DJHastings
Registered
Thread Starter
- Joined
- Oct 29, 2024
- Posts
- 1
I've been attempting to linearize the phase response of my headphone drivers while also eq'ing them using the following procedure:
1. Measure the driver
2. Create a pure impulse response using the "generate measurement from filters" button in the EQ panel
3. Trace arithmetic ->(eq target)*(pure impulse) = eq and phase target
as you can see, the problem is that the resultant EQ and phase target has a phase that goes crazy above 20 khz.
4. Trace arithmetic -> (EQ + phase target)/driver = final impulse response
you can see that same phase craziness carried over to the final result above, along with a vertical line at 20k that drops straight down to -60db? figured this could have been just a trivial artifact, but when I plugged the impulse response into my convolver, the gain was so low I could barely hear it with my amp turned all way up. When I examined the measurement files, I did see that the data rates were different between the generated pure impulse and the actual measurements. Could this be part of the problem and/or is there a way to align data rates between traces?
Any help on troubleshooting this issue is appreciated, thank you!
1. Measure the driver
2. Create a pure impulse response using the "generate measurement from filters" button in the EQ panel
3. Trace arithmetic ->(eq target)*(pure impulse) = eq and phase target
as you can see, the problem is that the resultant EQ and phase target has a phase that goes crazy above 20 khz.
4. Trace arithmetic -> (EQ + phase target)/driver = final impulse response
you can see that same phase craziness carried over to the final result above, along with a vertical line at 20k that drops straight down to -60db? figured this could have been just a trivial artifact, but when I plugged the impulse response into my convolver, the gain was so low I could barely hear it with my amp turned all way up. When I examined the measurement files, I did see that the data rates were different between the generated pure impulse and the actual measurements. Could this be part of the problem and/or is there a way to align data rates between traces?
Any help on troubleshooting this issue is appreciated, thank you!
Last edited:
