Measuring and understanding measurements and then A/B testing EQ is deceptively difficult and has many easy to make mistakes and pitfalls.
Some basic stuff you should correct.
If measuring off a sine sweep you should be EQing to a straight line that drops 10db from 20hz-20khz. Not to ‘flat’.
When A/B’ing you will need to adjust the volume so that the average db is actually comparable. See the massive gaps above the green line all the way to the red line? That is all db that you have lost through EQ. That is normal but it is also why it sounds lifeless. You are simply listening to the EQ’d output at much quieter levels. I estimate it needs around a 5-7db increase to get the average db to match again but your EQ software (I assume REW) can tell you what that is.
There are probably many other mistakes you are making through this whole process that aren’t obvious through a simple graph either. Again all normal, all easy to make, and people who have been doing this for years like myself are still learning and making new mistakes daily.
My overall point is that if the EQ’d result sounds worse to you, you are far more likely to be making mistakes in EQing rather than genuinely liking the non-EQ’d response better. Trust your ears as a sanity check to your process, rather than as an arbiter of preference.
Measuring and understanding measurements and then A/B testing EQ is deceptively difficult and has many easy to make mistakes and pitfalls.
Some basic stuff you should correct.
If measuring off a sine sweep you should be EQing to a straight line that drops 10db from 20hz-20khz. Not to ‘flat’.
When A/B’ing you will need to adjust the volume so that the average db is actually comparable. See the massive gaps above the green line all the way to the red line? That is all db that you have lost through EQ. That is normal but it is also why it sounds lifeless. You are simply listening to the EQ’d output at much quieter levels. I estimate it needs around a 5-7db increase to get the average db to match again but your EQ software (I assume REW) can tell you what that is.
There are probably many other mistakes you are making through this whole process that aren’t obvious through a simple graph either. Again all normal, all easy to make, and people who have been doing this for years like myself are still learning and making new mistakes daily.
My overall point is that if the EQ’d result sounds worse to you, you are far more likely to be making mistakes in EQing rather than genuinely liking the non-EQ’d response better. Trust your ears as a sanity check to your process, rather than as an arbiter of preference.