Archive: 7 posts

---

When doing bit compression Bit compression is when you compress multiple ON/OFF signals using analogue in a binary decrease and detect the same on the other side (EX: 50%, 25%, 12.5%, 6.25% with adding 50% and 6.25%, detecting the same number on the other side of a decompression system would output 1 and 4 in one wire) you are limited to 24 bits of information, when you use normal analogue values to compress an on/off signal because after 24 digits, entropy in digit values ensure So how do you get around this barrier? Well the thing is, is that you can't BUT there is a way to slightly bend the limit, by a factor of two digits The first digit you can add onto the limit of 24 is the digit of (+/-) Discovered by Seku (http://www.lbpcentral.com/forums/member.php?23190-Seku) So how to incorporate this into the compression, is to first make a % smaller then the rest (don't worry about entropy of the value, that % is just to piggy back onto, I'll explain why later) so just take your smallest percent and lower it further. Next make it so that activates when you activate the bit compression signal for +/- bit then that % activates AS LONG AS any of the other 24 aren't active WHY? Because you cant have -/+0 Then what you must do is just change the sign of the analogue value going into the other end and make some way to detect that change on the other side. No this wont change accuracy because it only adds the other percent in the case of no percent being existent at the moment of +/- bit encoding, and will keep the same values as when any others are activated at the same time. Next way (discovered by me, but I don't doubt that Seku (http://www.lbpcentral.com/forums/member.php?23190-Seku) figured this one out as well) So the only thing you have to ask yourself now is "we took the actual analogue value aspect from the information carrying capacities of the compression of numbers, and now we took the analogue sign, what's left?" and I would say The wire dimness So you just put in digital manipulation to the output and detect the bit of digital manipulation with just the input of a microchip and a battery on it, simple as that Seku (http://www.lbpcentral.com/forums/member.php?23190-Seku) has a great digital manipulation set in his Logic Kit (http://lbp.me/v/jy7b2e) If any parts of this is unclear, then please ask me to clarify, I'm not good at explaining things, also feel free to ask me anything you would like, I'm bored ATM and I don't know what tutorial to do next

2014-04-08 01:57:00 Author: amiel445566 Posts: 664

---

Could you try to reword it in a less technical way? It's a shame that the only people who will be able to understand this are the people who probably already know how to do it. And that defeats the purpose of a tutorial.

2014-04-11 00:16:00 Author: AmazingKittyCat Posts: 204

---

I could try to use pictures and diagrams, when I get on LBP ill take some pictures of wiring and diagrams, and try to clear up some terms sorry about the ambiguity of definition AmazingKittyCat

2014-04-11 07:13:00 Author: amiel445566 Posts: 664

---

Could you try to reword it in a less technical way? It's a shame that the only people who will be able to understand this are the people who probably already know how to do it. And that defeats the purpose of a tutorial. From what 445566 is saying the analogue signals in LBP have 24 bits of precision and you can get a couple of extra, conditional bits using the signed bit and the 'wire dimness' i.e. the signal as a percentage. My question is if all the analogue encoding and decoding circuitry is going to be cheaper component-count wise than just 24 counters with some muxing.

2014-04-12 04:41:00 Author: Ayneh Posts: 2454

---

(Couldn't get on my ps3 yet, but I'll try to explain my best without pictures) So what I mean is that, if you were to compress a signal by default you would be faced with inaccuracies after 24 bits per wire, unless you do certain things to overcome that limit (see above ) But explaining compression is hard, I'll do my best to show you what compression in LBP is generally Example: lets say you have 4 possible options (The actual limit is 24 without "conditional bits") but let's say you have 4 possible, and you activate 1 and 3, then all of that information could be kept on one wire, and unpacked to show 1 and 3 were chosen, but all of that information would be kept on one wire, instead of 4 separate wires How it works generally:Input one could be 50%, then the next one would be 25% (half) then 12.5% (half once more) then you detect the same things on the other end (activating first 2, would = 75%, then on the other end you would get 50%?=True so -50% then next you would do remainder of before which is 25% so 25%=True so -25% but then you get 12.5%=false so -0%, then the outputs that are true are 1 and 2 out of 3, like previously turned on, and the limit before it starts activating random ones due to error is 24 digits of numbers (12.5 is 3 digits back, and 6.25 the next one is interpreted as 4, then 5,6,7,8, and so on, and once it reaches 24 it starts to lose value) and this tutorial explains how to get past that limit (slightly) I hoped this helped without pics, still trying to get pics though, haven't been able to access ps3 yet

2014-04-13 04:05:00 Author: amiel445566 Posts: 664

---

Here's a diagram (http://ib.lbp.me/img/ft/28bd986e20437c27560407f66283a75a1ef1ffce.jpg) I put together to explain how it works while I got on LBP Hope it helps

2014-04-13 08:57:00 Author: amiel445566 Posts: 664

---

My question is if all the analogue encoding and decoding circuitry is going to be cheaper component-count wise than just 24 counters with some muxing. Just to clarify; you could still use counters / muxing with this kind of system to perform some kind of function since this system, in itself, does not really do anything beyond converting a signal between a binary or analog form, it's just a transformation of the information. The primary uses for it are to use in conjunction with the memoriser tool since it allows you to efficiently save a large volume of data in very few memoriser 'slots' which is beneficial to anybody playing your level(s) which utilise a memoriser. Some other applications of this can be found in using the least significant bits as some kind of 'identifier' perhaps to identify which of the 4 possible players in a level a signal may be referring to (since the inherent signal ownership within the game remains largely intangible to us with our current logic tools). This works because the least significant 2 or 3 bits have magnitudes equal to or downwards of (1/2)^23 or (1/2)^22 [respectively], meaning that for a large amount of applications involving things like health or currency systems which use signals of magnitudes mainly upwards of 0.001, the subtraction or addition of [(1/2)^22 + (1/2)^23 + (1/2)^24] won't make a visible difference. [although it's actually more thermo efficient to use the MOST significant bits in this way and scale the other bits' analog values up when displaying the health or currency value but it's much easier to explain the concept as above] Similarly, the bits involving the sign (+/-) or the actual digital state of the signal make literally zero difference for most numerical applications since things like health and currency are usually considered only by magnitude (players don't usually have negative health). As far as the thermo is concerned, the encoder is extremely simple and takes very little thermo, the decoder is somewhat thermo-heavier (but still relatively light) as you generally need to use 24 'inequality' or 'comparison' chips plus a few extra components to determine the sign and digital state of the signal. But for the primary application here (memorisers), for what you lose in thermo efficiency inside your level, you gain a large amount of efficiency in the way that you utilise every player's memoriser data who decides to play your level(s).

2014-04-17 07:03:00 Author: Seku Posts: 65

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.