What is causing my HF roll off?

[ThELiZ]

I only have one cal file from Dayton. It doesn't state whether it is a 0 or 90 cal file. That purple trace looks suspiciously like the issue I'm having!

 

[kejar31]

I only have one cal file from Dayton. It doesn't state whether it is a 0 or 90 cal file. That purple trace looks suspiciously like the issue I'm having!

I am going to guess that your calibration file is for 0 degree. point the mic directly at the speaker you are measuring to get proper results.

 

[ThELiZ]

Thanks. I'll try that later and report back

 

[kejar31]

Also if that works for you.. I went ahead and did a delta (simple trace arithmetic in REW of A/B) between my 0 and 90 degree files, you can probably use this and multiply it with your 0 degree cal file to make a 90.. or you can just shot me your 0 calibration file and I can do it for ya.

edit: better image spelling

 

[ThELiZ]

Thanks so much. If you wouldn't mind, I've attached my cal file

 

[kejar31]

Here you go...

edit: proper file in image lol

 

[kejar31]

btw if you have measurements that you have already take with your 0 degree cal file when while the mic was in 90.. you should be able to just change the calibration file on measurement tab

Note: this will not work with MMM using RTA..

edit: added not about MMM

 

[ThELiZ]

Cheers bud! I'll have to wait til tomorrow to do any more testing. Kids are in bed and I think everyone is fed up of hearing measurement sweeps. My wife is a lucky lady!

 

[kejar31]

Cheers bud! I'll have to wait til tomorrow to do any more testing. Kids are in bed and I think everyone is fed up of hearing measurement sweeps. My wife is a lucky lady!

You should be able to take the first measurement you posted and simply change the calibration file.. no need for a new sweep

 

[Kachda]

The Lintons do have a comparatively faster rolloffs in HF.

 

[kejar31]

But tomorrow you shoud validate everything.. AKA make a measurement using the 0 degree cal while pointing the mic dir at your speaker.. then take another one using the 90deg file while the mic is pointed at the ceiling and make sure the overall measurement looks similar..

Note: you will see variations in the measurements (peaks etc) and the only way to avoid that is to do a MMM measurements using the cal files. Regardless the overall curve from 1k up should look the same.

 

[VMAT4]

You are leaving the grills on, aren't you? The Lintons are made to have grills on when listening.

 

[kejar31]

You are leaving the grills on, aren't you? The Lintons are made to have grills on when listening.

no he is def using a 0 degree calibration file while taking a 90 degree measurement..

 

[ThELiZ]

Soooo... It seems like it's a room issue. I've tried the mic at 0 and 90... No difference in the results. Thanks to all for helping me out here.

I don't know what I might have done before when measuring at 1m, but now I'm measuring at 1m and the roll off looks as to be expected. Sadly I can't show the graphs as my PC is playing up.

Now the next question is... Should I be EQ'ing that HF roll off? I've used House Curve app to measure and upload FIR file to Roon. That file corrects to 500Hz, and from there I've made changes to the curve via PEQ. I'm now at the point where the measurement data looks good and the subjective result is good. When I turn off the PEQ, the high end sounds a bit more dull, as you'd expect. I'm only unsure about leaving it that was as I've read you shouldn't EQ the highs.

Appreciate any input on this.

 

[staticV3]

Now the next question is... Should I be EQ'ing that HF roll off? I've used House Curve app to measure and upload FIR file to Roon. That file corrects to 500Hz, and from there I've made changes to the curve via PEQ. I'm now at the point where the measurement data looks good and the subjective result is good. When I turn off the PEQ, the high end sounds a bit more dull, as you'd expect. I'm only unsure about leaving it that was as I've read you shouldn't EQ the highs.

Do what sounds best to you.

 

[ThELiZ]

Yeah, I that makes the most sense! Just when you hear you shouldn't mess about with the high end and let the speaker do its things...

I'm happy with the sound.

 

[levimax]

it possible for both tweeters to show the same issue?

Funny thing is yesterday I noticed something very similar with some DIY speakers using Morel CAT 308 tweeters. Both started rolling off over 8 KHz. I did the trouble shoot routine and surprizingly it was the tweeters (it would really help if you could try another pair of speakers to confirm). Reading more it sounds like as tweeters age the ferro fluid can dry out which starts effecting HF response first. I was under the impression that tweeters either worked or not but that is not the case. I wonder how many older speakers suffer from this issue? The speakers I used to confirm it was the tweeters were 25 year old Sonus Farbers and those tweeters had no roll off until over 16 KHz.

 

[kejar31]

I dont think the room is causing that much rollofff. You can EQ higher freq's too, just not based on room measurments.

I personally do EQ my speakers freq from ~250 and up but I do a gated measurement primary one diretly on axis pointed right at the tweeter... I then apply EQ for that gated measurment. See this thread on how to do that

https://www.audiosciencereview.com/...ents-spinoramas-with-rew-and-vituixcad.21860/

Please note: This is not an easy process as you will need to move your speakers to the middle of the room and make sure its lifted at least 4 feet from the floor and still 4 feet from the ceiling.. I personally put my speakers on thier side then on a stand for this, you can also always go outside if you have the means to get a signal to your speakers there.. Also if you are not well versed in REW you might run into trouble as well as you will have to combine both that higher filter from the gated measurment, and the lower filter based off room modes..

edit: added context to the note

 

[kejar31]

just a quick question though.. When you are doing these measurments you are sure you are using exclusive mode on both the mic and preamp right?

 

[kejar31]

Here is an example of my speaker before and after using a Gated measurement

Here is an image of the impulse response and IR window, as you can see, I was able to get a decent measurement resolution down to about 260hz using this method..

edit: updated the image to show the DB range of the before and after