If you want to understand the blog and how it's laid out, go here.
Feb 22, 2006
Feb 22, 2006: A lot of big major things have occurred since the last update. First, my MP3 player came and works, but with a bug giving a triple advantage and loads of experiments and processing. Second, an even longer tech support episode occurred and with a strange twist to it. Third, the kitchen is being redone meaning no sink or counters are available in there. Fourth, an experiment I recently ran helps me get my frame rate at least 5 times higher than it currently is in my 3D game (giving my first ever logarithmic scale graph). A major cold snap finally hit after 40 days since the temperatures haven't gone below zero. I lost 50 pounds in weight in an unexpected surprise. Finally, 4 new dreams have been added (342 is the current count), one of which has something all my other ones don't. An article has been written to explain how I process my music.
#1 MP3 player came: In short: yay, boo, tick-tock, and bonus. Yay as in my MP3 player finally came and works. Boo as in it has a bug in it. Tick-tock as in, to fix the bug meant having to run the longest noncomputer-related experiment I've done, over 5 hours. Bonus as in, the bug gives me 3 advantages: more capacity, faster loading of files, and faster conversions. For further details, my MP3 player finally arrived. Although I had no MP3's on my hard drive, I first tried it out copying from my 15-month old MP3 CD, the last one I made. I did this to check to make sure it worked. The software installed successfully (last time it wouldn't due to a virus on the disk). However, some of the files didn't play right (faster with a strange ticking sound mixed in) and others were slowed down a lot causing it to sound weird. Support told me that the player doesn't support the low sample rates. I'm using the standard ones, generally 8000 and 11025 Hz (although I could easily throw 13,259 at it and see what happens). 11025 Hz sounds correct without the distortions, but it plays back much slower than it should. Going to higher sample rates wouldn't be worthy as, at 16Kbps, the quality would be too degraded and considering I hate high frequencies, 11025 was sufficient enough. To work around the bug, I ran an experiment. I first did only a 40-minute run at it which gave me 3 significant figures of precision. That wasn't too enticing to me as, after 17 minutes of play, it falls behind by 1 second. I decided to jack up the time and go with a 322-minute-long experiment, the longest I've ever done not involving the computer (except to compute the results). This gave me 5 significant figures, so precise, it loses less than a second (about 0.9 seconds) on the day. The magic number I got was 37.804% as the offset after averaging out the results I had during the last 30 minutes of the experiment where precision is at the highest. This in turn gives me 3 advantages. The shorter file caused from speeding it up (not the sample rate as I usually do it by) means that the songs load faster and save faster. It also means faster WAV to MP3 conversions. With the shorter length comes a smaller MP3 file thus the MP3 player, in effect, has 37.804% more capacity. The player itself has 4,066,799,616 bytes capacity. This bug gives me the sense of having a 5,604,212,543-byte capacity (±40 kilobytes). So, instead of 22 59/60 days of nonstop music total, I now technically have 31 2/3 days of nonstop music, a whole month. The only downside is that, if I forget to do the speed adjust, I have to reprocess everything. I normally check after saving all the WAV files before converting to MP3 to minimize this. After giving a second thought to Audacity, the equalizer really isn't necessary and the equalizer in the player isn't worthy. Besides, with Audacity, I have, in a sense, a 1500-channel equalizer going out to +30 dB, 150% further than most equalizers do so I can really filter out the frequencies above 3KHz. That's far more precision than I really need but far better than any equalizer I have. The MP3 player does have a power adapter with it, but I'm not currently sure how to run on just the power from the grid, or at least I haven't bothered testing it yet. The player plays back very clearly. For further details on how I process songs and get them on my MP3 player, go here. #2 Record tech support got even longer: Remember before that I had a phone call lasting nearly 144 minutes long? This time, I had two consecutive hour-long calls, but the tech support episode actually lasted over 3 hours long, the longest ever, but with 2 hour-long phone calls. On top of that, I used an awesome, but weird program to perform this. At the start, I was typing a reply to the HWW forums that I rarely go to nowadays. I then was called about my case with the styles and colors. After a quick intro into a weird, but interesting program called Live Meeting (Net Meeting?), he then tried setting the colors himself. He couldn't do it and it looked like he was stumped on it too. I was then told that he'd call back 15 minutes later. While waiting, I searched into clues on the colors. MSPaint didn't get much much further. I then recalled the 3 bugs I encountered. I first began demonstrating the bug with Notepad (saving with wordwrap enabled). I then did the same for the bug in Wordpad (copying characters from character map into text-only mode in Wordpad causes the text to become formatted text (likely a bug with the character map). I mentioned the window switching bug as well, but couldn't directly reproduce it and I still don't, even though I have quite a few leads. The 15-minute wait turned into an hour. At the end of this wait, I began using the GIMP to check the colors which gave me a major clue. There were some odd gradiants used (at a diagonal even) which is also present in those radio buttons and in many other places. I get the call and got bad news. Only the classic style allows me to fully get the dark gray background I use constantly. When closing GIMP, GIMP decided to crash on me causing my computer to get really slow. This took an extra ten minutes to get going again with my computer running at half it's normal speed afterwards. I then mentioned the bugs I've found. I brought 2 bugs up in Excel as well (pressing // at the start of a cell causing the menu to be accessed; printing with the black and white option checked prints in the color used by the screen (green in my case, but could be yellow) and not actually black at all. I showed him, while on the phone again, how to reproduce the bugs I've found. The second of the 60-minute calls ended after showing the bugs I found. #3 Finally a major cold snap: It's been 40 days since temperatures have gone below zero, a record for a big chunk of my home state and only recently there was a major cold snap. Care for air temperatures of -20°F with wind chills to -60°F when a week earlier it was 20°F for the overnight low? It was like that for 2 days, a super freeze. It was as if winter was skipped. Considering that the temperatures are now back to normal for this time of year (about 15°F highs with about -5°F lows), the weather this year has been bizarre. Correction, beyond bizarre. #4 Kitchen remodelling: Pop quiz: how do you handle washing dishes or preparing food without counters and a sink? Seems tough, doesn't it? It just so happens that this is the case lately. The bathroom sink is used for this sort of thing as a replacement. #5 Polygon performance experiment: About the best thing I've done for my game was running an experiment to find out what effect there was for objects of different polygon counts against object quantity and processing speed. The results give a very strong light to my game development. First I thought 700,000 polygons per second was the best I was going to get. Now it turns out that I can get ten times that! I was using objects of 20,000 polygons or more when the peak, at least for me anyway, is a model of 2000 to 5000 polygons. Or, put into a graph of a logarithmic scale, you'd get this:
This is the first graph I've ever made using a logarithmic scale. They're not all that difficult to make either. The formula I used was this:
POS = log(VLU)*GS-(GO*GS)
POS is the position of the vertex on the graph, log indicates to find a logarithm of the stuff in parentheses, VLU is the value to be plotted, GS is the graph scale (the number of pixels per increment of 10, 200 in my case), GO is the graph offset (used if not starting from 1 (1 is 100), 1 in my case as I started from 10, which is 101). I have other experiments planned. One of which is what the peak is when multiple textures are used, a more precise peak finder from the 1000 to 10,000 range. Upon doing this, the 20 fps I've been getting will almost be a thing of the past. I could have far more detail and still get 50 fps! I lately seem to say the word "experiment" a lot after running this, well, experiment. As a side note, these results are from using my computer. It has a 3 GHz Pentium 4 processor, Radeon 9600 XT video card, 512 MB RAM, on Windows XP Pro done in a test environment that is completely empty other than the test subjects. #6 I lost 50 pounds - no diet or exercise: From a doctor appointment I had on Feb 13, 2006, I got my weight checked. I was shocked to find that, from 220 pounds 15 months ago, I'm now down to 170. What a big difference! However, what's been going on during that last 15 months is what makes this even more amazing. Considering I don't eat right (pizzas as my main meal, with snacks mainly as cereal and rarely anything else) and hardly exercise, it's a wonder how this happened. For the first 9 months or so, I've been drinking Powerade drinks, which have 700 calories per day for me. I switched to light lemonade (starting with regular lemonade at 120 calories or so per can) having only 8 calories per can or about 50 calories per day, a huge reduction. Could it be my fat-removal trick that I'm agressive with using to further minimize it? Perhaps that scale was off or not set correctly. I'm continually at my computer day in and day out with only sudden high-speed dashes across streets as my most intense exercise. Even more on top of this, TV ads say of losing that much in 8 to 10 weeks (or 2 to 2.5 months) and that's with exercise and diet change. It could also be that I ran some calculations as well. To lose a pound in fat (9 calories per gram, 453.6 grams per pound), you need to burn off about 4080 calories. To lose a pound of carbs (sodas!) or protein (4 calories per gram), you need to burn about 1820 calories. A pound is quite a bit and this isn't that big of a number (to me anyway). I haven't had a single soda, not even a sip, since January of 2004 (over 2 years ago), rarely even touching or messing with candy (even rejecting it). So many possibilities, it's a puzzle to how this actually happened. The numbers are right there. All you need to do is put them into formulas and work them out then plan things out from there. #7 4 new dreams added: I don't recall a dream where I had a super hero involved (like those in the comics - the classic cape and so forth). I just very recently had such a dream. I've had 4 others as well pushing my dream count very close to 350.
Feb 9, 2006
Feb 9, 2006: I'm now running on Windows XP, but with a long story behind it including the longest single phone call I ever made. I've noticed many benefits, but several bugs as well with it. 3 new dreams were added as well, including another fun dream 97% in tact and two occurring in one day.
#1 The story behind Windows XP: Wouldn't you know it? More problems occurred along the way, but this time nothing serious (unlike the motherboard/CPU issue 5 months earlier (see September 2005's blog for details on this event). I originally had five problems, most of which I was able to work out myself. I was down to only two problems. The first is that XP wouldn't activate saying that the key was activated too many times. The second involves my well-known dark gray background scheme and that one of the settings wouldn't work at all (not even a Microsoft official could). A third issue involved was that Winzip wouldn't register for some reason. That turned out to be that I had the wrong key, as I forgot to update my notes. A fourth involved installing MSWorks 4.5a where it crashed twice before successfully installing. The fifth and final issue was with the shortcuts as many pointed to the wrong path. The Winzip, MSWorks 4.5a, and shortcuts issues were easily fixed. I was forced to wait until Monday to call about the activation issue (the most important) as, not only was it the weekend, it was 2:00 AM (like anyone's there at that time of day). The call I made to Microsoft lasted for 143 minutes and 36 seconds. My previous record was 28 minutes (±2 minutes). This is 5 times my record. Almost all of it was dealing with the activation issue. Almost all of the total call time was dealing with the activation issue. Strangely enough, I ended up having to get a new key. After that, which lasted nearly 2 hours (about 15 minutes on hold as a total), it was then off to figuring out why I couldn't set one particular color. If you went into the control panel, double-clicked display, clicked on the appearance tab, then advanced, you can then see where I'm getting this from. I like the XP style and the default blue used, just not the bright white associated with it. If you were to select menu, or click on the banner with the menu, you can set the colors. From here, you can't set the background color as it doesn't work, but you can certainly change the color of the text. Even a Microsoft official couldn't get the color to get set. Other than this, everything is seemingly going well. #2 Benefits of XP I've experienced: So far, I've noticed that everything runs much faster. Even my game runs faster, 50% faster. One scene in Windows 98 was getting ~42 fps. The same scene from the same position was getting ~65 fps! I start up a program and it loads twice as fast as that of Windows 98 (such as calculator or the character map). Calculator isn't much different, but the character map is what surprised me. First, not only do I get the usual 224 characters as I had in Windows 98, but I get far more, most of which I don't recognize and weren't available in 98. However, I'm unable to be able to use these extras on my website as HTML is text-only and 8-bit, so I can't give a demonstration that easily. I can use them in Word and Excel though, supposedly anyway. As for my website, I'm still restricted to the same old 8-bit set. I otherwise very rarely use the other characters though so it isn't that big of a deal. The Windows Task Manager looks very decent and has several interesting features, covering things in far more depth than System Monitor did in Windows 98. I was right after all in that Firefox is using over 100 MB in memory just to load. It shouldn't be using more than 20 MB in memory. #3 Up next, more memory: Yep, I gotta wait another month. My next intent is to get the 1 GB in memory I was hoping for. It so seems that memory is cheaper than I thought it would be so I could afterall go with even 2 GB, the maximum the motherboard can take (or at least 1.5 GB). This way, I'd almost never run out of memory (without stuff going into swap file anyway). The only rare exceptions would be when working with huge graphics (such as 2048x2048 textures) or running complex experiments. #4 3 new dreams: Three new dreams have been added to my ever-growing dream journal. One of which, a fun dream involving water, is 97% in tact, so well-recalled, another story got created. A second involves learning the computer, which I find somewhat weird to have occurred, and a computer museum in a mall setting. The current count is now at 338. #5 Minor update on 2D game details: Although the 2D game itself has no progress, I made a minor update to my website to include details about it on category 9 including an updated screenshot of it (rather than the animation I had).
Feb 1: I finally figured out acceleration in my game! I've had a tough time figuring how to get it working on all frame rates. I took 5 attempts beforehand, but it didn't work properly. The closest I came was where high frame rate had higher acceleration rates than that of lower frame rates. Now, not only does it work on all frame rates equally well, it's mathematically accurate! It looks so nice and almost exactly copies how it is in my mind game! Due to a severely hazardous virus (BlackWorm), I ran backup just in case. The fact that it deletes ZIP files, files I use as backup, worried me so I took it safe. This took about 130 minutes to do, but it was worth it. Both XP and 98 are affected. Outside this, I've recreated the extra life icons in full outside scripting them. Scripting them should be very easy. Woke up at 3:52 PM and went to bed at 7:22 AM. Feb 3: Most of the day was spent on the forums. Although TV was slipped in often, mainly CNN Headline News, I didn't really do much else outside an experiment, wikipedia, and converting Works files to Word and Excel files. The experiment was to find out what effect my LOD system would have over just objects being invisible. 1728 objects all invisible took 1/5 of a millisecond to render versus 95 milliseconds if all were visible, a very considerable difference. This was better than expected, but the part of finding the distance between it and the camera, however, wasn't tested. As for the converting stuff, about half of the files were converted. One document, being 67 MB or so, much of which due to images, was causing MSWorks 4.5a to crash repeatedly when trying to save it as a DOC file. As for Wikipedia, I mostly just looked at things, but made a few minor edits and checkups on things. I should be getting XP when I wake up, or, less likely, tomorrow. Woke up at 4:58 PM and went to bed at 7:25 AM. Feb 4: XP came! I ran backup again as I made more changes then began the installation of XP wiping out 98 first. This took a long while to do, and since nothing was going on on TV, I played Bubsy. Nothing new or special happened while playing the game, just the usual jumping around, tossing Bubsy in the water (in the first level mainly, the best one), trying to pass 32 pixels per second using the eggs in several levels (I reached 30 pixels per second). Currently, I have two issues remaining (out of 5) with XP. The first involves activation and the second involves configuring the colors. For details on this, see news item #1. So far, everything is fine. I'm not all that acquainted with XP yet, considering I've been using Windows 98 for 5 years. I had 3 nosebleeds though, mostly moderately-minor ones (the most severe was 100 drips, the others totaled that). Boy do I hate my nose! Woke up at 5:25 PM and went to bed at 9:57 AM. Feb 5: So far, no progress toward fixing the two issues with XP (colors and activation). I can only activate on Monday and shortly before I go to bed. I got bored of Tiny Toon's theme song (from Buster's Hidden Treasure, a Sega Genesis game), and returned to Out Where the Lake Is again adding more onto the one million plays it has total. Winter Land still plays on my MP3 player. The new MP3 player hasn't arrived yet and I currently don't know the state of it. Toward the end, I began making the plan for the city world. That's going well so far. Most of the day was spent on the forums. The textures for the buildings are done as well. GIMP racked up a total of 4.2 million page faults, more than that of all other processes running combined! If I was doing those mountains from my 2D game, it would almost get to even 200 million page faults total! Woke up at 8:26 PM and went to bed at 9:16 AM. Feb 6: Activation is finally done, but not after about 103 minutes on the phone. The issue with color configurations was forwarded to the engineers to look into. See news item #1 for details on this. Outside this extremely long phone call, 5 times longer than my previous record, I was on the forums for most of the day. I did happen to run an experiment to find out what effect having multiple textures on a model has and the results were worse than I thought. It took 9.7 times longer to render the exact same object with 16 skins than with 1 (and with the same number of objects, 1728). I also made a small update to my website for my 2D game which is now included in category 9, but only briefly described. Woke up at 8:59 PM and went to bed at 2:45 PM. Feb 7: Cleaned out the forums, checked E-mail, then ate. After about 3 hours' TV, I returned to the forums. I began updating my website, the first update made on XP. I then thought of a way to try and fix the color issue - the registry. I found several clues and reported details on it to the case. I also encountered a bug in Notepad. Enable wordwrap, write something so it wraps on to another line, then save. Go to the beginning and type something to cause words to the right to wrap on another line. A strange effect occurs. This wasn't present in 98's Notepad. Then, in Wordpad, if I copied a character from the character map, special formatting is used and every time I save, it mentions saving into text-only format. This didn't happen in 98 either. It seems like I'm "beta testing" XP, even though I only have a half week's experience at it. Later in the day, I reran my experiment to find out what effect having multiple textures has on the frame rate. This time, I went a little further. Instead of 1728 1-texture objects and 1728 16-texture objects, I also tried 108 1-texture objects and 108 16-texture objects. This was more enlightening though as 108 16-texture objects was 60% faster than 1728 1-texture objects. 108 comes from 1728÷16. They both have the same number of vertices and polygons. As for the update to my website, only the major headlines have been done and it seems like I won't have any dreams to add. It also seems that "experiment" and "complex math formula" or "complicated math formula" are my favorite things to say. I tend to say them often and find it funny. Woke up at 11:00 PM and went to bed at 11:23 AM. Feb 8: I woke up with 2 recalled dreams, including a fun one 97% in tact. I was very rudely interrupted by my nose - another nosebleed. It has two pages in details but took 2 hours to right as my mind kept drifting off. It was clear until 1:10 AM that I got out of bed. Most of the day was spent on the forums. Someone on Howwhatwhy was in need of math help, simple stuff (to me) that I can work out in my head in under 20 seconds. I offered some assistance to this. From this, rather than just 10 minutes per day on Howwhatwhy, it was more around 130 minutes for the day (tredecupled). I otherwise didn't get anything on my website updated as I was way too involved with this. The other part of the day was the usual 3DGS forums and watching space stuff on TV, especially the show "parallel universes", of which I saw twice before. Woke up at 11:03 PM (±2; before beginning to write details on my dreams) and went to bed at 1:52 PM. Feb 9: Spent about 32 minutes writing details on another dream that I had. After cleaning out the forums, as usual, I updated my website adding my 3 latest dreams. I reported two bugs with WED and tested one someone else reported. No other known details are available for this day. Woke up at 2:23 AM and went to bed at 3:38 PM. Feb 10: Another day, another dream, a weird one this time. Most of the day was spent on the forums. Only TV (CNN Headline News almost entirely) was the big chunk of the day. Outside this, I made the numbers marked on the ruler and began continuing on with making the rule. It is about 60% done. I'm currently left with 3 puzzles - how to insert the digits (mostly solved - not tested), how to texture it (limitations of creating the texture make this very difficult), and how to best optimize the ruler (fewest textures - an experiment showed the effect it had - it was a big negative for many). I see how I can use 2 textures instead of 4, however. Woke up at 3:39 AM and went to bed at 5:14 PM. Feb 11: Another dream already! That's 5 in just 4 days! That's also the 340 mark. My MP3 player also came. Myself versus MP3 player round 3 has me as the winner! The MP3 player works, and it is surprisingly fast. It transferred 702 MB from an MP3 CD in only 5 minutes. Now to process all my songs. This'll take several days, even a month. Outside this, I watched Pink Panther for a long time as there was a marathon going for it. I also went to the recycling center and got $20 for my cans. I was given 70¢ per pound in cans, more than double what I thought I was going to get (which was 30¢ per pound). I also went to 2 grocery stores as well getting my usual: pizzas, cheese, drinks, and cereal snack). Right now, as for the current song played, it's World 7 from Super Monkey Ball 2. My old MP3 player has Out Where the Lake Is running on it. Woke up at 4:08 AM and went to bed at 9:09 PM. Feb 12: No dream this time. Most of the day was spent on the forums. I did, however, run an hour-long experiment to find the impact of polygon count to performance. Oddly enough, it peaks around the 2000 to 5000 visible polygons range. Another 3 hours were spent after that making a logarithmic graph of high accuracy. It was the first time I've actually made a logarithmic scaled graph. I got 7030 polygons per second rendering rate, nearly ten times as much as my current game does and 75% higher than my old record. That was with a textureless object, however. See news item #5 for further details. I have other experiments in store as well. Woke up at 7:49 AM and went to bed at 8:11 PM. Feb 13: I had a doctor appointment today, of which I finally got to go to. That went fine, better than expected even. I was also surprised that I lost a lot of weight. I went from 220 pounts to just 170 in about 15 months. See news item #6 for further details on this and how I find it strange. The doctor appoint was otherwise 95% questioning, 3% checkup (blood pressure, pulse, weight, etc.), and 2% other (These values don't include 40% waiting time (~20 minutes).). I then went to Kmart, the only department store I'm not blocked entirely by (except the clothing section). I went off and got shoes (average shoe life time for me - 9 months) while my mom got me benefiber and vitamins. Outside this, most of the day was spent on the forums. I also ran an experiment and found that a random mesh takes 50% longer to render (or 2/3 the frame rate) as before I used a regular grid. That's all I recall doing. Woke up at 7:59 AM and went to bed at 11:17 PM. Feb 14: Valentine's Day! I figured it'd be an ordinary day. I found a weird bug involving my new MP3 player. For details on the MP3 player episode, see news item #1. Most of the day was spent on the forums. A big chunk of it, about 5 hours, was getting a more worthy WAV to MP3 converter. Audacity seems to be my best choice. The issue is still not resolved in full, but considerably close. Hardly any TV was involved. Outside these, nothing else happened. It was an ordinary day after all. Feb 15: Another 2 long phone calls. One of which was for the MP3 player. I'd have to run a time-syncronization experiment to verify the effect of 11,025 which sounds just right. I'd like to try some odd ball like 13,287 just to see what happens. The second thing involved Microsoft's tech support. Details on this episode are in news item #2. I otherwise spent most of the day on the forums. Toward the end of the day, I watched TV and began reconverting my WAVs to MP3s. I need more tests though especially a time-syncronization experiment. After the call, I added the information to category 9 about my 2D game as it should also belong in there. Woke up at 11:15 AM and went to bed at 3:06 PM (±3 minutes). Feb 16: Finally another dream taking 15 minutes to write details on! After cleaning out the forums, I began getting more songs processed. I ran an experiment to fix those that use 8KHz and used 11KHz. I then tried them out and the speeds set don't match so I ran numerous experiments to accurately time the actual length and found that it takes 1.3715 times longer to play the song although it's still not right. If one thing don't work out normally, I have other work arounds. I thought that I should record it so I can compare the timing. Trouble was, I spent 2 hours trying to crop the recorded sound as nothing seemed to be matching up to mimic the original WAV file. The waves didn't closely match each other and were far different (one smooth, another rough). Either Winamp doesn't play in sync with actual time, my computer clock runs more slowly, or something else is screwy. I used an actual clock toward the end and found it was more around 1.3782 instead of 1.3715, a considerable difference. On 2000 seconds, that's about 13 seconds off, way too much. For better results, I need to run a much longer experiment lasting several hours, not just 40 minutes. I was too tired to run the test so I'll start it tomorrow. Other than the 4-hour-long troubleshooting events, another 2 hours were spent converting 2 new songs and their speeds to MP3s which led to my experiments. 3 or 4 hours of TV was mixed in and I don't know what else happened. Woke up at 1:39 PM and went to bed at 4:17 AM (±1 minute). Feb 17: I began that 5-hour experiment as I planned, the longest not involving computers. For details on this, see news item #1. I only had a short visit to the forums and I watched quite a bit of TV, mainly while waiting for conversions. For details on how I do these conversions and process my songs, go here (opens in new window). I can put a big red X on dBpowerAMP. It seems to be somewhat glitched and it cuts off my songs unexpectedly. Audacity is almost as fast and doesn't cut off my songs as far as I can tell. There's no batch conversion as far as I can tell however, but this isn't like the BMP to GIF batch conversion which takes a tiny fraction of a second to process, the conversions here take in the order of several minutes to do. Also, Audacity has a lot more to it than I first though it did including a feature that functions as a 1500-channel equalizer with a range 3 times that of most equalizers. Care for -20°F air temps on top of -60°F wind chills? Today and yesterday I've been getting this extreme cold. It makes a freezer seem like room temperature! This isn't atypical in my area, especially during the winters. It's been atypical this winter however. Further details are in news item #3. Woke up at 1:48 PM and went to bed at 3:35 AM (±1 minute). Feb 18: A huge chunk of the day was spent processing WAV files and getting MP3 files. 7 songs have been processed, over 170 tracks! Everything since that 5-hour experiment was processed has been going very well. I now have the idea to have an article on my website on how I process my songs. In short, I take the originals, loop them (attaching the intro if there is one) to 16:30, apply the speed up using the magic number 1.37804, use the hex edittor to change the sample rates, then convert the WAVs to MP3s and finally upload them to the MP3 player testing them afterwards. I only briefly visited the forums. I spent 80 minutes trying to get a 4x4 magic square for a post on the 3DGS forums. Nothing else happened worth noting. Woke up at 3:58 PM and went to bed at 5:52 AM (±3 minutes). Feb 20: The first part of the day was spent doing unexpected housework - kitchen remodelling. See news item #4 for details on this. After eating, I cleaned out the forums. I next processed "Tree climb" (level 13 on Bubsy) taking screenshots to explain my process I thought of writing about yesterday. Once completed, I wrote that article as I had intended earlier. Finally, I began processing Target Zone, but got too tired to do the WAV to MP3 conversions. At the end of the day, when in bed, I had a feeling that I forgot to apply the speed-up needed which made the processing I've already done worth nothing. Woke up at 5:22 PM and went to bed at 7:58 AM. Feb 21: Another dream, but not recalled entirely even though it has something unique among my already over 340 dreams. My original suspicions were right - Target Zone didn't have the speedup required. After cleaning out the forums, I reprocessed Target Zone then processed Mountains So Far Away, but made a dumb mistake when stitching the intro that went unnoticed. It was processed into MP3s, but due to the mistake, I have to redo everything. I did, however, process Out Where the Lake Is with the equalizer effect used on it to amplify the best parts and deamplify the bad or unliked parts, especially the frequencies above 3KHz. There's a lot of news items since that last update and updating my website is my plan for tomorrow. Woke up at 5:27 PM and went to bed at 9:28 AM. Feb 22: Website updated! 342 dreams, 158 more to the 500 prediction by the end of this year. Besides cleaning out the forums, I watched TV including "The Death Star" for the 9th (10th?) time. Like Out Where the Lake Is is to music, "The Death Star" is to a particular TV show. The two are close together (988.2 versus 971 (yes, 988.2 - it went down slightly) - Out Where the Lake Is is still about 6 times better, but at 300 times better than neutral, it's definitely worth it). Reported a bug for GIMP yesterday that was confirmed as being a bug and I also reported an enhancement for the select-by-color tool since it doesn't partially select a pixel. I've lately been thinking of a way I could process the clouds for my 2D game that a very recent dream depicted. I then messed around with Celestia to check the Moon phase on Dec 26, 2004 to see if the Moon and Sun put together had any effect on the tsunami, but from what I can tell, it was almost a full Moon where the gravity is cancelled out. While doing so, my main compact flourescent light bulb finally burnt out after at least 1 1/3 years of use. That is amazingly long! Woke up at 7:50 PM and went to bed at 9:46 AM. Feb 23: Added a new section to my mind game report, frequently asked questions about it. A huge chunk of the day was spent on the forums and watching TV. I made a suggestion to the National Geographic Channel for the "Seconds from Disaster" series. On Jan 18, 2002, in my home town, a train containing anhydrous ammonia derailed. One died and at least a hundred were injured. That was a scary event and given the style of "Seconds from Disaster", this seems like a perfect candidate. I found another bug in XP as well. To replicate it, run task manager (press control+alt+delete once to bring it up). Click on the performance tab and double-click one of the graphs. Drag them off the side somewhere then double-click them again. It's a maximized window, but it's weird compared to all others. That's now 4 bugs that I've found. The window-switching one has another lead - multiple windows of the same type are running (commonly Firefox). Also, I fixed and reprocessed my song "Mountains So Far Away", of which had the intro not cropped entirely. Finally, I fixed the transparency issue with the text for my start and goal platforms in my 3D game. Woke up at 7:58 PM and went to bed at 11:00 AM. Feb 24: A well-recalled dream - a whole page long! Most of the day was spent on the forums and watching TV, mainly CNN Headline News. Processed 3 more songs for my MP3 player - Battle Zone, Tiny Toon's theme (from the game), and Cosmos' ending theme song. Outside this, I don't recall anything else. Woke up at 10:57 PM and went to bed at 1:27 PM (±2 minutes). Feb 25: An second consecutive well-recalled dream also taking up one page in details. Ran a battery longevity test for my MP3 player. Went to the grocery store then Wal-mart where I had to wait outside. I also went to Pizza Hut. It seems that if I do go, I must only get the Full House pizza as that is close to that of what Domino's has and it's cheaper than the large is by $3, but still $3 more than that of Domino's. I showered beforehand. I processed a 20-lap edition of Desert Zone. I likely spent 3 hours watching CNN Headline News as nothing was on on TV at all. I doubt that it is a record, but I noticed that what was played seemed to be extremely repetitive even down to how it was said (timing, tone, relative position of the video to the sound, etc.), of which I also experience at night. There's no "just in", "breaking news", etc. at all during this time, seemingly only on the weekdays at times not fully certain. Even after 15 1/2 hours, the was still one mark left on the battery indicator, better than expected. Got Fruit2O as there was a sample available and tried it. The strength of the flavor is too low for me as I can barely even taste anything making it seem like pure water. Had a strange mind game event as well. I'm supposedly in the Olympics in a mountainous area skiing. I run up a mountain at 200 mph and when at the top of the mountain, I jumped too late so I jump more straight across than jumping higher. After reaching nearly 400 mph falling with the diagonal speed at 550-something, I bounce adding even more speed. Then, I land again only float-running supposedly becoming the world's fastest skiier, without skis. At the bottom, I whiz past the finish line at about 320 mph and I slip between a building and a tree (which was only a few feet from the building) then crash into a cliff at 320 mph. Woke up at 12:39 AM and went to bed at 5:00 PM. Feb 26: Woke up at 2:28 AM, but couldn't get back to sleep giving me an hour less sleep than I normally get, which isn't that bad. I didn't get any dream this time. I finally returned to my 2D game trying out the idea depicted in my recent dream. To my surprise, not only did the technique work, but the effect it causes almost entirely replicates what I saw in my dream. From 9-ish AM to nearly noon, I was involved with putting up sheetrock. Began the battery longevity experiment right when I went to bed. I generally sleep for ten hours and my previous test told me that 15 1/4 hours wasn't long enough I had it playing while I slept. I did go to the forums, but not much was going on. Going strong on Final Fantasy 1's World music. I got bored of it several months ago racking up over 20,000 plays. I racked up another 1500. Woke up at 3:03 AM and went to bed at 5:16 PM. Feb 27: Two recalled dreams in one night! Spent about 50 minutes writing details, 20 minutes actually wrting. When I woke up, I was 11 hours in the battery longevity experiment and there are still two marks on the battery. The battery lasted 17 hours and 20 minutes. The specifications for the product say 14 hours so I guess I've proven them wrong. This is a considerable difference. 2 1/2 more cloud layers are done. The scene now looks much better. I need more layers for better results. With 4 layers officially done, the effect seen in my dream most closely mimics how I'm seeing it in my 2D game within 3000 feet of the clouds. Here are screenshots as my 2D game was on this day: #1, #2, #3, #4. I also watched the final episode of Monster House. I've liked their series, but they kept moving their times like crazy which made it difficult for me to watch it. I don't understand why Monster House is done with, likely that a lack of interest occurred? I'm going crazy with the song "Mountains So Far Away" which sounds best from 46 to 64% of the true speed which means that I need to add more speeds to my MP3 player. Woke up at 3:35 AM (±10 minutes) and went to bed at 8:05 PM. Feb 28: No new dream this time. I redid the battery longevity experiment and this time it lasted 18 hours! 2 1/6 more cloud layers are done (6 officially done with the 7th 2/3 done). The effect is now even better-looking. When at the highest point, even at 1024x768 resolution, the scene looks natural, except that beyond the mountains of which haven't been done yet. I'm currently doing the clouds closer than the mountains. For the next round of the battery longevity experiment, I wanted to find out what effect volume the volume setting has. I'm testing 20 instead of 18. Since 20 is too loud (max is 25) for sleeping with, I had to put soft sound-absorbing things to deamplify the sound. 15 is too soft to hear over my very noisy computer fan. I theorize that it'd last a much shorter time period, 15 hours. Louder sounds take more energy to produce. Next up is March where my luck is slightly above average. Lately, I seem to be calling a "dot" a "spec". Details on this on the March 2006 blog entry (news item #3). Also, I finally got contact with a friend of which I haven't spoken with in nearly 50 days! Woke up at 6:18 AM and went to bed at 9:36 PM (±3 minutes).
The news entry headlines in bold are color-coded for a reason. They tell how big the event was for the headlines and how good my day was for the daily entries. From May 2006 onward, a slightly different system is used, of which has triple the precision. For dates prior to this that use color coding, from the white color, skipping two to red and so on is used. Here's what each color represents (for updates to before May of 2006):