Announcement

Collapse
No announcement yet.

BrightSpeech Voices, do you want them?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    #46
    And Here is AT&T's Crystal.

    If someone could upload a similar test with NeoSPeech, I would like to hear that to compare it. I do not hav the NeoSpeech voices myself.
    Attached Files
    Visit My Home Seer Site at:
    www.JohnWPB.com
    Created with LCARS

    Comment


      #47
      John

      Listening to your 2 clips, I can hear the differences in response time, but I don't think it is a good test. If you try Rupp's series of system.speak lines posted earlier, you will see just how big the difference is.
      ______________________________
      Skibumsplace - Locate Me

      Comment


        #48
        O dear, just been to Skibum`s place and he has changed from heather to some OLD BLOKEs voice, <sob> we just have to find her its not the same talking to Ski with anything other than Heather

        Comment


          #49
          <BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by JohnWPB:
          The inflection in the voice, almost displays emotion... <HR></BLOCKQUOTE>

          TRANSCRIPT OF SPEECH:

          "Oh, John, you are so handsome. Have you been working out?

          Oh, baby, nobody codes script for me like you do!"


          ~Bill

          Comment


            #50
            Bill.... Your starting to frighten my computer!
            Visit My Home Seer Site at:
            www.JohnWPB.com
            Created with LCARS

            Comment


              #51
              Wow- that response time is good. Count me in for a Heather.....

              Man- your VR is quite good- what are you using for a microphone/recognition engine?

              Your voice is actually very clear- ever consider radio? By you avatar- you have a face for it.... :-) JK- Samantha is a lucky woman.....
              HS3 Pro Edition 3.0.0.435 (Windows Server 8.1 on ESXi box)

              Plug-Ins Enabled:
              Z-Wave:,RaspberryIO:,AirplaySpeak:,Ecobee:,
              weatherXML:,JowiHue:,APCUPSD:,PHLocation:,Chromecast:,EasyTr igger:

              Comment


                #52
                I am using the standard MS VR engine. The microphone is a Telex M-560


                More information on the mic is &lt;Available Here&gt;.
                I bought it at Best Buy about a year ago for someting like $40 or so. It is not as good of quality as a lot of the ones made specifically for VR, but I have had very good luck with it for the price. I can be across the room (about 10 feet away) and speak in a normal tone and it picks it up pretty accurately.

                <BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>wpiman wrote:
                ever consider radio? By you avatar- you have a face for it.... <HR></BLOCKQUOTE>

                Would have to be radio, as this mug wouldnt work on TV
                Visit My Home Seer Site at:
                www.JohnWPB.com
                Created with LCARS

                Comment


                  #53
                  I see a LOT of posts comparing response times, but it is very apparent to me that people fail to realize they are not comparing apples to apples.

                  Neo and AT&T are concatenated voices, Heather and MS Mary are synthetic-concatenated hybrids, and some of the really poor quality voices out there are pure synthetic.

                  The speed at which a concatenated voice can go to the database to get the WAV audio data corresponding to the phonemes that make up a word are always going to be slower than the time it takes to synthesize it.

                  I have run both the AT&T voices and the Neo voices on test machines that were so fast and had so much RAM that the difference between them and the hybrid voices was neglegible whereas on my production HomeSeer system (800MHz PIII, 512M RAM) I have a very small (less than a second) delay.

                  The fact that Heather is liked so well AND it is a hybrid makes it a very desireable product, but in comparing the products you really should compare sound quality and price, not delays.
                  Regards,

                  Rick Tinker (a.k.a. "Tink")

                  Comment


                    #54
                    Rick, I think it's safe to say that we are comparing the 'end results', not the technology behind it (I do understand what you are saying, and you are right).
                    HSPRO 2.4 (ESXi 4.1) | my.Alert NEW | my.Trigger | HSTouch | ACRF2 | UltraM1G | BLWeather | BLLan | Rover
                    (aka xplosiv)
                    Do You Cocoon? Home Automation News, Tutorials, Reviews, Forums & Chat

                    Comment


                      #55
                      Well, I may be a little different and Rick will be the first person to say so but I do agree. But, in my analysis the order of comparison would be:

                      1) Sound Quality
                      2) Delay
                      3) Price

                      My rational is that if I've invested over $15,000 to date on my completely automated home I'm not going to quibble over $10 or $50 dollars in pricing to get natural sounding voices.

                      G

                      Comment


                        #56
                        it does not get much more apples to apples when you compare the response time of one product to the other. I could not care less what the underlying technology is.

                        &lt;IMHO&gt;
                        If you want to list quality, then Brightspeech wins again. Regardless of the technology.
                        &lt;/IMHO&gt;

                        Priority:

                        Sound Quality, including fidelity, and the "naturalness"

                        Delay

                        Price
                        ______________________________
                        Skibumsplace - Locate Me

                        Comment


                          #57
                          Mentioning the underlying technology was not saying it is a factor in the comparison, it is explaining why there is a delay with the concatenated voices that can never be compared to the speed of a synthetic (or hybrid) voice.

                          It is fortunate that you all like the Brightspeech voice because quite honestly, it is the FIRST every hybrid to even run close to the quality of a concatenated voice.

                          George, you have too big of a HomeSeer machine and too much money to even use delay and price in your analysis! You get to consider based upon sound quality only.
                          Regards,

                          Rick Tinker (a.k.a. "Tink")

                          Comment


                            #58
                            Agreed!

                            Make it so, Number One!

                            Warp factor 9.2. Engage!

                            Comment


                              #59
                              I think Heathers voice sounds better than Kates. If the delay is shorter than NeoSpeech's delay, I'm willing to buy!

                              Comment


                                #60
                                Again,
                                What kind of delays are you guys seeing. I posted my test and the delay is at or less then 2 seconds. This is very natural when a person is speaking to stop for a second or 2 between sentences. Are some of you guys seeing longer delays than a couple of seconds? If not how in the world did you ever notice it?
                                💁‍♂️ Support & Customer Service 🙋‍♂️ Sales Questions 🛒 Shop HomeSeer Products

                                Comment

                                Working...
                                X