My REW measurement

[zermak]

That's for sure. I was just trying to show him, shomehow, that a "standard" crossover can't work on a DIY design and his measurements nearfield and farfield (in another manner) confirm that. Plus to me the tweeter crossover point is way too high.

And to add on that we don't know the drivers nor the crossover network.

 

[ernestcarl]

View attachment 71065


I was happier before buying the Umik

I would not boost the bass dips too much. REW indicates 12.4 dB of headroom required around 51Hz! I think that’s not advisable... Post EQ does it even sound good with such an extreme boost applied?

Use test tracks with a lot of energy in that area and test tones with REW’s generator.

 

[oldsysop]

I would not boost the bass dips too much. REW indicates 12.4 dB of headroom required around 51Hz! I think that’s not advisable... Post EQ does it even sound good with such an extreme boost applied?

Use test tracks with a lot of energy in that area and test tones with REW’s generator.

Is the EQ okay or too aggressive?

 

[ernestcarl]

View attachment 71796

Is the EQ okay or too aggressive?

I still find your overall individual max boost and overall boost too high. I usually try to set mine to 9 and 0 respectively, if I can get away with it. Maybe 3 or 6dB for the overall, if absolutely necessary — and IF actual listening tests show no adverse effect. Some speakers (and rooms) at certain frequencies are okay with boosts here and there, while others, not so much... you might hear awful resonances or distortion... so do a sweep at, say, 96dB SPL after your correction and look at the distortion graph as well.

Flatness of 1dB for the entire range of 20Hz to 1kHz is also rather asking for a bit much, IMHO... I do use 1dB sometimes for a smaller, targeted ‘specified range’ that may be amenable to it.

In the end, you have to listen for yourself how adjustments/changes sound to you with test tracks and/or tones.

 

[zermak]

And as with every thing, you sometime can not fix something bad from start. My advice again is to revisit the crossover network.
And what are the drivers and the crossover network? And the baffle dimentions and the drivers positions?

 

[oldsysop]

so do a sweep

Umik pointing up in listening position, 3 meters from the speakers.
"1MMeasSweep_0_to_20000_48k_Float_LR_refL"
Playing from Roon, with and without EQ.

Without EQ :

With EQ :

 

[oldsysop]

"Pink_PN_64k_Full_48.0k_Float_LR"

Without EQ :

With EQ :

 

[zermak]

The sweeps are really confusing (use a 50dB scale please, that is a 340dB scale!). There is also too much noise floor and so you should raise the volume (75dB C weight at least) to have better readings.

 

[ernestcarl]

Bass looks a bit more linear but you also get more %THD distortion. How much so, is really, really hard visualize because of your scaling... But the snapshots of the RTA window does list the peaks... The added distortion with your EQ is overall high -- although, some reduction in some of the harmonics can be seen as you did cut down the peaks.

 

[CDMC]

I was happier before buying the Umik

I think there are stages when owning a UMIK, like the stages of grief:

1) The utter horror as you realize, yes my room has 20db variances in the bass range.
2) The wondering how you ever could listen to such an inaccurate system without realizing it.
3) Wondering what some of those really bad systems you have heard would measure like.
4) Really learning about equalization and room effects.
5) Getting past years of conditioning that using an equalizer is is somehow wrong.
6) Learning how to use the auto eq feature in REW.
7) Realizing that 2-4 PEQs will address 90% of the cleanup in the bass range.
8) Realizing that +/- 5db in the bass range is fine and can sound incredible.
9) Enjoying speakers that sound great on almost everything. No more, bass is too high, bass is too low.

 

[oldsysop]

The sweeps are really confusing (use a 50dB scale please, that is a 340dB scale!). There is also too much noise floor and so you should raise the volume (75dB C weight at least) to have better readings.

I do not understand

 

[Erici]

I do not understand

I think he was referring to your graph that went from -130 to +170 SPL. After that you added another graph with a better scaling.

 

[zermak]

I do not understand

I think he was referring to your graph that went from -130 to +170 SPL. After that you added another graph with a better scaling.

Yes and about the SPL. You need more power to at least reach 75dB with C weghting at the listening position.

And the other graph is pink noise and not a sweep and we can't see the distortions.

 

[dasdoing]

View attachment 71796

Is the EQ okay or too aggressive?

imo it is anything but agressive lol check the E1 note with this https://www.szynalski.com/tone-generator/
it should be right in your deepest null. and F1 is close to your highest mode.
PERSONALY I would set the target at 45ish in this one. discussion around this method: https://www.audiosciencereview.com/.../need-advice-about-rew-and-house-curve.12315/
and I would correct the midrange, too; while leaving the HF rolloff.
BTW: you can attach the MDAT of the meassurement in a post for people suggesting corrections

 

[oldsysop]

BTW: you can attach the MDAT of the meassurement in a post for people suggesting corrections

 

[QMuse]

Here, try the attached filter. It requires 6dB of attenuation so be sure to set it to avoid clipping. Here's the predicted FR, it looks pretty good but on the account of serious headroom loss, which unfortunately cannot be avoided with the response like yours.

 

[dasdoing]

this is what I would do. majority in this forum prefers a bass boost though

 

[ernestcarl]

this is what I would do. majority in this forum prefers a bass boost thoughView attachment 71960

At what distance is your listening position? The response should normally slope down in-room, anyhows, even without adding a bass boost/hump.

 

[oldsysop]

Thank you very much @QMuse.
Could you upload the .txt or .req to upload them to Room?
I don't know how to view the .wav in REW.

 

[dasdoing]

At what distance is your listening position? The response should normally slope down in-room, anyhows, even without adding a bass boost/hump.

this is OP's meassurement. about bass boost in rooms, I consider it destructive as it is the room mods producing those. in a bass treated room you wont have the boost. the Harman bass boost will mask the mids. HF (10kish) roloff on the other hand is "natural": https://www.audiosciencereview.com/...-favorite-house-curve.2382/page-9#post-444578