The Song Fight Bulletin Board

joshw wrote:How does this sound in comparison with using straight EQ in the same direction?

This raises another question which I don't think is worth its own thread (JB: feel free to veto that):

Why does parametric EQ give such a different sound than graphic EQ? When I use a graphic EQ, I get icky flatness and I can never reach my desired sound; when I use parametric, I get flashy results I didn't plan for.

What gives?

I might be totally wrong, but I believe a graphic eq has a fixed Q setting, meaning that if you look at an eq curve which is flat and boost up the 3khz "slider", the slope of the curve which meets 3khz on either side is the same no matter what slider you push. So boosting 3 khz by 3db means also boosting 2.5khz by 1.5db and 1.5 khz at .75 db etc...same on the other side of 3khz. So you might be boosting or attenuating frequencies you didn't want to touch.

And that's the beauty of parametric, you control not only the degree of Q but which exact frequency you want. It can be a very precise instrument when used properly. It also means you need to know a bit more what you are doing and why. One fun part is you can sweep frequencies at a certain boost or cut to find the sweet spot you are looking for, Using a larger than desired (so it will be obvious) boost/cut..it should be easy to find, or it'll give some new ideas about how to structure your mix.

Hope that helps more than it hinders...

TheHipCola wrote:And that's the beauty of parametric, you control not only the degree of Q but which exact frequency you want. It can be a very precise instrument when used properly. It also means you need to know a bit more what you are doing and why.

That really makes sense. I always assumed the graphic EQ bands used band-pass filters that were somehow flat-topped (pretty much equal gain across the band); you're saying they're more pointy (centered on a target frequency using a high Q). That would make it mathematically impossible to make a nice, smooth curve with a "Q" wider than a single fader's bandwidth. You could approximate it (as I always do by making little wavy patterns with the faders), but it won't give you what you want. But a parametric EQ will.

Question asked and answered. Thank you.

Oddly, I find I need to know less to use the parametric because I can easily hunt and peck for that "sweet spot" and I know it when I hear it. The graphic EQ makes me think things like "okay, which band controls that nasally-tinny part again?" before I start playing with it.

yeah i don't ever have a reason to use a graphic eq. you're just limiting yourself if you don't use parametric. another thing to think about for the Q on a parametric rather than just making it smooth is simply how much of that range you want boosted/cut. I find it better to use a more narrow cut/boost on low frequencies to fine tune things, and a wider cut/boost on upper mids and highs depending of course on the song and the desired effect. but remember the E string of a bass is around 41Hz so the low octave of a bass has a range of of 41 Hz because the next E up would be 82Hz. But the higher you go up in frequency there is a wider range of Hz to each octave. For instance with an A at 440Hz and the next one up at 880, that octave covers a span 440 Hz, so you can see why in general you should think about changing the Q specific to the range.

One technique I've used for parametric EQ is to boost the bass, mid or treble all the way up before you start the sweep of the frequency range. Then rotating the sweepable portion of the EQ all the way from the lowest portion of the range to the highest, you can zero in on the frequency that you want to cut or boost. After you have found that frequency, remember to zero the bass, mid or treble and try cutting or boosting to the desired effect. Mind you I'm talking about using this process on an analog board, but I assume the process would be the same for digital.

toddlans wrote:I find it better to use a more narrow cut/boost on low frequencies to fine tune things, and a wider cut/boost on upper mids and highs depending of course on the song and the desired effect. but remember the E string of a bass is around 41Hz so the low octave of a bass has a range of of 41 Hz because the next E up would be 82Hz. But the higher you go up in frequency there is a wider range of Hz to each octave. For instance with an A at 440Hz and the next one up at 880, that octave covers a span 440 Hz, so you can see why in general you should think about changing the Q specific to the range.

erm... Q defines the bandwidth in octaves. A Q of 1.4 will be 1 octave wide at 40Hz, 440Hz, or 10kHz. If your goal is to attenuate a smaller range of low frequencies as well as a larger range of high frequencies note that a constant Q does that inherently.

yeah you're right puce. I guess i worded that kinda retarded. what i was getting at is that since the lower ranges cover a narrower range of frequencies its harder to isolate certain frequencies for say kick or bass with a wider q, so i usually use a pretty narrow one. I'm not thinking in octaves here, i only mentioned them demonstrate that low frequency ranges cover less ground than high ones. sorry.