Jump to content
Sign in to follow this  
magentaline

Voice Control Specifics?

Recommended Posts

I have not tried P2ATC yet but have a question re voice control. 

Does it require exact phrases to be used by the pilot or does it accept the pilot hitting key words within requests and read-backs?  in the way that MCE does?

Thanks

Tim

Share this post


Link to post
Share on other sites

IN general, it is listening for exact phrases that contain key words in the right order.  That said, there are a lot of ways to say the same thing that it will accept, so the perception sometimes is that it's listening for key words. 

The downside is that you have to train yourself to say the right thing.

The up side is, that's easier than trying to train a machine to understand random phrases.

 

Thanks for the question.

Dave

Share this post


Link to post
Share on other sites

OK

Regardless of whether I get the phrase right or not, will a green text be displayed for what I did say: mine rarely is. So I am left wondering if I am heard at all.

 

BTW the practice page never works for me, I wear dentures....that said I have no problem with the default Windows  8.1 speech recognition software, it always gets what I say bang on.

 

Is not P2ATC using the same engine?

 

Just realised I hacve to adjust my mic, seems ther is quite an echo I must have set it up wrong with effects enabled/...

Share this post


Link to post
Share on other sites

Have you set  the voice option to en-us  and not  en-uk is the first place to check first?


I7-800k,Corsair h1101 cooler ,Asus Strix Gaming Intel Z370 S11 motherboard, Corsair 32gb ramDD4,    2  ssd 500gb 970 drive, gtx 1080ti Card,  RM850 power supply

 

Peter kelberg

Share this post


Link to post
Share on other sites

yes it is en_us

and I just did the "train your computer to better understand you" option (win 8.1 only I think)

 

At first it (micorsoft's speech recognition software test) was not understanding me all that well. but with a little training/practice the articulations are either better spoken, or better understood.

 

Will see later if this helps P2ATC as well

The difference with the microsoft speech recognition in its usual usage is that it looks from a list of likely command words whereas P2ATC is tending more to deal with dictation sentences to obtain sets of specific words. which is inherantly more difficult to establish that one-two-or-three-word commands (like "open program" instead of the lengthy read-backs of P2ATC).

 

And then there is the all to famous phrase:

"WHEN IN DOUBT MUMBLE" and we often are in doubt...?

 

Allen

Share this post


Link to post
Share on other sites

If you still having issues  than most likely it could be your head set which could be  the cause of it, read on another thread  that a similar issue till they got  a usb  head set which fixed it.


I7-800k,Corsair h1101 cooler ,Asus Strix Gaming Intel Z370 S11 motherboard, Corsair 32gb ramDD4,    2  ssd 500gb 970 drive, gtx 1080ti Card,  RM850 power supply

 

Peter kelberg

Share this post


Link to post
Share on other sites

I am using a cheap-as-possible headset via my onboard ausio outputs in the rear of my computer. Its so cheap the instructions were in Chinese "only". 19 bucks brand new.

 

How low can you go?

Share this post


Link to post
Share on other sites

I am having the same issue as Peppy197!!

 

Speech recognition works great (US setting) in all other cases, for example I can dictate a whole page of text and most is correct. The only speech P2ATC will recognize is "Roger" and numbers, so I can say Clearance Delivery 'til I'm blue in the face and nothing happens. I can say the same ting in Notepad and it appears no problem. That therefore eliminates mike problems, given it works fine elsewhere and even passes the Xsquawkbox test for clarity. Model is a Logitech USB of reasonable quality.

 

Running Windows 8.1 with P2ATC 1.300

 

Any help appreciated as I bought the software at the introductory price but so far have had nothing but frustration with the voice.

Share this post


Link to post
Share on other sites

Hello Dave (Deetee),

 

I have too a Logitech USB (G930). I don't know if you have the same model but just in case, shut down all effects and Surround sound option.

 

Btw my problem is not exactly yours. Controllers understand me (no problem with that hopefully) but curiously if I run the Grammar program, the "Practice Speech" understands nearly nothing !

 

Regards,

 

Richard Portier


Richard Portier

MAXIMUS VI FORMULA|Intel® Core i7-4770K Oc@4.50GHz x8|NVIDIA GeForce GTX 1080ti|M16GB DDR3|Windows10 Pro 64|P3Dv5|AFS2|TrackIr5|Saitek ProFlight Yoke + Quadrant + Rudder Pedal|Thrustmaster Warthog A10|

Share this post


Link to post
Share on other sites

 

 

 if I run the Grammar program, the "Practice Speech" understands nearly nothing !

 

Regards,

 

Richard Portier

 

Nearly nothing? You are lucky

In my case "practicing" recognises absolutly nothing

 

why cannot P2ATC's speech recognition software recognise my voice when Windows 8.1 engine does almost without fail.

 

I know I have a cheap headset but should I go buy a 120 dollar set just to pass go?

 

Question: if I get a cheap USB-connectivity headset will that be better than the 3.5mm plugin I now have>?

Share this post


Link to post
Share on other sites

All,

Pilot2ATC uses the Windows Speech Recognition Engine, so if your microphone works great with dictation in Notepad, it should be OK for Pilot2ATC.

 

Once you have the right mic setup and en-US language selected, the two most common reasons for not being recognized are not saying a set of words or phrases that can be recognized or not waiting a half-second after pressing  the PTT button and another half-second at  the end before releasing the PTT button to ensure the entire pharse is heard by the engine.

 

In order to know how to respond, Pilot2ATC listens for specific words and phrases.  It takes some getting used to in order to say the right words in a given circumstance.  Not all controllers can recognize all phrases and not all phrases are always recognizable.  There is a speech context that determines what phrases should be recognizable.  For example, if you switch to Ground Frequency, you'll notice some red text saying "Ground Ops" in the text window.  You have just switched to the Ground Ops Speech Context.  Only the phrases allowed in Ground ops will be recognized.

 

The phrases listed under All controllers, are always listened for by all controllers, so they should always be recognized.  Things like "Say Altimeter".

 

The Grammar Helper can assist in this.  You should be able to read the example phrases at  the top of each request, substituting your call sign for Beech One Victor Bravo and have fairly good success. 

 

All that said, we are constantly looking for ways to improve grammar recognition.  It will never be perfect, but it should certainly be better than what you apparently are experiencing.  If you have specific phrases that are not working with a particular controller at a particular airport, let me know.  We just recently chased an inability to get a response after a call for taxi to gate down to some missing parking paths in the database.  I'll be glad to test it and see if there is an issue that can be fixed.

 

Thanks,

Dave

Share this post


Link to post
Share on other sites

 

....If you have specific phrases that are not working with a particular controller at a particular airport, let me know.  ....

 

Thanks,

Dave

 

I am not getting any response regarless of phrasology from ground at EGHI (UK2000 freeware installed with OpenLC_EU )
 
Trying to get asn IFR clearance from a filed and validated plan
 
Tried ....xxx..say altimeter...no response
 
 
note: EGHI SoutHampton not EGNI as previously erred in

Share this post


Link to post
Share on other sites

Okay the practice part is definitely hosed, but I have made a little progress with the tutorial.

Fine until I got to the taxi readback which was partly unrecognized. Ground then said "Say again stating callsign".

After that I tried repeating a few times but my voice seemed to be no longer recognized in the sense that what I said was not displayed.

The tutorial says to expect no response after readback, but I did expect my speech to be shown.

So I figured I'd try taxiing anyway, but was not asked to cross rw30R, nor contact tower.

I tried requesting to cross 30R but no recog, no response.

So... i switched to Tower anyway and requested take off - idem, nothing.

 

Did I do something wrong along the way which prevented me continuing?

If so it would be great if one could skip by requesting stuff manually, otherwise it's impossible to known where the program has got stuck.

 

Cheers

Share this post


Link to post
Share on other sites

I just went out and bought a 140 dollar "non-returnable" headset with USB connectivity,

installed the updated Win 8.1 software, check all the resources corectly setup for it

Ran P3DV22

Comes through nicely on the nice headphones,

tried P2ATC at EGFF and only Tower available

Said my call sign as clear as day

 

and no response ???   try talking louder >

 

 

Had to talk quite loudly to get a response FINALLY

 

So far I am getting better interactions..... will see if it lasts now

 

well got my clerarance + readback accepted, got my taxi instructions, much better

 

Sort of frustrating though was told to contact center and I am not getting through to them on initial contact

 

OK got it working

 

Now its all back to OK, sort of

Not getting a response again, no green either

 

You never know

Share this post


Link to post
Share on other sites

John,

Glad you're doing a little better.

I went to the Practice screen and tried several calls myself, being careful to talk clearly and deliberately.  Didn't get recognized.....What???

Well then I tried talking a little faster and more fluidly...hard to type that way,,,but low and behold the recognition rate went way up!  Not 100%, but much better.

So I think one of the tricks is to talk clearly but confidently as if another person were on the other side of the radio beam instead of trying to go slow and clearly for the machine.  The practice helps you pick up a good cadence that should also work in the main program.

 

Also, made some changes to the Grammar Helper that should help it do as good a job at recognition as the main program.  That will be out in the next release soon.

 

In the main program, sometimes it helps to do an Initial Radio call to a controller that seems to have fallen asleep.  Something like "Ground, <Call Sign>.  If the Grammar context has got out of sync or something, this sometimes brings it back.  Another trick is to Swap frequencies back and forth one time.

 

Thanks for sharing your experiences.

Dave.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

  • Tom Allensworth,
    Founder of AVSIM Online


  • Flight Simulation's Premier Resource!

    AVSIM is a free service to the flight simulation community. AVSIM is staffed completely by volunteers and all funds donated to AVSIM go directly back to supporting the community. Your donation here helps to pay our bandwidth costs, emergency funding, and other general costs that crop up from time to time. Thank you for your support!

    Click here for more information and to see all donations year to date.
×
×
  • Create New...