already said that.

 

If you’re still “not sure” why it’s a bigger issue for some players than others then perhaps you shouldn’t be so quick to dismiss the idea that some really are playing in a different level. Of course I can avoid the unintended diagonal causing me to throw up in DKC2 by changing my play style. I don’t even have to hold left or right when I intend to throw left or right, but not holding left or right to avoid an errant diagonal WILL impact my abilities and hold me back in the game. I will have a lower time for a 102% speedrun. If you aren’t pushing the limits and are fine with adapting to the controller then you are like most people and the D-pad is fine. I’ve been saying that all along.

 

Need another example of our controller pickiness? Watch us fool around with SMB3 on a fleamarket vendor’s clone console:

https://youtu.be/ax4rhMLPcig

 

Now watch us fool around with a Classic Controller and Classic Controller Pro:

https://youtu.be/cgQsuGwZPu0

 

We still seem to be the only people online to notice and report this issue. Never for a second did I think we were examples of normal use. Heck, we wouldn’t normally even think of playing SMB3 with an analog stick! I was just goofing around when I discovered that issue.

 

Yes, looks like the original non-Pro Classic Controller is more pro than the so-called Classic Controller “Pro.” [emoji6] The Pro is really just a cost-reduced version of the Classic Controller, lacking the articulated mounting pieces and the analog shoulder buttons. Open up a NES CE or Famicom Mini controller and what do you find inside? The same “WCP” chip from a Wii Classic Controller Pro.

 

Point of that tangent is that I’m not under any delusions about how niche our level of controller enthusiasm is or how little our preferences apply to the majority of people. I already took pains to address that every time I discussed the issue and downplayed it by pointing out that you have to have unreasonably high standards for this to be a deal-breaker. Obviously, it wasn’t a deal breaker for me because I love it. If someone cares more than I do, we both agree that it’s unreasonable. [emoji6]

 

 

Nothing displays 720p really well other than 4K displays (there are no 720p displays in the consumer market, they are all 768p). And 480p isn't integer scaled on any consumer flat panel display, nor is 240p.

 

But agreed, Analogue hasn't really hit the NES market yet at an 'affordable' price let alone the JB cores.

 

(I'm speaking as an Analogue NT Mini owner, but I have a lot of disposable income)

I think you’re right about that.

 

I just compared HD Retrovision cables with SNES on various TVs in my collection that were always top of the line when they were first made. It looks surprisingly bad on my old 1080p XBR LCD that specifically detects and says “240p.” That TV was $5,000. It still looks amazing on my 1080i XBR CRT that was $2,500. It also looked great on my non-XBR Trinitron HD CRT sets, like HS510, HS420, etc. Even the random curved-screen Toshiba I tossed into the mix because it had component video looked better than the $5,000 LCD that explicitly supports 240p and passes 240p Test Suite.

 

On that same LCD I saw diagonal interference and noise all over. It was more subtle but it looked like the same diagonal interference you get from improperly wired S-video cables that use composite for either chroma or luma in order to work on the SNS-101. Of course, the scaling was soft and looked nothing like a proper integer scale. The XBR and Hi-Scan Trinitron HD CRTs do a better job scaling 240p, even over composite.

 

I'm happy to hear Kevtris is going to work on improving the scanlines. I think they look good, but maybe a bit too "perfect".

I wonder if there is a way to add a filter which makes the Super Nt look a bit more like a CRT.

 

For example, if you look at this image of Super Mario World running on a Sony Wega, via the excellent HD Retrovision YPbPr Component cable, you'll see some very small vertical "slices", along with the horizontal black lines. It would be neat if this could be replicated.

 

SNES-SMW-CRT-04-vgo.jpg

 

 

And compare to Super Mario World on the Super Nt (1080p, 5x, interpolation off, scanlines @200)

AnalogueSuperNt-vgo-20.jpg

Those lines come from the aperture grill pitch. The screen has hundreds of vertical wires that mask the electron beams from hitting the wrong color phosphors on the CRT’s inner surface. The electron guns are arranged horizontally so that each fires at a different angle through the vertical slit. This is why Trinitron displays are curved from left to right but not up and down. When Sony’s patents expired CRTs were nearing the end of their long life span so I wouldn’t consider them indicative of CRTs in general. That said, it was clearly superior to the “triad” kind of shadow mask. Edited by CZroe
  • Like 1
Link to comment
Share on other sites

So I was testing/playing some games and something came to mind.

 

In games without SRAM or have No Battery, how trivial would it be to have an alternative to a cheat menu be a way to preserve high-scores or even progress in certain games that are controlled by passwords?

 

Like the Tetris game in the Dr.Mario/Tetris, it saves your high scores on reset, but not on power off. Not like a save state (which saves everything, and requires more space.) eg 7E1E30-7E1E6D appears to be the high score table in Tetris.

 

Obviously someone would need to use a debugger to actually find the addresses, but once found they could be stored as a CSV file on the SD card. This seems more like a thing to do with the SD2SNES, but could also be entered as it's own type of game genie/pro action reply type of code, where instead of "patching" the memory, you preserve a specific address and size, and "save on reset/save-on-change/restore on cold boot/restore-after-boot/restore-upon-first read to address." Like with the Tetris game I watched the addresses, and the actual high score table is only manipulated when it's changed, so it could theoretically just be written at any point before the it's read, or it could just be straight up over written at any time before the game over.

 

Anyway that would be an interesting way to extend a feature that would otherwise only be useful for cheats.

Nintendo did something like this in the NES Classics GBA series but they did it the easy way and ended up kinda breaking some things. Rather than setting up the initialized memory like a freshly powered-on NES, the emulator essentially loads a save state and then resets.

 

It works in the sense that Super Mario Bros will retain your hi-score but you might also notice a cool bonus: You can hold A and press start to resume the last world you played, exactly like you would after getting “Game Over” on the original NES game. I used it to jump straight to the Minus World any time I wanted to demonstrate, but there’s a nasty side-effect: Beating the game changes all the Goombas to Buzzy Beetles and speeds the enemies up, so saving a hi-score after finishing the game will leave it stuck in this state until you clear the save memory.

 

Finding the memory addresses for the hi-scores of each game would have been trivially easy but they probably didn’t want to customize the emulator for any of them. That’s why they all have 2p link support that downloads the entire game into another GBA’s memory even when it’s a 1p game. That’s also why some of the boxes in the series have a sticker covering the number of supported players on the back. [emoji6]

 

 

The crt looks more like rectangles or squares to me then lines. Maybe put a option for rectangles or square also.

Is there any reason why you guys are using 1080p. MLIG already said you get the best results with 720p with scan lines with full picture since it scales better with 240p.

I hope for a cheat menu option in the future. SOmetimes I like playing around

Because integer scaling is more important than scanlines and it’s possible to get integer scaling with scanlines using a 5x vertical scale or a windowed 4x vertical scale.

 

Having scanlines at 720p is perfect for 720p displays and 4K displays but there are better options for 1080p displays.

 

Welp, soon that will hopefully no longer be an issue now that Kevtris is working on a solution.

The solution I saw him mention would still be a problem for displays that don’t support full sRGB, which is a lot. Most?

 

 

What we saw back in the 90s was largely just random. The viewing area of your TV was 4:3, but that didn't necessarily mean that it was accurately stretched to 4:3. You'd have a hard time accurately replicating what your specific TV did right and wrong.

 

Meanwhile, many games have objects that are clearly meant to be squares, circles, or other shapes where it's easy to tell when you're displaying it in a different aspect ratio than it was created. Often it turns out that 8:7 shows squares as squares, while 4:3 shows squares as rectangles. This isn't *always* the case, and one option is to change what you're using depending on how the game looks, but a lot of us feel that 8:7 is usually the right choice for getting graphics that don't have a bunch of ovals and rectangles that clearly should be circles and squares.

4:3 has always been an approximation of the even less clearly defined “Academy aspect” in old time cinema. Tubes come in various shapes and sizes. Heck, many early black and white TVs were circular and they could tune the same analog broadcast signals as the last CRTs, so it definitely wasn’t a rigidly defined aspect ratio. Heck, few CRTs have 90° corners and a lot have user adjustments that distort the horizontal aspect ratio.

 

That said, I wouldn’t always judge by the appearance of circles and squares in-game. It’s very clear that the devs themselves usually didn’t get what they were going for even back in the day and it all depends on what they did on a case-by-case basis. Heck, most multi-platform ports had drastically different aspect ratios for the same pixel art. Compare Street Fighter II character bios across console platforms. Heck, the other day I saw a particularly egregious example with Street Racer on SNES versus Megadrive (game wasn’t released on Genesis):

https://youtu.be/KRzmWBs02AQ

 

Compare Biff’s headshot:

3d77886597801aa9754d795e38bc1339.jpg

 

 

People claim an Everdrive has never fried a system, but that is untrue because one fried the PPU on my NES. I bought one for a friend and was setting it up and it destroyed it immediately. I won't put another one of those things (at least the NES one) in my systems.

Yes, it’s theoretically possible, but I’ve had PPUs spontaneously fry on me with no Everdrive so we can’t assume that was the cause. Also, the NES Everdrive N8 is not on the naughty list. What Everdrive did you use? I’m unfamiliar with earlier revisions of the NES ED but if you were using an EDN8 then you did not encounter the theoretical problem with Everdrives that we were discussing.

 

Here is one of my fried NES PPUs that failed in a particularly interesting way:

https://youtu.be/9gNHo9VLBro

 

Yes, the problem follows the PPU from console to console and putting a good PPU in that console fixed the issue. It also works fine with Hi-Def NES... and in full color. [emoji4]

 

I’ve also purchased NES consoles with “fried” PPUs and then had them suddenly work after socketing them (also never touched an NES ED). PPU failures happen and there’s a good chance that the ED had nothing to do with it.

 

 

Is your display 1080p or 4k? If it's 1080p and you're running the Super Nt at 1080p60 5x vertical, requiring interpolation to get even scanlines is a sure sign you don't have 1:1 pixel mapping enabled on your display.

 

It would be much easier to help people with display issues if everyone noted what display they were using and what settings.

 

The core settings that 1080p display owners should be using if they want clean scanlines are:

 

Resolution: 1080p

Height: 5x (1200)

Scaler: No scaler, vertical interpolation disabled

Scanlines: Normal or hybrid (whichever you prefer, depth to taste)

 

AND: Display set to 1:1 pixel mapping (zero overscan), which as others have said varies depending on your display manufacturer (1:1, Just, Full, etc.)

 

Horizontal dimensions can be whatever you prefer, with horizontal interpolation disabled if it's an integer multiplier (5x or 6x) because it's not necessary. Tweak vertical position also if you like; I adjusted mine to 45 because it clips a bit more on the bottom than the top and lines up the scanlines so that they fall right at the top and bottom of my display.

Yes. I’ve almost given up on telling people this, but it isn’t about specific settings for certain displays with certain native resolutions because those displays are almost never displaying signals with their native resolutions 1:1. It’s going to be a combination of settings between the Super Nt and the display device. The correct settings for the display are HIGHLY variable (when they are even possible) so any settings “for 1080p” or “for 720p/4K” will only work if the person sharing them and the person using them have managed to disable overscan simulation and other functions that would result in a non-integer scale.

 

IOW, It isn’t as simple as just having good settings for your Super Nt (as much as I wish it were). You also need good settings for your TV to support good settings for the Super Nt, and not nearly all TVs have those settings. It’s MUCH more difficult to share those settings too.

 

Sharing “good settings for scanlines on 720p” or whatever absolute requires that both users have achieved the correct TV settings for an integer scale.

Edited by CZroe
  • Like 1
Link to comment
Share on other sites

So, I have a weird problem with the Super NT and an SD2SNES.

The SD2SNES works fine in my original SNES. But when I put it in the Super NT, on first boot I get just a black screen. When I turn the Super NT off and on again (not reset) then it's fine. It'll keep working through reboots and on/offs until I leave it turned off for a while. After which, I have to turn it on twice again to get it working.

Nothing seems wrong with its performance in the Super NT other than this one weird quirk.

Link to comment
Share on other sites

Nintendo did something like this in the NES Classics GBA series but they did it the easy way and ended up kind breaking some things. Rather than setting up the memory like a freshly powered-on NES, the emulator essentially loads a save state and then resets.

 

It works in the sense that Super Mario Bros will retain your hi-score but you might also notice a cool bonus: You can hold A and press start to resume the last world you played, exactly like you would after getting “Game Over” on the original NES game. I used it to jump straight to the Minus World any time I wanted to demonstrate, but there’s a nasty side-effect: Beating the game changes all the Goombas to Buzzy Beetles and speeds the enemies up, so saving a hi-score after finishing the game will leave it stuck in this state until you clear he save memory.

 

Finding the memory addresses for the hi-scores of each game would have been trivially easy but they probably didn’t want to customize the emulator for any of them. That’s why they all have 2p link support that downloads the entire game into another GBA’s memory even when it’s a 1p game. That’s also why some of the boxes in the series have a sticker covering the number of supported players on the back. ;)

 

Because integer scaling is more important than scanlines and it’s possible to get integer scaling with scanlines using a 5x vertical scale or a windows 4x vertical scale.

 

Having scanlines at 720p is perfect for 720p displays and 4K displays but there are better options for 1080p displays.

 

The solution I saw him mention would still be a problem for displays that don’t support full sRGB, which is a lot.

 

4:3 has always been an approximation of the even less clearly defined “Academy aspect” in old time cinema. Tubes come in various shapes and sizes. Heck, many early black and white TVs were circular and they could tune the same analog broadcast signals as the last CRTs, so it definitely wasn’t a rigidly defined aspect ratio. Heck, few CRTs have 90° corners and a lot have user adjustments that distort the horizontal aspect ratio.

 

That said, I wouldn’t always judge by the appearance of circles and squares in-game. It’s very clear that the devs themselves usually didn’t get what they were going for even back in the day and it all depends on what they did on a case-by-case basis. Heck, most multi-platform ports had drastically different aspect ratios for the same pixel art. Compare Street Fighter II character bios across console platforms. Heck, the other day I saw a particularly egregious example with Street Racer on SNES versus Megadrive (game wasn’t released on Genesis):

 

Compare Biff’s headshot:

3d77886597801aa9754d795e38bc1339.jpg

 

Yes, it’s theoretically possible, but I’ve had PPUs spontaneously fry on me with no Everdrive so we can’t assume that was the cause. Also, the NES Everdrive N8 is not on the naughty list. What Everdrive did you use? I’m unfamiliar with earlier revisions of the NES ED but if you were using an EDN8 then you did not encounter the theoretical problem with Everdrives that we were discussing.

 

Here is one of my fried NES PPUs that failed in a particularly interesting way:

 

Yes, the problem follows the PPU from console to console and putting a good PPU in that console fixes the issue. It also works fine with Hi-Def NES... and in full color. :)

 

I’ve also purchased NES consoles with “fried” PPUs and then had them suddenly work after socketing them (also never touched an NES ED). PPU failures happen and it there’s a good chance that the ED had nothing to do with it.

 

Yes. I’ve almost given up on telling people this, but it isn’t about specific settings for certain displays with certain native resolutions because those displays are almost never displaying signals with their native resolutions 1:1. It’s going to be a combination of setting between the Super Nt and the display device. The correct settings for the display are HIGHLY variable when they are even possible so any settings “for 1080p” or “for 720p/4K” will only work if the person sharing them and the person using them have managed to disable overscan simulation and other functions that would result in a non-integer scale.

 

IOW, It isn’t as simple as just having good settings for your Super Nt (as much as I wish it were). You also need good settings for your TV to support good settings for the Super Nt, and not nearly all TVs have those settings.

 

Sharing “good settings for scanlines on 720p” or whatever absolute requires that both users have achieved the correct TV settings for an integer scale.

 

It was an N8. And I don't care what you do with your systems and you can believe what you want to believe, but I won't be using them.

Link to comment
Share on other sites

4:3 has always been an approximation of the even less clearly defined “Academy aspect” in old time cinema. Tubes come in various shapes and sizes. Heck, many early black and white TVs were circular and they could tune the same analog broadcast signals as the last CRTs, so it definitely wasn’t a rigidly defined aspect ratio. Heck, few CRTs have 90° corners and a lot have user adjustments that distort the horizontal aspect ratio.

 

That said, I wouldn’t always judge by the appearance of circles and squares in-game. It’s very clear that the devs themselves usually didn’t get what they were going for even back in the day and it all depends on what they did on a case-by-case basis. Heck, most multi-platform ports had drastically different aspect ratios for the same pixel art. Compare Street Fighter II character bios across console platforms. Heck, the other day I saw a particularly egregious example with Street Racer on SNES versus Megadrive (game wasn’t released on Genesis):

Yeah, I wasn't really trying to say 8:7 was "correct" in any way. Just that in some cases, game assets were clearly developed for that ratio. I find at 8:7 I run into less circumstances where shapes seem wrong than if I run at 4:3. But my point was just 8:7 and 4:3 both have justifications for being the "right" ratio for the SNES, and beyond that it's down to taste.

Edited by DarkkOne
Link to comment
Share on other sites

 

It was an N8. And I don't care what you do with your systems and you can believe what you want to believe, but I won't be using them.

Woah. Calm down, Francis. I go out of my way to avoid the improper 3.3v products too, even if I doubt that led to any real-world failures for Krikzz’s products (he did take SOME precautions). I usually prefer the non-Everdrive alternative in most cases for that reason and it’s why I went with a Mega ED X5 instead of a Mega ED X7. Being concerned enough to let it influence my purchase decisions doesn’t mean I think the X7 will blow up my console with improper level translation, so hopefully I’m capable of looking at it rationally and without being clouded by emotion. That’s what made the fallout from the article such a clusterf—k.

 

You later established that it was a defective Everdrive but Tapatalk wouldn’t show me the replies until I finished my post and refreshed. I was only establishing that it wasn’t the 3.3v vs. 5v issue that we were discussing when you weighed in. Of course defective stuff can ruin things. Of course EDs can be defective. I know that. Why would I say otherwise? I don’t deny that something here was defective and it definitely could have been the NES ED.

 

Ever see KevBot’s NES Killer? [emoji6]

http://nintendoage.com/forum/messageview.cfm?catid=22&threadid=168385

Edited by CZroe
  • Like 1
Link to comment
Share on other sites

So, I have a weird problem with the Super NT and an SD2SNES.

 

The SD2SNES works fine in my original SNES. But when I put it in the Super NT, on first boot I get just a black screen. When I turn the Super NT off and on again (not reset) then it's fine. It'll keep working through reboots and on/offs until I leave it turned off for a while. After which, I have to turn it on twice again to get it working.

 

Nothing seems wrong with its performance in the Super NT other than this one weird quirk.

A lot of NES carts behave that way due to failing capacitors but that would definitely not be the case here. Still reason to suspect a power issue though.

 

First, what version is it? I’m running both a Rev H (latest) and a Rev E1b with the Rev H mods (inductor, capacitor, and MSU-1 audio amp). After my mods they should be functionally identical and both work fine in my Super Nt. I’m not even using the 2A PSU (using 1A from a Fire TV stick). What PSU are you using?

Edited by CZroe
Link to comment
Share on other sites

Woah. Calm down, Francis. I go out of my way to avoid the improper 3.3v products too, even if I doubt that led to any real-world failures for Krikzz’s products (he did take SOME precautions). I usually prefer the non-Everdrive alternative in most cases for that reason and it’s why I went with a Mega ED X5 instead of a Mega ED X7. Being concerned enough to let it influence my purchase decisions doesn’t mean I think the X7 will blow up my console with improper level translation.

 

You later established that it was a defective Everdrive but Tapatalk wouldn’t show me the replies until I finished my post and refreshed. I was only establishing that it wasn’t the 3.3v vs. 5v issue that we were discussing when you weighed in. Of course defective stuff can ruin things. Of course EDs can be defective. I know that. Why would I say otherwise?

 

Ever see KevBot’s NES Killer? [emoji6]

http://nintendoage.com/forum/messageview.cfm?catid=22&threadid=168385

 

Yes, I can't confirm that's what it was (and really I don't care what it was because the end result of a fried PPU is the same regardless) because A) I wouldn't have the knowledge to determine that, and B) neither SAG or Retrofixes quite had an answer. That said, I've no need for one anyway since I have a Mini, so I'll be staying away. However, the claim that an Everdrive has never blown anything up is not true from my experience.

 

On a side note, thinking back on this (it happened March of last year, so been a lil while) I don't think I actually used the N8 on my Mini. I had an old toaster laying around that I ended up setting the new one up on. Not that there would really be any need to use an Everdrive on a Mini, but just throwing that out there for the sake of accuracy.

Edited by jamon1567
Link to comment
Share on other sites

 

Yes, I can't confirm that's what it was (and really I don't care what it was because the end result of a fried PPU is the same regardless) because A) I wouldn't have the knowledge to determine that, and B) neither SAG or Retrofixes quite had an answer. That said, I've no need for one anyway since I have a Mini, so I'll be staying away. However, the claim that an Everdrive has never blown anything up is not true from my experience.

 

On a side note, thinking back on this (it happened March of last year, so been a lil while) I don't think I actually used the N8 on my Mini. I had an old toaster laying around that I ended up setting the new one up on. Not that there would really be any need to use an Everdrive on a Mini, but just throwing that out there for the sake of accuracy.

 

I have an N8 in my mini because of the save state feature. Other than that there's no real point in using one in the mini.

Link to comment
Share on other sites

A lot of NES carts behave that way due to failing capacitors but that would definitely not be the case here. Still reason to suspect a power issue though.

 

First, what version is it? I’m running both a Rev H (latest) and a Rev E1b with the Rev H mods (inductor, capacitor, and MSU-1 audio amp). After my mods they should be functionally identical and both work fine in my Super Nt. I’m not even using the 2A PSU (using 1A from a Fire TV stick). What PSU are you using?

 

It's a Rev. H. board, and I'm using the power supply and cable provided with the Super NT.

Link to comment
Share on other sites

Yeah, I wasn't really trying to say 8:7 was "correct" in any way. Just that in some cases, game assets were clearly developed for that ratio. I find at 8:7 I run into less circumstances where shapes seem wrong than if I run at 4:3. But my point was just 8:7 and 4:3 both have justifications for being the "right" ratio for the SNES, and beyond that it's down to taste.

8:7 fanatics love to post pics of tile-based games like SMW and SMW2, where the 8x8 tile graphics create angles that only look correct in 8:7, but there are plenty of examples where the artists did take into account proper aspect correction, and it's not 4:3 either. It's the 256 * 8/7 formula I mentioned earlier that Kevtris correctly labels in the Super Nt for 1170x960 and 1462x1200. I disagree with his rounding down to an odd number on 3x scale with 877x720 though. It should be 878x720 because it's much closer to that and you avoid odd pixel counts for centering.

 

At any rate, there are some superb examples of SNES devs taking into account this formula. The Triforce in the beginning of A Link to the Past are only perfectly shaped when the correction formula is applied, and also the artists that made Chrono Trigger also took into account the correct formula. This is my favorite counter-argument image I like to post in regards to 8:7 lovers:

 

bLSZUjb.png

Edited by Karbuncle
  • Like 1
Link to comment
Share on other sites

8:7 fanatics love to post pics of tile-based games like SMW and SMW2, where the 8x8 tile graphics create angles that only look correct in 8:7, but there are plenty of examples where the artists did take into account proper aspect correction, and it's not 4:3 either. It's the 256 * 8/7 formula I mentioned earlier that Kevtris correctly labels in the Super Nt for 1170x960 and 1462x1200. I disagree with his rounding down to an odd number on 3x scale with 877x720 though. It should be 878x720 because it's much closer to that and you avoid odd pixel counts for centering.

 

At any rate, there are some superb examples of SNES devs taking into account this formula. The Triforce in the beginning of A Link to the Past are only perfectly shaped when the correction formula is applied, and also the artists that made Chrono Trigger also took into account the correct formula. This is my favorite counter-argument image I like to post in regards to 8:7 lovers:

 

Yeah, by 8:7 I was actually referring to an 8:7 grid of 8:7 AR pixels. The moon in Chrono Trigger was actually one of the particular things I was thinking of as "circles that don't look like circles." Square pixels for some games may also look better to some people though.

 

There's definitely no "right" just because it depends on each game, and whether or not the developers of that game took into account the display properties. Sometimes they did, sometimes they didn't. But I find the Super NT's "8:7" mode is generally a solid choice if you don't want to fiddle.

  • Like 1
Link to comment
Share on other sites

Probably showing my lack of attention.. but I have to say throughout my life I never noticed "scan lines" on any CRT, or knew of their existence/that they were a thing, until they were pointed out to me during discussions like these. And I've had CRT's pointing in my face a long time lemme tell ya :lol:

Link to comment
Share on other sites

Maybe it would be helpful if Kevtris broke down all the audio/video settings and offered a native option?

 

Resolution - native (whatever the game is coded for in real time)

Refresh rate - native (whatever the game is coded for in real time). I suspect variable refresh rate displays will help.

Aspect Ratio - native (256 x 8/7)

Audio - native (32 kHz). Hoping the upcoming DAC has an optical pass thru for our own audio DAC.

 

 

On a side note, does the snes/sfc output in the sRGB color space? We should be calibrating our monitors to the correct full range color. I dont think anyone has mentioned this.

Link to comment
Share on other sites

Maybe it would be helpful if Kevtris broke down all the audio/video settings and offered a native option?

 

Resolution - native (whatever the game is coded for in real time)

Refresh rate - native (whatever the game is coded for in real time). I suspect variable refresh rate displays will help.

Aspect Ratio - native (256 x 8/7)

Audio - native (32 kHz). Hoping the upcoming DAC has an optical pass thru for our own audio DAC.

 

 

On a side note, does the snes/sfc output in the sRGB color space? We should be calibrating our monitors to the correct full range color. I dont think anyone has mentioned this.

Link to comment
Share on other sites

 

Same as what the OSSC does in x5 1080p mode - it gives you a true integer scale on a 1080p display - I actually came to the same conclusion as Firebrandx on the shmups forum:

 

 

This results in a pixel-perfect scale to 1080p, albeit with the top and bottom cut off by 20 pixels each. Which really isn't noticeable, especially when it's smaller than the CRT safe-zone games were designed around:

 

 

 

It results in an identical image to what the OSSC produces in x5 mode.

 

If you want to play games in native 1080p as an integer scale, this is the only way to do it. And really, it's a fair trade-off.

Yeah, I have 0 Scan mode and 5x just cuts too much off for me. There plenty of games where valuable information gets cut. At least I consider it valuable information.

Link to comment
Share on other sites

I just have to throw it out there that I am so impressed with both Analogue & Kevtris. This is not a super large company or anything, but the product they've released is really class. From the satisfying click of powering the unit on, to the ever-changing intro graphics, to the wide variety of options present, the perfect emulation... it all really adds up to an A+ effort in my opinion. I feel that my money was very well spent and will be continuously following Analogue and supporting them from now on (this is my first Analogue product, I never did get the Mini).

 

I find too that the Super NT looks superior to my real SNES through the OSSC and controls just as well with all the games I've put it through. So I think when it comes to my LED TV, the go-to will always from now on be my Super NT while the SNES will get a bit less abuse. I do enjoy playing my older consoles on the Sony PVM, but as those units are aging, the reduced wear and tear should extend their lifespan.

 

Also, being someone who repairs consoles, I kept coming across broken SNES consoles that had bad chips or were beyond repairable... all of this when fixing the NES or N64 had a much higher success rate. So, that said, the Super NT was the most needed clone console imo. The one that I believe has saved my favorite console of all time from a sad extinction. Hope Analogue keeps producing these! And I am crossing my fingers for more excellent works of art from Analogue. Whatever they make, I shall buy. (Especially if Kevtris is involved)

  • Like 1
Link to comment
Share on other sites

Nintendo did something like this in the NES Classics GBA series but they did it the easy way and ended up kind breaking some things. Rather than setting up the memory like a freshly powered-on NES, the emulator essentially loads a save state and then resets.

 

It works in the sense that Super Mario Bros will retain your hi-score but you might also notice a cool bonus: You can hold A and press start to resume the last world you played, exactly like you would after getting “Game Over” on the original NES game. I used it to jump straight to the Minus World any time I wanted to demonstrate, but there’s a nasty side-effect: Beating the game changes all the Goombas to Buzzy Beetles and speeds the enemies up, so saving a hi-score after finishing the game will leave it stuck in this state until you clear he save memory.

Being a huge fan of the original Super Mario Bros, I remember encountering this issue on Wii VC. I've played Super Mario and rescued the princess more times than i can count, but the second quest always felt rushed and tacked on for some reason. It is not significantly more difficult than the first quest, although the buzzy beetles can be annoying to deal with in certain places. Beetles and Koopas also move marginally faster in second quest, but it just made it feel like the timings were off. VC lacked a mode to perform a hard reset, and I would long to play the base game in short order. Initially, I tried going into settings and manually deleting the save, but finally discovered the simpler workaround of manually cutting the power while in the game, which fixed the issue. This was before Nintendo released a statement in the support section of the website instructing prople to power down their consoles in game to hard reset the game. When I started collecting Nintendo in the early 2000s, I had the Mario/Duckhunt combo cart so never experienced this firsthand on real hardware, at least until I set out to collect every Mario variant, including both the ntsc and pal exclusive 3-in-1s. In fact I prefer the pal version running on ntsc hardware over "second quest" mode, if I desire a difficulty boost.

 

Hqaving scanlines at 720p is perfect for 720p displays and 4K displays but there are better options for 1080p displays.

The solution I saw him mention would still be a problem for displays that don’t support full sRGB, which is a lot.

720p mode works fine on 1080p displays, even with scanlines, but this depends upon the quality of the built in scalar. On my ASUS, they look gorgeous. If your "1080p" display cannot disable overscan, it takes 1080p input, scales it to approximately 110% scale, cropping 108 hd scanlines from the top and bottom of the image. This translates to roughly 21.6 snes scanlines in 5x mode, cropping the screen from 216 scanlines to 195 scanlines. At this point, the picture will have been cropped beyond the 15% safe area, possibly cutting off important hud information. The thin scanlines in 5x mode being scaled an extra 10% will cause ~11 pixels of screen for every 2 snes scanlines, roughly 5.5 pixel scale leading to an uneven look. In this situation, 4.5x mode will create uneven scanlines before scaling, resulting in more unevenness when scaled up 10%. Non-integer scaling is less detrimental when closer to a n.5 ratio as opposed to n.1 or n.9 ratios. So on 1080p sets where you cannot disable overscan, a 720p image scaled by 1.6x onscreen will be more desirable than a 1080p image scaled by 1.1. My two cents...

 

 

4:3 has always been an approximation of the even less clearly defined “Academy aspect” in old time cinema. Tubes come in various shapes and sizes. Heck, many early black and white TVs were circular and they could tune the same analog broadcast signals as the last CRTs, so it definitely wasn’t a rigidly defined aspect ratio. Heck, few CRTs have 90° corners and a lot have user adjustments that distort the horizontal aspect ratio.

 

That said, I wouldn’t always judge by the appearance of circles and squares in-game. It’s very clear that the devs themselves usually didn’t get what they were going for even back in the day and it all depends on what they did on a case-by-case basis. Heck, most multi-platform ports had drastically different aspect ratios for the same pixel art. Compare Street Fighter II character bios across console platforms. Heck, the other day I saw a particularly egregious example with Street Racer on SNES versus Megadrive (game wasn’t released on Genesis):https://youtu.be/KRzmWBs02AQ

Compare Biff’s headshot:3d77886597801aa9754d795e38bc1339.jpg

Yes ratios were all over the place, however Sega consoles output 320x240 native resolution which translates directly to a 1:1 pixel ratio. Nintendo's own NES, SNES, as well as NEC Turbografx / PCe output a 240x256 pixel image (though SNES blacked out the top and bottom 8 scanlines). Assuming both consoles display similar h-scan areas, 320 / 240 = 1.25. This is why I prefer a slightly wider 1.25x pixel aspect on Nintendo hardware. This is 640 pixel width for 2x vertical scale, 960 pixel width for 3x vertical, 1280 pixel width for 4x vertical, 1440 for 4.5x vertical (I recommend 720 3x in leui of this setting), and 1600 pixel width for 5x vertical scale. Turn horizontal interpolation on except for 5:4 pixel aspect (960p) and you should get clean scanlines (except for the factory default 4.5x mode with scanlines off).

 

I've noticed various developers used different settings for drawing circles. Sonic 2 Chemical plant uses circles that are actually skinnier than they are tall (on my crt sets), and circles and blocks in super mario world appear wider than they are tall, considering Nintendo's wider than square natve pixels. And perfect square blocks never looked right in Mario games anyway, but Nintendo always developed 2d Mario games to the grid regardless of native pixel aspect. And it always varied between manufavturers. Atari 2600/5200/7800/8-bit/XEGS used a whopping 1.8:1 pixel aspect ratio! Squarish pixels were virtually non-existent during the pre-crash era...

 

Yes, it’s theoretically possible, but I’ve had PPUs spontaneously fry on me with no Everdrive so we can’t assume that was the cause. Also, the NES Everdrive N8 is not on the naughty list. What Everdrive did you use?

I think that PPU was ready to die anyway and nothing was going to save it.
  • Like 1
Link to comment
Share on other sites

This is correct, particularly about the aperture grill retaining wires.
That said, I had a Diamondtron monitor with Shadow Mask and that was quite a nice monitor. I wish I still had it but downsizing got to me.
Yeah, truth be told, I prefer curved rgb tubes for old games, lol. I have a Panasonic BT monitor and a SONY PVM and I really can't call it between the two for which I really prefer. I think i'm just used to that curve, lol.

Where did 8bitdo ever claim that they were lag-free? I’ve seen some clueless users say that but anyone who knows how BT works knows that isn’t true. I wouldn’t measure it in “frames” though. [emoji6]

While certain games and certain play styles are fine, I disagree that you can simply practice and get used to it for any game/level of play. Again, most people would never notice the issue on, say, an SNES ASCII Pad but it is simply impossible to play something like Tetris Attack at that speed when pressing right five times in a fraction of a second can end up leaving you a row or two above/below where it should be despite never coming near the up or down button. It’s no different with this controller.

You would have to slow down to accommodate the D-Pad which will drastically impact your play options (and, thus, scoring). If you are pushing the limits of how fast you can possibly press a direction but you have to slow down to keep it from sending directions you didn’t press then it has already impacted your ability to play a game as well as you were capable of playing it. In this case it makes the game impossible to play at the level we’re easily capable of. I’m nowhere near as good as Ichinisan but even I can’t play with those D-pads. I never pretended it was normal. Heck, his ability to max out the score in under 10 minutes shows that he’s playing beyond what was reasonably expected. Even when they made the score 10x higher in Pokémon Puzzle League and later versions he maxed it out right away (all 9s).

There is no “getting used to it” that could enable a high-level Tetris Attack / Panel de Pon player to play with a controller like NES Classic Edition, 8bitdo SN30-SF, SN/SF30, SNES ASCII Pad, etc. There is only slowing down and that is not an option when “playing well” means going fast. The controller makes or breaks this, and in this case it breaks it. I’m not saying that the controller is bad (I LOVE it) and even I insisted that this is not terrible for other people, even pointing out that Nintendo has made controllers with the same flaw. You really don’t have to tell me that it’s OK for most people (casual or otherwise). I agree and already said that.

If you’re still “not sure” why it’s a bigger issue for some players than others then perhaps you shouldn’t be so quick to dismiss the idea that some really are playing in a different level. Of course I can avoid the unintended diagonal causing me to throw up in DKC2 by changing my play style. I don’t even have to hold left or right when I intend to throw left or right, but not holding left or right to avoid an errant diagonal WILL impact my abilities and hold me back in the game. I will have a lower time for a 102% speedrun. If you aren’t pushing the limits and are fine with adapting to the controller then you are like most people and the D-pad is fine. I’ve been saying that all along.

Need another example of our controller pickiness? Watch us fool around with SMB3 on a fleamarket vendor’s clone console:



Now watch us fool around with a Classic Controller and Classic Controller Pro:


We still seem to be the only people online to notice and report this issue. Never for a second did I think we were examples of normal use. Heck, we wouldn’t normally even think of playing SMB3 with an analog stick! I was just goofing around when I discovered that issue.

Yes, looks like the original non-Pro Classic Controller is more pro than the so-called Classic Controller “Pro.” [emoji6] The Pro is really just a cost-reduced version of the Classic Controller, lacking the articulated mounting pieces and the analog shoulder buttons. Open up a NES CE or Famicom Mini controller and what do you find inside? The same “WCP” chip from a Wii Classic Controller Pro.

Point of that tangent is that I’m not under any delusions about how niche our level of controller enthusiasm is or how little our preferences apply to the majority of people. I already took pains to address that every time I discussed the issue and downplayed it by pointing out that you have to have unreasonably high standards for this to be a deal-breaker. Obviously, it wasn’t a deal breaker for me because I love it. If someone cares more than I do, we both agree that it’s unreasonable. [emoji6]

I think you’re right about that.

I just compared HD Retrovision cables with SNES on various TVs in my collection that were always top of the line when they were first made. It looks surprisingly bad on my old 1080p XBR LCD that specifically detects and says “240p.” That TV was $5,000. It still looks amazing on my 1080i XBR CRT that was $2,500. It also looked great on my non-XBR Trinitron HD CRT sets, like HS510, HS420, etc. Even the random curved-screen Toshiba I tossed into the mix because it had component video looked better than the $5,000 LCD that explicitly supports 240p and passes 240p Test Suite.

On that same LCD I saw diagonal interference and noise all over. It was more subtle but it looked like the same diagonal interference you get from improperly wired S-video cables that use composite for either chroma or luma in order to work on the SNS-101. Of course, the scaling was soft and looked nothing like a proper integer scale. The XBR and Hi-Scan Trinitron HD CRTs do a better job scaling 240p, even over composite.

Those lines come from the aperture grill pitch. The screen has hundreds of vertical wires that mask the electron beams from hitting the wrong color phosphors on the CRT’s inner surface. The electron guns are arranged horizontally so that each fires at a different angle through the vertical slit. This is why Trinitron displays are curved from left to right but not up and down. When Sony’s patents expired CRTs were nearing the end of their long life span so I wouldn’t consider them indicative of CRTs in general. That said, it was clearly superior to the “triad” kind of shadow mask.
Link to comment
Share on other sites

 

When I get mine, I'll probably make an effort to compare the SNES and the Super NT, but I'll probably have to move the surround sound box to the PC to do it, as it's currently setup around the CRT.

 

 

I imagine the sample conversion wouldn't matter much. [...]

 

It looks like someone has already done some testing with SNES surround sound games and the Super Nt: "I'm running the Super NT into a Denon AVR-S930H via HDMI, which detected and decoded the Dolby Surround from Fatal Fury Special. I can't even describe how weird and cool it is to hear SNES games in actual surround. It's bananas. Highly recommended."

  • Like 1
Link to comment
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.
Note: Your post will require moderator approval before it will be visible.

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...