Scratch

note my analysis now added to above post

Last edited by sparks (2011-01-06 16:12:14)

On a MacBook pro, full volume, barely any ambient noise, no jack-to-jack cable, it only starts picking things up around volume 40

I'm thinking we should have a script that puts out a tone of known volume and then asks you to line them up until the icon is green. Gonna try that tomorrow. Too tired right now.

I think some more experimentation should be done, after all, I've found two different instruments that are of different physical volume on the same scratch volume setting.

Sparks: If someone wants to look at Scratch's ability to measure duration, e.g. smallest duration registered as correct volume, smallest difference noticed between durations etc.

"Herbert S. Zim, in his classic introductory cryptography text "Codes and Secret Writing", gives the English letter frequency sequence as "ETAON RISHD LFCMU GYPWB VKXJQ Z", the most common letter pairs as "TH HE AN RE ER IN ON AT ND ST ES EN OF TE ED OR TI HI AS TO", and the most common doubled letters as "LL EE SS OO TT FF RR NN PP CC".[1]"

Source: Wikipedia.

This might be worth using. The order of characters should let us put the most commonly used in the shortest duration, with the greatest accuracy.

The double characters and letter pairs might, providing the tone is still shorter than two of the shortest tones, allow us to speed transfer up!

This does not, however look at other ascii characters. (space might be used more than e for example)

this website claims there are 255 ascii characters in total. I suppose that makes sense as 255 is a value commonly seen in binary. I don't think Scratch supports all these characters though. Certainly, Some of them cannot be typed in, but some cannot be typed but manually pasted, and some appear as a small square but, when recopied, retain their original character image again.

Last edited by sparks (2011-01-06 16:52:19)

do you have a large family or some young siblings then? 

Ooh, can I help? I have a mic, speakers, AND an audio cable! 

I... Don't get anything, even knowing quite a bit about sound stuff.    Maybe it's because it's 1:00 AM here right now and I read everything quickly, lol. Indeed it sounds interesting, but I think there'll be many inaccuracies using note duration (lag while counting time) or loudness (particularly when using the mic-speaker combo). But then it'll take lots of time to transfer. Maybe somehow let the user find an equilibrium between "data quality" (less quality - more possible glitches) and time needed to transfer that data. 

I have a good amount of sound cables, a semi-decent mic and some audio understanding (from a musician's point of view at least    ), I'll see if I can help tomorrow. If I can't manage to understand how it works I can help testing 

Last edited by technoguyx (2011-01-07 13:22:27)

Let me give you some more maths.
For now, just imagine we are using these figures.
5 possible volumes (0-20, 21-40, 41-60, 61-80, 81-100)
2 possible tone lengths (0.3 seconds and 0.6 seconds)
A fairly quiet place
A calibration method
32 characters to transmit -- The string "hello, i am playing with my dog."

for every volume, there are two possibilities of tone length. That makes (i beleive) 10 combinations. to reach the total of 255 characters while excluding some useless ones just to round it nicely, we need a number of characters able to fit it perfectly.

Forgive my puny trial and error strategy, but I'm tired and cannot think properly.
1 tones per character: 10 possibilities
2 tones per character: 10² possibilities = 100
3 tones per character: 10³ possibilities = 1000 characters possible??

ok, forget the tone length then. Have it a single length so:
1 -> 5
2 -> 5² = 25
3 -> 5³ = 125
I'm happy with 125, just enough for the common letters and symbols.

How does it seem to you? 

Oh, and goodnight 

Last edited by LS97 (2011-01-07 16:32:11)

I like your bolded idea, Sparks!
I also like how you wrote "1, 2, 3, 4, and five" 

Not sure if this is relevant but my brother is a radio amateur..and i am sure data for computer programs has been broadcast via radio..before now..anyway that is all..
really.

Can I help? I'm a good "encoder"

P.S. This is completely off-topic, but I made a way to do you 'image to text' Panther project to Scracth. It's in the Misc. forums.

Last edited by Hardmath123 (2011-01-10 11:14:43)

Hardmath123, Could you link to it? it sounds interesting!