IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Digital filters - quality value limits, min/max for Q and bw in EQ implementation
Juha
post Dec 21 2013, 15:03
Post #1





Group: Members
Posts: 442
Joined: 14-February 07
From: EU-FIN
Member No.: 40610



Are there some "standards" for max/min of Q and bw values when used in EQ.

Would 0.001-100.000 for Q and 0.010 - 20.000 for BW be reasonable ranges?


Juha
Go to the top of the page
+Quote Post
saratoga
post Dec 21 2013, 20:37
Post #2





Group: Members
Posts: 4844
Joined: 2-September 02
Member No.: 3264



I think 0.1 to 10 is pretty standard for Q.

BTW, Q is the width of the filter, with Q=1 having bandwidth equal to the center frequency (so quite broad already). A Q of .001 would be so broad it would essentially just be a really complex way of doing a volume adjustment smile.gif

Go to the top of the page
+Quote Post
Juha
post Dec 21 2013, 22:12
Post #3





Group: Members
Posts: 442
Joined: 14-February 07
From: EU-FIN
Member No.: 40610



QUOTE (saratoga @ Dec 21 2013, 23:37) *
I think 0.1 to 10 is pretty standard for Q.
...


I was 'bout to select that range but, then checked the Electri-Q implementation which uses range 0.010-20.000.

As shown in the image (graph is for peak type filter), bw=20.000 corresponds Q=0.001:





In my EQ implementation, it's possible to switch the quality "type" to either Q or bw (or none (if not needed)).

In this implementation, the Notch type filter without given quality value corresponds a Notch filter with bw value set to ~0.048

(i.e. Q = 30.055) so if I select bw as min = 0.100 and max = 20.000 there's area left between those 2 type options that I can't control if needed.

I wouldn't want to set the Q/bw min/max by the filter type.


Juha
Go to the top of the page
+Quote Post
saratoga
post Dec 21 2013, 22:39
Post #4





Group: Members
Posts: 4844
Joined: 2-September 02
Member No.: 3264



I'm not sure I understand what you mean?
Go to the top of the page
+Quote Post
Juha
post Dec 21 2013, 22:46
Post #5





Group: Members
Posts: 442
Joined: 14-February 07
From: EU-FIN
Member No.: 40610



QUOTE (saratoga @ Dec 22 2013, 01:39) *
I'm not sure I understand what you mean?





http://www.sengpielaudio.com/calculator-bandwidth.htm




Juha




Go to the top of the page
+Quote Post
saratoga
post Dec 21 2013, 22:56
Post #6





Group: Members
Posts: 4844
Joined: 2-September 02
Member No.: 3264



I mean I don't understand what you're asking. Did you still have a question?
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Dec 22 2013, 11:57
Post #7





Group: Members
Posts: 3642
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Juha @ Dec 21 2013, 09:03) *
Are there some "standards" for max/min of Q and bw values when used in EQ.


Look at commercial parametric equalizers implemented in the analog domain for guidance in that area. The most recent implementations of real-world equalizers are parts of digital consoles.

Some software implementations are way beyond anything that anyone would actually use or need in real world recording, etc.

QUOTE
Would 0.001-100.000 for Q and 0.010 - 20.000 for BW be reasonable ranges?


Ne Plus Ultra, man.

This post has been edited by Arnold B. Krueger: Dec 22 2013, 11:58
Go to the top of the page
+Quote Post
AstralStorm
post Feb 12 2014, 06:02
Post #8





Group: Members
Posts: 745
Joined: 22-April 03
From: /dev/null
Member No.: 6130



In my experience, Q=0.01-40 are sufficient. There's no real reason to constrain it too badly - the main issue is numerical precision in the coefficients, especially at low frequencies.


--------------------
ruxvilti'a
Go to the top of the page
+Quote Post
markanini
post Feb 12 2014, 15:18
Post #9





Group: Members
Posts: 550
Joined: 22-December 03
From: Malmö, Sweden
Member No.: 10615



Above Q10 it's more of a phase effect than an EQ.
Go to the top of the page
+Quote Post
extrabigmehdi
post Feb 12 2014, 16:13
Post #10





Group: Members
Posts: 408
Joined: 15-August 09
Member No.: 72330



Yeah, the phase shift can increase with Q (assuming minimum phase eq).
And it's preferable to use use smooth curves in your eq anyways.
Unless you use a linear phase eq, but I think you'd hear other artifacts with big Q values.
Notice that the Q value can have different impact from a digital eq, to an other (not necessarily standard values are used).

Go to the top of the page
+Quote Post
AstralStorm
post Feb 13 2014, 22:53
Post #11





Group: Members
Posts: 745
Joined: 22-April 03
From: /dev/null
Member No.: 6130



High Q filters are almost necessary at high frequencies for precision, but I didn't find any use for Q=40 to be honest. Highest I've ever needed was 24 when equalizing headphones. Blame the darn s-plane for being so skewed.

This post has been edited by AstralStorm: Feb 13 2014, 22:54


--------------------
ruxvilti'a
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 22nd July 2014 - 21:57