I think most of us agree that G-Force rocks but I was thinking the other day (yes, it does happen) and I was trying to decompose what makes GF so cool.
Maybe this is futile exersize since, in my mind, GF is part art, part science and art as we all know can be subjective.
So... what, in your opinion, makes a certain visualization appealing?
For me, a large part of it has to do with how responsive the graphics are to the music. When something dynamic happens in the audio something dynamic has to happen on screen or I'm dissappointed. I'm rarely disappointed with GF. Part of it has to do with the number of ways the dynamics are rendered: elements change size, shape, position, and/or color.
It's kind of like the opening scenes from the original Fantasia only better (IMHO).
Then there's the things that change regardless of what the music is doing... global pallet, backgrounds, etc.
The other really cool thing about GF is the seemless transitions it does between the various flavors of visualizations. Lines break apart, bend, and some rejoin in totally new shapes while others fly off screen.
All in all, never a dull moment.
What Makes A Good Visualization (A Philosophical Discussion)
Moderators: BTT, andy55, b.dwall, juxtiphi
There are many different approaches.
Variety is very important in a visualization. It needs to change other the viewer will get bored. GF does this really well with its transitions from scene to scene and the many patterns that can be made with different combinations.
Moving to the music is important too. Here is where a few authors are divided though. I like my vises where the whole screen/more of the scene moves to the music rather than just the waveform and so do many others so in that sense the 'look' of GF I find ….ummh…I cannot think of the correct word I will get lambasted if I use the word ‘dated’ and ‘limited’
But others don’t like this approach at all I had a discussion with Jaw who wrote many/most of the sonique player visualizations (with CoR) about what I was doing in R4. And he complained of motion sickness when viewing some of the scenes like travelling/bouncing to the music down the tunnels etc.
I never really thought that way until recently.
Also in another online chat was discussed with a viewer why he liked GF over MilkDrop. And made some interesting comments about how MilkDrop looked like people just tried to cram the visually best stuff they could have in there and he preferred GForce as they it had more ‘direction’ and style of the visualization and all the scenes flowed into one another.
An interesting point of view and one that I will look at integrating more into any new visualization projects I embark on butt his gets more and more difficult if you have more variety and us more 3d technology for the different scenes within a visualization.
Sometimes I am disapointed in visualization if I see how it can be improved to make more variety, more happening, more complexity, etc.
Variety is very important in a visualization. It needs to change other the viewer will get bored. GF does this really well with its transitions from scene to scene and the many patterns that can be made with different combinations.
Moving to the music is important too. Here is where a few authors are divided though. I like my vises where the whole screen/more of the scene moves to the music rather than just the waveform and so do many others so in that sense the 'look' of GF I find ….ummh…I cannot think of the correct word I will get lambasted if I use the word ‘dated’ and ‘limited’
But others don’t like this approach at all I had a discussion with Jaw who wrote many/most of the sonique player visualizations (with CoR) about what I was doing in R4. And he complained of motion sickness when viewing some of the scenes like travelling/bouncing to the music down the tunnels etc.
I never really thought that way until recently.
Also in another online chat was discussed with a viewer why he liked GF over MilkDrop. And made some interesting comments about how MilkDrop looked like people just tried to cram the visually best stuff they could have in there and he preferred GForce as they it had more ‘direction’ and style of the visualization and all the scenes flowed into one another.
An interesting point of view and one that I will look at integrating more into any new visualization projects I embark on butt his gets more and more difficult if you have more variety and us more 3d technology for the different scenes within a visualization.
Sometimes I am disapointed in visualization if I see how it can be improved to make more variety, more happening, more complexity, etc.
-
JayPro
- Posts: 738
- Joined: Sat May 01, 2004 10:51 pm
- Location: Huntington Station, Long Island, New York
2¢
See, this is where I think Project X (along with the next generation of vis apps, especially some current iTunes visuals like Kaleidostrobe) IMO will succeed marvelously.
It may be in the minds of some that having a line, squiggle or figure act as the sole harbinger of audiovisual response is passé. I've seen color become an increasingly active component of a particular vis-app's response scheme; but only in rudimentary manifestations. The next step will be IMO ways to tailor pixelflow schemes or 3D-behaviors towards the same thing.
PS: Perhaps it would have been an interesting exercise to see whether the idea of reactive colormapping in G-Force could have worked. Could it *still* work?
It may be in the minds of some that having a line, squiggle or figure act as the sole harbinger of audiovisual response is passé. I've seen color become an increasingly active component of a particular vis-app's response scheme; but only in rudimentary manifestations. The next step will be IMO ways to tailor pixelflow schemes or 3D-behaviors towards the same thing.
PS: Perhaps it would have been an interesting exercise to see whether the idea of reactive colormapping in G-Force could have worked. Could it *still* work?
"God is syntax."
More brain farts
I'm not entirely sure 3D visualization is the be all end all but it does offer some interesting opportunities...
Currently changes in music dynamics / wave form can be reflected on screen by changes "element" size, shape, position, and/or color. With 3D you could add depth and rotation to the mix (change the Z value based on sound intensity). Maybe you could use additional motion blur as well... If the sound intensity changes quickly enough some element could quickly sping forward (Z axis), increase in brightness, and leave a blur behind.
As they say, "the proof is in the pudding", I think I'd have to see a cool 3D visualization in action before I could say it's the way of the future (the 3D vis that comes with JRiver Media Center isn't that cool IMHO).
Currently changes in music dynamics / wave form can be reflected on screen by changes "element" size, shape, position, and/or color. With 3D you could add depth and rotation to the mix (change the Z value based on sound intensity). Maybe you could use additional motion blur as well... If the sound intensity changes quickly enough some element could quickly sping forward (Z axis), increase in brightness, and leave a blur behind.
As they say, "the proof is in the pudding", I think I'd have to see a cool 3D visualization in action before I could say it's the way of the future (the 3D vis that comes with JRiver Media Center isn't that cool IMHO).
Kaleidostrobe is one of the few visualization I have not seen as it is MAc based although the screenshot look a bit to samey but that is only from having viewing the screenshots. But I would not call it next generation though?! I wonder how it would hold up against what is out there for the PC atm.
Arragh the great Mac/Pc divide. Sometimes I seem like the only PC user on these forums.
I have done some stuff with colour changing things movement to the music but it is
Talking about 3D stuff the first major ones to come on the scene where BEn MArsh's Tripex series at the same time Gordon Willaims was doing his R series.
But tripex is a few years old (version 3) now since it's last meaning full update.
http://www.tripex2.fsnet.co.uk/
http://www.tripex2.fsnet.co.uk/screenshots.html
This came out the same time as R2/Extreme back in 2002
http://www.rabidhamster.org/R2/index2.php
http://www.rabidhamster.org/view.php?subdir=/R2Extreme
And now R4 is the latest incarnation.
http://www.rabidhamster.org/R4/
http://www.rabidhamster.org/R4/screenshot.php
Screenshot though as you know don't really do vises justice.
There are various techinques to add to 3d movement spin, depth, etc. On to a a deltamaps type effect. In fact all that you say here is already being done so I am somewhat confused.
A mixture of 2d and 3d is the way I see it going forward. We have become sooo used to 3d now on computers it is hard picture it without it. At least in years to come.
Arragh the great Mac/Pc divide. Sometimes I seem like the only PC user on these forums.
I have done some stuff with colour changing things movement to the music but it is
Talking about 3D stuff the first major ones to come on the scene where BEn MArsh's Tripex series at the same time Gordon Willaims was doing his R series.
But tripex is a few years old (version 3) now since it's last meaning full update.
http://www.tripex2.fsnet.co.uk/
http://www.tripex2.fsnet.co.uk/screenshots.html
This came out the same time as R2/Extreme back in 2002
http://www.rabidhamster.org/R2/index2.php
http://www.rabidhamster.org/view.php?subdir=/R2Extreme
And now R4 is the latest incarnation.
http://www.rabidhamster.org/R4/
http://www.rabidhamster.org/R4/screenshot.php
Screenshot though as you know don't really do vises justice.
There are various techinques to add to 3d movement spin, depth, etc. On to a a deltamaps type effect. In fact all that you say here is already being done so I am somewhat confused.
A mixture of 2d and 3d is the way I see it going forward. We have become sooo used to 3d now on computers it is hard picture it without it. At least in years to come.
No Rovastar, you ain't the only PCer on the block. There are two of us. Wheeeee !!!
I'm afraid a lot of this 3D discussion is pipe-dreaming until Andy comes with a much more efficient engine for G-Force. I know that's in the works, but I'm not sure it will be enough unless it is a major efficiency breakthrough. My PC ain't the fastest machine on the block, but it's no slouch either, with a 2.2 GHz P4, a half gig of memory, and fairly fast graphics. When running about a 750x750 pixel window, my CPU is usually about 75-85% loaded, with occasional peaks into the high 90s, depending on what configs are running, and I'm getting a frame rate of about 18. All of this is quite acceptable. G-Force looks good, not jerky except on very rare brief occasions, and I can do pretty much anything I want that isn't too CPU or disk intensive. I cannot do any intense number crunching, and an Access search take forever, or even longer. If G-Force gets much more math-intensive it will take one hell of a computer to run it 'at speed'.
I don't think I am creative enough or have enough visualization experience to try to specify that would be a 'good' visualization... but I am opinionated enough to know what I like, and I love G-Force. It may not be doing everything that it could be doing, but it is doing everything that it does very well.
Andy has to be careful in what he does however. A genius level programmer I worked with 15-20 years ago, who did all of the grunt-work on a petrophysical software package we were building from scratch, used to say something to the effect that the bane of all software, and especially good software, is 'creeping featurism'. You add this little feature that someone wants, then that feature that someone else wants, and on, and on... and after a while your good, fast package has turned into a unusable, lumbering behemoth that can't even get out of its own way. But you can't just go back and take anything out, because everythin is now considered an integral part of the package that is supposedly critical to its ongoing success.
Can you spell Microsoft? Can you spell Windows? Do you remember when even complicated software would fit on a couple floppies?
You should be careful of what you ask for. You might just get it. And 3D could be the straw that breaks G-Force's back.
I'm afraid a lot of this 3D discussion is pipe-dreaming until Andy comes with a much more efficient engine for G-Force. I know that's in the works, but I'm not sure it will be enough unless it is a major efficiency breakthrough. My PC ain't the fastest machine on the block, but it's no slouch either, with a 2.2 GHz P4, a half gig of memory, and fairly fast graphics. When running about a 750x750 pixel window, my CPU is usually about 75-85% loaded, with occasional peaks into the high 90s, depending on what configs are running, and I'm getting a frame rate of about 18. All of this is quite acceptable. G-Force looks good, not jerky except on very rare brief occasions, and I can do pretty much anything I want that isn't too CPU or disk intensive. I cannot do any intense number crunching, and an Access search take forever, or even longer. If G-Force gets much more math-intensive it will take one hell of a computer to run it 'at speed'.
I don't think I am creative enough or have enough visualization experience to try to specify that would be a 'good' visualization... but I am opinionated enough to know what I like, and I love G-Force. It may not be doing everything that it could be doing, but it is doing everything that it does very well.
Andy has to be careful in what he does however. A genius level programmer I worked with 15-20 years ago, who did all of the grunt-work on a petrophysical software package we were building from scratch, used to say something to the effect that the bane of all software, and especially good software, is 'creeping featurism'. You add this little feature that someone wants, then that feature that someone else wants, and on, and on... and after a while your good, fast package has turned into a unusable, lumbering behemoth that can't even get out of its own way. But you can't just go back and take anything out, because everythin is now considered an integral part of the package that is supposedly critical to its ongoing success.
Can you spell Microsoft? Can you spell Windows? Do you remember when even complicated software would fit on a couple floppies?
You should be careful of what you ask for. You might just get it. And 3D could be the straw that breaks G-Force's back.
Well I'm pretty new to GF and I can't say that I'm up on whats being done "out there" by other developers. So your confusion it no doubt due to my ignorance or inability to describe things in the proper terms.There are various techinques to add to 3d movement spin, depth, etc. On to a a deltamaps type effect. In fact all that you say here is already being done so I am somewhat confused.
Thanks for the pointers. Some of those screen shots are quite beautiful. They also help to clarify some misconceptions I had of what 3D visualizations are all about. I was picturing texture mapped models that were rather sharp and angular, not the curvy, flowing models depicted in the screen shots. Nice stuff!
I should add that my acceptance to 3d visualizations is still gated by how well they react to the track being played. If they don't react well then I may as well watch a pretty screen saver.
I didn't mean to sound condecending or anything I have seen way too many visualizations and forget that most (allmysticeti wrote:Well I'm pretty new to GF and I can't say that I'm up on whats being done "out there" by other developers. So your confusion it no doubt due to my ignorance or inability to describe things in the proper terms.There are various techinques to add to 3d movement spin, depth, etc. On to a a deltamaps type effect. In fact all that you say here is already being done so I am somewhat confused.
) out there haven't seen as many as me. In my opinion most do 'react' better then GForce. Some of the R4 scenes need a bit more work but for the sound reaction stuff something like MilKDrop is probably classed by many as the best out there atm. There are more variables to choose from teh rate of change of teh bass one frame to the next and this is blended with other aspects of teh music and design your own beat detection, etc.mysticeti wrote: Thanks for the pointers. Some of those screen shots are quite beautiful. They also help to clarify some misconceptions I had of what 3D visualizations are all about. I was picturing texture mapped models that were rather sharp and angular, not the curvy, flowing models depicted in the screen shots. Nice stuff!
I should add that my acceptance to 3d visualizations is still gated by how well they react to the track being played. If they don't react well then I may as well watch a pretty screen saver.
http://www.nullsoft.com/free/milkdrop/
http://www.milkdrop.co.uk
The beauty of GF IMHO is that it all flows and has more linked direction to it all but the sound reactivity is a little basic as only the waveform move. Not for example the delta maps moving dynamically to the bass or whatever and every 4 beat change have a flash of colour zooming away from you.
(It is stupidily hard to describe visualizations at the best of times)
Some find it too much and thinking about it when I say it for the first time in June 2001 or something it was too much to be honest. Too much movement but now it is more common and work well.
"You should be careful of what you ask for. You might just get it. And 3D could be the straw that breaks G-Force's back."
I don't think so. there is always a bueaty behind GF. 3d waveforms could be blended into GF. Andy is confident of a marked speed increase (100+ times I think he said) this still has me curious
BTW you must running GF at a huge res for that slowdown.
What seems to be considered good visualizations today are those that either look the prettiest & most detailed 3D stuff by requiring high end systems or those that are fluid in changing to the music. I prefer the latter. If I want to look at something that's only pretty, I'll look at something like the sunset.
Variety definately makes for better visualization. The more times you see a certain shape, picture, or thing in a visualizer, the less likely you are to "Oooh" and "Ahhh" at it again.
I hope visualizations get better at reacting to changes in the music such as build ups, fade outs, breakdowns, or a sudden change to loud or quiet noise. These are the things that I would "choreograph" a visualization to, via direct commands or a script. If done yourself, it can look awesome, but it can take quite some time to make it happen. For example, I made a GF script for a specific song once. It took listening to it a few times, and pausing it countless times to get things to morph or particles to appear when I wanted them to. Also, I made a music video tribute of The Matrix years ago. It took a long time to choose 3 minutes of scenes from that 2 hour movie and put it to music in the way I saw fit, but the end result is something that I am really happy with.
Some day I would love to see a multitrack type of visualizer that could react in distinctly different ways to different elements of a song like Vocals, Guitar, Drums, etc. This however wouldn't be possible with live/realtime processing, because the visualization program would have to know how to differ between the instruments.
Variety definately makes for better visualization. The more times you see a certain shape, picture, or thing in a visualizer, the less likely you are to "Oooh" and "Ahhh" at it again.
I hope visualizations get better at reacting to changes in the music such as build ups, fade outs, breakdowns, or a sudden change to loud or quiet noise. These are the things that I would "choreograph" a visualization to, via direct commands or a script. If done yourself, it can look awesome, but it can take quite some time to make it happen. For example, I made a GF script for a specific song once. It took listening to it a few times, and pausing it countless times to get things to morph or particles to appear when I wanted them to. Also, I made a music video tribute of The Matrix years ago. It took a long time to choose 3 minutes of scenes from that 2 hour movie and put it to music in the way I saw fit, but the end result is something that I am really happy with.
Some day I would love to see a multitrack type of visualizer that could react in distinctly different ways to different elements of a song like Vocals, Guitar, Drums, etc. This however wouldn't be possible with live/realtime processing, because the visualization program would have to know how to differ between the instruments.