Archive: 6 posts

---

First an explanation, and sorry in advance i will add pictures some time later, i can't get to the console right now. I have rigged up a sensor to input range dependent analogue signal into a sequencer. And rigged the sequencer with batteries in a step sequence end to end in increasing 5% analogue intervals from 0 to 100. I've also inverted the signal output of the sensor. This (with a sensitivity of 5%, obviously) system basically acts as a conduit, converting 'range' into an equivalent analogue output. The process losses a large whack of granularity but it wooks more than smoothley enough for my purposes. I feed this output into an advanced mover on the same object as the senser as 'up/down' input, and along with a look at rotator the object will move towards a tag with an Inverse exponential distribution(?) on the speed at which the tag aproaches the ranges inner boundary. If the source (tag) is moving, say for instance is on an object being controlled with a DCS and an advanced mover of its own, and the object holding the sensor (Object X) is set to a maximum speed faster than that of the speed of the object holding the tag (object Y). Object X will intelligently maintain a range: the point at which the Inverse exponential distribution of the analogue signal inputting into the advanced mover on object X has diminshed to a percentage equivelent to the input required by the mover on object Y to move at maximum speed with 100% input while set to a lower speed. This relationship maintains regardless of speed or orientation. (Obviously to within certain relative thresholds between the two movers paramiters and inputs percentages) I am using it as part of Enemy AI for one of the ships in a twin stick top down shooter i'm working on. On a side note i have also added a similar input to the left/right paramiter of the advanced mover on the follower object (X). Only this time with an up>down cycling timer as analogue input with the feed being periodically inverted via a pair of direction splitters and a pair of toggles on alternate states being flipped in synk by a repeating timer set to an arbitrary value. (is there a cleaner way to do this? i don't like splitting the signal and then feeding it back into a single channel.) The problem: If multiple ships are emitted containing this logic, each ship will quite quickly fall into the exact same possition relative to the movement of object Y. As every object is made entirely of holo. For all intents and purposes, it may aswell be one ship. My request: I need to know if there is a way to feed 'random analogue anomolies' into the signal inputting into either or both of the paramiters of the advanced mover on object X, so eachship will still follow the same basic laws defined by the current logic, yet with 'random' periodic alterations to the signal percentages of the inputs in turn interfeering with the relative positioning of each ship uniquely, seperating them out as they follow ship Y. I have tried a few experiments but the signal seems to feed constantly (amongst other behaviours i am having trouble understanding as of yet) from the source, of which i am trying to intersperse sporadically into the main flow. I have encountered several odd characteristics within this experimentation, that i cannot really go into as i simply cannot draw corelation between (what i know of) the signal flow and the front end behavior. The movers will give wildly unpredictable outputs and the ship will basically act very oddly. I require some food for thought on how to approach this problem. I know its doable, easily, theoretically. I just don't understand some of the fundementals deeply enough as some of you to fully overcome the issue. Any advice or opinions on the best way to aproach this will, as always, be very greatly received and appreciated.

2011-02-16 00:09:00 Author: Epicurean Dreamer Posts: 224

---

Off the top of my head without fully understanding your need (I read it, but it hasn't fully sunk in what you're asking for), use a randomizer, with outputs to a series of microchips bearing various powered batteries, and add it to the analog signal. Depending on the need, you may be better off with quickly switching random choices, or select once/slowly switching. If you want to make the random event rare without increasing the complexity overly much, combine two randomizers, and AND one output from each. Two 4 output randomizers will pick the same output 1/16th of the time. 5 output 1/25th of the time. Or add more, 3 (4 output) will pick the same output 1/64th the time.

2011-02-16 00:45:00 Author: Tygers Posts: 114

---

funnily enough i tried exactly what you said right there. The problem is as the direction movement (be it +/-) is constantly changing due to the nature of how range dictates the analogue signal strength inputting into the mover. Any predefined 'amount' (from batteries) will not add to or subtract from the overal signal strength. Instead, as the only way to feed (that i know of) these external inputs into the feed coming originally from the sequencer outputing range value, is through OR gates, the only time the battery signal will effect the 'master signal', as it were, is if the percentage in the battery is above the current percentage outputting from the original source (range). It doesn't really do much on the front end atall, except 'slightly' change the speed at which the ships vector alters, in the very small window of time (and chance) the specific battery is in an on state at the specific time the constantly altering signal coming from the range output falls below the amount set in the battery. Visually it does nothing much atall. Using AND gates does the same but in reverse. Using selectors seem to output the highest analogue signal out of the batteries from all output gates regardless of which gate is 'on'. trying to cut a signal using a chip with a node in it, with the input powered off with an input from a selector doesn't seem to stop the analogue signal from passing through. I'm having serious issues making additional/subtractive analogue strength percentage sources 'appear' within the circuit and 'dissapear'. It seems if a source is there, with analogue... its there. Besides that, the other big issue is 'integrating' said analogue subtrations/additions from these seperate sources into the analogue signal in a way that effects the sum total instead of simply act as a state dependent variable. ... so far i'm stumped. RTM!!! TETSUJIN!!! SHADOW!!! I NEED YOU!!!! lol

2011-02-16 01:21:00 Author: Epicurean Dreamer Posts: 224

---

Have you read RTM's analog addition/subtraction blog? You don't want logic gates, you want the directional splitter, combiner. You can even handle the fact that it may be + or - just fine, though it gets more complicated. this one (http://www.lbpcentral.com/forums/entry.php?2107-Analogue-Logic-2-Sorting-amp-Addition) specifically. Specifically, this will handily sum two signals together. http://i600.photobucket.com/albums/tt82/rtm223/adder%20sorter/APhoto_55.jpg

2011-02-16 01:45:00 Author: Tygers Posts: 114

---

Yep, use RTM's logic blog for analog addition, overflow, etc. I learned in a single day what would have taken me about a week of testing to understand. Off-topic: When you make a very long series of overflowing additions, and attach a randomizer to the inputs of the signals, you get a very, very mesmerizing light show. Tip: If you run into an issue of negative values being created from the original input overflowing and acting negative, just run it through another combiner and put the negative output into the postive input (as another overflow section). Ensures all signals are always positive (gave me trouble for a few hours to find issue, not sure if relevant to anyone else. Also, sorry for the poor phrasing).

2011-02-16 03:55:00 Author: SSTAGG1 Posts: 1136

---

thanks for all the replies people. I have indeed read RTM's blogs. I read that one a while back mind you and since forgot allot of what it went on about (i read it before any of it had any practical application to what i was doing, so didn't attempt to let much of it 'sink in' to a helpfull degree.) Speaking of RTM, i've used his sine principle to overcome the issue of randomness within order though. I've made this: http://ia.lbp.me/img/ft/8dac4a44da53d15e6a2c6ae62345fa925c982453.jpg Its a random wave generator. Its quite a langthy process. The 6 chips on the left are 2 sets of 3. They each have in them a basic triangle wave generator. http://i0.lbp.me/img/ft/dc08f937403b12d7e58757c37c90f9d1100e336d.jpg The timings on the timers are 1, 2 and 4. The randomzer will select 1 of the 3 every (between 2 and 4 seconds). The two randomizers power 1 wave generator each, in parallel, but which one's are active at the same time intermingle. The chips output through OR gates into signal combiner, i'll talk about the top one first. The signal inputs into the + input on combiner, output goes into timer, timer set to speed, timer feeds back into - input on combiner. This creates a (fairly acurate) sine wave. Depending on how high you have the count setting on this timer depends on how 'smooth' the sine curve is at the ends of the input scale, the smoother the sine, the more % you loose off the ends of the input strength. So, the top 3 chips > OR gate > + input > timer do that. The bottom 'set' as it were does the exact opposite, well, does the same but with an inverted signal and the timer set to and inverted output. The timer is also wired into the + input, with the feed going into the - input. Its a negative mirror of the top set. Doing it 'split' like this allows the sine to traverse the entire spectrum. RTM did it in a way that makes you lose 50% of the DC spectrum. He used a single wave genertor and gave it a -50% DC offset as the last input on the chain. This simply shifts the 0 > 100 DC oscillation down 50%. Using 2 wave generators, one outputting an inverted signal, both signals run through a combiner, gives you a 'full spectrum' sine wave. ( all i've done is mess with the frequency of either side of the sine input which outputs a random wave.) The problem at this stage is the sine wave outputs are out of sink, running it through a combiner outputs the average, the difference between the two signals. It swings wildly about, and sometimes sits around the middle nudging about a bit. Its too chatoic so i needed to do a bit more. Output the combiner to 2 timers and invert one of them. This creates an in sink full spectrum triangle wave of the output difference between the two out of sink sine waves. Outputting these timers individually into the same sine creation process earlier in the chain and then combining them through a combiner creates a wave out of the signal except with smooth extremes. http://i8.lbp.me/img/ft/b28fc736dc3a5ad296590acb8e26802cfc7433b8.jpg this final combiner feeds into these: http://i9.lbp.me/img/ft/17982b90ef50eaa2cfa3c56667b892c75bacbada.jpg Each hold an advanced mover, the input going into the left/right input, and a sensor inputting into the up/down input. Each chip, randomly selected every (between 1 and 6 seconds) also has: http://i8.lbp.me/img/ft/488b180a39232402d589e7cacdb947e8133bf9f8.jpg 3 sensors, randomly chosen every (between 1 and 3 seconds), each sensor has a differernt rage, which due to them being set to strength, will give the ''settled' orbit different ranges with each overal range change. What this whole thing does is creates an initial state thats always the same. Yet each emitted ship will 'evolve' into a completely unique behavior pattern. It works perfectly Edit, if your interested, the OR gate going into the chip at the top is the ships damage sensing logic. The other output from the final combiner is simply a spectrum analyzer ( splitter into 2 sequencers set to possitional, the - input sequencer turned upside down)

2011-02-17 11:50:00 Author: Epicurean Dreamer Posts: 224

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.