Billy Mitchell's Direct Feed Recordings are MAME Generated

[Sqrlmonger]

I wanted to post this here because it contains a lot of good information from the TG dispute thread and I know some people here might miss it:

Note: All of the content below is from Asterra on the Twin Galaxies forum, credit goes to them.

Link to the first post.

Since this matter hasn't yet been clarified adequately, I feel I should step in. Let me begin by reiterating my previous conclusion (which should be very easy for Jace to test) that the recently-rediscovered Billy 1.05M tape is itself showing only 30fps gameplay, downconverted, for whatever reason, from 60fps (or perhaps MAME was told only to generate a 30fps video) by way of discarding every other frame. At the very least, this is what is shown in the two full-length Youtube videos of said score, which is where the above GIF is sourced. My conclusion about the true framerate of the DK gameplay in Jace's tape stems from the fact that there is absolutely no evidence of the missing frames present in the video he recently provided wherein he scrutinizes the tape frame-by-frame.

Now, with that having been stipulated, here is the short answer to the question about the disparity in board transitions: Donkey Kong runs at 60fps, and it does not conveniently begin board transitions on an even (or odd) frame every single time. If you're only seeing every other frame of gameplay, as is the case with every specimen of the 1.05M score, then sometimes you're seeing frames 1,3,5,7 of a transition, and sometimes you're seeing frames 2,4,6,8.

The proof is in the pudding. Here is the complete breakdown of how barrel transitions appear in the 1.05M score as seen on Youtube:

00:00:37 Single frame, split 75%
00:04:39 Single frame, split 25%
00:10:28 Single frame
00:16:35 Single frame
00:19:04 Single frame
00:24:01 Single frame
00:26:45 Broken beam, split 75%
00:29:03 Broken beam
00:32:01 Broken beam
00:34:43 Broken beam
00:37:07 Broken beam
00:39:53 Broken beam
00:42:10 Broken beam
00:44:16 Broken beam
00:47:24 Broken beam
00:49:52 Single frame
00:52:25 Broken beam, split 3%
00:55:29 Broken beam
00:58:00 Broken beam
01:00:37 Single frame
01:03:22 Single frame
01:05:53 Broken beam
01:08:22 Broken beam, split 10%
01:11:22 Broken beam, split 90%
01:13:57 Single frame
01:16:18 Single frame
01:19:54 Single frame
(death)
01:21:09 Broken beam
01:23:44 Single frame
01:26:21 Single frame
01:29:10 Single frame
01:32:00 Single frame, split 35%
01:34:23 Broken beam
01:37:00 Single frame, split 8%
01:40:17 Single frame
01:42:51 Single frame
01:45:53 Single frame
01:48:17 Broken beam
01:50:58 Single frame
01:54:01 Single frame
01:56:34 Single frame
01:59:17 Single frame
02:02:36 Single frame
02:05:14 Single frame
02:07:44 Single frame
02:11:10 Single frame
02:14:18 Broken beam
02:16:44 Broken beam
02:19:50 Broken beam, split 8%
02:22:38 Single frame
02:25:13 Broken beam
02:28:18 Broken beam
02:30:30 Single frame, doubled frame
02:32:17 Broken beam
02:35:03 Single frame
02:37:50 Single frame, split 85%

"Broken beam" means the second long beam is visibly broken at a specific spot (as seen in the above GIF). "Single frame" means the entire board appears in a single frame. "Split x%" refers to an anomaly of the screen being split vertically with the left portion appearing one frame ahead of schedule -- probably a quirk of the MAME version used, since the vertical split is not straight from top to bottom in every instance.

Let's compare this to a MAME Donkey Kong score that is also presented in 30fps with every other frame discarded:

MAME Donkey Kong World Record Dean Saglio 1,206,800 HD https://www.youtube.com/watch?v=wnhihX0A1_8
---
00:00:15 Broken beam
00:04:25 Broken beam
00:10:25 Single frame
00:16:14 Single frame
00:18:53 Broken beam
00:23:38 Broken beam
00:28:27 Broken beam
00:31:12 Single frame
00:34:17 Broken beam
00:36:46 Single frame
00:40:10 Single frame
00:46:48 Broken beam
00:49:35 Broken beam
00:52:29 Broken beam
00:56:27 Broken beam
00:59:12 Single frame
01:02:55 Broken beam
01:06:36 Broken beam
01:09:39 Single frame
01:13:15 Broken beam
01:17:29 Broken beam

The "broken beam" in this version of MAME has a different appearance (identical to MAME 0.115 as shown in Xelnia's original post, whereas Billy's 1.05M broken beam matches MAME 0.116 - 0.121). As before, some transitions show the broken beam, some don't.

Here is data from a 60fps MAME recording of Donkey Kong. I've taken the time to note the very first frame we see any portion of the barrel stage.

Donkey Kong MAME 60fps Mame32Plus!0.100 https://www.youtube.com/watch?v=63hIsjyxAGA
---
00:00:25:06 Even
00:03:40:37 Odd
00:06:37:02 Even
00:10:03:50 Even
00:14:35:09 Odd
00:18:12:44 Even
00:22:05:25 Odd
00:25:36:35 Odd
00:28:54:53 Odd
00:31:42:00 Even
00:35:17:50 Even

And finally, the 60fps direct feed video recently posted by Chris G.

Chris G direct capture via Nintendo Inverter Board and Composite Out https://www.youtube.com/watch?v=kkjt17XiPIs
---
00:00:22:24 Even
00:01:31:57 Odd
00:03:06:12 Even
00:05:19:04 Even
00:07:52:18 Even
00:08:45:57 Odd
00:11:10:36 Even
00:12:13:14 Even
00:13:13:58 Even
00:13:41:39 Odd
00:15:45:24 Even
00:18:11:22 Even
00:19:16:15 Odd
00:21:13:40 Even
00:22:50:13 Odd
00:24:04:40 Even
00:25:53:29 Odd
00:26:50:00 Even
00:28:21:29 Odd

In short, MAME faithfully reproduces Donkey Kong's unpredictability as to whether board transitions begin on an even or odd frame, and this phenomenon, when combined with the removal of every other frame, generates the discrepancy that has inspired some recent confusion. In every 30fps Youtube video of MAME Donkey Kong, one should be able to find examples of both types of barrel transition: Single frame and multiple-frame.

Link to the second post.

I should here point out that there's no question about whether the gameplay on the 1.05M tape is MAME. Every board transition does one of two things, either a broken beam (broken the same way every single time) or not. The specific way this broken beam appears is not only an identical match for MAME, but absolutely never manifests in any known actual direct feed or indirect video of Donkey Kong, including Chris G's recent 60fps direct. Meanwhile, direct/indirect videos of Donkey Kong perform barrel transitions with unique but 100% reliable visual quirks which never manifest in the 1.05M video. When I was obtaining the exact frames from Chris G's video, this became blatantly apparent. It is likely because all of this is so obvious that the gaming community at large has been able to form a solid conclusion.

Nonetheless, Jace is obligated to give the matter as thorough an examination as can be done, in order to nip precisely these kinds of potential obfuscations in the bud. This means figuring out the nature of Billy's recording, and whatever else comes along, regardless of how conclusive the evidence may already be.

In short, think of the MAME question as having been answered in brief by the DKF's removal of Billy's scores, and answered at length by TG's pending verdict.

Note: All of the above content is from Asterra on the Twin Galaxies forum, credit goes to them.

[Bounty Bob]

Even with a piece of video equipment that cost several thousands of dollars, they STILL cannot  confirm or deny (that's whole the plan in case your wondering ) if the tape loads the graphics like MAME or arcade.

 I don't care if the machine cost $6,000,000,000, from what I've seen, this "VERY EXPENSIVE" equipment is still incapable of confirming MAME loads or Arcade loads therefore the very fine, expensive piece of equipment is totally useless.

 If anything it raises more suspicions because there are NO load states to be observed in any way shape or form.

Jace just showed us the start of the tape so that we could see how accurately the player was able to show the frames. For what its worth, what Jace showed over at TG matched exactly to the youtube video at that same point, so I've no doubt that further analysis will just confirm what we see on the youtube videos.

As obvious as the cheating it might seem to us, they need to investigate properly because it would be pretty bad if they just struck the scores and they somehow turned out to be valid. At least they now have a proper source for one of the 1 million+ games rather than relying on digital conversions which may or may not have lost frames etc.

[homerwannabee]

Finally someone decided to compare an Arcade Donkey Kong Junior transition draw to the MAME Donkey Kong Junior transition draw.   And wouldn't you know it, Billy Mitchell's Donkey Kong Junior transition draw looks more like a MAME draw.  Props for Matthew Felix for taking the time to do this.

https://www.youtube.com/watch?v=QIcE-v7lvUc

[Josephjo]

Finally someone decided to compare an Arcade Donkey Kong Junior transition draw to the MAME Donkey Kong Junior transition draw.   And wouldn't you know it, Billy Mitchell's Donkey Kong Junior transition draw looks more like a MAME draw.  Props for Matthew Felix for taking the time to do this.

https://www.youtube.com/watch?v=QIcE-v7lvUc

Thats down to Robert Childs Jnr's Wonkavision camera set up.

He filmed Mitchell's DK and Jnr scores on the same day - if his DK boards transfered from Arcade to Mame then his Jnr boards were bound to.

 

[maximumsteve]

I want to thank everyone who came out the other night and contributed, with your admission and attendance, to the cause of the Marjory Stoneman Douglas GO Fund. Great atmosphere and a lot of fun was had.
We give a special thanks to "Apollo Legend" Youtube footage man for taking the time to come to the event, support the cause, and watch Billy Mitchell play Donkey Kong throughout the course of the night and get to meet myself, Robert Childs, Carlos Pineiro and discuss in a friendly and objective manner, many of the questions and aspects to the DK game footage that has been under scrutiny. Bill wanted me to relay that it was a pleasure for Bill to introduce Apollo to his family, show Apollo the DK machine, the inner workings in the back of the machine and field questions from him and his friend regarding the controversy, along with Robert and Carlos chiming in with their answers and information to his questions.
It is greatly appreciated that Apollo took the time to come out and experience our event and get to know each other and have fun, which is the true spirit of what gaming and community should be. He also came dressed like Bill, which was hilarious!! and here are some pics of Bill meeting Bill, lol. And Robert.

[homerwannabee]

Very glad Apollo Legend showed up.  It's funny that you never advertised that Billy was going to go for a score.  From now on I encourage anyone from the Donkey Kong community show up to any public event Billy Mitchell shows up to.  Apparently he likes to surprise people with announced Donkey Kong attempts.  So people from the community should come to these events just for the chance to see him play a game of Donkey Kong.

[Realzach2001]

So, how did Billy do on DK??

[Raven]

I transcribed the scores from Billy Mitchell's 1,047,200 and 1,050,200 point Donkey Kong games, board by board, level by level. Here are the basic stats from each transcription.

1,047,200 Points Game (+1 Life @ 7,000 Points):

Code: [Select]

Board Average Total
-------------------------
Barrel 10,500 583,200
Rivet 7,700 168,700
Pie 7,700 152,900
Spring 7,200 142,400

Code: [Select]

Level Average Total
-------------------------
1 7,700 15,400
2 8,000 23,900
3 9,100 36,400
4 8,900 44,100
5 9,300 55,700
6 9,800 58,400
7 9,600 57,100
8 9,500 56,500
9 8,200 48,900
10 8,900 53,000
11 9,000 62,500
12 8,900 53,200
13 9,600 57,200
14 9,800 58,600
15 9,000 53,500
16 9,800 58,600
17 8,200 49,000
18 8,000 48,000
19 7,300 50,700
20 8,200 48,700
21 9,700 57,800
1,050,200 Points Game (+1 Life @ 7,000 Points):

Code: [Select]

Board Average Total
-------------------------
Barrel 11,000 601,100
Rivet 7,800 156,000
Pie 8,600 162,400
Spring 6,900 130,700

Code: [Select]

Level Average Total
-------------------------
1 8,600 17,100
2 9,300 27,900
3 9,400 37,500
4 9,700 48,500
5 9,000 54,000
6 9,800 58,700
7 9,600 57,200
8 9,300 55,700
9 9,800 58,400
10 9,000 54,000
11 9,500 56,900
12 9,000 62,400
13 9,800 58,300
14 9,000 53,500
15 9,100 54,500
16 9,300 55,600
17 8,900 53,200
18 8,700 60,800
19 9,500 56,700
20 9,600 57,500
21 11,800 11,800

[LMDAVE]

So, how did Billy do on DK??

I saw the claim from Joel, "It was 760k+ change, over a 1.5M+ change pace." (Which is probably 1.05M is what he meant)

and another "Pretty decent from what I've heard. Was on a 1.1M pace ending at 700+K. "

[danman123456]

I transcribed the scores from Billy Mitchell's 1,047,200 and 1,050,200 point Donkey Kong games, board by board, level by level. Here are the basic stats from each transcription.

1,047,200 Points Game (+1 Life @ 7,000 Points):

Code: [Select]

Board Average Total
-------------------------
Barrel 10,500 583,200
Rivet 7,700 168,700
Pie 7,700 152,900
Spring 7,200 142,400

Code: [Select]

Level Average Total
-------------------------
1 7,700 15,400
2 8,000 23,900
3 9,100 36,400
4 8,900 44,100
5 9,300 55,700
6 9,800 58,400
7 9,600 57,100
8 9,500 56,500
9 8,200 48,900
10 8,900 53,000
11 9,000 62,500
12 8,900 53,200
13 9,600 57,200
14 9,800 58,600
15 9,000 53,500
16 9,800 58,600
17 8,200 49,000
18 8,000 48,000
19 7,300 50,700
20 8,200 48,700
21 9,700 57,800
1,050,200 Points Game (+1 Life @ 7,000 Points):

Code: [Select]

Board Average Total
-------------------------
Barrel 11,000 601,100
Rivet 7,800 156,000
Pie 8,600 162,400
Spring 6,900 130,700

Code: [Select]

Level Average Total
-------------------------
1 8,600 17,100
2 9,300 27,900
3 9,400 37,500
4 9,700 48,500
5 9,000 54,000
6 9,800 58,700
7 9,600 57,200
8 9,300 55,700
9 9,800 58,400
10 9,000 54,000
11 9,500 56,900
12 9,000 62,400
13 9,800 58,300
14 9,000 53,500
15 9,100 54,500
16 9,300 55,600
17 8,900 53,200
18 8,700 60,800
19 9,500 56,700
20 9,600 57,500
21 11,800 11,800

Good stuff! This may all be in the Kongtrackr site. It has a lot of info in it check it out here http://kongtrackr.herokuapp.com

Its great to have this here so others can see. Those spring averages are pretty darn good at 7,200 & 6,900? Wow... Mine was only 6,000 for my 1,059 game

[maximumsteve]

There's nothing "funny" about it George Riley. The emphasis of this event became raising money for families devastated by a local tragedy in the community. Bill was greeting people,roaming around and decided to play a few games, starting to practice for Banning next month. In case he happens to go deep on a game, he wants it recorded all the way through ofcourse. I made mention several times that Bill will be in attendance and the address of where this was taking place. The more the merrier on anyone that wants to come and watch Bill play, when I hold an event and he tells me he will be there. We are not hard to find.

[Sqrlmonger]

Doesn't really serve any useful purpose to be that hostile.

Indeed it doesn't, but it's nothing new. Regardless, it certainly bolsters my previous point about apparent hostility in the video game community, where it hardly creates the impression of a "welcoming" environment. And to think, I was briefly considering the possibility of dropping a few grand on a new arcade cabinet.

I'm not clear as to how these comments are any more helpful. 

Perhaps we can all focus on the issue at hand and ignore any slights, perceived or real, going forward?  The rhetoric, justifiable or not, is just not relevant to the topic and as such it really has no place here.

I think the easiest thing to do is to let any off comment speak for itself. Trust that the rest of the community can see such comments for what they are and save your replies for the topic at hand.

Anyways, I'm obviously not a mod so I'll let it be now, I was just hoping I could appeal to better judgements and cooler heads.  We all want the same thing, so lets try to give each other the benefit of the doubt that we would ask for ourselves.

Cheers~

[homerwannabee]

Steve, my point is you could have had a few more people show up if you announced that he was going to play.  He usually just shows up as a representative.  You announce he's going for a score and you get more money for the charity.  That's what I found strange/funny.

[maximumsteve]

I see what you're saying George, but oddly enough down here, on average, people care more about the game selection we have when we do it rather than who is going to be there. He is more of a novelty then a spectacle when he is here because a lot of people that attend have already met him numerous times or know him already through local business, restaurants, etc.. being that it is the area where he actually lives in. It's been somber down here from the Parkland thing the last couple weeks. That incident happened about 15 minutes away from where we are located, but I get your idea!

[Sqrlmonger]

Apologies if this has been covered before but I was just thinking about the IVGHoF presentation of the DK and DK Jr scores.  Specifically, I was considering that the claim coming from Billy and his supporters at the moment seems to be that the MAME artifacts are a result of the combination of:

1) The monitor/displays being used
2) The cameras and settings being used

And that what we see as MAME artifacts are not a result of:

3) The content actually present on the direct feed VHS tapes.

The reason I think this is important, and again, tell me if this has been pointed out already, is that there are multiple videos from multiple people of the IVGHoF event and they show both the DK and DK Jr scores.

Thus,
1) If multiple people with different cameras and settings are all getting the same artifacts it tells us that either the camera is not involved in the production of these artifacts -OR- that a wide range of (probably default) camera settings will recreate the effect with ease.  If this is the case then we already know we need a direct feed recording of DK on VHS and we know what display was used and that the VHS recording is a direct feed. So what could possibly be taking so long to reproduce when we know all three steps of the process?

2) Recently, as I'm sure many of you have seen, there has been demonstrations that the DK Jr footage is also showing MAME artifacts. However, the MAME artifacts in DK Jr do not appear, to my admittedly non-expert eye, to be similar in nature to the artifacts in the DK footage.  This of course makes me wonder, how can the same displays and cameras produce different artifacts?  To my mind, the only explanation is that the difference resides on the tapes themselves.

But perhaps I just don't know what I'm looking at (which I accept is possible).  Open to some thoughts or clarification on any of this.

PS - Am I wrong about the "Billy camp" claiming the "original" tape won't show these effects though? Part of my frustration here is that there is no "official" claims coming from them making it impossible to nail anything down. I can't imagine they think the original tapes show the problem given that they are out there saying the original tapes will vindicate them.  Of course, I suppose it depends on if we are talking about the direct feed tape or the room camera tape (which I doubt exists anyways).

Cheers~