shure e2's and e3's on rockbox

[Febs]

Nate, you might find this useful:
http://www.recordingwebsite.com/articles/eqfreq.php

[Nate!]

Perfect Febs, just perfect.  Thanks!

[soap]

Nate!, I think I plotted your EQ settings properly, and I don't fully grok them.
That cut you make at 500 is the one which catches my eye.

[Nate!]

soap: While I was listening, it seemed that that range was clipping or distorted.  So I was trying to compensate.  I tried playing with the precut setting, but it didn't seem to get rid of it.  I didn't try Febs advice that I posted earlier.  I'll try that later.

[Arj]

Nate!, I think I plotted your EQ settings properly, and I don't fully grok them.
That cut you make at 500 is the one which catches my eye.

Soap, I am trying to understand how you came up with that graph.  I assume you are using the rbj_eq.xls spreadsheet to get that?  If so, it looks like you are inputting the Q settings from Nate's cfg file as, for example, 10 for his cutoff at 150, rather than 1.0.  Similarly, you are using a Q of 5 rather than .5 for the band at 500.  Am I reading this right?

Again, if I am understanding how you did this, shouldn't the Q's be entered as 1.0 and .5 rather than 10 and 5, respectively?

[soap]

Arj, I believe you are correct on all counts.
It wasn't the Q values which caught my eye (though I get your point), but the gain values.

Code: [Select]

eq enabled: on
eq precut: 0
eq band 0 cutoff: 150
eq band 0 q: 10
eq band 0 gain: 0
eq band 1 cutoff: 500
eq band 1 q: 5
eq band 1 gain: -50
eq band 2 cutoff: 1500
eq band 2 q: 8
eq band 2 gain: 0
eq band 3 cutoff: 4000
eq band 3 q: 5
eq band 3 gain: 10
eq band 4 cutoff: 11000
eq band 4 q: 5
eq band 4 gain: 20

So is that -50dB, or -5.0?
If the former - I find it a bit odd.
If the latter, I can grok it, but that would make the 1.0 gain at 4000 to be a waste of CPU, IMHO.

[Arj]

Arj, I believe you are correct on all counts.
It wasn't the Q values which caught my eye (though I get your point), but the gain values.

Code: [Select]

eq enabled: on
eq precut: 0
eq band 0 cutoff: 150
eq band 0 q: 10
eq band 0 gain: 0
eq band 1 cutoff: 500
eq band 1 q: 5
eq band 1 gain: -50
eq band 2 cutoff: 1500
eq band 2 q: 8
eq band 2 gain: 0
eq band 3 cutoff: 4000
eq band 3 q: 5
eq band 3 gain: 10
eq band 4 cutoff: 11000
eq band 4 q: 5
eq band 4 gain: 20

So is that -50dB, or -5.0?
If the former - I find it a bit odd.
If the latter, I can grok it, but that would make the 1.0 gain at 4000 to be a waste of CPU, IMHO.

In the cfg file, what is listed as -50 is actually -5.0dB.  In band 4 that you listed, a gain of 20 is 2.0dB.

But back to the Q values for a minute.  I have been playing with that spreadsheet to see how it maps out some of the cfg's I am experimenting with and my Q values, as set from the graphical EQ screen, range from .5 to 1.0.  In the .cfg file these are values from 5 to 10.  When I put them in the spreadsheet, it turns out they cover a fairly broad range of frequencies and if you have 2 or 3 bands trying to control frequencies that are close together, you get an additive effect.   If I graph Nate's setting, for example, he is getting at least a 2dB gain from about 5000Hz and up.

I have been trying to surgically adjust the dips in response on the e2c's high end.  One approach has been based on reading the frequency response graph.  The other approach has been to use Saratoga's tone method.  Anyhow, I am just trying to reconcile what my ears hear with what my head thinks should be happening.  Understanding how to use that spreadsheet will help me with that.