It's About Time C

[Jazzy09]

yeah mine lasts 5 minutes easy

[saturnian]

Did anyone else notice that sound files provided by national site run slightly too fast on Windows Media Player?
I tried them on two computers and compared with stop watches and they seem to be off by about a second on a 100 second file. Other programs, like QuickTime or RealPlayer, play right on a mark.

[Camden SO 09]

how do you go about studying for its about time?
i know that we need to know some physics equations but what else?
but mostly, do you guys know how to calculate sidereal time and where can i find out how to learn about it?

~Leigh & Jeff~

[Sheogorath]

I'm pretty sure you don't need to know the equations as I do believe in the rules it says all equations and formulas will be provided.

[Jazzy09]

so we wont need to know how to convert to sidereal time? i think there was a practice question on that

are there any practice tests out there for its about time?

[fleet130]

Did anyone else notice that sound files provided by national site run slightly too fast on Windows Media Player?
I tried them on two computers and compared with stop watches and they seem to be off by about a second on a 100 second file. Other programs, like QuickTime or RealPlayer, play right on a mark.

See FAQs on soinc.org

[saturnian]

See FAQs on soinc.org

Yes, we called national S.O. to point this out.

[binary010101]

We just had an invitational on Thursday (Jan 8th). I notice that many of the other teams had very complex clocks: stuff like escapement clocks and water clocks. On the other hand, we just had a pendulum. We still beat them, though.

[fleet130]

And I brought it to their attention also. I'm don't know if the method used to generate the time intervals involved traceability to any time standard. Even if the tones were generated by a calibrated source, the method used to record them may not have been calibrated, and the method of playback certainly is not known for its timebase accuracy.

Without the ability to "trace" the accuracy back to a known calibrated standard, it's impossible to predict the probability for error. Keeping the time intervals accurate throughout the generation, record and playback processes can be quite involved.

[chalker]

And I brought it to their attention also. I'm don't know if the method used to generate the time intervals involved traceability to any time standard. Even if the tones were generated by a calibrated source, the method used to record them may not have been calibrated, and the method of playback certainly is not known for its timebase accuracy.

Without the ability to "trace" the accuracy back to a known calibrated standard, it's impossible to predict the probability for error. Keeping the time intervals accurate throughout the generation, record and playback processes can be quite involved.

I'm the National Event Supervisor for It's About TIme and was intimately involved in rewriting the 'old' rules for this year. This discussion thread was recently brought to my attention (yes we monitor this message board at a national level), and while we've posted an FAQ on the national website (which someone already linked previously), I'd like to provide some additional details and hopefully bring some closure to any concern over issues related to the audio files.

I synthetically created all of the audio files posted on the website and can assured you they are all accurate and far exceed the precision required by the event rules (0.1s). There was no 'source' nor 'recording method' and thus no concerns over the calibration of such. Of course since we can't control the playback hardware / software configurations, there may be errors introduced there, which was the main message in the FAQ we posted. As an aside, I don't have any problem playing them under Windows Media Player on my system, and have checked them against a stopwatch several times.

For those of you interested in the specific details of how I created them, I utilized a program called MATLAB (which I highly recommend to anyone doing any type of programming related to science and engineering - you can try out the open-source alternative called OCTAVE if you'd like). MATLAB does everything utilizing matrices. I created a program that for each audio file constructed a large matrix as follows:

-8khz sampling rate (i.e. every 8000 elements correspond to 1 second of time), which taking into account Nyquist frequency, technically produces a resolution of 0.25 ms. Put another way, to be off by 0.1 sec (the required resolution according to the rules), my program would have had to erred by 800 elements one way or the other, which would have been immediately noticable algorithmically.

-The initial 0.5 s (or 4000 elements) are comprised of a 440 hz 'gong', which is a mathematically derived waveform based upon a special formula involving sin, cos, and exp.

-This is followed by 1.5 s of complete silence (0s), and another 440 hz 'gong'.

-These are followed by a 0.25 s long 880 hz 'beep' (synthetically created the same way as the gong), the very first element of which signifies the start of the timing period.

-Next is a variable duration of complete silence (the target time period - 0.25s to account for the beep), followed by a stop 'beep' (0.5 s long, 880 hz). The actual 'elapsed' time is over at exactly the start of the stop beep.

-My MATLAB program constructed these large matrices, then wrote them out to a mp3 file using an API for the LAME encoder. While I could have directly written them to raw wav files, those are uncompressed and would take up a LOT of space. The mp3's are VERY compact because the majority of the files are complete silence.

-For those of you wondering about the accuracy of the mp3 files, they were also encoded at 8 kHZ (in order to ensure there was no sampling error introduced during that process). MP3s essentially convert the time-domain samples into frequency domain samples, typically 576 samples at a time. Since there is the majority of the source matrices consisted of 0s, the majority of the MP3 frames also consist of 0s.

I hope this helps clear up any concern over the audio files themselves. As indicated in the FAQ, I highly recommend you cross-check your specific hardware/software configuration against another accurate timing device to ensure you aren't introducing any errors inadvertently.

Good luck everyone and I hope to see you in Augusta in May!