| T O P I C R E V I E W |
| All Digital |
Posted - Jan 06 2007 : 06:30:45
Hello everyone,
This is my second post since joining the forum .
And I've been wanting to make a post about it since the first time I even saw the book. This is the first book that I've read cover to cover in 15 years, seriously!
I first worked for a web developer in 1998, During that time I not only got to make a website, (the owner never got it on line ), but I also had the time to spend many hours on line with UltraHal,but I also was able to talk to the man that created the Girl that is on the cover of the book. Her name is "Webby 2k", and she was created by Steven H. Stahlberg. She was the first virtual human to ever have a contract with a fashion magazine in Europe. Mr. Stahlberg works for a company in Malaysia, named Optidigit (Optidigit.com) He also maintained a website called Android Blues. He has also written a screenplay by that same name. Optidigit at that time also offered a MSAgent desktop assistants based on Webbie 2k. (I have both of them, for WIN98 & WIN2000.
I purchased Dr. Plantecs' book before Yapanda discovered the WIN2000 problem. That really wasn't a problem, because I had, "my very own" UltraHal by then. The emphasis Dr.Plantec put on developing VHs with personality really impressed me. So much so, I really want to create these wonderful characters, especially one of my own !
Oh, yes! On Topic! I found another book titled, "Voice User Interface Design, by Cohen, Giangola, and Balogh. By Addison-Wesley Publishing. ISBN 0-321-18576-5. $44.95 at Amazon.com.
While not as entertaining as Dr. Plantecs book, It's as if the authors created any entire book on developing personalities. The scope is from very simple created voices to hiring an entire sound recording studio to develope ones own telephony system.
What really impressed me the most was the emphasis not on emotion, but speech inflection, the rising and falling of the speech guiding the conversational flow of the system, in response to the needs of the caller.
This causes one to consider the possibilities of the use of VoiceXML and Aureal style sheet with the Pre-made,low-cost,low frequency, (8kHz)voices, until we all get rich enough to create our own sound studio. .
Thank you for your patience, and I'll be back next week with another book review .
Oh,BTW, there is a piece on Dr. Plantec about Virtual Teachers in The Feb. 2007 issue of 3D World. A quote from Dr.Plantec,"We fake real intelligence, we don't mess with AI" pg 39. What ever that horrid beast at the bottom of the page has to do with teaching children escapes me.
|
| 15 L A T E S T R E P L I E S (Newest First) |
| rowboatics |
Posted - Dec 16 2007 : 22:59:51
books are so hard to read i have to know everything every dead person thought about a subject be fore i can read the book,., but the humars real world fiction sure is fun |
| hologenicman |
Posted - Jan 07 2007 : 08:44:10
quote:
laura,
Great to have you here. we only have 2 ladies marielle, holgenics better half and me. you seem in tune and knowledgable. i am sure you are going to be a great asset to our group. ~~alady
Yes, I am pleased to have my better half, Alady, and now Laura as contributing memebers of this great forum.
The forum would be saddly lacking if it were merely dominated by men only... 
John L>
IA|AI
|
| hologenicman |
Posted - Jan 07 2007 : 08:40:43
quote:
Hey, I wonder. We have been quite and very little engaged over the last month (except for a sick raven that flew through) Maybe we stay healthy by staying involved with each other. Sort of a reverse virus.
yes, quite ironic, but a sence of community does work toward keeping people healthier. If that is true, then they should NOT call them communicable diseases... 
John L>
IA|AI
|
| laackejim |
Posted - Jan 07 2007 : 07:08:11
Hot Dog Diddley. We are alive again, even if sick. The fun begins again and all infirmaties heal!!
Hey, I wonder.  We have been quite(oops) quiet and very little engaged over the last month (except for a sick raven that flew through) Maybe we stay healthy by staying involved with each other. Sort of a reverse virus.
We'll give it a test as soon as we get back from hospitals and sick beds and trauma centers and other "fix the body" places.
|
| GrantNZ |
Posted - Jan 07 2007 : 06:07:26
Mmmm. It seems like a consolidation. The voices back then were far more flat and robotic than current day voices, which do a pretty good job, even if they've lost pitch/stress control. Oh well |
| GamerThom |
Posted - Jan 07 2007 : 06:01:26
It would seem for certain advances that take place, there
is a corresponding step backwards in some instances. |
| GrantNZ |
Posted - Jan 07 2007 : 05:55:58
quote:
It would be great to be able to simply control speed, pitch, and volume in a dynamic fashion in order to relay emotions and effect a realistic voice.
Exactly what I'm interested in.
Back in the days of the Amiga computer (two decades ago? Maybe three?) the speech synthesis was hardware-accelerated, and adding stress to syllables was as simple as inserting a number from 1 to 9 into your text stream. (I think 5 was the default.) So for example "I rea7lly lo9ve this" would accentuate the "rea" a bit and the "love" a lot. (You'd actually use phonetic language rather than English (or American ), but they're not too dissimilar ) Somehow, as far as I figure out, we've lost even this simple functionality. |
| aladyblond |
Posted - Jan 07 2007 : 03:20:36
laura,
Great to have you here. we only have 2 ladies marielle, holgenics better half and me. you seem in tune and knowledgable. i am sure you are going to be a great asset to our group. ~~alady |
| GamerThom |
Posted - Jan 07 2007 : 02:39:03
quote:
Originally posted by All Digital
Thank you, Mr. Rossi, the quote thing has been taken care of.
Those ActiveX Controls have been posted in Programming under the topic, quote:
SR and TTS ActiveX Controls
GamerThom, I'll post a reply to your question in the modelling forum.
OK. But please, just refer to me as Gamer. Most everyone else around here does. |
| All Digital |
Posted - Jan 07 2007 : 02:31:04
Thank you, Mr. Rossi, the quote thing has been taken care of.
Those ActiveX Controls have been posted in Programming under the topic, SR and TTS ActiveX Controls
GamerThom, I'll post a reply to your question in the modelling forum. |
| GamerThom |
Posted - Jan 07 2007 : 01:08:36
Yes, I've been working in Poser 5 for about 10 months now and have just gotten the
upgrade to Poser 7 which seems to do the job much faster, but it doesn't accept or
recognize many of the Poser 4 & 5 figures and props. Either that or they just need
to be installed into a different folder hierarchy for Poser 7 to find all the files.
I agree haptek is very limited as far as fullbody figures and is not very customizable
unless you want to, or have the money ($2,000.00+) to buy a commercial developement
license from them for the required fullbody editing and mesh construction software.
Have you given any thought to what a real-time animation using Poser and similar
figures might work like? We have a couple of members here that are working on
graphics engines that would create real-time (haptek like) animations utilizing
figures from Poser, Hexagon, Cinema, Blender and other mesh creation software instead
of pre-ordered animation sequences. In reference to this, you might want to take
a closer look at the post's under "Graphics" by GrantNZ and Guile3D. |
| vrossi |
Posted - Jan 07 2007 : 01:05:13
quote:
It still isn't showing the quote I came up with but It will come eventually!
What you write in your Profile under the sections Hobbies, Favorite Quote and Bio is displayed only when one clicks on your name and looks at your profile.
If you want something to appear under each of your posts, you must write it in the Signature section of your Profile.
|
| All Digital |
Posted - Jan 07 2007 : 00:52:24
It still isn't showing the quote I came up with but It will come eventually!
(Shes not too sultry is she?, all of the shorts ones were too provocative.)
Yes, The Haptek system is fantastic, However, like all of us it is just too limiting with those jack boots! I'm looking forward to getting a character into a 3D environment, much like a game but not so engaging. And a much higher resolution than VRML.
As I mentioned, I'm using DesktopMates' Dawne poses to build my animation sequences in Poser 5.
With MASH I was able to get a list of the poses that were used to simplify the coordination with a scripting language. Scripting is a weakness I'm currently working on overcoming. Then to get some database system for building an animation database. Facial animation and speech synchronization doesn't appear to be a major issue as regardless of mood the mouth can only make so many shapes. (Hee, Hee, Yes I have "Stop Staring" by Jason Osipa, too.) The principle of using visiemes as opposed to phonemes is much more practicle.
Yes, I'll get those ActiveX controls for the TTS and Speech Recognition and post them here. |
| GamerThom |
Posted - Jan 07 2007 : 00:24:31
Yes. I'm quite inscrutable with just a touch of the devil in me. 
Very cool choice of avatar you made there.  |
| hologenicman |
Posted - Jan 07 2007 : 00:24:17
Welcome Laura,
Everyone here seems to have cut out a little section of the greater virtual human endeavor. I am VERY pleased and excited that you are so interested in developing the voice capabillities. Voice can make quite a difference in the interactive experience that the end user has with a virtual human.
One of the more captivating experiences is to visit the Haptek site and interact with the virtual human there. This is so very captivating since they use voice actors for the scripting rather than the SAPI. If you can get the SAPI anywhere near this engaging, it will be an increadible advance in virtual humans.
quote:
Check it out if you get the chance, I haven't played with it much but it looks like it has some good potential.
Yes, I've played with these settings in the past in order to get the relatively inhuman SAPI's to sound acceptable. It would be great to be able to simply control speed, pitch, and volume in a dynamic fashion in order to relay emotions and effect a realistic voice.
John L>
IA|AI
|
|
|
