First REW Plots

[mrsnell1]

First,
Great site, great software, really good stuff.

After trying four different sub positions, I settled on the "best" location for sound and WAF reasons. One sub only and it is on the front wall between the TV and the right main; not in the corner but in about the 1/3 position along the wall. Room is about 20' wide by 15' deep w/ 10' ceiling, square, with a large opening on the back right corner to the living room.

Three plots follow, and I hope they attach as I am a rookie; 1st is the mains only, 2nd is the sub only, 3rd is the mains + sub.

Mains only:


Sub only:


Mains + sub:


How do they look? To me, the mains + sub looks Ok, not too bad, but maybe it can be improved. I have a BFD, but it is not in use unless I need it.

Question, I have the volume knob on the sub set about halfway. If I turn it down, say to the 1/3 position, will it lower the line in the 15hz to 80hz region of the main + sub plot? I know I can test this myself and probably will tomorrow, but I wasn't sure if that was part of the standard EQ calibration process.

Thanks, and again, great site.

[mrsnell1]

Ladies and Gents,

Here is the waterfall for the previous plot of mains + sub for two different time extension. The first is at 300ms, the second is at 500ms. The 300ms waterfall doesn't look so good in the lower frequency region it seems like. The 500ms is better of course. The floors in the room are tile, no area rugs, standard furniture, couch etc, and no room treatments. So I know that doesn't help any. Are these waterfall Ok, or is improvement needed? How would you improve these, room treatment, or some type of EQ?

Mains + sub, waterfall at 300ms


Mains + sub, waterfall at 500ms


Thanks

[brucek]

For subwoofer plots, always use the standard Vertical graph axis of (45dB - 105dB) and the Horizontal graph axis of (15Hz - 200Hz) using the Graph Limits button in the top right corner of REW. Waterfalls should have the same axis. Don't use any smoothing on subwoofer plots.

Your decay for lower frequencies is quite slow. Generally you hope to have your signal evenly drop to the noise floor by ~300ms.

brucek

[mrsnell1]

Oops, sorry about the scale on the previous plots, these are rescaled, please forgive my goof.

Sub only


Sub + Mains


Waterfall of Sub + Mains, to 300ms


Ok, here are my questions,
1. Does the Sub + Mains spectrum look OK, or do you think some EQ is warranted.
2. The waterfall for Sub + Mains doesn't fall off that well in the lower frequencies, but is this within acceptable norms? Or should I do something? Is this where room treatments come into play? If so, what type of treatments help best with the decay?

Thanks for your patience. I am new to all of this, an am trying to learn as best I can

Regards
Chris

[brucek]

The response graphs look quite good, although it's quite difficult to get a feel for the actual response when you have smoothing turned on. Subwoofer plots should not have smoothing applied.

It appears you have measured at quite a high target level (rather than the recommended 75dB), so again it make the waterfalls a bit more difficult to comment on, but they do appear to decay quite slowly at many frequencies.

It would be good to see a response with a 75dB target with no smoothing applied and the associated waterfall.

brucek

[mrsnell1]

Bruce,
Again, thanks for your patience, I am a little slow, but anxious to learn.
I did calibrate to 75dB, but, I think maybe my sub volume is turned up too high relative to the mains? Here is a figure of what I mean.



Should I turn the sub vol down when looking at mains + sub, or should I focus only on the sub?

Regards
Chris

[brucek]

Should I turn the sub vol down when looking at mains + sub, or should I focus only on the sub?

The level balance between the sub and mains can be adjusted (as you say) with the subs own volume knob.

Generally, before you start with REW you dial the subs volume knob to about halfway and zero all your AVR speaker trims and then use the AVR's test tones to balance all the speakers and sub.

Then when you use REW to take a measure of the sub and mains together, if you want the sub higher or lower, you can adjust the subs own volume know. Once you have the sub and mains balanced, then take a measure at 75dB and generate a waterfall......

brucek

[mrsnell1]

Bruce,
Ok, I am getting it now. Told you I am slow. I will do as you suggest with the AVR when the house gets quiet again, if ever. But, I did run a test of mains + sub with changes to the sub vol, and yes, it changes the level balance. I guess that should have been obvious.

Anyway back to the sub only. I ran sub only measurements, recalibrated to 75dB using the sub test tone, and eliminated smoothing. I think these are what you are looking for, if not let me know. The sub plots follow.

Sub only Spectra Plot, with REW calibration 75dB


Waterfall of same


The sub spectra seems Ok, higher than the 75dB calibration but presumably due to room acoustics and within limits? But the watefall does not seem too good at the lower frequencies, or is it? Is this waterfall within norms though? If not, are room treatments needed, and what kind.

Thanks for the help and patience.
Chris

[Wayne A. Pflughaupt]

mrsnell1 wrote:

But the waterfall does not seem too good at the lower frequencies, or is it? Is this waterfall within norms though? If not, are room treatments needed, and what kind.

Waterfalls are tricky to get a handle on. For instance, your first one (green) appears to be much worse than the second, with significantly longer signal decay times. But in reality, the only reason the second (purple) looks better is because the overall level you took the reading at was reduced ~6 dB (focusing on < 70 Hz, as the first graph also includes the main speakers).

So you can see, the goal to get the signal down to the noise floor by 300 ms is rather arbitrary without a reference standard for the level the measurement is taken at. You could easily "obtain" the goal merely by turning down the volume and re-measuring until the graph "looks" right. Not to mention, you’d have to know the level of your room’s noise floor as well - it’s going to be much lower in a dedicated room with extensive soundproofing than a typical family room.

There are really only a few reasons to be concerned with waterfall charts. One is to locate and identify any significant room modes, which will show up as a peak with a much longer decay time than average, so that you can apply an appropriate parametric equalizer filter and bring the decay time back into line with the room’s average.

Another is to gauge the effectiveness of low frequency room treatments – bass traps etc. In this case you’d compare “before” and “after” waterfalls to see if the decay times had been reduced. The lowest frequencies (< 60 Hz or so) are the toughest to deal with, as treatments for those are quite invasive in a room that wasn’t designed with built-in remedies.

Either way, your waterfall graph looks perfectly normal for an untreated room. You don’t appear to have any significant room modes.

Regards,
Wayne

[mrsnell1]

Wayne,
Thanks for the input. Later I got to thinking about the floor noise, which you also touched upon. The floor noise was relatively high during that session (~50hz), lots of neighbors still running A/C, plus my own house noise. So, given that, the decay may not be so bad.

Thanks for the input on treatments. That is also what I really wanted to know, that the <60Hz stuff is more difficult. I didn't want to spend a lot of time trying different things if the overall general consensus is that it is a difficult beast to tame. Thus, I wont worry about treatments for now.

Chris