Home Theater Forum and Systems banner

Need help understanding process of testing combined response of individual filters

2K views 5 replies 2 participants last post by  AudiocRaver 
#1 ·
I understand that I need to take measurements of each channel of my stereo system separately. I also have a decent idea of how to apply EQ to each channel using a house curve. I assume that once I create the EQ filters for each channel I will want to measure the combined effect with both speakers playing and make further adjustments to each channel's filters as required. If so, what I dont understand is how to do this. In other words, how do I measure the response of both speakers playing at once with their individual filters applied?

Thank you in advance.
 
#2 ·
That assumption (measure both channels at once) is not correct. The measurement is difficult and will yield nothing useful that you can't get with separate sweep measurements, and in fact is likely to give erroneous information from phase cancellations which could lead to nonsensical corrections and a worse sound.

If you insist, you will have to do it from recorded pink noise playing over two channels via HDMI (my assumption) and the two channels of pink noise must be offset in time so they appear random relative to one another. It is quite a pain.
 
#3 ·
Thank you for the reply Wayne. As you can tell I am a bit of a noob when it comes to this stuff. I will admit that it doesnt make any sense to me to not take measurements with both speakers on because ultimately that is what I will me hearing. How can I be sure that the filters I apply to each individual channel actually produces the response I want?

As a related point of discussion, it seems there is no way to verify that the EQ filter actually gives the response predicted by REW? For this to happen, there will need to be a way to do a measurement with the filters applied. Am I right to assume that REW cannot do this?
 
#4 ·
Thank you for the reply Wayne. As you can tell I am a bit of a noob when it comes to this stuff. I will admit that it doesnt make any sense to me to not take measurements with both speakers on because ultimately that is what I will me hearing. How can I be sure that the filters I apply to each individual channel actually produces the response I want?
Your question is valid. Conventional wisdom is to apply the same filters to the L and R through the range where the ears can detect phase differences (roughly 100 Hz to 3 kHz), assuming
  • the speakers match well
  • speaker positioning and room and treatment are symmetrical
The ears can detect phase differences in the soundstage & imaging (SS&I) through that range. Below 80 Hz, multiple subwoofers should also be EQed together as one.

Bottom line, best approach to set up and treat with maximum symmetry
  • so L & R FR measurements match well
  • so SS&I sounds as sharp and clear and specific as possible - listen closely - the measurements can look pretty good (until you dig deep) and still give lousy SS&I
then EQ them the same to keep them matching.

Fixing differences in phase & frequency response can SOMETIMES be done (an advanced topic). Better to not mess them up in the first place.

As a related point of discussion, it seems there is no way to verify that the EQ filter actually gives the response predicted by REW? For this to happen, there will need to be a way to do a measurement with the filters applied. Am I right to assume that REW cannot do this?
Correct, REW omly calculates the values and shows what FR should look like with those filters applied. But REW will not not actually apply the filtering.
 
#5 ·
Thank you again! I did learn from another forum that it is actually possible to apply the filters and test the output of both channels. You have to use JRiver Media Center to apply the filters and set the output driver in REW to JRiver. I havent actually tried this yet but it looks like a fairly straightforward way to apply individual filters and test the combined output.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top