This is something I’ve long pondered: when using a Parametric EQ or similar (I use Fabfilter Pro-Q 2), does the ‘Q’ number actually mean something, or is it completely arbitrary? Put another way, if you put in a bell shaped boost of +2dB, with a Q value of 1.4, with two different brands of Parametric EQ, will they both do roughly the same thing (putting aside the colouration of their individual algorithms). I’ve noticed that Q of 0.71 often seems to be neutral.
This might seem like a silly question to some people (“Just use your ears!!”), but I like knowing what things mean/how they work.
EQing - a question about Q values and what actually mean.
-
- KVRAF
- 5716 posts since 8 Jun, 2009
Q is a measure of the bandwidth of a filter: the number is the ratio of the centre frequency to the bandwidth. The higher the Q, the narrower the bandwidth - it affects less of the surrounding audio spectrum. With higher Q, you get more unwanted side effects, such as phase shift inside the filter region. So if you can get away with a lower Q, do that. This is partly why the rule of "boost wide, cut narrow" exists - if you boost with a high Q, you're amplifying a chunk of audio that's had its phase messed up. On a cut, it should be less obvious because, well, that bit's now quieter.
There isn't a hard standard for determining which point on the filter curve you measure to get the bandwidth so it can vary between EQs. If you need to confirm whether two EQs are doing the same thing and at what settings that happens, the best bet is to run white noise through them and match the bell curves on a spectrum analyser. Or you can do the phase-cancellation thing.
There isn't a hard standard for determining which point on the filter curve you measure to get the bandwidth so it can vary between EQs. If you need to confirm whether two EQs are doing the same thing and at what settings that happens, the best bet is to run white noise through them and match the bell curves on a spectrum analyser. Or you can do the phase-cancellation thing.
-
- KVRist
- Topic Starter
- 144 posts since 11 Mar, 2014
Thanks for the reply...I knew about Q being a measure of the bandwidth, the piece of information I was missing was about it being a ratio! So that explanation is really helpful.
In the process of doing this post, I noticed that Q-3 has been released, so that's part of my next pay packet already spent...!
In the process of doing this post, I noticed that Q-3 has been released, so that's part of my next pay packet already spent...!
-
- KVRAF
- 4590 posts since 7 Jun, 2012 from Warsaw
The Q has very specific definition in physics, but I'm not sure how close arbitrary filter designs follow it. Maybe not at all.
Blog ------------- YouTube channel
Tricky-Loops wrote: (...)someone like Armin van Buuren who claims to make a track in half an hour and all his songs sound somewhat boring(...)
Tricky-Loops wrote: (...)someone like Armin van Buuren who claims to make a track in half an hour and all his songs sound somewhat boring(...)
-
- KVRian
- 1274 posts since 24 May, 2004
To add to the good explanation already given by Gamma-UT, I'm just reading Synth Secrets and came across this:
Just to specify what "bandwidth" exactly means in this case. But as Warmonger said: Whether all plug-ins strictly obey that..I'm not sure.
Code: Select all
Center Frequency (in Hz)
Q = –––––––––––––––––––––––––––––––––––––––––––––
0.5 * EQ width (in Hz) at half maximum gain
