Archive: 21 posts

---

They exist. While experimenting with making a score display, I stumbled across a small discovery... I was curious what would happen when it passed the score I set on the sensor. What happened next was unexpected, and after some experimentation, I discovered that I was looking at a signal stronger than 100%. With some further experimentation, I also found that many devices can deal with this signal in the expected way. Timers can tick FASTER than their setting. Lighted materials can be brighter than their settings. Directional combiners happily add/subtract the values. I didn't check NOT or other logic gates. How is this useful? So far I've just come up with 2 ideas. Score sensor for scores > 100,000 The score sensor is limited to 100,000, but the analog output always returns actual score / target score, even if that is >100%. Since timer division works, use a timer to divide the value by the number of times the score you want. Theoretically I could also extend my score reader to 7 digits of precision, although it would be a pain. Timers with an interval finer grained than 0.1s This one is a little less obvious and a little bit messier to set up. But to get a timer running at, say, 0.13333s, at the start of the level, give the player 30 points. Use a score sensor set to 10 points, and use an OR gate to sample the value. Cut off the input from the score sensor (Using an AND gate), and take the points away from the player if desired. Pass the OR gate output into a timer's speed input. Set the timer to 0.4s duration, and it will count out 0.13s whenever reset. Neither of those are spectacularly useful though. So, anyone care to try to come up with better ideas what to do with a signal >100%? Anyone care to try to figure out where all the signal works, and if there's any other way to get a supersaturated signal like that? Edit: Signal probe to display large values available here (http://lbp.me/v/x6rgs9).

2011-02-23 10:51:00 Author: Tygers Posts: 114

---

Whoa, that is very intriguing. You say the directional combiners can add and subtract these values without problem. Do you mean that +120 - 50 would be 70, thus allowing +/- of values greater than 100 as inputs. I assume that the output must lie within +/- 100, but being able to use these previously unusable values would be interesting. I'm trying to think of a way to use this for a damage system, allowing for more than '0.0333`` per pulse, and thus opening up the variability of these systems significantly. Some things to test: Does it travel through tags? Effect on sequencer? How these values influence sackbot movements?

2011-02-23 11:20:00 Author: SSTAGG1 Posts: 1136

---

I don't know that combiners have the limitation that the output must be 100% or less; they are just doing math after all. It's only timers (Dividers) that have that limitation since the result isn't directly the division, it's the percent the timer gets filled. It won't be useful for allowing anything faster than 30hz though. That is an inherent limit in the underlying simulation. However, it will allow stronger damage than before potentially.

2011-02-23 18:49:00 Author: Tygers Posts: 114

---

In terms of combiners, a very practical use would be overflow handling in analogue addition.... Being able to offset the system by 1500% or so and then subtract from that would be convenient... Assuming, as you say, that combiners don't put limits on the signals... The do put a wraparound on the negative limit of -100%, so who knows. More importantly, unless we can find another way to generate such signals, or a method to store them in create mode, it's gonna e bloody difficult to use them in any real sense

2011-02-23 19:02:00 Author: rtm223 Posts: 6497

---

I ran across that last night too, but don't remember how I created the condition to get >100%. Can you describe what you did? It is strange to look at a probe reading 120.

2011-02-23 19:07:00 Author: tdarb Posts: 689

---

Couldn't you trigger a score giver whenever you end up with more than 100%, and use a score sensor and store it that way? I dunno I guess that'd be hard since you have to express your number as a percentage of the score sensor, but that at least seems like the avenue you'd want to look down, right? The only problem of course would be that the player could see your level doing its math since their score is jumping around for no reason.

2011-02-23 19:10:00 Author: Speed Racer Posts: 156

---

Couldn't you trigger a score giver whenever you end up with more than 100%, and use a score sensor and store it that way? I dunno I guess that'd be hard since you have to express your number as a percentage of the score sensor, but that at least seems like the avenue you'd want to look down, right? The only problem of course would be that the player could see your level doing its math since their score is jumping around for no reason. You know, I haven't done any experimentation along these lines - but I wonder if you could tie this specifically to player 4's score... 'Cause sensors provide player identification... If you captured player 4's signal in create mode and stored it in a D-latch, and used that signal as part of the input to a score giver - would the score giver still give the points to player 4 even if he's not present? If it does, and if player 4's score in that case is still tallied and sensed, but not displayed, then maybe it'd be possible to do this stuff in a way that's transparent to the players - as long as there's not four of them. If you can capture the >100% signal in create mode, you should be able to keep it stored. (For instance, feed your signal into one terminal of an OR gate, feed the output of the OR gate back into the other input - the OR will store the greatest analog magnitude it receives...) But then, will the score sensor even work in create mode? And I wonder if this is a "bug" or a "feature"? I look forward to playing around with it when I get the chance... Thanks for sharing!

2011-02-23 19:35:00 Author: tetsujin Posts: 187

---

In terms of combiners, a very practical use would be overflow handling in analogue addition.... Being able to offset the system by 1500% or so and then subtract from that would be convenient... Assuming, as you say, that combiners don't put limits on the signals... The do put a wraparound on the negative limit of -100%, so who knows. More importantly, unless we can find another way to generate such signals, or a method to store them in create mode, it's gonna e bloody difficult to use them in any real sense Yeah, the lack of being able to produce such signals in create mode is why I don't have more information on what works and what doesn't. That's a good idea on being able to use it as a basis for subtraction. But given the difficulty of using it, and the fact that you can just scale your values down and get the same math, it's still of questionable usefulness there. I ran across that last night too, but don't remember how I created the condition to get >100%. Can you describe what you did? It is strange to look at a probe reading 120. I set a score giver to give 40 points on level start, and a score sensor to sense 10 points. The resulting signal is 400%. I think I need to add a decimal point to my probe, so it can display numbers > 1.0 (100%).

2011-02-23 19:38:00 Author: Tygers Posts: 114

---

@tetsujin: score sensors don't activate in create mode... Yeah, the lack of being able to produce such signals in create mode is why I don't have more information on what works and what doesn't. That's a good idea on being able to use it as a basis for subtraction. But given the difficulty of using it, and the fact that you can just scale your values down and get the same math, it's still of questionable usefulness there. Well we don't know how hard ti is to use - it might be moderately useful and if we can produce it in some way in create mode and it has some use in other components like ANDs / ORs etc. then we should be able to store a constant in an OR gate pretty easily.

2011-02-23 19:53:00 Author: rtm223 Posts: 6497

---

ohhh, right, I forgot that you can't generate scores in create mode. That definitely limits what you could do with it, but even so, couldn't you have multiple score givers that get triggered by some mechanism that checks for overflow, with each giver correlating to a different analog value above 100? It's messy, and definitely not as flexible as it would be if you could simply create and store high analog values at will, but it still might be something? I dunno I am nowhere near as good with this stuff as some of the folks in this thread considering that these ideas are still barely a week old for me.

2011-02-23 21:10:00 Author: Speed Racer Posts: 156

---

Well for getting them in play mode, you just need at most two different values larger than anything you'll need. The rest of the values you can get through division and subtraction.

2011-02-23 21:20:00 Author: Tygers Posts: 114

---

You could just hit a score giver in play mode as the player enters the level to give points and use some score sensors to detect some massive analogue values and store them in an OR gate, then remove the scores and isolate all of that initial circuitry, so the scoring system can continue as normal. Your logic would only work in play mode, so if we can find a really good reason for this then it'll be worthwhile.....

2011-02-23 21:21:00 Author: rtm223 Posts: 6497

---

Right, I don't see this as being useful unless someone comes up with a "killer app". But nearly everything I can think of that this could be used for, you can do other ways easier. Only the two I gave can I think of no other way to handle; even rtm's suggestion of using it to not have overflow, could be handled by scaling the values down to begin with. (Though that may be a case where it is simpler if you have a large value to begin with. It would significantly simplify a signed analog addition circuit.) If anyone wants to play with this, I'm going to modify my probe to handle numbers above 100% properly tonight, and I'll publish that along with a few chips to help people generate and use these in play mode.

2011-02-23 21:54:00 Author: Tygers Posts: 114

---

Can someone confirm if the signal passes through tags. It'd be interesting to be able to adjust a player's weapon, or even an enemies weapon, to detect and use a wireless value >100. Though there isn't really any point, since signals could just be divided to reach the right ratios for, in my case, a health system.

2011-02-23 21:54:00 Author: SSTAGG1 Posts: 1136

---

Can someone confirm if the signal passes through tags. Yes, it does. I have two 10% light cubes being driven by a score sensor (one directly and one via tag-and-sensor pair) and they appeared to be at the same light level (and they've both apparently gotten up to their maximum light level, well beyond the barely-visible 10%). I was thinking about how else it might be possible to create an out-of-bounds signal like this... One possibility that comes to mind would be to build a USB game controller that says its joystick output will be in the range [-63,63] but then output a higher value. It's been a while since I've built a USB device but I think you can do things like that... Then the relevant question would be whether either the PS3 or LBP would filter the signal to be in the proper range before it got into the Controllinator... One thing I wonder about this feature is whether it might make it possible to break some of the usual rules of the game - go faster than the top speed with a mover or whatever...

2011-02-24 05:25:00 Author: tetsujin Posts: 187

---

I can't recreate this, so take it for what it's worth. Last night I was getting pulses of 120% in create mode. At the time I thought it was odd, but thought "nah, that can't be right". Now, I'm not so sure. I wish I had saved what i was doing before I tried to "fix" it.

2011-02-24 08:11:00 Author: tdarb Posts: 689

---

I'm sure it's going to be stumbled across eventually. We're only 1 month in (well, most of us are), so I'm sure we're going to get some more 'glitches' in the coming months. I've tried most other sensors, and nothing else seems capable of actually going beyond its measurement. Mind me asking what it was you were making that you thought created a >100 signal.

2011-02-24 09:28:00 Author: SSTAGG1 Posts: 1136

---

So I discovered a bug with timers while updating my logic probe to handle large numbers; if I set a timer to divide by 30, it divides by 29 instead. This is true of any multiple of 5 or 7. (Dividing by 15 divides by 14, dividing by 135 divides by 134 instead.) Any other value you can divide by with a timer works correctly. So I had to work around it and divide by 21 (Which divides by 20), 1701 (Which divides by 1700), 8001 (which divides by 8000) and 50001 (which divides by 50000). Of course, if they ever fix the bug it will make the precision worse. However, division is already pretty low precision on LBP2. For numbers 100% or less, my logic probe can accurately produce 7 digits of precision of the internal signal representation. For numbers over 100%, it varies drastically, sometimes as low as 2 or 3 digits of precision. Anyway, I published a level with 2 versions of my logic probe - one with 7 digits, and one with 12 or so (To be able to show the true value of alleged 0 numbers). Both can handle numbers up to 1,000,000% (Or 10,000.0 in floating point terms) with the caveat that large numbers aren't precisely displayed. The level also includes a chip which will capture a 500% signal as the level starts. It is actually barely noticeable that the score sounds as the level starts, and you never see the numbers. Get all three as prize bubbles in http://lbp.me/v/x6rgs9.

2011-02-24 14:48:00 Author: Tygers Posts: 114

---

I was messing with overflow handling in an adder. I don't remember exactly where it was that i was getting the pulse, but i believe at that time i was trying to sample the output using an OR gate, and a combination of XORs and ANDS to combine that with the overflow when it was non zero. I went through so many variations trying to get that to work that i can't be sure though. I don't even remember what element of the circuit i had the probe hooked to at the time. I did manage, after many attempts, to pause the create mode with the 120 displayed so I know it wasn't just two rapid flashes that I was seeing as 120. It could be some weird condition that was confusing the probe. I don't know.

2011-02-24 18:32:00 Author: tdarb Posts: 689

---

I did manage, after many attempts, to pause the create mode with the 120 displayed so I know it wasn't just two rapid flashes that I was seeing as 120. It could be some weird condition that was confusing the probe. I don't know. It depends on the probe, how it works. It could be that when the probe in question jumps from some other signal to 100%, the "1" is displayed before the other two digits are cleared out...

2011-02-24 19:41:00 Author: tetsujin Posts: 187

---

Dude, your level has freaked me out! Up is down! Black is white! Analog values are greater than 1.0! How are we supposed to go back to the way things were?? http://www.lbpcentral.com/forums/attachment.php?attachmentid=31561&d=1299011176

2011-03-01 20:27:00 Author: tetsujin Posts: 187

Posts: 1077139    Threads: 69970    Members: 9661    Archive-Date: 2019-01-19

Aus dem Archiv wurden alle persönlichen Daten wie Name, Anschrift, Email etc. - aber auch sämtliche Inhalte wie z.B. persönliche Nachrichten - entfernt.
Die Nutzung dieser Webseite erfolgt ohne Speicherung personenbezogener Daten. Es werden keinerlei Cookies, Logs, 3rd-Party-Plugins etc. verwendet.