Dedicated listening room setup. Personal blog like thread

[ObjectiveSubjectivist]

Just came today.
I need some time to save some pennies to buy speakers (currently using q acoustics 3030i), but for now I'll do some measurements where we at....

Ps. Im not sure when I'll be able to afford d&d 8c, so I'm aiming for something like KEF R7 meta

 

[ObjectiveSubjectivist]

Just done first measurement of my room. I'm not sure if it looks ok but I'll investigate more in my free time.
I wonder if reverb that I have in my room could affect fr graph? I'm still during setting everything in new place, so room is still pretty empty and very reverberant. RT60 shows big values - as my understanding goes, but as I said it's going to change in the next month (s?)
Actually it sounds pretty bad (for some music it's okay for other a messy sound, overall too 'lively' for my taste) - but that's for now. Can anyone check if graph looks legit?
(1/3 smoothing, 90 degrees calibration file applied, umik placed in sitting position where my head is, maybe a bit lower, facing celling )

 

[Keith_W]

At first glance you have a huge room mode at 60Hz. This is good news because it's a peak, it can easily be DSP'ed out. But what I would really like to know is whether this was a good quality measurement. I can't tell without looking at the other graphs. Would you be able to post the .mdat please?

Your graph label says "L + R sty 9" - is this both speakers playing together?

Whether reverb can affect your FR: absolutely yes, but it depends on how long your window is left open. I'll explain a bit more when I look at your .mdat.

 

[ObjectiveSubjectivist]

At first glance you have a huge room mode at 60Hz. This is good news because it's a peak, it can easily be DSP'ed out. But what I would really like to know is whether this was a good quality measurement. I can't tell without looking at the other graphs. Would you be able to post the .mdat please?

Your graph label says "L + R sty 9" - is this both speakers playing together?

Whether reverb can affect your FR: absolutely yes, but it depends on how long your window is left open. I'll explain a bit more when I look at your .mdat.

Yes I measured both speakers at the same time. And of course I can attach *.mdat file (added it as zip)
Yes the peak around 60hz seems to be huuge.
Also sound is kinda bright (brightness feeling could be because of big reverb?) and sucked somewhere in lower midrange/higher bass, at least that's how I hear it.

 

[audiofooled]

Looking at the file, couple of comments:

That mode at 60Hz is quite nasty. Have a look here https://amcoustics.com/tools/amroc
If you enter your room dimensions you may notice that exactly at that frequency, right in the middle of the room you will have this peak and it will ring like a bell. I would advise adjusting speaker positioning and main listening position by taking into an account that if your speakers are close to the front wall, the more the mode would be excited. Maybe try a smaller listening triangle (bring the speakers closer together, away from corners) and sit closer.

Another thing to notice is that the impulse response looks bad. Try pointing the mic towards the speakers and measure them separately and then together. Finding the spot where the impulse response is at it's best would most likely be the best listening position, with a higher D/R ratio. Keep in mind that at the ceiling you will pick up yet another mode close to the 60 Hz so this seating position and mic pointing will result in messing up the measurement.

Feel free to experiment, maybe even try a close field desk setup until you have dealt with room acoustics. I'm sure someone more knowledgeable would have some advice.

 

[DavidEdwinAston]

My room is set up just like your second proposal, right down to a PC on a desk in the rear.View attachment 283163

Can you just put a brick through that tv screen, please!

 

[Keith_W]

Thanks for posting your MDAT. Firstly, it is standard to take separate measurements of the left and right speaker, then both together. This limits what we are able to see. Having said that, I see quite a few problems.

First, why does it sound harsh and bright? Answer: because you have a lot of early and loud reflections.

This is your energy-time curve. According to Haas, all reflections arriving within the first 20ms are integrated with the main sound, which will both smear it in time and change its tone. To avoid this, the target is to have all reflections at least -15dB to the main sound, preferably -20dB. I have marked the main impulse with the green arrow and the reflections in red. You can use the timing of the reflections to calculate where the reflection is coming from. For example, 2 is 2.5ms delayed to the main impulse. According to d = t/1000 * c (c = speed of sound, 343m/s), the reflection had to travel an extra 0.86m to the microphone.

1 is delayed by 0.66ms to the main sound. Because I have seen the step response, I know that it is the result of group delay from the woofer (a heck of a lot of group delay!).

Your RT60 shows the same problem. I took the room volume from your first post (I assumed ceiling height of 2.4m) and calculated the RT60 target in Acourate. I then drew in the graph over your measurement with MS paint. You can see that the reverb in your room is way too high.

Your frequency response is a bit off. Again, I took your measurement then drew a Harman target on top of it with MS Paint. Ideally, everything above the blue line should be trimmed to match the target, and everything below should be boosted to match the target. The blue line really helps you see where the problems are. That 60Hz bass peak is 10dB too loud. Worse, it is preceded by a dip, and followed by a dip, which will make it stick out even more than it should.

I don't know what speakers you are using, but there is severe lack of bass output below 50Hz. The shape of the FR curve makes me think lack of speaker output is the problem, rather than a room mode. The dip that you see at about 160Hz is definitely a room mode. It can be DSP'ed out if you are super anal about it, but you could just as well leave it alone.

And finally, here is the step response. We should not be looking at it since it is a measurement of two speakers together rather than just one, but let us assume for a moment that it is a single speaker. You can see that the tweeter and the woofer is not time aligned. As for the woofer, there are two peaks, which suggests a ported speaker or very early woofer reflection.

Recommendations:

1. Repeat measurement of left and right speaker individually. Ensure the measurement is taken properly - mic on a mic stand please. Do not use the mini-tripod that came with your UMIK-1.
2. The reverb is way too high. You need more furniture or maybe even room treatment to bring the RT60 down to target. DSP will not help you here.
3. Consider DSP to smoothen out the bass.

 

[ObjectiveSubjectivist]

Thanks for posting your MDAT. Firstly, it is standard to take separate measurements of the left and right speaker, then both together. This limits what we are able to see. Having said that, I see quite a few problems.

First, why does it sound harsh and bright? Answer: because you have a lot of early and loud reflections.

View attachment 420038

This is your energy-time curve. According to Haas, all reflections arriving within the first 20ms are integrated with the main sound, which will both smear it in time and change its tone. To avoid this, the target is to have all reflections at least -15dB to the main sound, preferably -20dB. I have marked the main impulse with the green arrow and the reflections in red. You can use the timing of the reflections to calculate where the reflection is coming from. For example, 2 is 2.5ms delayed to the main impulse. According to d = t/1000 * c (c = speed of sound, 343m/s), the reflection had to travel an extra 0.86m to the microphone.

1 is delayed by 0.66ms to the main sound. Because I have seen the step response, I know that it is the result of group delay from the woofer (a heck of a lot of group delay!).

View attachment 420042

Your RT60 shows the same problem. I took the room volume from your first post (I assumed ceiling height of 2.4m) and calculated the RT60 target in Acourate. I then drew in the graph over your measurement with MS paint. You can see that the reverb in your room is way too high.

View attachment 420043

Your frequency response is a bit off. Again, I took your measurement then drew a Harman target on top of it with MS Paint. Ideally, everything above the blue line should be trimmed to match the target, and everything below should be boosted to match the target. The blue line really helps you see where the problems are. That 60Hz bass peak is 10dB too loud. Worse, it is preceded by a dip, and followed by a dip, which will make it stick out even more than it should.

I don't know what speakers you are using, but there is severe lack of bass output below 50Hz. The shape of the FR curve makes me think lack of speaker output is the problem, rather than a room mode. The dip that you see at about 160Hz is definitely a room mode. It can be DSP'ed out if you are super anal about it, but you could just as well leave it alone.

View attachment 420044

And finally, here is the step response. We should not be looking at it since it is a measurement of two speakers together rather than just one, but let us assume for a moment that it is a single speaker. You can see that the tweeter and the woofer is not time aligned. As for the woofer, there are two peaks, which suggests a ported speaker or very early woofer reflection.

Recommendations:

1. Repeat measurement of left and right speaker individually. Ensure the measurement is taken properly - mic on a mic stand please. Do not use the mini-tripod that came with your UMIK-1.
2. The reverb is way too high. You need more furniture or maybe even room treatment to bring the RT60 down to target. DSP will not help you here.
3. Consider DSP to smoothen out the bass.

Thanks for the answer. I really really appreciate that. I need some time to absorb new information and maybe ask some further questions.
For now I have something to read and think about for the weekend

 

[ObjectiveSubjectivist]

ps. The speakers currently are Q Acoustics 3030i (used with q acoustics dedicated stands). Also room height is ~275-277cm

 

[audiofooled]

I don't know what speakers you are using, but there is severe lack of bass output below 50Hz.

Q Acoustics 3030i Quasi-anechoic Spinorama and other measurements

Hello! I recently published my review of the new Q Acoustics 3030i. You can look that up if you want more text, but I'm here for measurements! The 3030i were recently announced, and retail for $400. So we have the quasi-anechoic spin, with nearfield bass spliced to the listening window at 385...

audiosciencereview.com

Good points. From the measurements above, directivity wise it seems that the optimal horizontal angle would be 15 degrees off axis. Also careful positioning because of the ports, room modes etc, along with smaller listening triangle. Stay on the tweeter axis vertically and treatment of the floor/ceiling as suggested in the thread above. More soft furnishing to improve RT60...

 

[ObjectiveSubjectivist]

More soft furnishing to improve RT60...

Yes, more things to come, as I said we basically just moved and whole house is still in quite pending state Whole floor is already covered with ~2cm carpet

For some music (andreas vollenweider for example) big "echo" sounds quite nice, but overall... yuck.
My previous small apartment where I used the same speakers sounded so much better - but at that time I had no desire to do the measurements, a bit pity I would have some comparison/starting point, oh well.

 

[audiofooled]

Just take your time, experiment, take a lot of measurements and you're going to get there slowly.

I've built my system myself which sounded great at the workshop and then I was given an x amount of space to make it work in the living room, so yeah, it's a process. But also, a hobby, after all.

 

[Keith_W]

Q Acoustics 3030i Quasi-anechoic Spinorama and other measurements

Hello! I recently published my review of the new Q Acoustics 3030i. You can look that up if you want more text, but I'm here for measurements! The 3030i were recently announced, and retail for $400. So we have the quasi-anechoic spin, with nearfield bass spliced to the listening window at 385...

audiosciencereview.com

Good points. From the measurements above, directivity wise it seems that the optimal horizontal angle would be 15 degrees off axis. Also careful positioning because of the ports, room modes etc, along with smaller listening triangle. Stay on the tweeter axis vertically and treatment of the floor/ceiling as suggested in the thread above. More soft furnishing to improve RT60...

Thanks for posting that. This graph here provides another explanation why @ObjectiveSubjectivist finds the sound too bright:

You can see that the speakers are injected a lot of sound power into the room between 2kHz - 7kHz. There are a lot of things the on-axis measurement does not tell you. This is one of them. There is not much that can be done about this, it is a flaw baked into the speaker design. You can TRY to introduce a -2 to -3dB shelf between 2kHz - 7kHz with DSP which might improve the tonality.

I have also recalculated the RT60 target for the room dimensions of 3.5 x 5.6 x 2.7m according to the DIN 18041 standard.

 

[audiofooled]

You can TRY to introduce a -2 to -3dB shelf between 2kHz - 7kHz with DSP which might improve the tonality.

Yes, the speaker loses directivity in this region but I don't think that tonality would be improved this way. IMHO the perceived brightness would be in the region where it starts beaming higher up in frequency. I suspect this because when I applied 15 cycles FDW and VAR smoothing to the measurement and compared it to one of the measurements of my system (which I perceive to be tonally neutral), this is what I get:

So, at 12kHz or so, there's only 4 dB difference, with 6,5k being 25 dB down. Add to that the fact that at 60Hz there's practically a one note bass that swamps everything but the highest frequencies and to me this is clearly the case that these speakers are not fit for this listening distance. They lose too much SPL in the direct sound at this distance (apart from the beaming region), and center of the room is bad for bass.

If I try to align the SPL, some things may be even more obvious:

I should note that in no way my system should be used as any benchmark, it's one of a kind and certainly not the best, so take it with a grain of salt. I just say that it sounds neutral to me and it was just convenient to use as a comparison.


But, generally speaking, if we take into account early reflections, the larger the listening distance, the more SPL you lose at the frequencies that are high but considerably wider than those that are beaming, if the room acoustics are bad. The region that is too wide will also interfere with reflections of the lively room in a pattern that won't do any service to perceived clarity and definition. Maybe this is why @ObjectiveSubjectivist experienced that in a smaller room with better acoustics, the same speakers sounded much better. Maybe the listening distance was smaller and wide directivity even added to the perception of width, which many people prefer. But this would only happen with more neutral FR, because after all SPL is what we hear.

What I would suggest in a current situation is a desktop setup where the listening triangle would be small enough, with 15 degrees off axis as far as aiming (maybe also extreme toe in, so 15 degrees inwards). Another thing is that, positioning closer to the front wall may even reduce SBIR and yield a better bass. Worth a try.

I'm no expert, but this is what I would try, prior to any DSP. So, measurements, correct positioning, spacing, aiming and then DSP, acoustic treatment, etc.