I'm working with the new Spectral Layers audio program. Quite amazing.
It's not a plugin.
It seems that it will only take about 3 minutes of audio at a time. In other words, if I have a 10 minute wav file, it will only load up the first 3 minutes or so of it.
Anyway, that leaves me with 7 minutes of audio that needs to be worked on.
Does anyone know if it's possible to save it in such a way so as to apply the settings I used to remove unwanted audio to another file?
It's going to come in real handy for me I think. We shot a program recently where the crickets and cicadas were unbelievably loud. I've been able to remove literally all of that background noise yet still leave in the occasional sounds of birds chirping.
I had to do ADR a while back because of water fountain noise and I'm sure this would have been able to remove that as well.
I wish the documentation were better though that comes with the program.
I've been working on some distant traffic hum from an interstate that is in my background audio trying to remove it.
I can remove it but it makes my voice sound a bit like I'm talking through a tube. As you can see by the picture, the hum and my voice are pretty well mixed together and can't be separated very well.
It worked perfectly with some audio to remove cricket sounds but this one I think may be beyond the scope of the program to fix unless I figure it out more.
Sure hope some others join in so we can learn from each other on this program.
What's a bit crazy with this program is that it allows you to make sound that you extract louder but does not allow you to make them softer.
So, I can totally eliminate audio of crickets in the background but I can't then just soften them which leaves me with a sort of sterile background.
The only way around it is to make a wav file of just the crickets and put it in Vegas as another track and lower it there.
The video I saw said you copy the elements you don't want, or whatever they're called, to an new layer and then invert the phase, effectively cancelling it out. I don't have it to work with, but it should work.
That's correct but there is no way to save it to use in another project. Any corrections can only be used in the project your working on.
The only reason I thought that would be good is because I might have an intro of say a minute and then come back to that footage later on in the program which would also need an audio fix.
I guess there would be a workaround by just rendering those two sections together but it would be more of a problem.
To make a sound softer just invert the phase of its layer, and lower its layer volume too.
It's like you start with 10, then you copy this 10 in another layer which makes 10+10=20 (double the volume), you invert the phase 10-10=0 (you don't hear your noise anymore) and then lower the volume of this layer 10-8=2 (the noise is softer but does not disappear).
Also about your cricket and voice sounds (and any other sounds) it's VERY important to choose the right spectral resolution for each of your targets (default is 2048, choose what looks crisper if needed).
I've been working on editing and haven't had a chance to get back to Spectral Layers until now.
I'm loving the program but have a few questions. I'm mostly using the "Register Noise" feature.
1. The "frequency range" allows to swipe a wide path after registering noise but does it have another function?
2. What does the "time range(sec)" do?
3. What does the "hardness" do?
4. What does the tolerance do?
I think I could do a better job at removing sounds if I understood this but the help files don't explain it very well at all.
Thanks Robin and anyone else who has offered explainations.
1. and 2. define the size of your brush.
1 in the frequency scale, 2 in the time scale.
No other function.
3. define if you want a hard brush (plain circle) or a soft brush (gradient circle, allowing smooth extractions); it's working exactly like Photoshop brushes.
4. define a tolerance against the noise print you have previously registered. More tolerance will capture a bit more noise, at the risk of removing real frequencies.
Feel free to play with the other tools as well, they can all help and complement each other during or after extraction. Play with as much layer as possible to decompose and refine your work; and sometime think reverse: deleting a noise may be easier by extracting what you want keep...
Can you explain the tolerance thing a bit more when you say,"define a tolerance against the noise print you have previously registered. More tolerance will capture a bit more noise, at the risk of removing real frequencies."
I'm confused about the part where you say define tolerance AGAINST the noise print previously registered. Does that mean that I capture some noise, test it to see if I got everything and if not I increase the tolerance?
Also, is there an easier way to move around the playing field here without sliding the arrow with the mouse. It seems like a double click on the area would have been much better. Even the grey slider on the bottom is almost the same color and hard to locate sometimes.
About the registered noise tolerance, yes it's basically that.
Are you talking about the play cursor and how to move it around ?
You have 4 options:
grab and slide
click the wanted position on the top bar
slide with the arrow keys on keyboard
select the time cursor tool (shortcut: c) and click the wanted position