View Full Version : One Day Sale - Text2Go converts text to iPod speech mp3s


Bob Russell
10-09-2007, 08:01 AM
Today's one day sale (http://www.bitsdujour.com/index.aspx?review=88) on BitsDuJour is for Text2Go. It's reduced to $5 from the regular price of $25.

"Text2Go is an innovative application that lets you transfer information from the web to your iPod. Listen to any text content on-the-go and utilize moments that would otherwise be wasted. Make productive use of those 'dead' times like commuting and exercising. Text2Go integrates with Internet Explorer and iTunes, making it the quickest and easiest way to get information from the web onto your iPod." There is an mp3 sample of the text conversion here (http://www.text2go.com/portals/0/SpeechSamples/Text2GoBlurb.mp3), which is based on the Microsoft speech engines.

For someone that wants to do this "the easy way", it might be worth looking into. But I bet that hard core tech fans can find many other ways of doing the same thing.

NatCh
10-09-2007, 11:12 AM
Well, that sample's pretty impressive, certainly worth $5 I'd think. I'll have to look and see if they offer other voices. The sampled one is fine, but I'm always curious about such things. :nice:

kovidgoyal
10-09-2007, 03:15 PM
There are free text to speech converters. festival + mbrola for one. Dunno how easy it is to get working in windows though.

NatCh
10-09-2007, 03:41 PM
I was less excited when I realized that it seems to require Internet Explorer to work. :shrug:

jbenny
10-09-2007, 07:00 PM
Before anyone buys this software, thinking that they are going to get the voice that was used in the sample, note that the program comes with no voices and uses the TTS voices that come with Windows (Mike, Mary, Sam). The program is "compatible with the incredibly natural-sounding RealSpeak speech modules", which means that you can buy these nicer sounding voices, as an option. This is true of just about any Windows SAPI5 TTS program - you can buy additional voices elsewhere.

Although $5 is probably a good price for this program, with only the default SAPI5 voices, you may as well save the $5 and get one of the free TTS programs that are available. One that I would recommend is called DSpeech and can be found here http://dimio.altervista.org/eng/. It also allows you to create WAV/MP3 files.

If you later wish to buy additional, better sounding TTS voices, the best place that I know of is Nextup.com, which sells AT&T, NeoSpeech, Acapela and RealSpeak voices (they may also still sell Cepstral voices, but they aren't listed on the web page).

vivaldirules
10-09-2007, 07:53 PM
Let me second that suggestion for caution. I've used similar software and can attest to the importance to use the right voice which also determines the price you'll pay. If you're interested in a comparison, go here http://www.naturalreaders.com/sample.htm# and try the demos for NaturalSoft's software. Their free version comes with Microsoft's Mary voice which is awful. An AT&T voice which I found I can't listen to for more than a few minutes before giving up will require that you buy the $50 version of NaturalSoft's product. For the really good NeoSpeech voice, you have to buy the $100 version. :(

jbenny
10-09-2007, 11:44 PM
Not to sound like a commercial for Nextup.com, but you can buy NeoSpeech and other good voices from them for a fair price. I got both Kate and Paul (NeoSpeech) for $35.

BTW, the Microsoft voices Mike, Mary and Sam are built-in to Windows XP. I understand that Vista has an additional, better sounding voice. I haven't heard it. If you have an earlier version of Windows, or for some reason don't have them on XP, you want to install the SAPI 5.1 package, which you can download from MS. This package includes those three voices, as well as the speech API files that other programs use.

vivaldirules
10-10-2007, 07:09 AM
Not to sound like a commercial for Nextup.com, but you can buy NeoSpeech and other good voices from them for a fair price. I got both Kate and Paul (NeoSpeech) for $35.

That's good to know. At that price I may try this again. I'd like to use my hour per day commute doing something more useful than just getting to work and back. Thanks. :)

Nightwing
10-10-2007, 11:05 AM
The program that take text to speech to wave/mp3 are not hard to to create... But the voices are the key... The Mac has and will be getting some better voices. The PC has had cummey one and just now getting decent ones...

A grand find..

JSWolf
10-11-2007, 03:10 PM
So someone summarize please. What is the best text to speech program for Windows and what is/are the best voice(s)?

jbenny
10-11-2007, 04:44 PM
So someone summarize please. What is the best text to speech program for Windows and what is/are the best voice(s)?

You can't make a blanket statement that one vendor's voices are better than another, because individual voices from each vendor sound better or worse than voices from other vendors. You have to listen to each voice from each vendor and decide what you like. To facilitate that, here are the web pages from each vendor, where you can enter your own text and hear what the voices sound like.

http://www.research.att.com/~ttsweb/tts/demo.php
http://www.neospeech.com/demo/demo_text.php
http://www.acapela-group.com/demos/demos.asp
http://212.8.184.250/tts/demo_login.jsp
http://cepstral.com/demos/

I'll give you my opinion, but only you can decide which sounds best to you. I think the NeoSpeech voices are better sounding than the AT&T Voices, even though I bought some of each. Lots of people like some of the new voices from Acapela and RealSpeak. I admit that they generally have better inflection, but sometimes, even that inflection sounds funny. Also, I think that these vendors voices have too much silibance. Again, my opinion.

One other issue is how accurately each voice pronounces words. In my testing, I found that the AT&T and NeoSpeech voices were the most accurate with pronounciation.

As to which TTS software to use, I tried several and again, offer my opinion. If you want a free program, DSpeech is very good. For a commercial program, I would recommend TextAloud from Nextup.com. It has several features that the free programs (and many commercial ones) don't have.

As for actually buying voices, the only place that I have found that has a wide selection and a good price, all in one location is again, Nextup.com. One important caution - the Acapela voices sold by Nextup.com only work with their software, as required by license. I know the AT&T and NeoSpeech voices work with any TTS software. I'm not sure about the others.

If you really want a recommendation, I would say you couldn't go wrong buying the NeoSpeech Kate and Paul voices and TextAloud. The only drawback with NeoSpeech is that they don't offer a wide selection of other voices. Their non-English voices are all oriental.

vivaldirules
10-11-2007, 05:44 PM
I found two disturbing aspects with the software I bought a few years back. One was inflection. The software I bought did not change inflection much. So questions sounded similar to statements. Another was that it wouldn't stop or even pause at the end of a sentence. Those sound like pretty glaring faults and they were. But I discovered them only after I had paid for it. Bottom line: test the program as much as you can before you pay for it.

jbenny
10-11-2007, 06:37 PM
I found two disturbing aspects with the software I bought a few years back. One was inflection. The software I bought did not change inflection much. So questions sounded similar to statements. Another was that it wouldn't stop or even pause at the end of a sentence. Those sound like pretty glaring faults and they were. But I discovered them only after I had paid for it. Bottom line: test the program as much as you can before you pay for it.

Inflection is up to the voice, not the software. The pauses at end of line, commas, etc. can be changed in the software, if the software supports it (TextAloud does).

Even the best sounding voices are not going to be confused for a real person. Given that, several of the newer voices sound very nice and are very usable.

Just as an example for TextAloud: in the Pronounciation Editor, I entered a regular expression to change the amount of pause at a comma (followed by a space).

Word: {{re=(,\s)}}
Pronounciation: {{Pause=0.1}}

Using the Pronounciation Editor affects all installed voices. If you want to only change the pronounciation of certain words in a specific voice, AT&T and NeoSpeech each have dictionary editors that only affect their own voices. I assume that other vendors have something similar.

markgladding
10-11-2007, 06:42 PM
Firstly the speech sample in Bob's post is produced using the RealSpeak Samantha voice, NOT the built in Microsoft voices as noted by jbenny.

As many people have commented, you really need to use high quality voices or the results will sound terrible.

Text2Go's normal price is $25 and RealSpeak voices are $35 each. However I strongly encourage people to purchase Text2Go and at lease one RealSpeak voice. Therefore I offer a package deal that includes Text2Go + 1 RealSpeak voice for $45, effectively pricing Text2Go at $15 and a voice at $30.

As to which text to speech product is best, it depends on your needs. There are now quite a few products available.

Some focus on reading text while at your computer and others are specifically targeted at the visually impaired.

Text2Go is targeted at iPod owners who want to listen to text from the web while out and about. It tries to make the process of converting text to speech and transferring it to your iPod as quick and easy as possible.

If you're interested, I encourage you to take advantage of our 30 day trial. You can also download the RealSpeak Samantha voice for use during the trial.

If you have any further questions, don't hesitate to ask here or email me directly.

Regards,

Mark Gladding
Director
Tumbywood Software
http://www.text2go.com

jbenny
10-12-2007, 01:27 AM
Firstly the speech sample in Bob's post is produced using the RealSpeak Samantha voice, NOT the built in Microsoft voices as noted by jbenny.


If you're going to quote me, then get it right. I did not say that the sample posted was the using a Microsoft voice. In fact, I said that your program, did not come with any voices and by default, used the MS voices. I knew perfectly well that the sample posted used a RealSpeak voice, which you charge extra for. This was not a slight at your software, as this practice is common (which I pointed out in my original post). However, in the case of the BitsDuJoir posting, the impression was given that the sample voice came with the software, as evidenced by comments from other posters in this thread.

My point in posting was to clear up the confusion, so that people who bought your program did not think that they were getting the voice used in the sample at the $5 or even the $25 price. I have never used your software and have no bias against it. I was simply tying to clear up the confusion created by the posting at BitsDuJoir.

When others here asked for a recommendation, I mentioned Nextup.com, since of the several TTS programs that I have tried, I was pleased with their software and thought it was a good deal. Not having used your software, I obviously could not recommend it.

markgladding
10-12-2007, 07:32 AM
Sorry jbenny,

My previous post was poorly worded. I was actually trying to credit you as the first to pick up the inaccuracy in Bob's post.

In regards to your previous posts, you have provided a lot of really useful, factual information for people. Anyone who knows how to use regular expressions is a power user in my books.

Regards,

Mark Gladding.

HarryT
10-12-2007, 07:46 AM
Hi Mark,

A few questions:

1. Can this software only be used with web sites or can it, for example, be used to read aloud an RTF eBook loaded into Microsoft Word?

2. Is it "real time"? ie, if I have an eBook which takes 16h to read, will it take 16h to create the output?

3. If I have a long file (such as an eBook) is there provision to automatically split the output into multiple tracks?

Thanks!

markgladding
10-12-2007, 07:25 PM
Hi Mark,

A few questions:

1. Can this software only be used with web sites or can it, for example, be used to read aloud an RTF eBook loaded into Microsoft Word?

2. Is it "real time"? ie, if I have an eBook which takes 16h to read, will it take 16h to create the output?

3. If I have a long file (such as an eBook) is there provision to automatically split the output into multiple tracks?

Thanks!

Hello Harry,

In answer to your questions.

1. Text2Go can convert other sources of text to speech. You just copy it to the clipboard and then use Text2Go's 'Add from Clipboard' command. If you want a document in Word or Acrobat read aloud to you, you're better just to use either Word or Acrobat's built in speech commands. Note that you if you purchase a RealSpeak voice for use with Text2Go, you can use that voice with other programs.

2. The speech conversion is much faster than real time and depends only on the speed of your PC. For example my PC which is a couple of years old will convert text to speech at least 10x. With Text2Go, once you've started a conversion, you can browse to a new web page, select some text and begin another conversion. It will queue up individual conversion jobs and process them one at a time in the background.

3. There is no provision in Text2Go to do what you suggest. However it's a good idea and I will look at how to add that for a future release. What I tend to do is to convert a chapter at a time which will result in one track per chapter. I group them all together by creating a Play list for eBook title.

Regards,

Mark.

HarryT
10-13-2007, 03:55 AM
Thanks, Mark.

Bob Russell
12-20-2007, 06:46 AM
FYI, today (Dec 20, 2007), there is a one-day giveaway of TextAloud...
http://www.giveawayoftheday.com/