Correct high frequencies "holes" in EQ

[BenToronto]

Ignore the phase graphs. I don't know what they can help with.

The REW impedance "X-ray" sweep plot shows how much the woofer cone is moving. Then you can compare that to sound in the room. It is also a rough diagnostic for the speaker revealing if there are impedance matching issues matching impedance to your amp (relevant to tube amps, planar drivers...) and helpful for BR cabs and other tricky cabs.

 

[Dox82]

Ok, where is the graph you talking about? I think I have to do a specific impedance mesurement...
In general I have no tube amplification, my amp is a Sudgen a21SE Signature and speakers are Neat Ekstra.

Ignore the phase graphs. I don't know what they can help with.

The REW impedance "X-ray" sweep plot shows how much the woofer cone is moving. Then you can compare that to sound in the room. It is also a rough diagnostic for the speaker revealing if there are impedance matching issues matching impedance to your amp (relevant to tube amps, planar drivers...) and helpful for BR cabs and other tricky cabs.

 

[ddude003]

Dox82, I am guessing @BenToronto is referring to frequency vs Impedance tracking curves and how amplifiers respond to such impedance curves... See http://www.atma-sphere.com/en/resources-paradigms-in-amplifier-design.html for some insight... For instance, my Martin Logan ESLs have an unorthodox impedance curve, dips to under 2 ohms at high frequencies, which seems to baffle most folks and amplifiers... I use a McIntosh amp which has "autoformers" that deals with the impedance issues very well... It will be interesting to see what the frequency vs Impedance looks like on your speakers... It might also be interesting to see what a Klippel spinorama looks like... BTY, using some EQ filters may cause phase changes... Again, read through Serkan Gur's method instructions... There is a section on using RePhase(freeware) to correct phase... Crossover Frequencies at 80Hz and 3.5kHz...

 

[Dox82]

Ok, there are surely a lot of interesting things to learn but I can't reasonably believe that the 80-100hz hole can be an amplifier<->speakers problem.
I mean, I have a pair of neat ekstra with 6ohm of impedance with 86db of sensitivity (they are classic speakers not ELS), driven by a sudgen a21se signature with 40W into 4ohm and 30W into 8ohm in pure class A. I don't overcome 70db of SPL because my listening point is at 185cm of distance to the speakers. I can't imagine there is a bad hardware interaction in those conditions. I'm in wrong?
80-100hz is a problematic acoustic node in the room, like the Room Sim shows. I think I need a specific acoustic treatment for that but I'm waiting for a response from my room acoustic expert about panel to adopt.
Simply, I'm looking for the different phase response in the graph but if isn't a problem I'll look for the specific acoustic treatment mentioned above.

 

[ddude003]

I did not say it _was_ the problem Dox82... It is something to consider and _could_ be an issue if your amp tries to double its power output if/when it is seeing low dips of impedances at or around 80-100Hz... And you wouldn't know if you did not have a frequency vs impedance map of your speakers and understand how your amp responds to that map... Remember that the "system" includes hardware, software and the _room_... With virtually a cube for a room you are most likely fighting room modes... Only bass absorption targeted at the offending frequencies will help... However, understand that boosting via EQ or thru speaker/amp impedance mapping won't help things... The specs you quote above don't accurately represent what happens to the impedance across the entire frequency range... 6ohms is RMS...

PS - It would be fun to hear what your acoustic expert has to say...

 

[Dox82]

I'll surely update the post when I'll get a response from the expert. They have my last mesurement...crossfingers...but I already know they are worried about the "boom" in the bass range (which is "easy" recoverable with room correction via sw), conversely I'm more worried about the "world famous" 80-100Hz hole hahahaha.
In the meantime I have totally recovered the 135-150Hz hole, basically moving my vinyl forniture in the left corner behind the listening position. This is act as trap in the corner and the 135-150 hole disappears, but the 80-100Hz hole does not move from there in any way. I'm fighting against room modes as you said...

PS - It would be fun to hear what your acoustic expert has to say...

 

[John Mulcahy]

Can someone clarify this graph difference with some hypotesis and tell me if it can be a real problem or not?

There is nothing odd about those phase responses.

You have a very strong reflection 3.35 ms after the direct sound in both the left and right measurements, from the desk, perhaps?

The room is well behaved about 250 Hz or so, but there are very severe resonances at 52 Hz and 58 Hz. Your best option to tackle those is to try different positions for the speakers and the listener. A quick way to experiment with that is to play the Pink Periodic Noise signal and watch the result on the RTA (with the settings shown in that help section) while you move the mic or the speaker to different positions.

You can use single 256k sweeps for measurements, there's no need to use longer sweeps or multiple sweeps for this kind of work.

 

[Dox82]

@John Mulcahy the peaks in the bass range are well know and they are not removable with speakers and listening position changes (basically they depends on the ceiling) . Luckily the peaks can be well managed by the digital room correction. Anyway, I'm working to try to minimize this problem also in the room with additioanl panels. The main problem, for me, is the hole in 80-100Hz and maybe this one is a "side-effect" of the peaks around there.
Normally, for my mesurements I'm using 512k sweep with 2 iterations.

 

[ddude003]

Soffit Bass Traps with range limiter... You could test by stacking 3 or 4 panels (need thickness) that you may currently have? BTY we still don't have a good idea of what's in your room......

 

[Dox82]

Yes, I'm waiting the "acoustic-expert final word" but a previous analysis showed a problem in the ceiling about the peaks cited by @John Mulcahy. Probably I'll have to use a tuned trap also for the 80-100Hz hole but I hope a final acoustic adjustment project will come in a couple of days...

Soffit Bass Traps with range limiter... You could test by stacking 3 or 4 panels that you may currently have? BTY we still don't have a good idea of what's in your room......

 

[Dox82]

Ok, basically the acoustic expert said what was assumed: I'm fighting agaist room modes until 80Hz and this is the main cause of peaks and residue hole. He say the modes are emphasized by the door (with some glass inside) and maybe by the speakers itself because the isobaric system at floor level. These last considerations, however, are of minor importance (because modes there are and with a lot of energy) even if some test with different speakers (pheraps of stand type) would not be a bad idea.
As first step we need a new bass trap pair to stack up on the current ones. After that w'll evaluate something on the ceiling but I'm not sure to do the "ceiling's job" because some "architectural" difficulties.
Anyway, the current situation with the digital correction by REW+Roon is a good point which I did not expect to get.
Thanks for your suggestions!

 

[ddude003]

Way back near the beginning of this thread @Dox82 had said... "I have no sub. I have a pair of Neat Ekstra in an HiFi setup." It appears that you have 2 stand mount 2 way speakers sitting on top of 2 subwoofers... The ribbon tweeter is crossed over at 3.5kHz to the midrange and crossed over at 80Hz to the subwoofers... An interesting design... I wouldn't give up on them unless you just don't like the sound... Speaker, along with subwoofer, placement will definitely be very important... Good luck with teasing out that last little 80Hz null... I still think it would be worth the time it would take to try out Serkan Gur's method...

 

[Dox82]

@ddude003 yes the ekstra is a "weird" design defined as "2.5 Ways" with an isobaric section in the bottom of the cab. I was superficial in my initial description, this is not a "real" sub but effectively this cab is capable to go very low. It seems that the manufacturer does not state a clear frequency response, but we are around 35Hz as my mesurement shows.
I like the timbre of this cab and I'm fighting (as you see) to save it. The upper range is clear and smooth at the same time, but in my current room looks oversized in the low range. In general they are a really good pair of speakers, it stay around 4500$ (not so cheap). Anyway I don't want to sell those speakers before trying others. Will be a nice idea to try a pair of classic 2 ways speakers just to have an idea in terms of room-interaction differences recpect to the esktra. Inverse math is in my TODO list and I have to say thank you to share with me this concept.
By the way, in a room (now "partially" acoustically treated) with the dimensions I mentioned, which pair of cabs would you have reccommended? Assuming my listening is focused on contemporary genres like prog, rock, contemporary jazz, indie/alternative rock, funk, blues...

 

[ddude003]

@Dox82, I don't think your speakers are a "weird" design... Unorthodox maybe and so are many others... I am partial to electrostatics and horns... If you haven't heard Martin Logan electrostatic hybrids I would suggest you try to find a shop that knows what they are doing and have a listen... I don't think horns, like JBL or Klipsch Vintage, would do well in your "cube" however... Maybe make that "cube" into an anechoic chamber and stick some Corner Horns in there... LOL... I also wonder how an open baffle speaker would do in your "cube"... And I don't have enough information about your integrated amplifier to suggest if ESLs would have the right kind of synergy, cough, impedance handling... I know that a ~40 watt tube amp can drive them... Horses for courses...