Is it necessary to AUTOMATICALLY change the convolution filters in accordance with the changes in the sample rate of the reproduced material?

[ZolaIII]

The range must be so narrow (30-300 only?)? What about the rest (30-17k)? Make it separetdly?

A shoulder point. IIR PEQ Q filter if going to what I see as standard 50 will remain to be pin point precise to uper mids F range (1 KHz). FIR tap based convolution will be more accurate in highs (and less in lows) as it's linear tap based. It only makes sense to do it if highs and mids overtones are already fairly good focused (and averaged) already. For REW PEQ range don't go over +2 individual and +3 over all for deeps. For impuls reversion calculate it down to max +2 also (based on video I am lazy same as everyone else). For the lowe of the rodents pleas address room fundamental harmonic first, put that PEQ in active chain and mesure with it in place to even get started with correction.What's exact listening device again (with proper measurement preferably) and zipp your own one's and attach them here.

 

[SashaR]

I do not correct above 1kHz

It makes no sense to ask you where you learned to read instructions so carefully - and it doesn't even make sense to ask you how you decided to try not to let REW amplify the signal: smart people know how to be smart and know how to have smart results. BUT TELL ME, PLEASE - why does the bass sound so fantastically correct and beautiful after applying the 0-booster???

I tried the PEQ with 20-300 booster-free filters, and I did the same with 300-18k.. I'm lucky - I have (almost) German Speakers too (I even hope they're not much worse. than yours, although nothing is written about mine anywhere, unlike your speakers) -- BUT I understand. that I will never get what you hear at home - and a simple ("simple") PEQ made under your guidance proves it.

And one more "theoretical" question (I've been tinkering with booster-free options for almost a day (and trying to compare them by ear, without mic), if you can answer in your spare minute: do you not adjust above 1 kHz because you not only have good speakers, but also a well-treated room? (I do not know how to compare my VELA 407 with your speakers, but probably mine are not much worse than yours, but I have such horrors - I have a hole in the MLP, and next to the desktop I have waves in the upper part of the range).


I almost forgot to tell you again: probably the EAPO does NOT affect all devices, but only those for which you are asking to install the EAPO driver. If this might interest you and lighten the load on the equipment, then I will check it and send you pictures.

 

[ZolaIII]

@SashaR really? Give me raw response. What mic do you use and how you do SPL calibration? I need readings aligned in mids to 76 dB taken properly and accurate. Please get and use cal file if it's UMIK. - 15 has a bit of preringing and -25 rings as hell. At -15 you over done room harmonic and on another one you missed it underneath (34.8 Hz). What you did in between that deep at 119~120 Hz disappears is question for you.
Edit: -15 is more to speakers behavior and easy to work with (PEQ).
Found only partial and not very good quality measurements for those overgrown EVO's.

ELAC Vela FS 407 Loudspeaker Lab Report

Lab Report The FS 407 is compact as floorstanders go, so even with a 4ohm nominal impedance its specified 88dB sensitivity looks optimistic. Our measurements recorded almost 87dB as a simple FFT mean but 86.1dB SPL for 2.83V at 1m on the industry standard pink noise input, and less on a typical...

www.hifinews.com

 

[boxerfan88]

It makes no sense to ask you where you learned to read instructions so carefully - and it doesn't even make sense to ask you how you decided to try not to let REW amplify the signal: smart people know how to be smart and know how to have smart results. BUT TELL ME, PLEASE - why does the bass sound so fantastically correct and beautiful after applying the 0-booster???

I learnt from experimenting. I tried before -- having lots of boost&cuts to achieve even frequency response in-room, it does not sound good to my ears. After much experimentation, my ears like it when I only do cuts in the 30-300Hz region to achieve a reasonably tight/even frequency response band (not straight line) inside the room.

Just to share for your reference, these are the FR measurement at the listening position of my system. Notice the measurement is not flat at all, but pretty flat within a band. Sounds very good to my ears.

I tried the PEQ with 20-300 booster-free filters, and I did the same with 300-18k.. I'm lucky - I have (almost) German Speakers too (I even hope they're not much worse. than yours, although nothing is written about mine anywhere, unlike your speakers) -- BUT I understand. that I will never get what you hear at home - and a simple ("simple") PEQ made under your guidance proves it.

Glad you like it. Listen to it for a few more days to confirm if you like it. Sometimes we just like "something different" ... it takes several days to adjust and confirm that one really like the change or not.

And one more "theoretical" question (I've been tinkering with booster-free options for almost a day (and trying to compare them by ear, without mic), if you can answer in your spare minute: do you not adjust above 1 kHz because you not only have good speakers, but also a well-treated room? (I do not know how to compare my VELA 407 with your speakers, but probably mine are not much worse than yours, but I have such horrors - I have a hole in the MLP, and next to the desktop I have waves in the upper part of the range).

My room is not well treated. Just a couple of bass traps at the front sides. No other "treatment".
I use open shelves, cupboard, iKEA chair, carpet, thick window curtain, general room furnishings as my "room treatment".
It works quite well.

I almost forgot to tell you again: probably the EAPO does NOT affect all devices, but only those for which you are asking to install the EAPO driver. If this might interest you and lighten the load on the equipment, then I will check it and send you pictures.

Thank you for the information.
I do have EQ APO installed on the same PC, but not used.

foobar2000 is my main HiFi playback app, and fully customized to my requirements, driving either Topping DM7 or Soekris R2R DAC.

Plus, nothing beats a 27" streamer GUI screen fully driven by foobar2000 to enjoy with HiFi music

ps: the attached mdat file is an average of multiple measurements, not too useful. you may want to post the original measurement mdat. and also post the PEQ you are using. then we can review and give you suggestions.

pps: if you post a measurement mdat (with no EQ, L & R taken separately), if time permits, i might be able to review and post a suggested room correction PEQ set for you to try.


.

 

[SashaR]

What mic do you use

These measurements were taken at different points in the room - the one with the 112Hz dip ("-15dB") was taken in MLP, and the second one, with the waves in the high frequency range ("-25dB"), was taken very close to the speakers, also between them.

I just got the second UMIK-1, after replacing the first microphone. I already managed to check it - but found out that the problem was NOT in the microphone, but in the USB ports of my laptop. In these two months, I got back the NUC i3, which I am going to use for Tidal streaming - and with this computer, the UMIK-1 works correctly. BUT the pattern of measurements that I will try to do at night, in a quiet environment, should not change (room mods).

 

[SashaR]

I need readings

Thank you VERY MUCH!!!

 

[SashaR]

if you post a measurement (with no EQ, L & R taken separately), if time permits, i might be able to review and post a suggested room correction PEQ set for you to try.

Thank you VERY MUCH! just today i got the replacement for the UMIK-1 - and realized that the problem was not the mic but my PC USBs. I'm afraid that the slow i3 2Mhz 16GB NUC comp with only 20 GB free disk space will not suitable to be the RAW host - and, in the addition, it is NOT more accurate to use the same ports for measuring and to playing, right? BUT if I don't find a way to deal with this USB problem, then I'll have to rely on the endurance of the little NUC...

I took this screenshot a month and a half ago. Today, a few hours ago, I realized that the microphone wasn't to blame. But what can be done to solve this problem???

Edit: I've just looked inside the microphone configuration files... Is THIS ok??? When I will perform the MMM with strictly and perfectly aligned to 90 degree mic?!

"Auto-generated 90-degree calibration file"

 

[boxerfan88]

Thank you VERY MUCH! just today i got the replacement for the UMIK-1 - and realized that the problem was not the mic but my PC USBs. I'm afraid that the slow i3 2Mhz 16GB NUC comp with only 20 GB free disk space will not suitable to be the RAW host - and, in the addition, it is NOT more accurate to use the same ports for measuring and to playing, right? BUT if I don't find a way to deal with this USB problem, then I'll have to rely on the endurance of the little NUC...

I don't quite understand what you're trying to convey.

View attachment 447447

I took this screenshot a month and a half ago. Today, a few hours ago, I realized that the microphone wasn't to blame. But what can be done to solve this problem???

Sorry, I'm not familiar with that chart. I don't remember REW having such a chart. What software is that chart from?

 

[SashaR]

I don't quite understand what you're trying to convey.

Excuse me! My measurements were taken in three days, during which I had UMIK-1. The SPL graph, made using the SPL meter of REW, demonstrates the instability of the microphone: the SPL level is constantly falling. I returned it. I got another UMIK-1 today. After checking it on another computer, I realized that my first UMIK-1 was working properly. The malfunction is caused by the USB of the computer used.

I have already used the new microphone and placed it in the MLP chair (slightly below my ear level). It became clear that I did not do the PEQ for my room.

I have already used the new microphone and placed it in the MLP chair (slightly below my ear level). It became clear that I did not do the PEQ for my room.
I was not able to make all the measurements I wanted to make (I need measurements for two more listening positions. But I will have to do them in a noisy environment: I did not pay due attention to the fact that the measurements from L+R are much louder than from one speaker - and, unfortunately, without the participation of neighbors, this was not obvious to me).

Here are the measurements I just made. I sent the reference signal from the recorded speaker.

Thank you very much!!! And please forgive me for the vague and unclear wording!

 

[SashaR]

PS. For the same listening point (MLP), I managed to make another series of measurements - BUT WITHOUT the computer chair standing between the speakers (this chair is my second listening point (even the first point in terms of time of use)). I managed to see that the dips in low frequencies changed when measuring without the work chair. I thought that maybe this difference can help to understand how to get rid of (or reduce) these dips in this complicate room - and therefore decided to send these measurements too. Thank you very much for your help!!!

 

[Keith_W]

I have already used the new microphone and placed it in the MLP chair (slightly below my ear level). It became clear that I did not do the PEQ for my room.

Please read this thread and make sure you are taking measurements properly. In particular: no mini tripod, and do not place your microphone on furniture when taking measurements. Use a boom tripod, and no furniture between mic and speakers.

 

[SashaR]

Please read this thread and make sure you are taking measurements properly. In particular: no mini tripod, and do not place your microphone on furniture when taking measurements. Use a boom tripod, and no furniture between mic and speakers.

Thank you! I read both your Post and the article. But since the ordered tripod has not arrived yet, and the replacement microphone has already arrived, I decided not to do anything approximate this time and placed the microphone where my head is. After all, if it is acceptable to perform MMM with a microphone IN YOUR HANDS, then pressing it to the wooden back of a chair should not produce more interference than my breathing, etc. artifacts when performing MMM, no?

In addition, how can I solve a new problem that arose at night, in a quiet environment: I cannot take measurements at two more points at night. Can I take measurements at more than 75 SPL to increase the SNR? Second question: is it possible to take a measurement (NOT MMM!) with a microphone at 90 degrees? One of my listening positions is almost between the speakers, and perhaps it would be wise to position the mic vertically rather than horizontally at that point (although either way it is almost 90 degrees from the speakers (less than 40 cm from the line connecting the speakers, and the distance between them is 190 cm)?

 

[SashaR]

no mini tripod, and do not place your microphone on furniture when taking measurements. Use a boom tripod

P.S. I made myself a BOOM without a tripod, as you advised.

Because of this tripod, the microphone during the measurements was not at the same height as during the first series of measurements, which I published a short time ago.
Is it possible to see the difference created by using a stick (using BOOM)? In the ZIP archives there are two measurements for L+R, R and L.
Thank you!!!

PPS. Unlike the previous one, this series of measurements was performed in a NON-quiet environment (I no longer have the opportunity to take measurements at night). Can you advise me on what rules and techniques should be used in such situation in order to have the most useful measurements possible? Thank you again!

 

[boxerfan88]

The measurement shows some quite wierd stuff going on. Maybe other folks can chime in. Allow me to highlight my observations.

• SPL compare
  • A - why is L+R dip far worse than L | R independently?
  • B - even weirder - why sudden 10kHz HF spike when the L | R independently is pretty smooth?
  • I really have no clue what is going on.

• Phase compare
  • A
    • why is L-channel (blue) phase rotating so much faster than R-channel (green)
    • why at the major dip point (114Hz), the L+R (red trace) show a sudden phase discontinuity?
  • B
    • L & R phase track each other pretty well, and stays along the 180 degree path from 3kHz-20kHz
    • but L+R (red trace) phase - goes haywire - starts rotating from 4kHz-20kHz (total approx 550 degrees)

I really have no clue what is going on ...

Quite difficult to start work on room correction without understanding what is going on... maybe someone else can shed more light...

.

 

[boxerfan88]

Crazy distortion <40Hz.

Pre-ringing way back in time -- 100ms before time-0 ?? Definitely weird.
Both speakers step response don't seem to be well aligned in time.

.

 

[boxerfan88]

Oof ... that 40Hz-30Hz ringing ... like wanting to go on forever ... still ringing like crazy near 2 seconds!

Is there a giant motor in the surrounding area?

 

[Keith_W]

@SashaR yes you can use your mic attached to a broom handle the way you pictured. In fact that's what I did when my tripod broke and I had to wait for a new one. I duct taped a broom handle to my camera tripod, then taped the mic to the broom handle. But the whole reason you use a tripod like that is so that you can move furniture away and let the mic sit in "free air". Because if you don't do that, this is what happens:

The measurement shows some quite wierd stuff going on. Maybe other folks can chime in. Allow me to highlight my observations.

View attachment 447616

• SPL compare
  • A - why is L+R dip far worse than L | R independently?
  • B - even weirder - why sudden 10kHz HF spike when the L | R independently is pretty smooth?
  • I really have no clue what is going on.

The reason for getting mics away from furniture is because reflections contaminate your measurement. 10kHz has a wavelength of 3.4cm. If the distance between your mic and chair is about 3.4cm, then all measurements below 10kHz are contaminated by reflections. These artefacts were created from improper measurement technique.

All those summation artefacts (i.e. strange things appearing on L+R which are not seen on individual L and R measurements) are due to phase issues - summation or cancellation when both speakers are played together, but not necessarily when they are played individually.

@boxerfan88, re those distortion measurements. Please compare the distortion measurement to the SPL of the speaker. You will see that his speakers do not reproduce those very low frequencies where "distortion" is very high. Ergo, it's not distortion - it's noise. Maybe something vibrating or rattling in the room. FYI, I have a lot of trouble taking subwoofer distortion measurements because my whole house rattles when the subs are played too loudly. I live in a glorified tent! So I can only obtain SW distortion measurements at very low SPL, much lower than the 75dB reference.

 

[SashaR]

Thank you all, @boxerfan88 @Keith_W @digitalfrost @ZolaIII !

I was able to make good PEQs for two of my three listening points: the point between the speakers and the point outside the room - but I have a hard time making a practical PEQ for MLP because of the room-induced frequency response flatness.

Please teach me the basic procedure: how to import a text file of filters into REW WITHOUT TYPING it manually in the Filters window? I am trying to make a PEQ with a deviation of 1 dB from the target by SMOOTHLY gluing together the PEQs made by REW for narrow diapasons - and to do this I MANUALLY rewrite the last significant filters of the adjacent PEQ into the REW filter of the diapason being created (before asking REW to create filters for this diapason). But this rewriting takes a huge amount of time and effort! How to import a FEW filters into REW automatically - preferably from a text file?

Am I breaking religious norms when I listen to music with two individual PEQs - for L and for R, and not for L+R? Or is further work with the room only possible with L+R?
How to create a correct audio file that will allow REW to check the quality of PEQs performed by EAPO, @staticV3 ? And how and what to check if I use TWO simultaneously executed PEQs - for L and for R?

I hope that at the end of next week I will get a tripod and will be able to set the microphone at ear height - but I do not understand why it is necessary to falsify the measurement of the room by taking furniture out of it??? Real and ACCURATE measurements MUST measure what will influence during real listening - i.e. not only cabinets and sofas, but also the chair together with the listener! Isn't it? And it is ALL THESE influences that must be taken into account and neutralized - they ALL, and not those that DO NOT correspond to the real situation - isn't it? I am not designing a concert hall, and I do not have to work with an empty room that will be filled with an average statistical number of people... In my situation there is no statistical audience - just one determined listener with a precisely known percentage of the concert hall seats filled (100%) - and even the hairstyles of the ladies are known to me reliably - and all this SHOULD have been taken into account when processing the room, right???

The photo shows the unbearable conditions in which the left loudspeaker is located and still tries to work. Next to it you can see a window - its width is equal to the width of the room and it is this that hums at the beginning of a long (2 MB) sweep. The main (?) problem is probably the 60 cm wide cabinet along the left wall of the room - I drew it on an incomplete drawing of the room and it is visible in the photo of the wooden boom.

In the middle of the photo you can see the LA90 amplifier, which I replaced with two B100 in the hope of improving the sound quality - BUT they are impossible to use due to deep dips in the frequency response, requiring -24 dB of pre-amplification (pre-attenuation!). Without your help I will not be able to cope with them. Thank you all for your advice and help!!!

 

[SashaR]

The reason for getting mics away from furniture is because reflections contaminate your measurement. 10kHz has a wavelength of 3.4cm.

I tried to get rid of the artifact created by the window resonance with a powerful measurement sweep.

How can I edit the measurements (erase the hump, in my case)? Is it acceptable to take measurements with a low sweep power - in the hope that the window will not notice the sweep and will not resonate?

In the set of measurements (attached) it is clear that the frequency response changes extremely strongly when the sweep length changes (from 256k to 2M - from 5 seconds to 47 seconds). In addition, the frequency response is less jagged with the window open.

What parameters should be assigned to REW to take measurements in the most accurate way (and, of course, removing the microphone by 6.8 cm from even upholstered furniture)?
@boxerfan88

 

[boxerfan88]

What parameters should be assigned to REW to take measurements in the most accurate way (and, of course, removing the microphone by 6.8 cm from even upholstered furniture)?
@boxerfan88

For the 20-20k range, I normally use 512k sweep. I also sweep L|R separately.

I avoid long slow sweep like 2M as it can trigger unnecessary resonances (in my room, slow loud sweep sometimes rattle an unlatched cupboard door)