Hi Hi,

How wonderful to find this Forum, very clear.
I have a urgent question, as Thursday I have to present my final Pure Data Project after three months of PD-course every Thursday.

I need to understand more about the Textfile object.

Basically I need a number output that I determine: example:
1 99 7 6 67
78 2 9 56

These numbers trigger a sound. But the sounds really have to follow one another (as I am simulating making words with sounds/phonemes (which are activated by the numbers 1 99 etc)

I have made a basic textfile, with a number on each line, but I cannot get the patch to read the numbers one after the other - I have tried many variations...
I always only get it to read the first number, or the consequtive numbers that I have to bang each time.

Can somebody explain the textfile somewhat more?

I have put my patch here:
http://www.lo-bat.be/wendy/PZI_Pure_Data/
The main file is called Super_syll.pd and it is in the subfolder Assignment Wendy
thanks
w
(oh, yes, I use Linux -> Pure Dyne)

Hi wendy,

I think the issue is the use of "cr" in read message. I don't know my way around PD that well, but it looks that lists are semi-colon terminated.

I included an example of what I'm talking about.

http://www.pdpatchrepo.info/hurleur/textfile_read.pd

Hi wendy,

I think the issue is the use of "cr" in read message. I don't know my way around PD that well, but it looks that lists are semi-colon terminated.

I included an example of what I'm talking about.

http://www.pdpatchrepo.info/hurleur/textfile_read.pd

Hi, digi-D,
Thanks for your reply!
I understand textfile somewhat more now, but I still have a problem in the sense that my patch needs to go through all the items in my textfile without handgiven bangs.

So basically it needs to go from item 1 in the textfile, automatically to item 2, to item 3 -> hence playing sounds in the order of the list of the textfile.

Pfew, quite abstract to explain, but not so hard to solve I think. But how???

bye!
w

Welcome Wendy,

Your patch is almost working as you hope. The textfile object is reading the list of numbers you want.

If you put a [print] after the

object to debug you will see the problem.

All the numbers are being sequenced very fast (in zero time) so you only hear the last one.

That's because you are using [del] wrong.

Add a [t b] object to only bang the delay and give [del] an argument for the delay between phonemes, eg [del 100].

Of course, a fixed interval between each sample will still sound wrong, but you have everything in place to fix that, just take the next trigger from the right outlet of [tabplay~] instead of using the delay.

Obiw - You are a Genious!

It works!!!

After trying every possible variation I could think of last night, till my eyes were getting bloodshot and only grunts of communication came out (and slight swearing now and then, combined with dramatic exclamations of "why is it not working") I put my last grains of energy towards posting on this forum..

Ahhh,
You will get a honorable mention tomorrow when I have to present my patch.

Yeay!

YVW, please don't mention it.

Thankyou for the phoneme samples (which I shall mot use elsewhere of course, but they are useful)

Good luck with your exam.

Could we get a final patch to look at??

Yes
Still busy though - it will be for tomorrow (it is a totally crazy patch - if you ask me - I have more than 200 arrays!)

good night
w

200? Jeezuhz!

I've often thought of doing something similar to this, except with entire words, and the repetition of arrays was what got me. My theoretical solution was this;

1. Append each sound, one after the other into one giant array (or soundfile)

2. Record a "keyframe" for each sound: it's beginning, end, and length.

3. Use a textfile to index the array; for example

##sound 0 begins at 0 and is 44100ms long, therefore; ends at 44100
0 44100 44100 44100;
##sound 1 begins at 44101 and is 100ms long, therefore; ends at 44201
1 44101 100 44201

Then, you could lookup each line as you needed and index the array using the values;

(int)
|
[goto $1, this(
|
[msgfile]
|

| |
N/A |
|
[$1, $3 $2(
|
[line~]
|
[tabplay~ junk]
|\ |
[dac~]

The problem is, how to get PD to stitch everything together and index it all at the same time for you!

I know

I actually talked about this with my teacher - and it is he who said that the 200 arrays are the only option. (also I already had all the seperate files..) We'll see - just some more numbers to change...

I think I'll have a pure data dream tonight

Must.. Patch...