REW Next Steps

[c1n3ma]

Even though (to me) my speakers already sound incredible, it seems like every small improvement I make makes a noticeable step up in quality. What should I be looking to do next with my setup? I have a decent DAC, Nikko-TRM600 Amp, and B&W Nautilus 804 speakers. I have 2 sound absorbers but theyre sitting behind the speakers due to space constraints - 10ftx12ft bedroom. Any suggestions on what to do next would be awesome. This is after EQ.

These are all L+R, when I try to just eq one at a time I lose the brilliant center stage. 1/12 smoothing

 

[sam_adams]

Welcome to ASR.

Now that the pleasantries are out of the way . .

First off is your presentation. You have the SPL graph set to a 200 dB range. Even the worst response looks good when presented like that. Adjust the vertical scaling to 5 dB above the highest unsmoothed peak and the bottom of the scale 50 dB down. In this case, the top of the scale should be 85 dB or 90 dB and the bottom 35 dB. The frequency range does not have to extend past the high and low points of the stimulus. Measure 20 Hz to 20 KHz, present 20 Hz to 20 KHz. Additionally, there is no need to include the mic calibration file response in the graph.

Waterfall graphs that extend to 20 Khz are of little value. You should use them to analyze the modal region of the response from 10 Hz or 20 Hz on the left to 300 Hz on the right. What is of interest here is ringing caused by room modes. The vertical scaling should match that of the SPL graph—again a 50 dB range.

When presenting the impulse response, the best presentation is the 'Envelope' plot as it makes it easier to see the reflection peaks from room boundaries. Scaling of the plot only needs to show 10 ms before the impulse peak on the left and—at most—100 ms on the right. When displaying 1.1 seconds, it is almost impossible to discern any early reflections in the plot. Vertical scaling does not need to extend above +5 db or below -65 dB.

*Edit for clarity.

 

[c1n3ma]

Gotcha. I really appreciate you taking the time to write that up for me. I’m pretty new to all this so every bit helps. I’m out of the house for the day but I’ll fix do it first thing when I’m back.

 

[sam_adams]

Gotcha. I really appreciate you taking the time to write that up for me. I’m pretty new to all this so every bit helps. I’m out of the house for the day but I’ll fix do it first thing when I’m back.

Remember to reply or quote—even a single sentence—so that an alert is generated for your response so we know you're responding.

 

[c1n3ma]

S

Remember to reply or quote—even a single sentence—so that an alert is generated for your response so we know you're responding.

Here are the updated (and more readable) graphs. Is it good practice to include the bottom bar of the graph as well?

SPL

Impulse (Envelope ETC -42dBFS)

Waterfall

Also for reference here is my room layout. Its far from perfect, but its what I got. Got out the measuring tape and official dimensions of my room not including the hallway is 12ftx13.5ft. My speakers are about 1.5ft off the back wall and 1ft off the side wall for the one in the corner. I have unobstructed view of the full speaker in my sitting position. The sound panels almost reach the ceiling and are behind the speaker.

 

[HarmonicTHD]

S

Here are the updated (and more readable) graphs. Is it good practice to include the bottom bar of the graph as well?

SPL
View attachment 243483
Impulse (Envelope ETC -42dBFS)
View attachment 243488
Waterfall
View attachment 243484
Also for reference here is my room layout. Its far from perfect, but its what I got. Got out the measuring tape and official dimensions of my room not including the hallway is 12ftx13.5ft. My speakers are about 1.5ft off the back wall and 1ft off the side wall for the one in the corner. I have unobstructed view of the full speaker in my sitting position. The sound panels almost reach the ceiling and are behind the speaker.
View attachment 243490

Yes include the bottom bar so it becomes evident, which smoothing you have applied. Common is eg VAR or 1/12th.

 

[sam_adams]

Here are the updated (and more readable) graphs. Is it good practice to include the bottom bar of the graph as well?

Yes it is. I've edited my original posting to clarify the Impulse response scaling. Should have been, "to show 10 ms before the impulse peak on the left and—at most—100 ms on the right." Include the legend at the bottom to show smoothing and what plot is being displayed.

Is the SPL plot of both speakers at once? If so, measure Left and Right separately to eliminate combing effects in the response plot and to make it easier to see the room effects on the individual speaker caused by placement.

 

[c1n3ma]

Yes it is. I've edited my original posting to clarify the Impulse response scaling. Should have been, "to show 10 ms before the impulse peak on the left and—at most—100 ms on the right." Include the legend at the bottom to show smoothing and what plot is being displayed.

Is the SPL plot of both speakers at once? If so, measure Left and Right separately to eliminate combing effects in the response plot and to make it easier to see the room effects on the individual speaker caused by placement.

Here are the SPL plots of Left and Right, with no EQ applied. I tried EQing them based on this, and prior tries, but end up with the central soundstage off-center.
Left

Right

And the Impulse graph with the correct boundries

 

[kemmler3D]

Looks like you have some pretty typical bass problems in the 100-200hz range, if you can find a way to squeeze some membrane / limp mass absorbers in the room, you should do that.

There also seems to be some resonance just below 300hz that would be worth sorting out with some absorbers...

Lastly and least expensively, you can at least try and EQ out the peak at 89hz. The big peak around 150hz might be worth a try also.

 

[sam_adams]

Waterfall

The area below 35 Hz shows some ringing, but because the signal level is dropping there this could be low frequency background noise. This area is very hard to mitigate with room treatments. A way to fill that in would be with a subwoofer(s)—if you are inclined to do so.

Left

The left speaker is very close to a boundary so the regular pattern of peaks and dips could be SBIR. The response indicated by the red line could be a power response known as the Alison effect.

Right

The right speaker is showing some response similar to the left that may be SBIR and Alison effect.

And the Impulse graph

The reflection peaks from ≈ 5 ms to 35 ms are low enough as to not have discernible effects on imaging. See below, from Olive and Toole via Everest.:

Some of these issues could be mitigated with a change in placement, although judging from the space they're in that might not be an option for you. The other option would be treatments to reduce the SBIR effects from the boundaries. Wall and ceiling treatments might involve the combination of diffusion and absorption. This could be very challenging—both in time and money—to resolve some of these issues.