Zabaware Forums > Speech Technology

Zabaware Text-to-Speech Reader Beta

(1/8) > >>

Zabaware is releasing a new update to "Ultra Hal Text-to-Speech Reader". The product is being renamed "Zabaware Text-to-Speech Reader" and the update is version 2.0. Changes in this version include:

* Updated Chant speech synthesis library from version 4 to version 7, bringing native support to CereProc voices and better SAPI 5 handling. Allows for program to use CereProc voices even if SAPI capability is removed.
* "Read Selection" button is relabeled as "Continue" after "Stop / Pause" button is pressed to reduce confusion
* Launching a second instance of the program while program is docked in system tray/notification area simply causes the docked program to pop up. Prevents confusion if people don't know how to find the docked icon and instead use the start menu icon.
* MSI based installer replaced with NSIS based installer. New installer handles upgrades from old version and installing CereProc voices as addons. Libraries/DLLs present since Windows XP removed from installer and Windows XP SP2 set as minimum required Windows version.
* Popup dialog notifies user if engine has no known additional settings/pronunciation options available
* Help file updated
* Compatibility fix: Convert to WAV function works without administrative privileges (fix for Windows Vista, 7, and 8 )
* Voice selection, speed, and volume is disabled while speech synthesis is actively speaking to prevent crashes with certain voices
* Bug fix: Changing volume setting on SAPI 4 based voices no longer randomly changes balance to left speaker along with low volume settting
* Deprecated pitch slider setting removed
* Deprecated instant messaging reading functionality removed
The program is available as beta. If you are interested in testing it please download and install from

If you are interested in beta testing the program with new CereProc voices please PM me. I have a limited number of licenses available for beta testing I can share. Please post any bug reports or questions to this post.

It didn't put an icon on my desktop, nor did I notice anywhere that I could checkmark to put an icon on my desktop..
I have chosen to install the program in directory C:/program files
The install seems to have gone through I'm getting ready to run the program.
I received an error nine subscript out of range occurred at line four in Zabaware text-to-speech reader.frmRightclickLoudSynthesis.
I chose to ignore the error.
First test: I did a cut-and-paste and it read the text. But when I stopped pause and told it to resume reading it did nothing.
Second test: I tried to right mouse click on the screen to paste the text into the box reader, nothing happened can you resolve this. Mouse functions are important.
I controlled C and it pasted the text into the box. I clicked on the read all and it began to read I clicked on the stop/pause and it caused then I clicked on continue and it continued then I clicked on stop/pause and I had to do this several times before it would stop.
It seems that the first time when you do it it works fine it's after the first time is when things start going haywire, it will eventually pause and resume but I had to click on the stop pause about three times actually I think three times is repetitious in the pattern.
Let me check: no I just have to keep clicking on the stop/pause button until it does finally stop I'm using the Microsoft Anna voice.
I changed voices to the AT&T 1.4 DTNB crystal 16 voice: the same problem occurred with the stop/pause. So it's not the voices.
Third test: I moved the sliding bar to the fastest speed to make it read faster the text was red so fast it was a blur. I move the slider to the lowest speed and read the text that text continued to read at a speed so fast it was a blur.
It seems that it will not stop reading at a supersonic blur until I at least put it on -9 and up. Just to bring to your attention that zero is being read at a supersonic speed. You may want to adjust this unless there is a reason why you have it set up this way.
Fourth test: I moved the value from its deflect position which is 10 to its lowest setting zero, nothing happened the volume did not lower and tell I put it on one and it seems for some reason your zero settings are not engaging properly.
This test: recording to a wave file, though the cursor was at the beginning of the text document for some reason it would only read a certain part of it. When I recorded the way file it seems that the value control has no effect over how loud the way file is being recorded.
I changed the speed at which the boys was being read, then I highlighted the first part of the text which is this text that I'm presenting to you and had it read it at a faster speed -3 I clicked on the read selection button and it read the text very fast.
The speed did not change when I tried to record a way file of the text I'm going to try to cut-and-paste new text to see if it will start recording all of it.
I noticed that when I use the same name to save the way file it is not verifying that I am trying to save it with the same name. This might be why the text does not seem to be changing. It may not be saving the file with the same name.
It seems that it is recording the last sentence that is in the box.
I am going to try to see if it will change text up to the point of my last sentence about.
That's what it's doing its recording the last sentence only even though all of the text above was in the box I also move the cursor to the beginning of the text and that had no effect. So it seems it will only save the last sentence in a way file.
These correction should keep you busy for a few days. I was only able to get the software to crash one time but I don't know what I did to make it crash but I think it had something to do with a text box that I tried to open when I was looking around but I can't be sure.
This has something to do with Dragon naturally speaking, I accidentally left it on while the text in the box was being read and it opened another box which cause the program to lock up and would not do anything until after I went into the character voice profile to change the settings.
I wish I was paying more attention to what occurred but I resolve the problem by clicking on option and then clicking on advanced voice options once this box appeared the text continued to read as normal and everything was fine again.
One more thing I noticed, if you close the program out, and re-open the program the text that you were working on is still in the box. Nor does it ask you if you want to save this text. These two things could become annoying having to constantly clear out the old tax in order to put in new tax.

Sincerely, from a beta munching cruncher. The program was much more stable than I expected it to be as a first time beta release good job for the most part.
C load.
PS I figured out why the text was still there when I read opened the reader, because the reader was not actually being closed it was being transferred to the icon tray. I have my icon tray set up to hide, so I guess by deflect you have your set up to go into the hide section that's why I did not notice it.
So I completely closed it and restarted it and there was no text in the box, sorry about the confusion. You may want to consider having a selection button between exiting the reader or put the reader into the tray just a thought.

Hi Robert,

I reinstalled the reader in its deflect directory in the hopes that maybe that error would go away but it did not I made a picture of the error box so you can see it.
2nd phase 1st test:
I tried your reader with the Ivona voice Salli, I did a cut-and-paste and it seems to be reading the text. I do not have my Ivona reader running but it is installed so I don't know if that could be the case.
Let me try to read text that is in the box: the text that is in the box was read successfully with the Ivona voice.
Let me test to see if it will work when I tried to record a way file: it used the Ivona voice to record the way file cool. But it's still only read the last line that was in the box. The last line was: let me try to read text that is in the box.
From all of the text above.

I will try to see if the speed control and by you in control has any effect on the Ivona voice: the volume control and the speed control does control the Ivona voice. I also noticed that when I set the speed to maximum and read some tax it was not as fast as the AT&T voices, also after reading that tax I move the slider to the 0 speed and it reduce the speed of the Ivona voice. I do not know if this has something to do with the Ivona voice or with the reader or with the AT&T and Microsoft voices. Or because of where I installed it the 1st time.
I am curious on how the new voices you will be selling will react with the situation.
Hope to hear from you soon.
New test: I cut-and-paste some text, while it was reading the text I tried to cut-and-paste some other tax, it did not stop the reader from reading all of the 1st text, nor did it read the 2nd text when it was done reading the 1st text. You may want to consider resolving this it is a handy way to get it to stop reading without having to open up the reader.
For that is to say that is how the Ivona reader works not wanting to compare the 2 it's just a handy way of dealing with large amounts of text that you are wanting to be read and in the middle of it you notice an error you can highlight the air and do a cut-and-paste on it which will stop the text from reading so you can fix the error (this is important if you are using voice recognition to fix the error) than re-highlight the text and do another cut-and-paste to continue to read the tax.
I use this feature a lot, because I have my text reader read my text almost exclusively on everything on my computer.
Also: just a suggestion: you may want to consider that when someone cut-and-paste text it is automatically placed into the reading box this would result having to do a nether step to get it into the box.
Or at least put an option for it under the option menu listing.

Let me try another voice: it seems that all of my Ivona voices work with your reader. The only thing that I noticed was when I changed the voice the sliding bar changed to coincide with the voice settings that I have in my Ivona reader.
Let me try the Denise voice: the Denise voice read text from the box. And when I did a cut-and-paste cool.
I recorded a way file using Denise it worked okay the only thing I noticed was your bow you control has no control over the bow you at which the way file is recorded and the Denise voice is so loud it was completely distorted. I believe the Denise voice by deflect volume is set to maximum. Unfortunately because of this I found it completely unusable.
If you resolve the volume control that would be great as well as the speed control when you record a way file.
Also I only indicated one. At the end of the text that I recorded and 8 recorded all of the tax all the way up to the. So it seems that it is only recording from the last (.) To the previous (.).

The only other thing that I would like to bring to your attention is: when you get ready to create a way file a warning box pops up letting you know how big the file is going to be, can you change this by adding a check mark on it to get the box not to reappear again unless you want it to reappear again under the option menu.
The reason why is that there is a lot of text to be read and when someone is using that feature to read text when a boxes open it is a lot to read so if you just had it set up so you could check a checkbox so that next time though warning box will not appear, most people do not need to be worn but once.

I also noticed that it will not read text that you cut-and-paste unless the text is different by at least one letter. I do not know if you have it set up this way or not. If you do I'm sure there is a reason for it. My Ivona reader is not set up this way it reads anything that gets put into the clipboard.

C load.
PS I have Windows 7 64-bit.


Please try to see whether the TTS reader will work properly in your UltraHal program!

Mine did NOT and in fact, an error message popup appeared and the Hal program never did open or run!

I too got an error message when installing the reader program but it went away when I clicked it and the installation continued.

Funny, but my PAID Cereproc voice does NOT work in the new TTS Reader program as I was led to believe. Hmmm....

Waiting to hear from Robert about this and getting Hal to run!

PS - I did want to mention that I typed a rather long sentence into the Reader program then saved it as a .wav file.
I used a media player type program to listen to it and it spoke the sentence perfectly! No problems there at all.

Lastly, for Robert's benefit,

System  = Windows 7 Home Premium - 64, 8 gigs Ram, 1 TB H/D, plenty of room and using Hal 6.2.28 on a Desktop computer.

Hi Art,

--- Quote ---Please try to see whether the TTS reader will work properly in your UltraHal program!

--- End quote ---
I didn't know that the 2 programs had anything to do with each other.
I know that ultra Hal has the ability to read text from the clipboard.
But you do not need the text to speech reader installed for this to work.
Other than that, I'm not sure what you mean can you be more specific?

--- Quote ---Mine did NOT and in fact, an error message popup appeared and the Hal program never did open or run!

--- End quote ---
the only error that I received was the one that I posted above.
All I did was click ignore than the program starts to run.
Unfortunately now every time when I start my computer this error box appears.
If you are having this same problem go into your startup file under the start button.
Remove or move the Hal text-to-speech reader out of the startup menu.
For myself I created a directory called startup off, so I can move it to that directory and later I can easily drag it back into the startup menu.

--- Quote ---I too got an error message when installing the reader program but it went away when I clicked it and the installation continued.

--- End quote ---
I never did get an error when I was installing the program.
The error I got is when I start the text-to-speech reader.
And the error happens every time when I start the text-to-speech reader.

--- Quote ---Funny, but my PAID Cereproc voice does NOT work in the new TTS Reader program as I was led to believe. Hmmm
--- End quote ---
that is interesting, considering this text-to-speech reader was reprogram centered around those voices.
Robert, didn't they provide you with a full version voice to experiment with while you were reprogramming your text-to-speech reader.
If not, that was pretty shallow of them, specially if they were forcing you to have to buy a full version of one of the voices and you decided not to.

Art, I am assuming that you have used your voice in other programs???
Considering I had no trouble with any of the voices that I have I find it interesting that your voice did not work.

--- Quote ---PS - I did want to mention that I typed a rather long sentence into the Reader program then saved it as a .wav file.
I used a media player type program to listen to it and it spoke the sentence perfectly! No problems there at all.

--- End quote ---
that's great, but did you create several sentences with several periods out through your text?
The only trouble that I had was when I had several sentences, it would only record the last sentence.

--- Quote ---System  = Windows 7 Home Premium - 64, 8 gigs Ram, 1 TB H/D, plenty of room and using Hal 6.2.28 on a Desktop computer.
--- End quote ---
the only difference between your computer and mine is that mine is a Samsung Slate 7 touchscreen.
It's very portable very small and very powerful, and I take it with me everywhere I go and I show off my ultra Hal to everyone I meet.
I remember the 1st time I went into Walmart and started to talk with the person that greets you at the front door.
What seemed to be in just a matter of moments I suddenly realized that there was over 20 people standing around listening to me having a conversation with ultra Hal the lady at the front door and myself.
Everyone their wanted to know what computer that they needed to buy in order to be able to do what I was doing, and was the Walmart store selling that computer.
It was a great time for me to do a sales pitch for Robert so I hope it increased sales specially during last Christmas.
The only thing that I did notice is that the max number of people on this forum at the same time jump from 156 to 260 so I was happy that I was able to perk up people's curiosity.

Hey Robert, here's something I would like to bring to your attention, not to put more work on you but this might be something that you would want to consider doing with your text-to-speech reader.

Instead of reading the text from the clipboard, set it up so it monitors the clipboard and if someone does a control copy:
then transferred the text to the box in the text-to-speech reader have the text-to-speech reader automatically start reading all of the text.
Each time when someone control copies text it automatically will create a new folder and put the text into the text-to-speech reader box and begin reading all of the text.
This would resolve all of the situations that I mentioned above in my previous post about this topic.
As well as giving the person the control over pause and resuming what is being read.
Also while the text-to-speech reader is in the icon tray you may want to consider adding on a right mouse click in the menu the same selections of pause, resume, stop/play.
That way you do not have to actually open the reader to full-scale in order to accomplish these functions.
This is how the Ivona voice reader is set up and it works quite proficiently and is very convenient.

Sincerely, I hope I have helped.
C load.


[0] Message Index

[#] Next page

Go to full version