REW Beta Release 1st. crack at Frequency Response curve

[dan garza]

Love the program, I have alot to learn! Using the Umik-1 mic with calibration file, 1/48th smoothing on the graph, room dimensions are 14' 6" 18' and 8' ceiling.
This is a home system in my apartment. I've done some acoustic work already.
More importantly: how do I do an impulse response and a waterfall? there's so much to learn about the program!
Thanks

 

[John Mulcahy]

For impulse responses and the like you need to make sweep measurements rather than using the RTA - use the Measure button.

 

[dan garza]

For impulse responses and the like you need to make sweep measurements rather than using the RTA - use the Measure button.

Thanks! Measured to day, left side only so far:

 

[Matthew J Poes]

You did an impulse response. Those measurements you shared were derived from the impulse response you took.

An impulse response is kind of hard to read on its own so I suggest looking at the waterfall. You generate that. It's a tab across the top ribbon. It's just before the spectrogram you posted.

In order to make that useful I suggest people give a 600ms window and some reasonable amount of dynamic range that bottoms with the noise floor. Maybe start with 25db to 100db for now. Also you only want to look at this between your lower limit and 200-300hz. It's not all that useful above that point.

 

[Matthew J Poes]

Woah things appeared while I was typing. Never mind!

 

[Matthew J Poes]

Ok so this didn't load right originally and my message doesn't make sense inight of the new context. Sorry.

I suggest adjusting the waterfall the way I mentioned. Shorten the window and expand the dynamic range (top and bottom db limits) to say 25db and 100db. It makes interpretation easier. Further I would reduce to just low frequencies. As you can tell the stuff above 300hz decays so quickly it isn't really helpful. RT60 and EDT are better for that range.

 

[Wayne A. Pflughaupt]

Waterfalls are best looked at with a decay time in the 3-500 ms range.

Edit - just noticed that MJ already covered that...

Regards,
Wayne

 

[Matthew J Poes]

Waterfalls are best looked at with a decay time in the 3-500 ms range.

Edit - just noticed that MJ already covered that...

Regards,
Wayne

No worries!

And so there is no confusion. I said 600, Wayne said 300-500. All are right. Anything in that range is reasonable. Anything out into the 1000ms range extends too far to be reasonable in a small room. Decay in a home environment is quick so we want to zoom in on a narrower window. I use 600ms because it creates a time standard to compare against. There are a number of people online who used 600ms and others followed. To make comparison against these online posted waterfalls it's helpful to use the same time scale. 300ms is fine too, I find it a little short for very low frequencies (where decay tends to be longer) but we don't need such windows to identify ridges which are indicative of ringing. In other words there is no true standard, it's just about making the graphs readable.

Hope that helps.

 

[dan garza]

How's this look? This program is amazing!

 

[Matthew J Poes]

How's this look? This program is amazing!

View attachment 5171

Looks good now.

To interpret this you want to to look at the ridges. Most of them, you will notice are associated with peaks in the response. That is because in a minimum phase system the amplitude peaks will also ring. Knock down the peak and the ringing is knocked down as well.

Because the subs are against the wall that roughly 15hz mode appears to line up with a rear wall longitudinal mode. Nothing to be concerned with but you could EQ if you like. The ones at 60hz and 79hz don't line up with my estimate of SBIR effects but I may be misunderstanding your precise speaker locations. In any case it looks like some EQ and possibly more significant treatment would help. 40hz appears to be a rear wall peak. The closest predicted mode I come up with is 46hz assuming subs against the wall. My guess is that the room is acoustically larger than it appears. Could be wall construction, stuff in the room, etc. Rooms often shift mode positions a little due to stuff like that. If the room opens to another space that might help to explain it.

You have good decay above 80hz. I think it looks like your treatments are working relatively well there. If you shared what you have and wear, that would help. It is possible that you are seeing good decay but with modal anomolies due to placement. Absorbers always absorb no matter where they are, that's why RT60 drops in the same way regardless of treatment placement. Special strategic placement both above and below schroeder can change the response however. It's possible that moving your treatments might improve things above 80hz. It's also possible that those are anomolies from the sub and speaker interaction.

See how these estimates line up with your peaks and dips. Remember that these can shift from prediction due to all sorts of things the models can't account for. Look for close not exact match. If you have effects that aren't even close to the modal peaks and dips, it may imply it's speaker related, including speaker and sub integration.

 

[dan garza]

Looks good now.
To interpret this you want to to look at the ridges. Most of them, you will notice are associated with peaks in the response. That is because in a minimum phase system the amplitude peaks will also ring. Knock down the peak and the ringing is knocked down as well........

Thanks so much for the info, you were spot on about the room dimensions, it's a "bachelor apt" one large room that leads into a kitchen and small hallway, see additional pics and room sim pic for exact placement of the speakers.
System info: B&W matrix HTM L, C, R monitors, full range, out of phase with the subs, two 12" Velodyne active FR1200 (with the feedback accelerometer) 70kz x over 12db per oct, Sony STRV555, Crestron A/V amps, bridged 200w per side for the B&W's.
As for the wall treatment, I have "ACOUSTICAL R55" 2"x26"x72" on the back wall, a sheet next to the door in the front right corner, 36"x26" sheets behind the speakers, and 8' (floor to ceiling) foam corner killers in both rear corners (Mybecca 10"W x 12"H x 24"L TRIAMP Corner Bass Trap Acoustic Foam), carpet/rug covers almost the entire floor and I figured my bed also acts as a giant bass trap!
I mostly got rid of the 80hz bump with the foam, there's still a substantial bump @60hz and a monster bump starting @ 40hz going all the way down to DC!

PS, I have experience as an audio engineer from the 90's, so I have a fundamental/basic understanding of room acoustics.

 

[Matthew J Poes]

Yeah the only way to attack that stuff below 80hz would be to switch to larger denser bass traps and/or add membranes.

A cheap trick that works well is to buy a roll of adhesive carpet protection film. Apply this to the face of your bass traps. It will reduce some of the high frequency absorption but boost absorption below 80hz. Similar you could buy denser fiberglass or minerwool and construct corner traps floor to ceiling. You can again add membranes as described to boost LF absorption.

This trap is a square profile 24" by 24" trap made from slabs of 4" thick 6lb minerwool. The inside is hollow and I lined it with the plastic membrane. I then loosely filled the interior with more material.
I paid $30 for the minerwool and had leftover. I paid $1 a yard for the burlap and maybe $10 in wood total.

You might look into something similar and see if that helps. You don't have to make it so big. You could make a triangle profile instead too. Won't work as low.

Another trick that I've found works in the cheap are wool movers blankets. Around here we have Harbor Freight Tools who have cheap 1/4" thick wool moving blankets. Hang these behind curtains or wrap your bass trap and gain extra LF absorption.

Speaking of which, a less permanent way to increase LF absorption of the foam corner traps is to wrap them in dense fabric. 7oz canvas is still breathable but dense enough. Thicker would enhance the effect of more Lf absorption but with more HF reflection. Just a balancing act. You can pin it to the foam using t-pins and do a test measurement. I recently bought some nice blue canvas for like $5 a yard.

 

[dan garza]

Yeah the only way to attack that stuff below 80hz would be to switch to larger denser bass traps and/or add membranes.....
.

Nice!!
I'll try the carpet film on the corner killers.
I could also build a trap like the one you've shown, I have some empty space against the back wall, would making it taller help with damping, I could make it 24" x 24" x 4' tall or more....

 

[Matthew J Poes]

Nice!!
I'll try the carpet film on the corner killers.
I could also build a trap like the one you've shown, I have some empty space against the back wall, would making it taller help with damping, I could make it 24" x 24" x 4' tall or more....

Yes bigger is better. Just don't make it solid if you make it with dense fiberglass like I did. It will be too dense and reflect low frequencies. I made mine that height due to using wood scraps. It was for an event to highlight what can be created cheaply. I actually plan to build more to go floor to ceiling which is ideal.

The best location is a corner of course. You could remove the corner traps and move them to the ceiling wall intersection.