IST 199 - Fundamentals of Sound Engineering - Course Resource Page

 

In week 1 we spent a lot of time on wave interference, most notably the comb filter. In week 2, we talked about waves arriving out of time with each other, but not creating a comb filter

This should work if you have flash - won't if you don't

 

The High Pass Filter

The first track is unfiltered pink noise. Notice the change when a high-pass filter is applied at 221 Hz. Track 3 features unfiltered vocals, while track 4 is the same thing with a high-pass-filter at 100 Hz. Not much changes, but if there was a kick drum right behind Mr. Pellowe while he was giving his speech, we would hear a lot less of it with the HPF. Putting HPFs on vocals and acoustic guitars is very common in live sound to reject low frequency "mud" that isn't completely necessary to capturing the character of the instrument. The last tracks show what happens to music when it is high-passed. The bass frequencies drop out, significantly changing the character of the song.

 

The Low Pass Filter

Again, we start with unfiltered pink noise, and follow with pink noise LPFd at 850 Hz. Most of the hiss we think of as "noise" disappears. Can you think of a useful application for such a filter based on this? We follow with unfiltered speech, and speech LPFd at 5,000 Hz. There isn't much of an audible difference. HPFs and LPFs are used in conjunction to limit the bandwidth of telephone transmissions in a similar way. The last tracks are unfiltered music and music LPFd above 2.5 kHz.

Low shelf filters

Same as before - 1st pink noise, then pink noise with a low shelf cut, then a low shelf boost. Both boosts and cuts are 6 dB at 200 Hz. Next we have unfiltered music, then music with the same low shelves used on the pink noise.

High shelf filters

Same as before - 1st pink noise, then pink noise with a low shelf cut, then a low shelf boost. Both boosts and cuts are 6 dB at 3,000 Hz. Next we have unfiltered music, then music with the same low shelves used on the pink noise.

Shelf filters generalized:

While HPFs and LPFs are primarily used to cut out something unwanted, shelf filters are more often used to change the bass/treble balance of an instrument.

Parametric EQ:

This is where we start having some more flexibility with our filters. These filters will be easier to describe graphically. Here's the breakdown:

1st, unfiltered pink noise for comparison, then pink noise with the following EQ curve:

After an unfiltered bit of John Pellowe's lecture, I applied the same filter shown above. On the next two tracks I do the same thing to music. This is a very extreme filter; one you're not likely to find a use for in practical application, but it is important to be able to hear what a lack of midrange sounds like. I think the effect is most pronounced on the vocals, which have the bulk of their frequency content in the midrange just below this filter and are consequently colored and attenuated.

Here is a slightly more reasonable EQ to apply to speech:

The HPF is in place to reduce mud and rumble. The narrow parametric cut at 123 Hz is to reduce plosive sounds (pops) caused by rushes of air from the speakers mouth. The parametric peak at 3.5 kHz is to increase the clarity of speech by giving a lift to consonant sounds, and the LPF is to reduce the chances of high frequency feedback if the presenter gets too close to a loudspeaker. It's a little drastic here, and I don't have a good audio file for comparison, but will try to post one soon. I call this the "corpo EQ." It's an attempt to fight the atrocious mic technique most non-musicians exhibit.

For music, you're very likely to see this type of EQ:

The intent is to add more punch to the kick and bass. It's not always a good idea, but it's done quite a lot to bring the bass instruments up in the mix without interfering so much with the mid and high frequency voices.

The take-home message on EQ:

Discretion, discretion, discretion!