Voice Reader Speech Central

What TTS voices does it use? Are they any good?

Thanks for the response.

Hello,

please, would it be possible to make the Find feature which lets the user search a text in the content of an opened book accessible with VoiceOver? It would be great if the feature, instead of the current implementation, displayed a list of found items where each item would contain a snippet of the text the search term was found inside. Then double-tapping on some item would move the reading position to that occurrence in the book and started playing. Basically, to make it work as in Voice Dream Reader.

The main reason why this could be helpful is that plain text e-books in TXT format typically use some symbols, such as "#" or "X" to denote chapters, and therefore by searching those symbols, one can get a list of chapters. in the book. Or perhaps even better, this feature for listing chapters in TXT file could be separate from the Find feature., so that one could use both the Find and Chapters list feature independently. In any way it would be nice to make the find feature accessible in the first place, similarly as I described.

Best regards

Adam

I did have a plan to add the "Find All" feature in the future which works as you have described, so I'll put it at a higher priority. This was inspired more by apps like Microsoft Word and Adobe Reader though I guess it works in a similar way in all apps that have this feature.

But I think that there might be also better ways to support what you have suggested (titles in txt books). Can you provide some samples to me? If those are copyrighted and can't be sent in full then please just send some samples of those titles.

I don't have any TXT e-book in English which uses such markup. But here is a simple example how it may look like:

# This is the book title
## Introduction
Here could be some paragraph with the introduction about this book.
## How I went to America.
In this chapter, the author may write about his trip to `America. Some very intersting text could be here.
A new paragraph starts here just to demonstrate this example.
## Conclusion
This is the end of this example. I guess no more nonsense is needed to demonstrate my point.

Hello Labsii,

if the "Find next" and "Find previous" functions, that is the possibility to move to the next or previous search occurrence in the book content, were available for custom assignment to buttons in settings, it could also easily solve the problem of navigating by chapter names in a TXT e-book, as described above.

Please, what do you think about this feature?

Adam, I have implemented a function based on your description that should display titles from TXT files in the regular Titles section in Navigation panel. It would be great if you could text if it works as I don't have a file to test.

Hello Labsii,

I have created a simple plain text file with the example contents which I posted in one of my previous comment here, that is, denoting chapters with the # (hash) symbol, but if I display the titles for this text file as you instructed, there is only an empty list. Therefore, How I am supposed to display the list of such chapters? It would be also good to have an option per document to specify which symbol is used for denoting chapters because every e-book uses a different symbol.

Anyway, thanks for implementing this feature. I believe it would be a great addition to Speech Central, something which I quite miss in the Voice Dream Reader app.

Hello Adam,

This sample clearly won't work as it is artificial material that largely misses usual statistics/ratios of the book that are used in the detection. If you have some real sample please check it.

Unfortunately this isn't such a wide spread use case that I would add a setting for it. However I can further improve it based on the feedback.

Could you please explain in other words why this sample is not working?

Actually, I don't have any sample for this function. THe books which I use denote chapters not by the "#" (hash) sign, but using the "X" letter followed by space. It would be nice if user could choose at least between those two characters.

The book shouldn't have more titles than the text.

As this is based on guesses there are always chances of both false positives and false negatives. There are some rules to ensure that false positives are minimized and they can't pass on this sample.

I see. So could you please modify the function to recognise as title the "X" letter instead of "#" (hash) symbol so I can test it?

Adam,

technically it can be done, currently it was set to two characters that I would expect to appear "*" and "#".

However I am not sure if this particular implementation will have the future - I made it with assumption that it is something fairly common in .txt books. While I still have just an assumption now that assumption is opposite. At this point I am more likely into removing this code and implementing Find All feature (which would be implemented anyway sometimes later).