Jump to content
Asmusr

CALL SAY("GOOD")

Recommended Posts

When I issue a CALL SAY("GOOD") from XB it sounds something like "illip" although "good" is supposed to be a word in resident vocabulary. Is this a know issue or are my ears playing tricks on me?

Share this post


Link to post
Share on other sites

This is an interesting effect. My father once said exactly the same. Me, I can hear "good", although it is not well pronounced.

 

The individual perception can sometimes be surprisingly different. I remember that discussion on the net where people saw a dress either as blue and brown, or as white and golden.

Share this post


Link to post
Share on other sites

It's also interesting that in the resident phrase "GOOD WORK" that I used in TI Scramble the "good" sounds perfectly fine to me. I don't think you can say this combined phrase from XB, if you use CALL SAY("GOOD WORK") it comes out as "Illip work" :) so it uses the individual words.

Share this post


Link to post
Share on other sites

Take a look at the video below...

 

When "GOOD" is played by itself, I can definitely hear the "illip" thing in there... The hard "G" sound is not enunciated well.

 

When you put it into a sentence (second part of the video) it seems clean and clear as day. I don't know. :) What do you think?

 

 

http://www.youtube.com/watch?v=O4u9fc631LA&feature=youtu.be

Share this post


Link to post
Share on other sites

Try CALL SAY("#GOOD WORK#"). ;)

 

 

What in the....

 

That is something I had never done before. Sounds 10x better!

  • Like 1

Share this post


Link to post
Share on other sites

Try CALL SAY("#GOOD WORK#"). ;)

 

Cool, so there is a way to say the resident phrase.

Share this post


Link to post
Share on other sites

The multi-word vocabulary entries require at least a "#" in front (I never used the trailing #). These are completely new LPC samples and are not related to the single word entries.

 

My favorite, by the way, is "#NICE TRY". Also, "#TEXAS INSTRUMENTS" is well done.

  • Like 1

Share this post


Link to post
Share on other sites

When "GOOD" is played by itself, I can definitely hear the "illip" thing in there... The hard "G" sound is not enunciated well.

When you put it into a sentence (second part of the video) it seems clean and clear as day. I don't know. :) What do you think?

 

 

I don't hear the "illip" thing at all in either version, to be honest.

  • Like 2

Share this post


Link to post
Share on other sites

The multi-word vocabulary entries require at least a "#" in front (I never used the trailing #). These are completely new LPC samples and are not related to the single word entries.

 

My favorite, by the way, is "#NICE TRY". Also, "#TEXAS INSTRUMENTS" is well done.

 

You only need the trailing pound sign if you are including additional words in the string. Like: "#NICE TRY# PLAYER"

 

Remember you can reduce the delay between words too, by using + -- compare "PLAYER ONE" to "PLAYER+ONE".

 

(Caveat: I can't actually check right now if those are in the resident vocabulary, but any two words that work ;) )

  • Like 2

Share this post


Link to post
Share on other sites

 

I don't hear the "illip" thing at all in either version, to be honest.

Me neither. After all these years, the speech synth remains my favourite thing about the TI :)

 

I can tell you a little story about that TI speech synth. Back in 1989, I tought my girlfriend to read using the TI speech synth. She had been brought up in children's homes and had had a rough upbringing and relatively little access to education. She had real trouble with reading and writing, though it was clear to anyone that spent any time with her that she was very intelligent indeed, just hadn't had the access to education. She was rather ashamed of her reading skills, and didn't like to sit one-on-one to practice (not reading, anyway! :-o ) so I wrote her a program in TI-BASIC that taught her to spell using the speech synth. It would say a word and she would have to spell it out. If she didn't get it after three tries it would spell it out character by character. It kept a score, and recorded her scores to disk so she could see her progress.

 

She used the hell out of that program. She would sit with headphones on and go right through it. Once she had a basic grasp of reading, a basic nucleus, so to speak, her reading abilities just exploded - her natural intelligence just took over and she did the rest herself. About three years ago she went back to college and did her exams. She passed them all. She publicly credits me with teaching her how to read. Bless her.

  • Like 9

Share this post


Link to post
Share on other sites

Surely I'm doing something wrong, but my MESS just won't talk at all. When I type

.

CALL SAY("GOOD")

.

in Ext BASIC all I get is silence. :( The speech synthesizer does work with Parsec, though. Options are like this:

.

$ mess64 ti99_4ae -peb:slot2 32kmem -peb:slot3 speech -peb:slot8 tifdc

.

Sound sliders are all OK, still nothing in Ext BASIC. :???:

 

Was there an update to ti99_speech*.zip? My MESS is at version 166, but I'm still using the 161 BIOS.

 

EDIT: Yes, I see there was ... Never mind then, time to update!

Edited by ralphb

Share this post


Link to post
Share on other sites

Wait wait wait wait ... currently bisecting git.

 

You can make it speak if you first do a CALL SPGET, say that string directly, then do the usual CALL SAY.

Share this post


Link to post
Share on other sites

Wait wait wait wait ... currently bisecting git.

 

You can make it speak if you first do a CALL SPGET, say that string directly, then do the usual CALL SAY.

 

Yep, that works:

.

CALL SPGET("GOOD",A$)
CALL SAY(,A$)

.

If you do

.

CALL SAY(A$)

.

though (without the comma), the result is hilarious: it'll keep repeating "uh-oh" over and over and over ... :rolling:

Edited by ralphb

Share this post


Link to post
Share on other sites

Found the culprit:

 

 

commit 12d07f91741615bb811e122ffc15ce8cf39f6266
Date:   Thu Aug 27 19:08:56 2015 -0400
    TMS5220: implemented talk status state machine properly as shown by patent. Got rid of m_target_* hack in favor of loading data from ROM as needed. Fixed ZPAR logic. Fixed pitch zeroing to match(?) patent.

 

I just contacted the person who submitted that commit, we'll see what he can do about it.

  • Like 1

Share this post


Link to post
Share on other sites

Problem has been fixed; please wait for the upcoming 0.167 release.

 

By the way, if not already noticed, MAME releases are scheduled once per month, so you have a good chance to get the latest version in at most four weeks.

  • Like 2

Share this post


Link to post
Share on other sites

Problem has been fixed; please wait for the upcoming 0.167 release.

 

By the way, if not already noticed, MAME releases are scheduled once per month, so you have a good chance to get the latest version in at most four weeks.

 

Nice. Glad you found this culprit!

Share this post


Link to post
Share on other sites

Nice. Glad you found this culprit!

 

Yes, but I did not fix it. It was the person who broke it ... just as it should be. :)

 

At least I was able to provide him with an operation sequence to reproduce the problem, and that is already halfway to go.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...