How to add more voice engines in win-xp-prof


1) How to add more voice engine in MS-Windows-Xp-prof, free of cost? At present it has only MS-Sam installed.
2) Will it help in generating mp3 audio output using the "Daisy Pipeline>creation>simple>narrator " procedure?
3) The output quality of my projects generated using the aforementioned route with MS-Sam TTS voice engine is not as good as is available in the Amis demo project of "Three little Pigs". In this latter project, do they use machine generated audio or it it human dubbed audio?


1) You need a third party TTS engine. Microsoft maintains a non exhaustive list of solutions compatible with the Microsoft Speech API.

2) The audio files produced by the Pipeline Narrator are indeed using the voice configured in the default Windows speech control panel. Both the wav files and mp3 files are generated from this TTS engine (the mp3 are actually encoded from the wav files).

3) Sam is known to not sound very natural. You will get a much better result by using a more professional TTS voice, although I don't know if you can find a good one for free on Windows.
As for the Amis demo project I don't know how they produced the audio files, feel free to ask on the AMIS forum.



The Three Little Pigs book that we have available for testing AMIS was recorded by a human being! I don't know of a TTS engine that is that good (smile).


Ok, Guys, thanks a lot.

It has helped clear a lot of dust in my head.

Daisy Pipeline is good, so is Amis. My complemts to both the teams. Please keep it up.

This combination is also a good tool for my family. It is good for kids who now have a better attention span. It is good for me too, I can now translate all the reports I have to review as audio and keep listening to them as they play in the background, while I attend to other things. In case, a part needs my attention, I can immediately locate it and arrange the needful.

I request you to please add a few good male and female voices as loadable TTS voices in Win-xp. This may not be a very demanding work for you guys, but it will help us a lot.

I have added MS-Mary and MS-Mike through "Sp5TTIntXP.msi". While MS-Mary is bearable, she is too abrupt at times. MS-SAM has a snoring voice that keeps resonating in my head, he eats most of the words. The machine generated audio experience is still poor.



Well! If not demanding, Coming up with new TTS voices is definitely sweating work. Had that been very simple Microsoft would have given lot better quality built in talking characters for their OS. Good news is minds around the work are working on given better quality audio solutions at cheaper rates. So lets wait and hope and contribute if we can. While working on generating automated Daisy with audio production making use of different synthetic voices, I had experience on different voices from difference vendors like AT&T, ScanSoft and Loquendo. The voices by these and many others are much much better in quality as compared to Microsoft voices of Sam and Mary. Still they all have limitations. Interestingly, different voices from the same vendor might not behave in different manner for a specific sentence or even word. For the same rate some of the voices might speak faster than others. Context of a word or a phrase has its own problems. Modern and classical text which contains new slang or obsolete words have different pains, and a lot more ... Generating DAISY with human voices is definitely a quality reading experience but on much higher cost in terms of time and resources. You can have a look at DAISY titles produced using automated process using synthesized voices at You will see that although these are not equivalent to a nice sounding human voice but voice quality for these titles is not bad either as they are applied upon different processing techniques. After hearing them you will surely feel that machine generated output is not very poor either. Enjoy! Zubair Younas.