Python & Windows prep for Dragonfly

67 views
Skip to first unread message

Gregory Tippett

unread,
May 13, 2019, 11:37:37 AM5/13/19
to Dragonfly Speech Recognition
I'm a relatively new developer, working so far almost entirely in Javascript and in the browser front end. Having developed a bit of RSI, I'd like to get started with voice coding as a preventative measure.

Wondering if someone can point me in the right direction as far as resources for learning Python, and also Windows, which would be most appropriate for using Dragonfly. There are loads of general tutorials available of course, but I'm hoping for recommendations to help me "learn what I need to learn" in the most efficient way possible.  Software is a vast sea, so I'm looking for some help getting oriented to the most relevant terminology and topics for voice coding in particular. 

Thank you for any thoughts you may have!

Emily Shea

unread,
May 13, 2019, 11:45:47 AM5/13/19
to dragonf...@googlegroups.com
Hi Gregory, are you using a Mac? If so, consider using Talon... its much easier to setup and won’t require windows vm, etc. I’ve been using Talon for about a year now and do almost 100% of computer work by voice due to RSI.


There’s a slack group that’s very active and helpful.

Sent from my iPhone
--
You received this message because you are subscribed to the Google Groups "Dragonfly Speech Recognition" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dragonflyspee...@googlegroups.com.
To post to this group, send email to dragonf...@googlegroups.com.
Visit this group at https://groups.google.com/group/dragonflyspeech.
To view this discussion on the web visit https://groups.google.com/d/msgid/dragonflyspeech/ee540e93-a947-4f05-993a-3c10249647ed%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Shervin Emami

unread,
May 13, 2019, 6:10:59 PM5/13/19
to dragonf...@googlegroups.com
Hi Gregory,

It's a shame you're developing some RSI but good that you're aware of it and trying preventative options like voice coding. (Another thing I highly recommend is to make sure you physically take regular breaks, such as setting a timer and every 30 mins you walk away to the other side of the room from the computer, even for just 30 seconds). I'm not much of a Python expert myself, but I don't think you need to learn any specific area of Python, just becoming more familiar with general Python will definitely help. Dragonfly does use Python Dictionaries a lot but that's already going to be covered by any Python tutorials. Dragonfly does often including importing Python files at runtime and possibly unloading the code. That's something I haven't dealt with much in other projects, so you might want to make sure after you've learnt some Python, you could learn a bit more specifically about runtime loading & unloading of Python.

Cheers,
Shervin Emami.


--

alexander15w

unread,
May 14, 2019, 12:13:49 AM5/14/19
to Dragonfly Speech Recognition
Gregory,

It looks like to get up and started you just need to pip install


there is a lot of documentation:

although you may want to check with the more active gitter for advice before diving in.


you probably only need a pretty basic understanding of Python to get your dragonfly set up up and going (just Google any introduction to Python that includes functions, lists, tuples, and dictionaries and you should be pretty set).  If you are actually interested in the Python language, I for one read through the O'Reilly learning Python book to get a reasonable introduction.

if you don't really want to get into developing in python there are some simpler alternatives, the one I use the most is vocola which is a rather elegant language specifically built for voice programming, which still lets you access the full power of Python should you need it.


another alternative which is built on dragonfly is castor, which comes with a lot of built in functionality out-of-the-box


in either case the following documents are worthwhile reading or scanning through if you're getting started


a getting started document on things to think about as you are setting up your environment, with a focus on things that will make the biggest difference:



there has been a lot written about voice programming so I would definitely research and ask questions if you're running into any roadblocks.

Also it looks like somebody has already replied about talon, which is a relatively new entry and currently Mac only.  One thing to be aware of is that on Windows you will likely need a version of Dragon NaturallySpeaking unless you want to be intrepid with Sphinx or some other open-source voice recognition engine.  That said everything else is open source.  With regards to Talon it sounds like the developer has built on an open source voice recognition framework  (wav2text iirc) given that nuance no longer supports that platform, although the core engine is closed source but free. I'm sure someone who knows Talon could give you details there.

best of luck in your journey to reduce your RSI and many sympathies.

alexander15w

unread,
May 14, 2019, 12:33:16 AM5/14/19
to Dragonfly Speech Recognition
Also well worth reading David's series on getting started which gives a nice overview of open source packages available at the time.  A bit dated but still relevant.


I'll let you find parts 3 and 4 on your own :-)

James has some great stuff too

Gregory Tippett

unread,
May 14, 2019, 5:50:02 PM5/14/19
to Dragonfly Speech Recognition
Thanks, I'm on Windows but I'll check it out for sure if I switch! It's encouraging to hear you can get so much done by voice :)


On Monday, May 13, 2019 at 8:45:47 AM UTC-7, Emily Shea wrote:
Hi Gregory, are you using a Mac? If so, consider using Talon... its much easier to setup and won’t require windows vm, etc. I’ve been using Talon for about a year now and do almost 100% of computer work by voice due to RSI.


There’s a slack group that’s very active and helpful.

Sent from my iPhone

On May 13, 2019, at 8:37 AM, Gregory Tippett <gntsk...@gmail.com> wrote:

I'm a relatively new developer, working so far almost entirely in Javascript and in the browser front end. Having developed a bit of RSI, I'd like to get started with voice coding as a preventative measure.

Wondering if someone can point me in the right direction as far as resources for learning Python, and also Windows, which would be most appropriate for using Dragonfly. There are loads of general tutorials available of course, but I'm hoping for recommendations to help me "learn what I need to learn" in the most efficient way possible.  Software is a vast sea, so I'm looking for some help getting oriented to the most relevant terminology and topics for voice coding in particular. 

Thank you for any thoughts you may have!

--
You received this message because you are subscribed to the Google Groups "Dragonfly Speech Recognition" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dragonf...@googlegroups.com.

Gregory Tippett

unread,
May 14, 2019, 6:04:46 PM5/14/19
to Dragonfly Speech Recognition
Hi, thanks for your thoughts an encouragement! Yes I have various routines to help with RSI (Alexander Technique has been particularly useful), but I'm heading into a period of my life that I anticipate to be keyboard-heavy, so I'm doing this as a precaution.

Cheers,
Greg


On Monday, May 13, 2019 at 3:10:59 PM UTC-7, Shervin Emami wrote:
Hi Gregory,

It's a shame you're developing some RSI but good that you're aware of it and trying preventative options like voice coding. (Another thing I highly recommend is to make sure you physically take regular breaks, such as setting a timer and every 30 mins you walk away to the other side of the room from the computer, even for just 30 seconds). I'm not much of a Python expert myself, but I don't think you need to learn any specific area of Python, just becoming more familiar with general Python will definitely help. Dragonfly does use Python Dictionaries a lot but that's already going to be covered by any Python tutorials. Dragonfly does often including importing Python files at runtime and possibly unloading the code. That's something I haven't dealt with much in other projects, so you might want to make sure after you've learnt some Python, you could learn a bit more specifically about runtime loading & unloading of Python.

Cheers,
Shervin Emami.


On Tue, May 14, 2019 at 1:37 AM Gregory Tippett <gntsk...@gmail.com> wrote:
I'm a relatively new developer, working so far almost entirely in Javascript and in the browser front end. Having developed a bit of RSI, I'd like to get started with voice coding as a preventative measure.

Wondering if someone can point me in the right direction as far as resources for learning Python, and also Windows, which would be most appropriate for using Dragonfly. There are loads of general tutorials available of course, but I'm hoping for recommendations to help me "learn what I need to learn" in the most efficient way possible.  Software is a vast sea, so I'm looking for some help getting oriented to the most relevant terminology and topics for voice coding in particular. 

Thank you for any thoughts you may have!

--
You received this message because you are subscribed to the Google Groups "Dragonfly Speech Recognition" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dragonf...@googlegroups.com.

Gregory Tippett

unread,
May 14, 2019, 6:27:02 PM5/14/19
to Dragonfly Speech Recognition
Thank you so much for all this information! A huge help in getting started. I am studying it now... will almost certainly have questions! :) 

Hawkeye Parker

unread,
May 16, 2019, 3:56:40 PM5/16/19
to Dragonfly Speech Recognition
Hi Emily,
I was curious about Talon, but it looks to me like DNS is discontinued for Mac at this point. Am I missing something?
Thanks,
Hawkeye


On Monday, May 13, 2019 at 8:45:47 AM UTC-7, Emily Shea wrote:
Hi Gregory, are you using a Mac? If so, consider using Talon... its much easier to setup and won’t require windows vm, etc. I’ve been using Talon for about a year now and do almost 100% of computer work by voice due to RSI.


There’s a slack group that’s very active and helpful.

Sent from my iPhone

On May 13, 2019, at 8:37 AM, Gregory Tippett <gntsk...@gmail.com> wrote:

I'm a relatively new developer, working so far almost entirely in Javascript and in the browser front end. Having developed a bit of RSI, I'd like to get started with voice coding as a preventative measure.

Wondering if someone can point me in the right direction as far as resources for learning Python, and also Windows, which would be most appropriate for using Dragonfly. There are loads of general tutorials available of course, but I'm hoping for recommendations to help me "learn what I need to learn" in the most efficient way possible.  Software is a vast sea, so I'm looking for some help getting oriented to the most relevant terminology and topics for voice coding in particular. 

Thank you for any thoughts you may have!

--
You received this message because you are subscribed to the Google Groups "Dragonfly Speech Recognition" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dragonf...@googlegroups.com.

alexander15w

unread,
May 17, 2019, 1:59:10 PM5/17/19
to Dragonfly Speech Recognition
it looks like he has been hooking in open-source voice recognition engines, Although I'm not sure which one he's at right now:


I have seen librispeech, and wav2letter mentioned.  He also presented demo over here of Dragon six versus his other engines:

To be clear I have not tried Talon at all yet but it sounds like it might be a viable option for Mac.

Caspar

unread,
May 18, 2019, 1:36:14 PM5/18/19
to Dragonfly Speech Recognition
Regarding Dragon for Mac being discontinued - the standard reply which the author of Talon gives is along the lines of "buy it anyway: it will continue to be a viable option for the medium term because I will continue supporting it; I have already patched numerous bugs in it including ones preventing it from working on the latest MacOS. Also, make sure to update Dragon to 6.0.8 before trying to use it for anything because it's very buggy without that."

Talon also already has a built in engine that apparently works reasonably well (powered by Mac's speech recognition). My understanding is that it doesn't handle freeform dictation as well as Dragon, and sometimes gets into a buggy state where made up words are not recognized; AFAIK that's why the author is experimenting with wav2letter++ and others.

Source: I hang out in the Talon Slack and this comes up all the time. Feel free to join and ask the Talon author directly :)

Dane

unread,
May 30, 2019, 5:36:24 AM5/30/19
to Dragonfly Speech Recognition
Hello,

I just wanted to say that this is a really nice summary of what's available at the moment, plus some useful resources! :-)
Unimacro (http://qh.antenna.nl/unimacro/) is another extension worth mentioning. It comes bundled with Natlink.

Dane Finlay
Reply all
Reply to author
Forward
0 new messages