Newbie here

[bnh]

I can help with some general directions.

I use Vegas and Adobe AfterEffects/Premier. Also have a bacground in classical music. Have done some of the things you're trying to do. I acutally ran 3 versions of GForce on three computers during a live concert where each version/computer was confined to just one instrument -- a trio obviously. (Yes, I contacted SoundSpectrum and paid their commercial fee.) You will have to use a video mixer to get that kind of separation-between-generated-images and specific instruments/sounds because GForce can't do it all alone. Just too much demand on the processor...any procvessor. GForce was designed to generate images in response to live audio stimuli. That's it. It wasn't intended to do compositing or chroma keying, luma keying, matting, etc. plus generate the visual image in real time. It seems to suggest that you can do so, but the processor is just too busy trying to translate the audio input into a visual output without stuttering.

GFoirce is very unflexible once basic parameters are set. And you change parameters via scripting or config files, which is another learning curve. Editing visual effects with scripts using GForce is like going back to linear video editing before they even had puch-points or edit cues stored on the second audio channel of 2-inch tape, linked to an external switcher/effects controller. With GForce, you always have to start at the beginning of the piece (script) to see what happens when the new visual effect starts. Playing around, investigating, changing your mind, experimenting is very time consuming because there's no "pick up at 1:05:29" It's maddening and an extreme wast of time.

I've posted some examples of keying and layering under jerohm's famous "coding with waveshapes" post. Don't ask me to find them. It's been quite a while. It helps to learn how to write config files, which will help "tame" some of the GForce responses. You'll need to understand how color maps work so you can assign specific colors (usually black) to different moving parts of the visuals, and then learn how to covert that color (usually black) to an alpha channel for tranparency.

To capture GForce you'll have to invest in a screen capture program. Camtasia is okay, but a little heavy on the processor side. There are free FRAPS programs that are good, and I think Windows 7 - 10 has its own screen capture program, but I'm not sure it captures video.

I like to use GForce to generate effects that are simply intercut or keyed under/over standard video. They look great and are easy to build.

I assume you have isolated your instruments and have not yet made a mix of the tracks. I'm assuming you hve one instrument or group of instruments per track. You can work with the entire group or a completed mixdown, but your control will be very limited... even though GF says it is responding to frequency and amplitude. If you have your music mixed onto a single track, then the stimuli are the sum of those frequencies and amplitudes. In other words, you won't be able to isolate instruments or bands of frequencies very well. You won't be able to isolate the flute part from the obe part or the violas from the cellos. You'd be lucky to isolate the bass part and assign it to a particular visual that responds to bass peaks, but even that response will be very iffy if its a mix down of all instruments. And even more tricky if its classical music that doesn't have a pronounced drum track contaminated with amplitude spikes.

Yes, the GForce visuals can respond ive to the overall mix, but the consitency, flexibility and predictability of the response will be difficult to maintain.

You will have much better results and options -- creatively --if you try not to work in real time, i.e. have your audio recorded and un-mixed. Then assign a GForce visualization to one instrument (or group) for a certain length of time and video capture/record that visual. Then assign that captured video to a particular track in Vegas and go on to the next instrument/musical stimuli. In this manner you'll have much more control over the final visual output. You can use Vegas or Adobe's products to do the layering and keying, and you won't have to invest in a video mixer.

Why? Because all the difficult visual layering stuff, the compositing, masking, matting, keying, track layering, etc. in video is not done in real time in the non-linear editing system. Yes, you can watch a much-reduced-version of the video and effects in real time, but the actual final product in all its glory is accomplished during rendering, which is done frame by frame over a considerable amount time.

GForce works in real time. The non-linear video editors depend on a lesser quality image to view in real time, but actualy do the real rendering in variable time.

There are three elements at play at all times in GForce you'll have to master:
1) the config files, i.e. the waveshapes, color maps, flow fields, etc. how do they respond to certain stimuli and how to modify their reaction to certain stimuli.
2) the scripts, i.e. the way to link combinations of waveshapes and color maps and flow fields. In the script, you call specific config files at specific times, e.g. at 4:01 you want config waveshape file X to fade in over Y seconds while config waveshape file Z is fading out.
3) the underlying parameters file (I forget what its actualy called) that is loaded during the GForce boot. This underlying parameters file determines what happens when certain colors meet, among several other things. This is a tricky file to master, but very important. Those parameters can be changed by copying the file to a text editor (e.g. word pad), edit the file, save it back to a text file, and then load it into the appropriate boot folder in GForce.

There's a hell of a learning curve here because the GForce system is not well documented. It was designed to be enjoyed without undeerstanding how to modify the various elements. This forum is basically populated by control freaks who love to take things apart, and it's good there are such individuals. It's a lot of trial and error. This forum will be your main source of detailed information. There's some very bright people here.

Give me a shout if you have any more general questions. The time invested will depend on how much control you want to exert over the visuals. If an already constructed waveshape config is good enough, you'll move faster than trying to learn how to construct wave shapes with math formulas, which also means you'll be accepting what is available rather than creating what you conceive. It's the difference between assembling a music track using digital effects with loops as opposed to learing how to write notes on a score and performing that score. I'm sure you get the idea.

[FrankN]

Hi bnh,

Thanks for writing such a detailed and helpful response! I appreciate it!

[g_randybrown]

Yes thanks very much bnh!
What I later did was (as you suggested) feed GF with a single instrument audio track (ie piano, guitar, etc)...recorded that with Bandicam (full version does a nice job with no dropped frames).
I then brought those individual clips into Vegas and color-keyed out the black background (so the other video tracks were visible too) and then added background video and/or generated media under those tracks.
Kinda cool but less so than I'd hoped for.
Thanks again bnh,
Randy