I had a look at gizmoz in april last year for the first time. It allows you to create avatar presentations from a single picture of yourself. This is pretty interesting for Web design or your marketing activities in the social web to get a better acceptance for your products. Meanwhile the next generation of gizmoz tools is online – and the results are much better. But, you gess it: it’s still a long way for such tools to be perfect.
The Virtual rainwebs
Let me introduce you to the avatar I created from my picture – you can find it in the footer of this blog – in April 2009:
Although you may recognize me this image and also the animation is full of trade-offs. I had to choose for example a hair that let me look a bit like Antonio Banderas. This is far away from reality 
.
The Next Generation
For the new avatar I took the same image again. I was amazed how good the result was:
The texture looks nicer and works better on the mesh. The head movement is pretty cool now – if the head isn’t moving up (I don’t know why the eyes become strange with this). Nevertheless, I became 20 years younger by default. But, I can’t really recognize myself. Pretty interesting: my 2.5 years old daughter can’t recognize me, too.
The problem with this is the head morphing. If the shape doesn’t work the original texture doesn’t help much. So, I tried to become a bit older and get a face morph that suits better to me. I tried some combinations, but the morphing tools didn’t allow me to make my avatar more real.
Did I mention that the selection of male hair is still a challenge? The female avatars have a lot more stuff to choose from. Wow.
Use Cases
The presentation of the avatar is much better and you can use it in more contexts. Although the gizmoz system allows you to create some nice greeting cards in different flavors I really miss the possibility to create a professional talking head. You can tweak some of the offers, but I’m interested in a more professional context to create and use avatars, like you can have with CrazyTalk for example.
In comparison to CrazyTalk gizmoz delivers a licence free text-to-speech creation. I tested the male voice and the quality is pretty cool. If you do this on your own you have to buy a special licence for the voice you wanna use before you are allowed to publish the created wave file on the net. This is valid even for private use.
Some of my point and click results from the gizmoz greeting cards:
The Future
If you read my blog for some time now you may remember that I’m using Poser to create photo-realistic avatars. One of my content providers is DAZ3D. Their Victoria 4 and Michael 4 models deliver excellent meshes and are state of the art. A lot of DAZ partners already have created textures that let you create still images you can’t differ from real photo shots.
Why I’m mentioning this here? DAZ3D and gizmoz prepare a merger at the moment. They plan to deliver the best avatar animation platform in the net. This is pretty cool. DAZ has exellent content and 3D modeling tools, even a mimic processor, and gizmoz delivers cool online animation tools. If you have a look at the gizmoz modeling tools you may already recognize DAZ content. Although it is rare at the moment (and not always of the highest possible quality).
So, for short, we can expect better animation features and excellent content in the near future. Maybe someone let me create cool talking heads in minutes with the current text-to-speech engine then. This would be something I would pay for, because none of the current desktop tools I had a look at delivered something better. And I also could have all the advantages of an online tool.
Oh I forgot to mention: gizmoz is free to use .
