Hello,

We have several clients using DragonSpeech medical voice recognition. Running in straight Chrome or the previous (and thankfully no longer used!) Awesomium version of our software, it worked very quickly. You would speak and the text would show up very quickly, in blocks.

With the current EO browser (3.0.89.1, plus the early single DLL versions) the text comes back letter by letter and slow enough that clients are asking about it. It is marginally faster in the multi-dll version than the single DLL version.

Clients using Dragon360 (the web-based voice recognition) do not have this issue (we do have that related issue with the Silverlight DLL having issues in the latest single DLL which you are working on - thanks!)

Any suggestions would be greatly appreciated!

Craig

The latest 3.0.95 single DLL appears to correct this issue: speech to text is now displaying in blocks versus word by word.

I've not tested the multi-DLL version as we prefer the single version.

I also tested this against the 1 environment we have that gives the rare run32DLL issue, and this release appears to correct that as well.

Edit: and this version also works with the Silverlight Dragon360 DLL (previously this was unable to load in the single DLL, but was loading in the multiple DLL version)

Thanks!

I apologize: we tested the wrong Dragon...Dragon360 works as expected, but DragonSpeech Medical is still slow.

To reiterate, Dragon360 is the online subscription model that uses Silverlight to process Nuance's Dragon Medical speech recognition. This works as expected.

DragonSpeech Medical is a stand alone piece of software installed on the user's PC, and can be used anyplace you enter text.

In our Windows application we use EO to access the web-based client our providers use. For some reason, the DragonSpeech Medical software is slower on the EO browser. Normally entire sentences just come up fairly rapidly; with EO the individual words come up and things are slower.

Sorry for the premature sign off!

Edit: Note: we have users using both speech recognition pieces. The users who have been with us previous to the web client usually have the speech software installed on their computers; the newer users all use the web-based Dragon360.

Yes, the previous versions also exhibited the slower response with Dragon Medical.

I also retested with the single and multiple DLL in case there were any differences. There did not appear to be any.

Thanks!

Thanks for offering shared screen demo. However that probably won't be enough. We need to be able to see the problem as well as to be able to debug the problem. That requires us having access to a development machine that can demonstrate the problem. I would also expect the troubleshooting process to take multiple trial and error attempts, in which we add some debug code, run it, observe it, then add some more debug code, run it and observe it again, over and over multiple times. This will take significant longer than a short demo session. As such if you can provide one of such environment and give us remote desktop access, it will be very helpful.

Yes. That would be great. I assume that you know our email address. You can email there directly and we will continue the discussion there.