Let's go ahead and open up some more examples. Again, we will be in our Media Resources folder. This time we will go to Spectra Examples and we will just select everything.
Let's take a look at the Drum Groove first. Now let's take a closer look at the very colorful Spectral Display and the information it gives us. I trust that when you get used to audio displayed spectrally, you will wonder how you ever survived without it. Instead of being a two-dimensional representation of our sound like the waveform view above it, the Spectral Display actually gives us three dimensions of information. You can consider color to be the third axis as a way to represent depth on our flat two-dimensional screens.
The X-axis serves the exact same purpose in the Spectral Display as it does in the waveform, it represents time. This makes it incredibly easy to resolve which blotches of color correspond to which bumps in the waveform, as they will be directly above and under one another. The Y-axis in the Spectral Display corresponds not to Amplitude but to Frequency. Lower frequencies appear at the bottom of the Display, higher frequencies at the top. The highest frequency displayed in this case just over 20000 hertz is directly related to the sample rate of the current clip being edited.
For example, this clip's Sample Rate is 44100 hertz. The highest tone that can be reproduced by any file at a given Sample Rate is half of that Sample Rate, which is why we find our ceiling here to be just above 20,000 hertz. More specifically, it would be 22050 hertz. We can see if we pan down here a little bit that we do have a clip recorded at a 192000 hertz, a much higher Sample Rate. That clip is the Gamelon Hi Res. Let's come up here, select it, take a look now.
Our highest frequency that's represented here in the Spectral Display is now just over 40000 hertz. You may notice that 40000 hertz is actually nowhere near half of a 196000. This file as it exists would be able to reproduce tones up to 96000 hertz that being half of a 192000. However, our audio hardware in the computer is only capable of reproducing sound that is recorded at up to 96000 hertz. This means that any sound recorded above that Sampling Rate will be played back, but it will be played back at the lower rate.
Soundbooth's Spectral Display reflects this as well. It takes into account the maximum Sampling Rate that your hardware is able to play back and limit itself to that range of audio, just so that it doesn't display information that you are not able to use. The Z-axis in the Spectral Display corresponds to Amplitude and it's represented by color. Blacks, blues, and purples are low values. Reds are medium values and yellows are high values.
Pitched signals as we have seen will typically look like cascades of horizontal lines. While our Noise looks more like blotchy regions of color. I am just going to move up here to the Editor pull-down and choose Drum Groove. Understanding what's in the Spectral Display can give you an insight into your audio beyond what's possible otherwise. It's easy to see relative pitch such as when tones are ascending and descending and isolate problem areas for clean up with much greater accuracy than before. But one thing that's harder to see is the overall combined volume. The sum total of the whole signal. Thankfully, Soundbooth gives you the waveform view on top. So all of that data is readily available to you as well.
Soundbooth's ability to actually make selective edits inside the Spectral Display is incredibly useful as you will see in later lessons, giving you ultimate control over pinpoint accurate noise reduction and other frequency dependent tasks. There are a few further options that Soundbooth offers that allow you to tell how to display the spectral information.
Over here in your Tasks Panel select the Remove a Sound option. We will look first at the Resolution here which we have currently set to Medium. Clicking on it we find three options available to us Fast, Medium and Slow. Let's change to Fast and see what this does.
You can see that with Fast, we actually get more resolution in the X-axis but less resolution in the frequencies in the Y-axis. You can especially see it down here at the bottom where the Y-axis looks like it's blurred up and a down a little bit. What this option gives us is the ability to see more exactly things in the Spectral Display that are time dependent. Really fast transient attacks, clicks, pops and that type of thing. It tends to be less precise when it comes to the actual frequency components of what's going on. Let's come over here and select Slow.
Slow, as you can see gives us essentially the opposite result. It blurs the X-axis which means that it doesn't show what's happening in times effectively. However, it actually gives us a lot more Resolution along the Y-axis. It shows the frequencies much more precisely. This is a better option for dealing with issues, such as hum, bad frequencies that you want to get rid of and that type of thing.
As the timing is not as important and the actual frequency range of the signal affected is what you want to see. Finally, we have this Vertical Scale option here. We as people hear sound in the lower ranges of your hearing ability, a little bit more easily than we hear sound in the high frequencies. It's easier for us to differentiate between frequencies between 20000 or 3000 or 4000 hertz than it is for us to discern between frequencies between say 8K and 20K.
This being given Soundbooth allows us to scale the frequency response of the Spectral Analyzer. If I grab this Vertical Scale option and bring it way up 95 being the highest it goes, you will see that we get a lot more Resolution down here at the bottom. There is a significant amount of space between 1K and 0, between 10K and 20K is really crunched up here at the top. This more accurately reflects the way that we as people hear audio.
However, if you want to do more precise editing in the higher frequencies, pull this Vertical Scale down. When you pull it all the way to 0% you get a linear response across the entire frequency display. You will notice a very similar gap between 10K and 0 as there is between 10K and 20K. You will also see a lot of the information that we as people would find very useful, almost lost here at the bottom of the display.
I typically like to keep my settings at about 75%. Let's look at some examples of spectrum. First we will come up here and select White Noise. Let's take a listen to what this sounds like.
It's a pretty harsh sound, so turn down your headphones if you haven't already done so.
[Demo]
White Noise is what we know as full spectrum noise. It sounds very harsh. This is the basic sound that we get when all the various frequencies are on at roughly the same amplitude. Let's take a listen to what hiss sounds like. You might want to turn your headphones back up here as the Hiss is a little bit more difficult to hear.
[Demo]
You can hear that the Hiss is similar to the White Noise and that it doesn't have a pitch, but it's only the high frequency component of it. That's what it sounds like when you have only these very top frequencies playing and really nothing happening in the lower end. Hiss is a very common audio problem that people find themselves trying to deal with a lot specifically in live recordings, recordings with open microphones and any piece of electronic equipment that generates its own noise.
Let's take a listen to the opposite of Hiss, which is Rumble. You see that he Spectrograph looks very much the same as the Hiss one except it's inverted. The only frequencies playing here are the ones at the bottom of the graph. Let's take a listen.
[Demo]
Rumble is the kind of unwanted noise that you get when there is an air conditioner running in the background, traffic noise on a highway or a bridge behind you, jet engines taking off in the distance, that kind of thing. That's all a really low frequency noise that needs to be filtered out from the bottom of the signal. Let's take a look at what some clicks and pops looks like. Go ahead up here to your Editor pull-down and select
Transcription by:
Scribe4you Transcription Services