Voice Recognition Software


Senior Member
I'm looking for opinions on Voice Recognition Software. I know there are some folks who are heavily into HAL as well as some who have tried using VR in Homeseer.

Can anybody provide a basic comparison of HAL .v.s. HomeVoice v.s. Microsoft based VR (like PowerHome, Homeseer, etc.)? I'm trying to figure out if there is a dramatic difference between how well they each work or if it is just a matter of minor feature differences. (Like HAL being a stand-alone HA package while HomeVoice is designed as a front end to be used with any system).

Also, on the response voice, I was wondering if HAL is limited to the available "HAL Voices" or if you can use any SAPI 5 voice engine with it?
HAL is speaker independent. the ms based VR requires training and different profiles for different users. my wife & i can use hal (and it will occasionally recognize my 4 year old's commands) but w/ ms VR it will not recognize my wife when using my profile.

HAL will only work w/ HAL voices which is annoying since i already have at&t natural voices.
Here's a recent thread of a new user picky HS over HAL. With HomeSeer you can switch your VR profile with a little HS script and several users can use VR. VR in general terms is not ready for prime time unless you are willing to throw a ton of hardware and money at it as Paul did then you can get something that works reasonably well.
Not that many years ago I evaluated HAL, Homevoice and HS all at the same time. The only thing I was interested in at the time was how well each did VR. In my opinion (not worth much these days! :blink: ) the VR engine (Verbex) used in Homevoice was the best as far as being able to recognize commands in a fairly noisy environment (TV on, background discussions, etc). One of the more enticing things about their VR was the ability to train individual words and phrases rather than stock training (like MS VR training).

At the time I was using the Houselinc box from Smarthome and the Homevoice software had a hard time with the box (kept getting really bad errors). Since support from Homevoice was pretty spotty, I started looking elsewhere and settled on Homeseer since the support was there and the message board support was stellar. HAL was just too restrictive at the time as far as interfaces. And too dang expensive. And had no web interface (at that time anyway).

Just my $0.02

Thanks for the input. When I get around to playing with VR I'll probably try HomeVoce first. That in combination with one of those fancy noice cancelling conference mixers sounds like my best shot at a reliable VR setup.
my 0.02 cents -- in the new vista edition (if it is released this decade), Microsoft has redesigned the VR engine and a new voice synthesizer. If I had not purchased all of the voices (at&t, neospeech) I would wait and see what microsoft is releasing. Perhaps somebody on the board is beta testing vista and could ring in on MS new system. Hopefully it will be compatible with SAPI 5 voices.

oh yeah, back to the topic. I have no problem with homeseer/microsoft's speech recognition. That being said, I am the only one that chooses to use it in my home. My wife just wants the lights to turn on when she comes in the room and go off when she leaves. Everything else is just a toy to her, my toy (she doesn't like to play).
I am running the latest beta of Vista, I never bothered to check VR/TTS, didn't realize they included this already.
Unfortunately the advanced DRM features in VISTA require each microphone to be licensed to each person who might talk into it. The licensing is also linked to PC hardware and the application being controlled and it is rumored that per word licensing is in the works.
Wow, according to the Homevoice docs, a voice can trigger an ascii command to a hardware controller. Does anyone see a reason this would not work with an M1 via an XSP?
Is HomeVoice still being actively supported and/or developed? I haven't heard of anything new from them in years and I thought that they had closed up, but their web site is still there.