Trying to fix possible bass null in my room - help please

[DL_Can]

Hi all.
I am a 2-channel listener. I've built up a fairly decent system over the past couple of years.
My room is approx. 24' long x 22' wide, with 9' ceilings that slope from, back and one side down to about 6' high.

I have the following gear:
Magico A3 tower speakers with Gaia feet
Electrocompaniet ECI6 Mk2 integrated
Gustard R26 DAC
Volumio Rivo streamer with external LPS
Blue jeans cables
2 x REL T/7X subwoofers

I am fairly limited to where I can have my listening chair due to the position of the TV and the couches in the room.

I have done the subwoofer crawl and spent a fair amount of time adjusting the gain/crossover and phase on the subwoofers.....to the best of my abilities. While I am not unhappy with the sound, I do feel like I am in a bit of a null and can improve the bass. I do plan on upgrading the subwoofers at some stage, but I'd rather do a little more work first.

I used REW a few months back and the results didn't make sense to me. I decided to give it another shot today. By looking at the graphs, it appears that my speakers and subs hardly put out any bass. However, when I play anything with good low end (Ratchets or Money for Nothing etc.), the bass does kick pretty hard.

I've attached the .mdat files as well as the room layout, and hoping someone can let me know what I am missing, any idea what's going on with the readings. All comments appreciated!

Thanks.

 

[Wayne A. Pflughaupt]

The room diagram certainly helps. A big part of the problem is that your listening position is in the dead-center of the room. The center of any square or shoe-box room is essentially a “bass hole” that exhibits the weakest perceived bass. Move from that position towards any boundary, and you will notice the bass level will increase.

Regarding your graph, if you say you get good bass with some songs, I can’t help but think you have something set wrong in REW.

Regards,
Wayne

 

[JStewart]

Regarding your graph, if you say you get good bass with some songs, I can’t help but think you have something set wrong in REW.

Or with Windows if using it. Windows can automatically apply noise cancelling to the mic input effectively rolling off the bass.

Post in thread 'I have (almost) no bass!'
https://www.avnirvana.com/threads/i-have-almost-no-bass.9988/post-76114

 

[DL_Can]

Thanks for the replies.

@wayne, I have moved my seat to different spots. Yes, the bass does increase when I move closer to the back wall, but that is about 14' to 18' from the speakers, so not ideal. I can't say the sound is better at that spot though, just more low bass. If I sit that far back, I will then need to move my speakers very far apart to stick to the Magico placement guidelines.

@JStewart, I did make the changes in Windows (after your recommendation), but this did not change the readings at all.

I have to believe that I have settings in REW that are incorrect?

 

[DL_Can]

Well, a little more playing around with the noise cancelling setting and.....voila! Thank you @JStewart!
Much better readings now. Looks like I may need to adjust the settings/position for the left sub?

 

[trobbins]

Do you have a cheapo sound level meter that can be set to C-weighting, and whilst playing a steady tone then move the meter away from your listening head position and note the change with distance in a few directions. That could provide some insight as to room modes and what to expect from a single point listening position (before your ears start helping).

 

[DL_Can]

Do you have a cheapo sound level meter that can be set to C-weighting, and whilst playing a steady tone then move the meter away from your listening head position and note the change with distance in a few directions. That could provide some insight as to room modes and what to expect from a single point listening position (before your ears start helping).

@trobbins , I don't but will buy one.

 

[JStewart]

Much better readings now. Looks like I may need to adjust the settings/position for the left sub?

Worth a try but impossible to calculate outcomes with the data at hand. To model with REW you would need a separate measurement of each speaker and subwoofer with a timing reference. With those, resulting measurements from changes to delay and/or relative spl can be calculated. Then to effect the changes one would need to be able to independently set delays. Without this ability I think you're left with changing the phase on a sub and re-measuring to see what change it made. Maybe there's a better way. I just don't know what it might be.

While not an exact predictor without a rectangular room and precise inputs, REWs room simulator appears to at least directionally agree with your measurements and I had to eyeball the speaker and subwoofer locations. I also assumed the subs are separate left and right channels as opposed to dual mono.

Left Speaker/Sub

And overlayed

Right Speaker/Sub

And Overlayed

I'd guess the setup as is sounds very good. There's very good SPL uniformity between the left and right throughout the frequency range, so imaging should be very good. The area of divergence in the lowest bass probably doesn't matter much. The right has higher SPL and the left lower, but together they are probably just about right and it's debatable if there's any ability for us humans to perceive bass directionality in that range and there's extremely limited recordings with directional bass anyway.

The RT60 decay time looks quite uniform accross the frequency band too and the time appears in the range I've seen suggested by industry experts for you size room.

I'm just another hobbyist, not an expert by any means, but my less than expert opinion is you can try to do better but any real improvements will probably be marginal. Maybe add a gentle touch of EQ to the 75hz peak on left channel if you have the ability, but whether it would sound better, or sound better with some but not other material, would be for you to determine. Oh, and I did see a little of what looked like pre-ringing on the impulse response. No big deal, but curios if you're using any EQ or DSP now?

 

[ddude003]

You might try this tool for a non-rectangular room... https://amcoustics.com/tools/amroc-pro

 

[DL_Can]

Worth a try but impossible to calculate outcomes with the data at hand. To model with REW you would need a separate measurement of each speaker and subwoofer with a timing reference. With those, resulting measurements from changes to delay and/or relative spl can be calculated. Then to effect the changes one would need to be able to independently set delays. Without this ability I think you're left with changing the phase on a sub and re-measuring to see what change it made. Maybe there's a better way. I just don't know what it might be.

While not an exact predictor without a rectangular room and precise inputs, REWs room simulator appears to at least directionally agree with your measurements and I had to eyeball the speaker and subwoofer locations. I also assumed the subs are separate left and right channels as opposed to dual mono.

Left Speaker/Sub
View attachment 87321

And overlayed
View attachment 87323

Right Speaker/Sub
View attachment 87322
And Overlayed
View attachment 87324

I'd guess the setup as is sounds very good. There's very good SPL uniformity between the left and right throughout the frequency range, so imaging should be very good. The area of divergence in the lowest bass probably doesn't matter much. The right has higher SPL and the left lower, but together they are probably just about right and it's debatable if there's any ability for us humans to perceive bass directionality in that range and there's extremely limited recordings with directional bass anyway.
View attachment 87325
The RT60 decay time looks quite uniform accross the frequency band too and the time appears in the range I've seen suggested by industry experts for you size room.
View attachment 87326

I'm just another hobbyist, not an expert by any means, but my less than expert opinion is you can try to do better but any real improvements will probably be marginal. Maybe add a gentle touch of EQ to the 75hz peak on left channel if you have the ability, but whether it would sound better, or sound better with some but not other material, would be for you to determine. Oh, and I did see a little of what looked like pre-ringing on the impulse response. No big deal, but curios if you're using any EQ or DSP now?

Wow, thank you so much for this! It is so good to hear from someone with tons of REW experience and I appreciate your help and comments.
I think you have given some positive feedback and I feel so much better about my current setup and speaker/sub placement.

Yes, I think it sounds great, but always trying to get the best out of the system, without having the upgrade bug. My only future change will be better subwoofers (probably the REL S510 or S820) as I believe these speakers deserve a better match.

I have the left sub fed via the left channel and the right sub via the right channel. Both connected via high level.

I don't have any DSP or EQ, but I sometimes make small EQ tweaks via Roon, depending on the recording and whether or not I'm listening at low volume. I do have a bunch of DIY absorption panels: three 2x4 (3" deep) and two 2x2 (3" deep). I also have triangular 20" DIY bass traps in each corner. One corner of the room is rounded so no trap needed. I have a DIY diffuser panel and plan to make a few more in the near future.

I ran separate measurements for each sub and each main speaker (attached). I want to make sure that I have the phase set correctly for each sub.

I have a lot of reading to do and videos to watch so that I gain a batter understanding of REW and what it can do.

 

[Keith_W]

To model with REW you would need a separate measurement of each speaker and subwoofer with a timing reference.

@DL_Can I took a look at your MDAT. You missed this important detail. Throw away those measurements and start again, and please be sure to carefully read what is suggested.

 

[DL_Can]

@DL_Can I took a look at your MDAT. You missed this important detail. Throw away those measurements and start again, and please be sure to carefully read what is suggested.

I have no clue what timing reference means as I am new to REW. I will do some in-depth reading up on REW before I post here again. Thanks.

 

[Keith_W]

When you press the "measurement" button, look for this in the dialog box.

 

[DL_Can]

View attachment 87430

When you press the "measurement" button, look for this in the dialog box.

Thanks Keith. I will give this another go in the morning.

 

[JStewart]

@Keith_W , I advised the need for timing reference before knowing the OP’s set up. The subs are connected to the left/right channels with high level inputs. That’s going to make acoustic timing reference more challenging. Also there doesn’t seem to be DSP in play that can effect independent delays for the 4, so REW can see what’s possible when supplied a timing reference, but any changes can’t be effected, assuming there’s any for the better.

Since the OP is also trying to integrate left sub to left main and same for right, I was thinking if there was a desire to obtain the knowledge, taking the left sub and main with the right as timing reference and vice versa for the right would work.

Might you see any other possibilities with this setup?

Would also like say it’s great to see you here on the forum! And nice work with your REW guide. Gotta good laugh too from the smoke alarm activation suggestion.

 

[Keith_W]

The subs are connected to the left/right channels with high level inputs.

[/QUOTE]

D'OH!!!! Sorry @DL_Can I withdraw what I said!!! Looks like it was me who missed that little detail! Please accept my apologies. I'll take another look at your measurements.

Since the OP is also trying to integrate left sub to left main and same for right, I was thinking if there was a desire to obtain the knowledge, taking the left sub and main with the right as timing reference and vice versa for the right would work.

Might you see any other possibilities with this setup?

Well, I am sure you know as well as I do that if you want to properly integrate a sub, you need independent control of sub and mains. Meaning each speaker needs to be on its own DAC channel.

Would also like say it’s great to see you here on the forum! And nice work with your REW guide. Gotta good laugh too from the smoke alarm activation suggestion.

Thank you! I am working on the third eBook at the moment, "Phase". I am really struggling to make it accessible for the layman. I already worry that the book is too long. That's why I split it up into sections and I direct beginners to skip some sections.

And it was a bit naughty of me to tell people to smoke a cigarette to activate the smoke alarm. I mean, i'm a doctor

 

[Keith_W]

Anyway, first things first. Let's look at the controls available to you:

All you have available is (1) volume (2) XO point (3) phase 0deg or 180deg, (4) find a better position for your subwoofer. It is fortunate that this particular REL subwoofer has a low level input unlike some utterly despicable REL subs of the past which were designed by imbeciles, because this opens the door to DSP if you decide to do so. I'll only focus on answering the first 3 questions, and i'll show you how to do this yourself.

I imported opened both sets of measurements into REW and renamed them for clarity.

I can see that you set your XO at about 60Hz and you are running your mains full range. Looking at where your mains roll off, I agree with this decision.

The first step is to see what the bass looks like when all four speakers are playing together. In the "All SPL" tab, select all four measurements, then right click and choose "Vector Sum". We immediately see a problem - there is a huge null between 40Hz - 70Hz.

Let's try inverting the polarity of one sub and see what happens. In the "SPL & Phase" tab, I selected the left sub, right click, and chose "Invert". Then back to "All SPL" and vector sum again. This is the result (in green). It looks worse!

... so back to the drawing board. I went an inverted the right sub. So now both L and R subs are inverted. Vector sum again, and the result is in purple. IMO it looks better. The dip isn't as broad, and in fact it almost disappears when you choose psychoacoustic smoothing. Meaning you likely won't hear it.

The last question is whether subwoofer gain has been set properly. Using the instructions on how to overlay a target curve in my eBook, I generated a target curve. Please read that eBook on how to interpret the frequency response so that you can see with your own eyes what is going on. We can see:

- there is a touch too much bass. Recommend reduce subwoofer level by about 2-3dB.
- there is huge treble roll-off above 5kHz. The eBook gives you a list of causes, it can range from measurement error, wrong cal file, using the mic with the wrong orientation, or failing to toe-in your speakers. Magico A3's are superb speakers, it is highly unlikely that they roll off like that.

And just for fun, let's see whether you would benefit from DSP. The answer is "yes". Look at that, the subwoofer starts about 28ms before the main impulse. Once again, look at that eBook for tips on how to interpret the step response.

So overall, the recommendations are:

- invert the polarity of your L, R subs.
- reduce the SPL of both subs by 2-3dB
- check why you have treble roll-off by reading the guide
- consider DSP for even better integration.

And BTW, my goal here is the same as it is on ASR. I'm not here to spoonfeed you or tell you what to do. I'm here to show you how to read those measurements. What you do is your decision, not mine.

 

[DL_Can]

Hi Keith. Thank you so much for all of your work on this! BTW, I have downloaded the eBooks and will start reading through them tonight!

I inverted the polarity of both subs.
I reduced the output for each sub by one click, hopefully that will reduce the db to what you recommended.

When I initially set up my speakers, I didn't toe them in much as I was a little worried about reducing the soundstage. I had another look at the Magico manual and adjusted the toe in to be more in line with what they recommend. I do notice an immediate change....better center imaging and the good thing is the wide soundstage didn't change. Thanks for this recommendation as well.

I tried to take measurements again this afternoon. In the past I was getting a crackling when trying to take measurements in REW. This happened most of the time. For some reason, it didn't happen with the measurements last week. However, the crackling is back. It happens when I play through the computer speakers and also when it output from the computer headphone jack into my amp and through my main speakers. Any idea why this is happening?

 

[Keith_W]

You're welcome. Re: crackling. You need to make sure it's crackling you are talking about, and not stuttering.

Stuttering = digital. It's due to lost packets / buffer starvation, etc. A lower sampling rate usually helps.

Crackling = analog. The problem is somewhere downstream of the DAC. Faulty cables, poor connection, faulty equipment, etc. can all cause crackling. Those computer headphone jacks are sometimes connected directly to the motherboard, especially on laptops. I hate using them, because any stress on the cable is transmitted directly to the motherboard. If that solder joint starts to fail and lose contact, it may crackle. Since the headphone jack seems to be the common denominator here, try using an outboard DAC.

 

[sm52]

Look at that, the subwoofer starts about 28ms before the main impulse.

How did you determine this? There are no published measurements of speakers with the same time synchronization in this thread.

 

[Keith_W]

How did you determine this? There are no published measurements of speakers with the same time synchronization in this thread.

The subwoofer and main speaker come off the same DAC channel and are therefore, synchronized.

 

[sm52]

This is a misconception.