Love_In_Rio
Veteran
Assassin's Creed - New Interview
- Thread starter scooby_dooby
- Start date
Fu3lFr3nzy
Regular
Not yet, though GameTrailers just put the same video up 
I'll spare my usual jade worshipping and just say the Eagle shadow thing @ the end of the video was very cool!
I'll spare my usual jade worshipping and just say the Eagle shadow thing @ the end of the video was very cool!
Not yet, though GameTrailers just put the same video up
I'll spare my usual jade worshipping and just say the Eagle shadow thing @ the end of the video was very cool!
Jade looks like she had a sandwhich or two in the interem. Looks better!
Game looks ok too :smile:
I hope they're holding back on cities with entirely different color schemes..the one gray city is getting a bit old.
The art in this game is absolutely amazing! 300 unique character models? Wow!
Didnt oblivion have 1200 different character models or so??
I don't think they were unique. And the major differences would be things like facial features. This looks like it might have the actual scale of the city. I think the actual cyrodill story was supposed to have thousands of people in a city but in the game, there were maybe 40.
Oblivion has 1500 NPCs, but they're definately not unique. Fallout 3 sounds like it has more unique NPCs with a better radiant AI system.
Yes, may be you have 50 differents caracteres in Oblivion
True, but there are hundreds of unique characters (guestimate ~600+). Dont remember excatly how many as I didn't count them precisly when i made a leveling mod.
Gradthrawn
Veteran
The number of unique characters never really bothered me in Oblivion, but the extremely limited number of voices did. I doubt the problem is space, so, that leaves me to believe that recording dialogue is time consuming, and probably expensive. I wonder how effectively mask or filters could be applied (if at all) to help with (believable) variation in voices.
Last edited by a moderator:
It is very time consuming to record voices and the large amount of different phrases or words. They are tehn mixed together to form a conversation. Then you also have to find people to do the voices. But yes to little voices in Oblivion and some odd mixes to, like a young person speaking with a mix of young voice and old voice!
Play the french version, we have, I think, around 15 differents voices…
Why they can put some synthetisal speaking system? The technology have make a lot of improvement this last years…
Filters, especially realtime pitch alteration, is a technology that has recently made some big breakthroughs. There were some fun experiments showing that just by changing the pitch, George Michael sounds *exactly* like Kylie Minogue and vice versa. Very interesting.
Hope for a integration in game for this consoles generation?
Even if you can't do it real-time (which I think you probably can, but may not want to because you don't want to use the processor power for it - could be really cool in something like SingStar though, which had some effects included already on the PS2 version) then I think it would definitely be helpful in sample creation process. You need fewer voice actors and they'll be able to do more voices of both sexes convincingly without sounding the same. A lot of games come with all sorts of stock phrases that all characters use, and you could easily create variation by changing their pitch but nothing else.
Realtime pitch shifting shouldn't be too demanding, but I don't know what the best quality results are like. Pro equipment can do it in realtime readily enough, and have been able to for a while AFAIK. It's an interesting notion. I don't think speech synthesis its up to it with purely synthetic audio, but how's about speech morphing? There was a great (but buggy) audio app on Amiga that morphed between audio samples in a slow audio-rendering process. I haven't seen anything like since, but things on moved on so far, that should be easily realtime on something like Cell. You could then tween sample sets.
Still, they haven't got that working to great effect on graphics yet, and graphics tech is way in advance. If they can't mix up characters to make them not look like clones, why expect them to mix up audio to remove likewise repetition? Audio is ordinarily bottom of the list in features. It's got no representation in stills!
Still, they haven't got that working to great effect on graphics yet, and graphics tech is way in advance. If they can't mix up characters to make them not look like clones, why expect them to mix up audio to remove likewise repetition? Audio is ordinarily bottom of the list in features. It's got no representation in stills!
I have been wondering this for 6 years. At the time one of the faculty members demoed their tool for voice adaptation (not just pitch). It basically takes a few training samples from the target voice, and than transforms whatever you say to that target voice. I was totally shocked by quality.
It is really unbelievable that technology didn't make it to game companies.
Who cares about the realtime part, disk space wasn't a real issue all this time.
Oh! Amiga time! My Golden Age!
It's seem that Mac OSX Leopard have a great synthetized english voice name Alex…
http://www.apple.com/fr/macosx/leopard/features/accessibility.html
Good, but you wouldn't want that in a game though! With games making great leaps and bounds in capturing human emotion in the graphics, having a robot acting your voice parts would really grate!Oh! Amiga time! My Golden Age!
It's seem that Mac OSX Leopard have a great synthetized english voice name Alex…
http://www.apple.com/fr/macosx/leopard/features/accessibility.html
betan's got a good point. If you can morph audio offline, you can save a lot of production time recording a few voices and generating alternatives, rather than sampling 300 different people. Still, capturing audio can't be that expensive. In games costing millions, audio recording of a load of voice actors can't be more than a few % of those costs.
