|
EQ'ing Terms: How to Memorize??
|
View this Thread in Original format
| TranceZoner |
Hey Guys!
I've got a question:
What the hell do terms like 'high pass filter', 'low pass filter' etc. etc. mean? And how do i memorize this? All the tutorials with the subject EQ-ing have these terms in em. what is a high pass filter and how do i memorize them? |
|
|
| dj avary |
HP = High Pass filter,
setting HP = abc Hz means that this filter would cut all frequencies below this 'abc' Hz (i.e. pass only that frequencies, that lying above 'abc' Hz)
LP = Low Pass filter, is vice versa to HP, just change above to below :)
hope this'll help you a little, I'm not to sure that I do not mix up LP and BP but you can see what this filters do in any decent program :)) |
|
|
| robin |
i had question about this too, i understand what it does but not how i do it :D
reason has the ecf-42 evelope control filter but that doesn't seems to do much (or i don't know how to use)
now i just set the vocoder to equalizer and use that to cut out the stuff i don't like (i beleave you use reason too TranceZoner?)
*edit, zo basicly the question was, is there an other way to do this or.. this works good enough so not really important anyway) |
|
|
| TranceZoner |
Yeah i Use Reason too Robin.. By The Way How Are You m8?;)
Use The PEQ-2 Equalizer, then you can experiment with hp filters etc etc. |
|
|
| Phantax |
high pass and low pass filters are self explainatory.
high pass means that it lets all the high frequencies pass through
low pass means all the low frequencies pass through
hi cut means the high frequencies are being 'cut' or removed.
low cut, the low frequencies are being 'cut' or removed.
i dunno what shelving means so i cant help there... or what parametric eq really means, im not sure yet either.
but those other ones pretty much explain themselves and are simple to memorize because of what they're called... it's easy to remember "pass" and "cut" and what that means in eq terms. |
|
|
| dj avary |
| quote: | Originally posted by Phantax
high pass and low pass filters are self explainatory.
high pass means that it lets all the high frequencies pass through
low pass means all the low frequencies pass through
hi cut means the high frequencies are being 'cut' or removed.
low cut, the low frequencies are being 'cut' or removed.
i dunno what shelving means so i cant help there... or what parametric eq really means, im not sure yet either.
but those other ones pretty much explain themselves and are simple to memorize because of what they're called... it's easy to remember "pass" and "cut" and what that means in eq terms. |
aha nice
So I was rite, thank you |
|
|
| Dj Thy |
Well really the names say everything, you don't need to be a genius to know what they do. Just have some kind of graphic representation in your head if it helps.
But, for the sake of it, a little rundown on filters.
High pass : it let's the high frequencies through, it attenuates the low frequencies. In other words it cuts the low frequencies. That's why a lowcut filter is another word for a high pass filter. If you want to be completely right, low cut would be more exact as the filter is made to attenuate low frequencies. The cutoff frequency is just what it says : the frequency where it starts to cut (Well not entirely right, in reality the cutoff frequency is the frequency where there is an attenuation of 3 dB).
Low Pass / High Cut : the opposite of above, cuts the high frequencies.
Now you need to know, there is a certain slope at which the cut can be done. To keep it theoretical, the simplest slope is 6dB per octave (it means that each octave you look further from the cutoff frequency has 6 dB more attenuation. An example. Lowcut with cutoff at 10 kHz. 5 kHz will have lost 6 dB in comparison to the cutoff. 2.5 kHz (one octave below 5 kHz) will be at -12 dB, 1.25 at -18 dB, etc...
We talk of orders or poles (the latter one usually with synths), which is basically a multiple of the slope of that simple filter). A first order filter is that 6dB/oct filter. A second order filter will have a slope of 12 dB/oct. 3d order : 18 dB/oct... Of course this is theoretical, in the real world, complex circuits will be able to use different values, but the main laws still apply).
Band pass : kind of a combination of the two. A certain frequency range is kept "untouched" the rest is cut. So you understand a bandpass filter has two cutoff frequencies (an lower one and a upper one). The range defined by those two cutoff frequencies is the bandwidth. Usually it's given in octaves. For example a bandpass filter with the lower cutoff at 1 kHz, and the upper one at 2 kHz would be a filter with a 1 octave bandwidth.
Now you must know that the transition zones will be pretty abrupt. In a perfect world you would have a slope up to the cutoff (well forget about that 3dB attenuation at the cutoff to keep it simple), then immediately a straight line (no cut at all) and at the second cutoff frequency a slope again. It's possible to make the transition one nice curve (so no straight line at all). Then we'll talk about a bell filter (cuz the shape ressembles that of a bell, duh).
Band cut / Rejection filter : the opposite of the bandpass filter. There is a "dip" between the cutoffs. If the bandwidth is VERY small, we'll talk about a notch filter. Those are usually used to cut an annoying element (like electrical hum) without affecting the useful material too much.
Now about shelved filters. You must look at the action of a "normal" cut filter as going on infinitely. So for example if you have a low cut filter at a certain frequency, it starts cutting at that frequency with a certain slope (defined by the order) and it keeps cutting from there. So for example : you use the lowcut with cutoff frequency at 10 kHz, 4th order (24 dB/oct, the classic Moog filter is a 4th order filter). You won't have much bass left, as it keeps always attenuating the signal more the further you move from the cutoff frequency.
Shelved filters have a limited action and then go at unity gain again. For example : again cutoff at 10 kHz, lo shelved. Say -10dB (here you have to specify an amount of cut). Depending on the electronic build of the filter circuit, it will take some time to reach that 10 dB attenuation (again, it's in fact the slope of the filter). So let us say it reaches -10 dB after an octave. Once it reached -10 dB, it stops cutting. It returns at unity at that point again, which DOESN'T mean it returns to 0 dB after that cut. No with returning to unity I mean it just stops cutting and then gets to a straight line again. So in our example : cutoff 10 kHz, - dB 10, after an octave. So at 5 kHz we reached -10 dB. If you look at 2 kHz, it will still be at -10 dB, if you look at 100 Hz, still -10 dB, and so on. That's just the difference between pure cut and shelved filters.
Low shelf acts on the frequencies below the cutoff, hi shelf on the frequencies above it.
You might have noticed that I haven't talked about boosting, only cutting. That's because the above filters are passive filters. They can only cut. They don't require powering, but they cannot boost.
Active filters can boost, so they involve some gain stage. Basically most active filters can be roughly made by taking one of the above passive filters, and adding a gain control somewhere.
So you got the active filters :
Lo shelved EQ : a lo shelf filter with gain, you cannot only cut, but also boost, with the same effect as explained above, only mirrored.
Hi shelved EQ : dito
Shelved EQ's are usually the Hi and Lo eq's you find on dj mixers or basic production mixers.
Bell EQ / Peak-Dip EQ : remember I talked about bell filters before. Again, we can boost now. But we make it simpler... We don't talk about two cutoff frequencies, but about a center frequency : the frequency where the most gain/cut takes place.
This is usually the EQ you find as Mid EQ on dj mixers.
Now you need to know one more parameter. It's the Q factor. Basically it's the "steepness" of the filter. It's the ratio center frequency over bandwidth. A high Q means the filter will be steep, a low Q means it will be wide. If you compare with the notch filter mentioned above, a notch has a very high Q (as it's bandwidth is very little). Why is this important? Because you can choose to EQ on a very small frequency range (to do precision corrections for example, high Q) or on a wider range (more musical, low Q). And most importantly Q and the choice of center frequency will determine the difference between a normal bell EQ, a semiparametric and a full parametric EQ.
With the peak-dip EQ explained above, you can only change the gain (cut or boost). Everything else is fixed for you.
Semi-parametric EQ : you can change the gain, and select the center frequency where you will operate the EQ. You can't change the Q factor
(Full) parametric EQ : you can adjust everything here : gain, center frequency and Q factor (width of the filter). The most versatile EQ, the most expensive too. Sometimes you may see terms like quasi parametric too. This only means you only have a couple of discrete selections for the Q (Hi and Lo Q settings for example). A real parametric has full control.
Now a little word about other filters you might hear about :
Graphic Eq : those devices with lot's of little sliders you see at concerts a lot. Like the EQ in Winamp! In fact it's nothing more than a combination of a lot of peak/dip EQ's (each slider being one little EQ). Pro graphic EQ's have normally 31 bands (frequency ranges). It means the full frequency spectrum (20 Hz to 20 kHz) has been sliced in 31 parts. The difference between each slider is in fact 1/3d of an octave (simpler versions with 15 or 9 or 5 bands exist also). This amount of control is necessary to compensate for room accoustics in live PA for example. Why is it called graphic? Because the position of the sliders roughly give an image of how the frequency spectrum/correction looks. If lot's of sliders on the left side are turned up, it means you boosted lots of lows...
Paragraphic EQ : with a little common sense, you might already conclude what this is all about. A real paragraphic EQ is a graphic EQ with parametric EQ's instead of peak/dip. Very expensive to make, and frankly you can't find much of them. Usually hardware "paragraphic eq's" are not really paragraphic. Usually it's a marketing joke (for example a graphic EQ + a parametric EQ behind).
There are other filters that are more specific, but they are basically one of the filters (or a combination of) above.
Like allpassfilters. Like the name says, they leave roughly the signal untouched, with the exception of a couple of notch filters inserted. You hear the full signal with some very small frequency "holes".
Formant filters are filters that are made to simulate human voice (like making synths doing oh, ah, eh, uh,... sounds).
Hope this has helped a little,
Thy |
|
|
| robin |
| quote: | Originally posted by TranceZoner
Yeah i Use Reason too Robin.. By The Way How Are You m8?;)
|
not too bad, but will probably get flamed bad in holland forum again :D |
|
|
| Vert |
| quote: | Originally posted by Dj Thy
The giant ass awesome article thy wrote. |
We want you for our site. Alanzo pm'd you but I don't know if you got it so I'm posting here. http://tranceproduction.com/status if you are interested :D
es |
|
|
| TranceZoner |
| quote: | Originally posted by robin
not too bad, but will probably get flamed bad in holland forum again :D |
And why is that m8?? |
|
|
|
|
