I have a pair of HTM-12 v2 which have been recently measured at Erins Audio Corner at: DIYSG HTM-12v2 Speaker Review
We know from the Klippel measurements that there is very impressive directivity index (DI) and thus the speaker is prime for EQing as all one would need to look at is the direct on axis sound as the radiated sound will follow and be identical based on the phenomenal DI of this speaker.
Because of this, I was thinking of adding an active layer of EQ / DSP or potentially making the speaker fully active but I had a few ideas and questions that I wanted to get peoples opinion on.
EQ ideas / options:
Idea #1:
Use the raw data from the Klippel measurements (if they get posted) and import it into REW and used to generate a very accurate EQ for the mid/high end and then use in room response below the Schroeder frequency for individual room EQ. The downside of this approach would be that it is not based on my specific HTM-12 speaker which has a slightly different speaker box size and bracing.
Idea #2:
It does seem there are potentially some resonances and with the version1 there were driver consistency concerns and thus I am wondering if one should do gated measurements of their own HTM-12 (as I made mine with slightly different dimensions, used constrained layer dampening throughout the box itself, and used different dampening material) and then EQ the high end based on that data and then EQ the low end based on in room response. Do people think that gated measurements on axis would give high enough accuracy for the higher frequencies based on ones specific speaker that this would be the ideal way to EQ their own speaker (resting assured based on the Klippel measurements that the off axis sound and directivity are very well controlled and thus one only needs to worry about the direct sound for generating an eq)?
How low of a frequency could one get with high resolution with gated measurements and would this be enough to bridge down to the Schroeder frequency for in room measurements and EQ?
Does this seem like a reasonable approach?
Idea #3:
Once again relying on the Klippel measurements to assure us there is very impressive directivity could one use their own measurements (similar to idea 2) and make the speaker fully active and bypass the passive crossover. Unfortunately I do not know enough about making active speakers and thus I assume the crossover point should be kept the same (as the DI seems to match up remarkably well) but as for slopes at the crossover etc that would be out of my current knowledge base. Are there any good guides for making an active speaker? Worth pursuing? Would it have benefits over idea #2?
Any thoughts would be much appreciated!
We know from the Klippel measurements that there is very impressive directivity index (DI) and thus the speaker is prime for EQing as all one would need to look at is the direct on axis sound as the radiated sound will follow and be identical based on the phenomenal DI of this speaker.
Because of this, I was thinking of adding an active layer of EQ / DSP or potentially making the speaker fully active but I had a few ideas and questions that I wanted to get peoples opinion on.
EQ ideas / options:
Idea #1:
Use the raw data from the Klippel measurements (if they get posted) and import it into REW and used to generate a very accurate EQ for the mid/high end and then use in room response below the Schroeder frequency for individual room EQ. The downside of this approach would be that it is not based on my specific HTM-12 speaker which has a slightly different speaker box size and bracing.
Idea #2:
It does seem there are potentially some resonances and with the version1 there were driver consistency concerns and thus I am wondering if one should do gated measurements of their own HTM-12 (as I made mine with slightly different dimensions, used constrained layer dampening throughout the box itself, and used different dampening material) and then EQ the high end based on that data and then EQ the low end based on in room response. Do people think that gated measurements on axis would give high enough accuracy for the higher frequencies based on ones specific speaker that this would be the ideal way to EQ their own speaker (resting assured based on the Klippel measurements that the off axis sound and directivity are very well controlled and thus one only needs to worry about the direct sound for generating an eq)?
How low of a frequency could one get with high resolution with gated measurements and would this be enough to bridge down to the Schroeder frequency for in room measurements and EQ?
Does this seem like a reasonable approach?
Idea #3:
Once again relying on the Klippel measurements to assure us there is very impressive directivity could one use their own measurements (similar to idea 2) and make the speaker fully active and bypass the passive crossover. Unfortunately I do not know enough about making active speakers and thus I assume the crossover point should be kept the same (as the DI seems to match up remarkably well) but as for slopes at the crossover etc that would be out of my current knowledge base. Are there any good guides for making an active speaker? Worth pursuing? Would it have benefits over idea #2?
Any thoughts would be much appreciated!
Wait for raw data from Erin.
Dip at 150Hz is most probably standing wave not being supressed enough (not enough damping material inside the cabinet or not on the right places), and i expect/presume that dip at 500Hz is a product of port resonance. Only detailed mesurements can tell and Erin wants to measure it with ports plugged to find out where the issue lies.
I'd always use passive crossover and some aditional eq-ing for a 2 way loudspeaker with crossover frequency as high as that. Since you are dealing with high sensitivity transducers, electronics behind it really needs to be low noise to avoid hiss and that can get expensive really fast.
Dip at 150Hz is most probably standing wave not being supressed enough (not enough damping material inside the cabinet or not on the right places), and i expect/presume that dip at 500Hz is a product of port resonance. Only detailed mesurements can tell and Erin wants to measure it with ports plugged to find out where the issue lies.
I'd always use passive crossover and some aditional eq-ing for a 2 way loudspeaker with crossover frequency as high as that. Since you are dealing with high sensitivity transducers, electronics behind it really needs to be low noise to avoid hiss and that can get expensive really fast.
If the current crossover is doing its job properly, then EQ on top will be a good way to go. It is important that EQ isn't allowed to interfere with the cross itself, and the good thing is that you will be doing that globally so it's OK.
How you EQ can be complicated so keep an open mind.
How you EQ can be complicated so keep an open mind.