It's probably too late for this suggestion, but the authors of the SM64 1 Key submission recently realized that Tyler Kehne, master of the parallel universe glitch which allowed such time to be saved for any%, lives only 8 hours from the location of AGDQ 2017. If possible and wanted by others, we would love for the 1 Key TAS to be shown off with Tyler himself present. We don't really know where else to ask or mention this, so if someone who knows this process and the people associated could point us in the right direction, we would be grateful!
It's probably too late for this suggestion, but the authors of the SM64 1 Key submission recently realized that Tyler Kehne, master of the parallel universe glitch which allowed such time to be saved for any%, lives only 8 hours from the location of AGDQ 2017. If possible and wanted by others, we would love for the 1 Key TAS to be shown off with Tyler himself present. We don't really know where else to ask or mention this, so if someone who knows this process and the people associated could point us in the right direction, we would be grateful!
PLEASE MAKE THIS HAPPEN
I had two reactions when I saw that new "1 key" TAS:
1) It's amazing!! (obviously ^^)
2) (after a short while) Ah wait... It would have been great to keep it secret and reveal the whole thing at AGDQ, to make more jaws drop at the same time. Especially since it is console verified.
Moderator, Senior Ambassador, Experienced player
(898)
Joined: 9/14/2008
Posts: 1007
To say that this update is past-due is an understatement, but it's for a very good reason: there's been an insane amount of activity in the last few weeks even if it hasn't looked like it. I've been so uninvolved on the site itself due to external TAS oriented distractions that I even forgot to change my profile update, but I digress (but not before mentioning that Mediamagnet has raffle prizes pretty much singlehandedly covered - thanks!). Here's a summary of where we are at:
Per the AGDQ 2017 schedule, the TAS block will be on Saturday, January 14th at around 3:45 PM EST (subject to change, usually for the later). Two items have been officially accepted and include a mystery game and DKC2. There are challenges with both which I'll describe below but we're looking to fill about 35-45 minutes of content, potentially more depending on the outcome of donation incentives.
- Mystery game: Truth be told, we really, really want to do something with the NES Classic. It's very popular, it's right up our alley, and there are a lot of possibilities with the platform. The problem is that no one I know was able to secure an NES Classic Mini console yesterday. Believe me, I reallytried. We have other ideas but the NES Classic Mini is the best option, so if you were able to secure one and can lend it to us for a while we promise to be (mostly) gentle to it.
- DKC2: I am remiss on streaming this in part due to needing to rebuild my desktop computer after an SSD went on the fritz but I've proven that the ceramic crystal clock used for the sound SMP board on the SNES has drifted over the years. For my console, the 24.576 MHz crystal is clocking more like 24.573 MHz likely due to age and being a cheap part to begin with. Why does this matter? If the sound board takes the wrong amount of time to play certain sounds the SMP will spinlock against the main CPU and the result will be a different number of frames to perform certain actions, i.e. it will cause a desync.
Now that we've measured that the clock is not as accurate as it was expected to be when the SNES was released, the plan is to replace the crystal with a quartz one that is far more accurate, but this requires replacing adjacent capacitors with ones of a different size (from 3pF to 18-22pF). The capacitors I was able to get my hands on will work but are the size of a gnat; I'm fortunate that a co-worker with extremely good soldering skills will be taking this on on Monday and I hope we can make some progress with testing out our theory.
There has also been one other major thing that's kept me busy - preparing for international travel!
GeekPwn 2016: I flew to Shanghai, China complements of the GeekPwn 2016 security conference on 2016-10-24 and I was fortunate to have Jaku and a couple of his co-workers attend. I spent a fair bit of time with geohotz and other talented hackers while I was there, which was quite a treat. Before the event I worked with total, micro500, fuzyll, Ownasaurus, Ilari, p4plus2 and others to to create something that has never been seen before on an NES. I have not received any video of the event itself but you can watch an overview of what we demonstrated in the following video:
Link to video
The talk was incredibly well received and won an award for Geek Thinking that's larger than TASBot!
There is the possibility of a monetary reward but it is unclear how large it will ultimately be. I discussed with everyone who helped out and we have agreed to withold a portion to cover actual equipment costs only and donate the remainder of the reward during AGDQ 2017. Expect a healthy donation during the TASBot block if all goes well
I attended the annual Day of the Devs event and spoke briefly with Tim Schafer about collaborating on something. More on that for a future GDQ but remember you heard it here first. This isn't Tim but it's still a fun picture:
Last but not least, if you are in the bay area, I will be presenting at the SF Bay Area chapter of the ACM (Association for Computing Machinery) on 2016-11-16. The talk will be a fusion of what I presented at DEF CON 24 and what I presented at GeekPwn.
Anyway, lots going on, more to come - please let us know if you can spare an NES Classic and expect an update from me in a week or two. Thanks for the support!
Wow you are doing really amazing work DwangoAC!
This is really cool stuff and I think yours and all the others involved hard work is very impressive. Looking forward to whatever you have in store for AGDQ!
Moderator, Senior Ambassador, Experienced player
(898)
Joined: 9/14/2008
Posts: 1007
I've been streaming at http://twitch.tv/dwangoac very frequently over the last few days but I wanted to pause for a moment for some updates. First, I wanted to share a couple of quick pictures:
This is a frequency counter showing the detected clock speed of my unmodified SNES's sound board clock, the ceramic clock that runs at 24.576 MHz. We calculated the jitter at over 120 ppm and the clock rate itself was lower than we anticipated, at 24.5613185 MHz. To put this into context, the original SNES when released had brand new ceramic clocks but that type of clock is known to degrade over time. In my case, 25 years later the clock has degraded substantially enough to cause problems with repeatability (that's an oversimplification but let's stick with that definition for now). So, with some parts selection help from micro500 and p4plus2 I ordered a 24.576 MHz 30ppm quartz crystal and installed it.
There was only one minor wrinkle - the original ceramic part needed a 3pF capacitor on each leg, but these required 18pF to 22pF capacitors. I only had small capacitors available, so I got.. uh.. creative:
Yes, that's a capacitor soldered on top of another capacitor; the original 3pF caps now have 18pF caps on top of them for a total of 21pF per leg. So, what was the difference?
The new quartz crystal is well within 30 ppm jitter and the clock speed is measured at 24.5772283 MHz. I took a number of measurements and the last two digits ranged from 83 at the lowest to 91 at the highest, but on average were in the 86 range. The lsnes emulator only measures down to the digit 24.577228 so the last portion being slightly different is hopefully of no consequence.
Now, here's the interesting part - if we tell lsnes to use that exact value, runs created with the default value of 24.607104 (the audio frequency of 32040.5*768) no longer sync. In fact, micro500 was able to show that manually changing a movie file to use that new speed causes the same desync we see on a real console, so this is good. The bad news is we've tried the brute force method of trying to get the same RNG result on the console and we have had little success even by removing up to 10 frames or adding 20. Masterjun is actively working on creating a new run with the new timing for us to test with.
Thanks to some outside-of-the-community-but-quickly-being-assimilated help from folks like arkun and the now full-fledged mediamagnet we have some parsing scripts that can take the output of the logic analyzer I now have attached to the console (more on that later with better pictures). I'll try to get a link up in my next post, but I just wanted to state somewhere here that we've done that work.
Finally, we have an NES classic! Here's the reveal video that's fun to watch and shows some emulation differences compared to the original NES:
Link to video
And with that, I'm off to stream more and see if we can't get a syncing run of DKC2 finally.
Sorry for offtop and everything. I know this is tasvideos and all... Guys, it's really cool what you can do with the games trough controllers ports... But maybe it's time to make like insane smb3 romhack or even new game from scratch? Like using raspberry pi or other modern hardware parts instead of usual cartridge boards. And make like huge nes game with a lot of insane levels without reusing graphics, with insane music and all. It would be so cool if you put all that work into new game or romhack instead of making this things trough controller ports. Just my opinion.
Bad english
Moderator, Senior Ambassador, Experienced player
(898)
Joined: 9/14/2008
Posts: 1007
Koh2fds wrote:
But maybe it's time to make like insane smb3 romhack or even new game from scratch?
Well, keep in mind that whatever we do is going to have some element of "Tool-Assisted" and "Speedrun" to it or it wouldn't be a GDQ TAS block. :) We have pitched original games in the past but they were not accepted by the GDQ submission committee. We may pitch Mitch's Kaizo Mario 3 romhack at some point in the future, however. Thanks for the ideas!
It would be so cool if you put all that work into new game or romhack instead of making this things trough controller ports.
The whole idea of TASing is to play an existing unmodified game on an existing unmodified console, using its controller, and achieve superhuman feats by that alone.
Just creating a custom cartridge that contains a custom game is a completely different realm, and has little to nothing to do with tool-assisted speedrunning. It's not really what this is about.
I mean - if those guys put all that work into creating romhack or new game it can be much more exiting than making those crazy tases. We already know that you can feed any program code you want trough controllers ports and then use controller ports like data bus or something. So it become like "what crazy thing we programmed". So basically it's the same thing every time - just different code has been feeded. I don't exactly sure, but when you do this - there is can be like obvious restrictions. Like you can use only console ram and can't switch memory banks or something for nes. If you made a cartridge from raspberry pi or something there is can be like infinite memory, infinite "ways to bank switching" etc. So we can't even imagine how game made like this gonna looks like. I don't know why you stick with TAS when you can make much more if drop that category and move to romhacking / game creation.
edit: And after you make a game it can be tased or speedrunned too.
Moderator, Senior Ambassador, Experienced player
(898)
Joined: 9/14/2008
Posts: 1007
Koh1fds wrote:
We already know that you can feed any program code you want trough controllers ports and then use controller ports like data bus or something. So it become like "what crazy thing we programmed". So basically it's the same thing every time - just different code has been feeded.
I don't argue this at all, and it's something we've discussed a few times. It's one of the reasons why I submitted Super Scribblenauts so many times, because I'd like to do something that isn't just taking over the game or the console more often. However, taking over games in unique ways is sort of what makes the TASBot block so popular, and part of that is finding new and interesting ways to present and integrate the payloads we've developed. We may not always have an ACE at SGDQ events (which are focused more on normal TAS playback and lower key aspects) but we'll generally have something like that for AGDQ events because it's what the audience wants to see.
Koh1fds wrote:
If you made a cartridge from raspberry pi or something there is can be like infinite memory, infinite "ways to bank switching" etc. So we can't even imagine how game made like this gonna looks like.
I think I finally realize why you're on such a different page than the rest of us. You literally want us to make our own hardware, or literally, physically make some abhorrent cartridge that breaks all the rules. Here's the deal, though - we only modify the consoles as much as is absolutely necessary, and only when necessary. Brain Age required some invasive cabling to integrate with the touchscreen, but we in no way interfered with how the cartridge communicated with the board. Now, accessing the cartridge pins is something we may do if pressed or if we have a really good reason to, but the whole point is how far we are pushing the console with such a narrow window. Don't expect us to ever try to create a new console, because audience feedback to that would be extremely poor.
Koh1fds wrote:
I don't know why you stick with TAS when you can make much more if drop that category and move to romhacking / game creation.
Because that's not what this is about. If you want to see what's possible on an SNES, have a look at the Pouet.net Demoscene SNES page under Productions. We *do* have elements of what we show off that are this complex by virtue of having complete control over the SNES, but the method we use to get to that point, the journey of getting there via only exposed ports or just the controller ports, is what gives what we do its uniqueness and what puts us on the map with places like Ars Technica.
Koh1fds wrote:
And after you make a game it can be tased or speedrunned too.
We've done this before by putting SMB on top of SMW and *then* TAS'ing it (thanks, Nach!), but it was emulated so well that the pesky 21 frame rule gave us trouble. :)
I do appreciate your feedback, but in general I think you're looking at this from the wrong angle. We do everything we do to entertain a nostalgic audience that loves seeing the games they grew up with get destroyed. If we took that part out of the process it just wouldn't be the same thing. Still, some of your ideas do have merit, so keep your eyes peeled at future GDQ events.
I think I finally realize why you're on such a different page than the rest of us. You literally want us to make our own hardware, or literally, physically make some abhorrent cartridge that breaks all the rules.
Yes. I would like to see new cartridge hardware based on modern parts and new game made specially for this hardware that uses everything that hardware can give. Without console modifications. It have nothing to do with tases or agdq. I post it right here, because this is only place i know where you can read this.
dwangoAC wrote:
We do everything we do to entertain a nostalgic audience that loves seeing the games they grew up with get destroyed.
But peoples never see something like new game based on new powerfull cartridge hardware. So they don't know if they like it or not.
The thing is: what would be the point?
TASBot is fun because it is bound by the original, unmodified hardware and software, which everyone already knows. And it only uses the same input ports that you would playing the game, and does a lot of crazy stuff with it.
When you make your own hardware and software this is lost. No one will be familiar with; no one will know its capabilities, and no one will know whether that thing that just happened is
a glitch or something intentionally placed there to look like a glitch. Heck, it might not even be a game being played, but a game-like demo.
But why stop there; why bother with making it a SNES cart at all? Just make a console-in-a-chip and ignore the SNES entirely; since you are making a monster cartridge that can do things way beyond the capabilities of the SNES, the SNES is worthless baggage that will add nothing.
I don't think it would be well received on GDQ events, and I don't think it would even be accepted. But even worse, people would (justifiably) feel it was cheating, from the ROM hackers making it to the audience. As a Genesis ROM hacker myself, I would refuse to do such a thing because half the fun is figuring out how to do something within the confines of the stock hardware, and I imagine SNES hackers feel the same way.
There is like Battle Kid game that made from scratch. Is it complete lost?
As i understand basically making new games it's lost.
Every nes game that uses memory mappers are bad because "it do things that nes can't do"? So basically on NES you can play only like Super Mario Bros and Pinball? How about Star fox and Doom for SNES?
edit: i can't see a difference between putting MMC3 mapper in the cart that maps like 512 kb and Rasberry pi that maps 512 mb of rom for NES
I think I finally realize why you're on such a different page than the rest of us. You literally want us to make our own hardware, or literally, physically make some abhorrent cartridge that breaks all the rules.
Yes. I would like to see new cartridge hardware based on modern parts and new game made specially for this hardware that uses everything that hardware can give. Without console modifications. It have nothing to do with tases or agdq. I post it right here, because this is only place i know where you can read this.
How is dwangoAC even related to romhacking or homebrew? If your demand doesn't relate to TASing or GDQ, then it clearly isn't supposed to be in this thread.
Koh1fds wrote:
dwangoAC wrote:
We do everything we do to entertain a nostalgic audience that loves seeing the games they grew up with get destroyed.
But peoples never see something like new game based on new powerfull cartridge hardware. So they don't know if they like it or not.
It's because using outdated hardware with limitations that look stupid nowadays is a bad idea. It's way more effective to pick an existing engine. Unless you actually intend to make a retro homebrew, but this doesn't belong here. If you want new hardware, well, errr... there's a fucking ton of that already. And it involves marketing you know? And all that business that scuppered SEGA one day. Please no.
marzojr wrote:
I don't think it would be well received on GDQ events, and I don't think it would even be accepted. But even worse, people would (justifiably) feel it was cheating, from the ROM hackers making it to the audience. As a Genesis ROM hacker myself, I would refuse to do such a thing because half the fun is figuring out how to do something within the confines of the stock hardware, and I imagine SNES hackers feel the same way.
True, the most terrible spoiler that can be invented is usually "Nah, TAS sucks, see, it's just cheating and hacking, any imbecile could do this". We've seen it for too much to get those ignorant people actual reason for such claims within TASing in general, or within GDQs.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
There is like Battle Kid game that made from scratch. Is it complete lost?
As i understand basically making new games it's lost.
If no one at the GDQ ever heard or played the game before it is showcased, yes; no one will have any clue how it is supposed to play normally, so a TAS of it would be meaningless.
Koh1fds wrote:
Every nes game that uses memory mappers are bad because "it do things that nes can't do"? So basically on NES you can play only like Super Mario Bros and Pinball? How about Star fox and Doom for SNES?
Yes, because pieces of hardware created by Nintendo for the NES/SNES at the time they were commercially available is exactly the same as creating a monster cartridge with custom modern hardware. /sarcasm
There is like Battle Kid game that made from scratch. Is it complete lost?
Reductio ad absurdum. Battle Kid is awesome. But compare it to latest NES games. Not even close to their level. Also, go ahead and list all NES homebrews that are as awesome/popular as BK.
Koh1fds wrote:
Every nes game that uses memory mappers are bad because "it do things that nes can't do"? So basically on NES you can play only like Super Mario Bros and Pinball? How about Star fox and Doom for SNES?
Lagrange Point used YM2413. There's literally no limit in expanding the hardware. As I said, it's a goal for serious companies, your own personal private custom console will be completely lost among those that are up to today's standards.
Otherwise, go and invent one.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Just want to throw this out there its not really anything planning related or anything of the such but the combined effort of everyone that has been in the twitch chat is an amazing sight to behold.
"Do not meddle in the affairs of Dragons, for you are crunchy and taste good with ketchup!"
Moderator, Senior Ambassador, Experienced player
(898)
Joined: 9/14/2008
Posts: 1007
There is a lot to update on but not enough time for me to do one of my proper update posts with fun images or proper spelling so you'll have to forgive me for this one. There has been a ton of stuff going on, with a fair bit of it happening on http://twitch.tv/dwangoAC and some of the more boring things happening off-screen but still in chats. It's been a bit scattered, so here's a summary:
We had to give up on Donkey Kong in every single regard. We originally pitched it based on some discoveries in Donkey Kong Country and ran into some timing and consistency issues, so we switched to Donkey Kong Country 2 but still had problems, went through the previously mentioned work to replace the SMP clock and determined that we could not get reliable results even with that change in place. Separately, I had borrowed DK and DK Jr. for NES but both console verifications failed (and updating the console verification tests Wiki page has been quite low on my list of priorities). So, we've moved on to LoZ:aLttP instead. It's looking promising albeit still unreliable on occasion. We're hoping we can make that work a little better in the future.
NES Classic is giving micro500 a lot of trouble and he's definitely getting frustrated but he hasn't given up yet. He's had some unfortunate setbacks but there's still hope on this one. I'll let him post an update if he feels up to it.
Tonight (12/13) at 7:30 PM PST I'll be presenting at http://NBLUG.org and streaming to http://twitch.tv/dwangoAC to show the remote setup we've been using for TASBot development. Feel free to swing by! More updates soon.
Aww that's a shame, I was looking forward to seeing what craziness was going to arise from DKC. Maybe with a bigger time horizon it could be possible in the future? Or is it just too impractical?
Good luck with everything else though!
Moderator, Senior Ambassador, Experienced player
(898)
Joined: 9/14/2008
Posts: 1007
Alyosha wrote:
Aww that's a shame, I was looking forward to seeing what craziness was going to arise from DKC. Maybe with a bigger time horizon it could be possible in the future? Or is it just too impractical?
We hope to bring back DKC/DKC2 at a future time.
I presented at NBLUG on the TASBot remote development environment the team has been using. The talk goes through how we can remotely power on a shared development computer using wakeonlan, how we can power on video game consoles, how we can view a live video feed from the console and also show a camera pointed at the visualization boards, and how we can capture controller data using a logic analyzer all within Linux Mint 18:
Link to video
I am happy to announce that I will be personally donating $1,000 USD plus an additional amount during the AGDQ 2017 TAS block on behalf of the team that helped me with TASBot projects over the past several months, especially the GeekPwn project for which we won the Geek Thinking award which is the source of this donation. I'm also happy to say that I have personally covered every expense I have incurred while representing TASVideos in my capacity as Ambassador at TWiT, SGDQ 2016, SVLUG, DEF CON, NBLUG, GeekPwn, SF Bay Area ACM, and NBLUG again. However, I would not turn down help covering AGDQ 2017 expenses which stand to be about $2K.
That's it for tonight, more as we make progress on the remaining items though!
Man... TASBot is officially complete rocket science now!
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
I know it's rather late, but this may be an idea for 2018 I guess?
Post #445289https://www.youtube.com/watch?v=SL_Zuc0tlvo
"Pokémon R/B/Y: Bringing arbitrary code execution to other games"
It even has a proof of concept.
We're only using the NES Classic, the original NES, and SNES with no SGB. We've previously exploited Red and the technique they are employing of cartridge swapping doesn't seem like a good idea (it feels "cheap" to me). Having said that, there might be something here for a future GDQ if done carefully. Thanks for the suggestion!
Imagine a TAS on calculator for sgdq or for the next year Kappa.
This seems to be hard to do something entertainig but there are very good remakes like Portal, Doom, Mario or Pokemon.
Projects:
Genesis Crack Down 1 Player - Work In Progress
Moderator, Senior Ambassador, Experienced player
(898)
Joined: 9/14/2008
Posts: 1007
Back in the AGDQ 2016 postmortem phase I set a deadline for AGDQ 2017 of December 31st. I'm now staring down that deadline and I admit that it's going to be a bit hard to hit it even with the huge team that's now helping including micro500, total, Ilari, p4plus2, arkun, MediaMagnet, Tompa, Ange, fuzyll, Ownasaurus, endrift, Mothrayas, and even several GDQ tech team members at this point. Everyone is working on different things, some of it undercover but a lot of it is stuff I can at least hint at. Here is the current plan, with a few redactions:
The NES Classic is a bit of a mess as it filters out things like pressing left on one frame and then right on the next frame so it's really difficult to get anything super accurate or complex. In general, the NES Classic is *not* very good at all, and it's the kind of thing TASBot would look down on for not being perfect. So, uh, we're going to kill two birds with one stone - Since the NES Classic has 30 games on it and a subset of those games are available in all territories, we'll pick that subset of games or a smaller subset and do a death tour, playing the game long enough to do a fastest death and hear the death music then move on to the next game by pressing the Home button to get back to the game selection menu.
Going this route might sound a bit strange, but it means that if there is a desync, it will "re-synchronize" at the edge of each game. It'll give us an opportunity to quickly cycle through several games with times ranging from 5 seconds to 20 seconds, although more for some really hard to die in games in which case we may have to skip them. The whole thing will take between 7 to 10 minutes. The alternative is to pick some other arbitrary short goal or use a simple game like Galaga to do a complete TAS with very conservative button presses of which would not be very interesting. I personally think the fastest death tour seems more interesting but we'll have to see how it plays out and I'm open to suggestions.
Following that, we'll show SMB3 total control and the GeekPwn payload but with AGDQ graphics and different NSF music - the amazing level 1-1 takeover is too good to pass up and most people haven't seen it. We'll transition away from that console and start playing [redacted] which is when things will really get interesting. Fortunately, [redacted] is a very quick hop into complete control and it has the advantage of [redacted] which means we can easily show [redacted]. At one point, we'll allow Twitch Chat to [redacted] which could get kind of crazy. It's going to look like utter randomness because it's Twitch chat but it should still be fun.
From there we'll move to LoZ:LttP which is just long enough for some decent commentary, as in, there's some actual gameplay involved (about 3 minutes), and during that time there's all kinds of crazy things going on thanks to work from Tompa. At the end, we'll trigger a glitch at the Sanctuary which will give us total control, and we'll use it to trigger the mirror effect used to change between light and dark worlds. Instead of transitioning to the dark world, we'll instead transition to [redacted] and show [redacted] which will definitely be the highlight.
I'd love to say more, but giving away the payloads would be a horrible spoiler. If you want to know more and you want to volunteer, contact me - the best way to reach me is on #tasbot in Freenode IRC, accessible easily at http://chat.TASBot.net as needed. There's still plenty of work left to do, especially if you can help out with some Linux toolchain needs involving data transformation. With the current team we'll be able to flesh out a rough pass by the December 31st deadline but we'll be polishing until my hard, hard deadline of Friday, January 13th at 9:00 PM EST, the night before we present.
In other news, I've been informed by GeekPwn that bank delays and new polices by the Chinese government have held up the award payment. I'll give the $1k I committed to even if the award is not paid on time, but I could *really* use help with the $2k it will cost us to get to AGDQ 2017 as we still have $1,880 to go to cover those expenses. (Note that I've personally covered all expenses for the many previous events this year and I won't be asking for help with those costs.) As a silver lining, I have a commitment from my employer (Ciena) to match up to $500 of the AGDQ 2017 donation, so the effective donation I'll be making on behalf of TASVideos and the team supporting these projects will be at least $1,500 going to the Prevent Cancer Foundation. At any rate, if you can help with a donation either via Bitcoin at1CknXDyb5jTG1Nw1hs437r6fnz7npBSEfA or through this PayPal donation link I will gladly give you any game here as a token of my appreciation.
More updates in a few days, possibly even including an update to the very neglected first post in this thread. Thanks as always for the support, this is a great community to represent!
There is the possibility of a monetary reward but it is unclear how large it will ultimately be. I discussed with everyone who helped out and we have agreed to withold a portion to cover actual equipment costs only and donate the remainder of the reward during AGDQ 2017. Expect a healthy donation during the TASBot block if all goes well
I attended the annual Day of the Devs event and spoke briefly with Tim Schafer about collaborating on something. More on that for a future GDQ but remember you heard it here first. This isn't Tim but it's still a fun picture:
Last but not least, if you are in the bay area, I will be presenting at the SF Bay Area chapter of the ACM (Association for Computing Machinery) on 2016-11-16. The talk will be a fusion of what I presented at DEF CON 24 and what I presented at GeekPwn.
Anyway, lots going on, more to come - please let us know if you can spare an NES Classic and expect an update from me in a week or two. Thanks for the support!
This is a frequency counter showing the detected clock speed of my unmodified SNES's sound board clock, the ceramic clock that runs at 24.576 MHz. We calculated the jitter at over 120 ppm and the clock rate itself was lower than we anticipated, at 24.5613185 MHz. To put this into context, the original SNES when released had brand new ceramic clocks but that type of clock is known to degrade over time. In my case, 25 years later the clock has degraded substantially enough to cause problems with repeatability (that's an oversimplification but let's stick with that definition for now). So, with some parts selection help from micro500 and p4plus2 I ordered a 24.576 MHz 30ppm quartz crystal and installed it.
There was only one minor wrinkle - the original ceramic part needed a 3pF capacitor on each leg, but these required 18pF to 22pF capacitors. I only had small capacitors available, so I got.. uh.. creative:
Yes, that's a capacitor soldered on top of another capacitor; the original 3pF caps now have 18pF caps on top of them for a total of 21pF per leg. So, what was the difference?
The new quartz crystal is well within 30 ppm jitter and the clock speed is measured at 24.5772283 MHz. I took a number of measurements and the last two digits ranged from 83 at the lowest to 91 at the highest, but on average were in the 86 range. The lsnes emulator only measures down to the digit 24.577228 so the last portion being slightly different is hopefully of no consequence.
Now, here's the interesting part - if we tell lsnes to use that exact value, runs created with the default value of 24.607104 (the audio frequency of 32040.5*768) no longer sync. In fact, micro500 was able to show that manually changing a movie file to use that new speed causes the same desync we see on a real console, so this is good. The bad news is we've tried the brute force method of trying to get the same RNG result on the console and we have had little success even by removing up to 10 frames or adding 20. Masterjun is actively working on creating a new run with the new timing for us to test with.
Thanks to some outside-of-the-community-but-quickly-being-assimilated help from folks like arkun and the now full-fledged mediamagnet we have some parsing scripts that can take the output of the logic analyzer I now have attached to the console (more on that later with better pictures). I'll try to get a link up in my next post, but I just wanted to state somewhere here that we've done that work.
Finally, we have an NES classic! Here's the reveal video that's fun to watch and shows some emulation differences compared to the original NES:
Link to video
And with that, I'm off to stream more and see if we can't get a syncing run of DKC2 finally.
