DSPRelated.com
Forums

Audio vs Video

Started by Vladimir Vassilevsky July 24, 2007
  Just an observation: the largest part of a modern computer game is the 
library of the miscellaneous sound samples. Announcements, dialogs, 
nature, music, etc.

Isn't it strange that the video part of a game is produced by the real 
time rendering of some kind of a simple model of the reality, whereas 
the audio is stored as the samples of the live sounds?

Why the fairly sophisticated synthetic audio is perceived as disgustful 
whereas the primitive synthetic video is OK?


Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

Vladimir Vassilevsky wrote:
> > Just an observation: the largest part of a modern computer game is the > library of the miscellaneous sound samples. Announcements, dialogs, > nature, music, etc. > > Isn't it strange that the video part of a game is produced by the real > time rendering of some kind of a simple model of the reality, whereas > the audio is stored as the samples of the live sounds? > > Why the fairly sophisticated synthetic audio is perceived as disgustful > whereas the primitive synthetic video is OK?
Audible distortion is generally more objectionable than visual distortion. Do you think that may be related to your question? Jerry -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
"Jerry Avins" <jya@ieee.org> wrote in message 
news:3audnYW9vtzs_TvbnZ2dnUVZ_uXinZ2d@rcn.net...
> Vladimir Vassilevsky wrote: >> >> Just an observation: the largest part of a modern computer game is the >> library of the miscellaneous sound samples. Announcements, dialogs, >> nature, music, etc. >> >> Isn't it strange that the video part of a game is produced by the real >> time rendering of some kind of a simple model of the reality, whereas the >> audio is stored as the samples of the live sounds? >> >> Why the fairly sophisticated synthetic audio is perceived as disgustful >> whereas the primitive synthetic video is OK? > > Audible distortion is generally more objectionable than visual distortion. > Do you think that may be related to your question? > > Jerry > -- > Engineering is the art of making what you want from things you can get. > &macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;
Jerry - what is the wrapped line of funny looking A symbols supposed to be? I view everything in plain text, maybe that's why I don't get it. Just curious

Jerry Avins wrote:


>> >> Why the fairly sophisticated synthetic audio is perceived as >> disgustful whereas the primitive synthetic video is OK? > > > Audible distortion is generally more objectionable than visual > distortion. Do you think that may be related to your question?
This is not it. A low bitrate vocoder produces the speech of the toll quality, despite of the distortion of somewhat 30% if compared to the original sound. But the full synthetic narrator always sounds like rotten. Why it is so difficult to build a model based audio? VLV
On Tue, 24 Jul 2007 16:22:05 -0500, Vladimir Vassilevsky wrote:

> Jerry Avins wrote: > > >>> >>> Why the fairly sophisticated synthetic audio is perceived as >>> disgustful whereas the primitive synthetic video is OK? >> >> >> Audible distortion is generally more objectionable than visual >> distortion. Do you think that may be related to your question? > > This is not it. A low bitrate vocoder produces the speech of the toll > quality, despite of the distortion of somewhat 30% if compared to the > original sound. But the full synthetic narrator always sounds like > rotten. Why it is so difficult to build a model based audio? > > VLV
Model based audio from musical instruments isn't that bad -- it's the human voice that's hard. For a video counter example, do a web search on "Uncanny Valley". -- Tim Wescott Control systems and communications consulting http://www.wescottdesign.com Need to learn how to apply control theory in your embedded system? "Applied Control Theory for Embedded Systems" by Tim Wescott Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html

Tim Wescott wrote:


>>>>Why the fairly sophisticated synthetic audio is perceived as >>>>disgustful whereas the primitive synthetic video is OK? >>> > > For a video counter example, do a web search on "Uncanny Valley". >
Straight to the point as usual, Tim! In the short, if something resembles a defective human, it is very unpleasant. The clearly non human objects are perceived better then that. I wonder which side of the valley is closer to me :-) VLV
BobF wrote:

   ...

> Jerry - what is the wrapped line of funny looking A symbols supposed to be? > I view everything in plain text, maybe that's why I don't get it. > > Just curious
It's called "macron" and serves as an full-width high dash. When it displays properly, if forms an underline for the line above it. It's character U+00AF; alt+0175. Jerry -- Engineering is the art of making what you want from things you can get. _______________________________________________________________________
Vladimir Vassilevsky wrote:

> ... Why it is so difficult to build a model based audio?
Ignorance? Jerry -- Engineering is the art of making what you want from things you can get. &macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;
Jerry Avins wrote:
> Vladimir Vassilevsky wrote: > >> ... Why it is so difficult to build a model based audio? > > > Ignorance? >
Accurate physical modelling is still quite demanding in CPU terms, and if the best outcome is that it sounds (to the average ear) just as good as sampling, the easier and cheaper option will win. Physical modelling for games is a very active research topic (e.g. for sonically realistic collisions, impacts, etc effects); its appearance in games may depend on factors such as availability of high-level tools. Also, frankly, real-life sounds are very dull with little or no 'emotional impact"; does anyone believe for example that the sounds one hears in films are true-to-life? Even the dialog is re-recorded, 99% of the time, an inch from a microphone in a clean dry acoustic. All the ambience is artifical, to say nothing of the vast amount of Foley that is used. I read a comprehensive account of the sound design for "Titanic" (i think, in Audio Media, might have been Studio Sound); great care was taken to develop the sounds of the creaking of the ship, progressively, so that they sounded more aweful and frightening as the film reached its climax. Some 200 different sounds were used. Those sounds were sampled and processed from a variety of sources, none of which (IIRC) originated on any ship. The rhetorical effect of sound is far more important than how "realistic" it is. And the truth is that most people are not audile to the extent that they notice such things with discrimination. Nobody even bothers complaining about explosions and whoosh sounds in space any more, because, as was stated very simply by someone on TV not very long ago, "silence is boring". Actually, very little on either TV or film is realistic in an objective sense; clearly games are no different in principle. The danger of literally accurate physical modelling is that the result will be just as dull as the real thing! Richard Dobson
"Jerry Avins" <jya@ieee.org> wrote in message 
news:DbidnR4Dj81MMTvbnZ2dnUVZ_ryqnZ2d@rcn.net...
> BobF wrote: > > ... > >> Jerry - what is the wrapped line of funny looking A symbols supposed to >> be? I view everything in plain text, maybe that's why I don't get it. >> >> Just curious > > It's called "macron" and serves as an full-width high dash. When it > displays properly, if forms an underline for the line above it. It's > character U+00AF; alt+0175. > > Jerry > -- > Engineering is the art of making what you want from things you can get. > _______________________________________________________________________
OK. Funny, this particular post shows the underline as it's supposed to be. But only this one. Thanks