Success with REW!!!

[Anthony]

Last night BoomieMCT and I calibrated his setup. He was having problems with his wife's Mac (I think you can see the problem from that sentence ) so I brought over my laptop, MobilePre and ECM8k for some calibrating fun. Come to think of it, I didn't even get a beer out of the whole experience.

Anyways, after a few false starts, we finally got things rolling. Here were the steps we took:

1. Baseline Sweep
2. Analyze Peaks/ Filters/ Optimize
3. Run another sweep to confirm -- we noticed a broad peak in the whole sub response, so we set a new bank of filters to deal with this separately.
4. Analyze Peaks/Filters/Optimize
5. Confirmation sweep
6. Run Audessey Setup on his Denon (EQ, levels, etc)
7. Run another sweep
8. Boost the low end of the sub a bit (single filter at around 25 Hz, +6db, wide Q)
9. Run final sweep to confirm
10. Enjoy'd some scenes from Matrix and Star Wars.

It ended up being VERY flat from just above 20 Hz throughout the range. there were the usual peaks and nulls in the higher frequency region, but with smoothing it looked pretty good.

He was going to play around with boosting the sub levels to get just a bit of extra bass. But overall it sounded really good.

A side experiment: we removed one of his gigantic absorbers and found it made zero difference in the response. So he got some extra floor space out of the deal. We also found that raising one of his corner panels helped -- so REW is not just for BFD tweaking!!!

Hopefully Boom will chime in here with his own thoughts, but overall the evening was a success. Not bad for only about an hour of work (most of that moving stuff around).

Anthony


Big thanks to John for this. I don't think the DIY community thanks him enough for such a great (and FREE) tool.

[atledreier]

About the absorber.. it might not have done a whole deal on FR, but how did decay look with and without?

[Anthony]

It's not an echo'y room, so too much resonance wasn't a problem. he thought it was sounding a little too dead. So if the panel wasn't doing anything for the FR, he wanted to take it out. If it was doing great things for the FR, then it would be a tough choice (FR vs. RT60).

Sorry, should have mentioned that earlier.

[Wayne A. Pflughaupt]

A side experiment: we removed one of his gigantic absorbers and found it made zero difference in the response.

That would have showed up with a waterfall plot (assuming you’re talking about a bass absorber).

Regards,
Wayne

[Anthony]

The FR and waterfall plots were virtually identical (well within the measurement to measurement error) with the panel removed.

[BoomieMCT]

First off a big thanks to Anthony for helping me out here. I have a ECM-8000, a XENYX mixer and my wife's Mac, the latter being the weak link in the chain. Mac soundcards seem to chop off the top and bottom octaves with whatever they do - I've found Macs to be horrible for room EQ, music recording, pretty much anything. So much for being the ideal platform for artists. I'd say Anthony's setup listed above is basically ideal for this kind of work.

I'll respond to the comments about the panel absorbers. I like to experment with different materials to use for acoustic treatment. This has left me with a small surplus of acoustic panels, most of which end up in my underground theater room. I had the back wall of my HT mostly covered with treatment and the result was wasted space and a bit of over damping. The tests we ran showed that one of the panels simply wasn't doing anything beneficial so I removed it. This was one of my first tests made from RT9 fiberglass compressed in a wood / chicken wire frame. It's been a good trooper before and performed well before I got other materials to work with. All that said, I'm not too sad to see it go.

I'll make a final comment on the Denon AVR-2807's Audessy EQing. I love it. I think it does a great job, especially in difficult rooms like mine (concrete floor, can't fit treatments on the ceiling) and with systems that can't fit proper floorstanding speakers. Without sounding too much like a commercial I think other auto EQ systems I've heard sound good but Audesssy beats them all. There are two things I'll point out though. The first is that if you do not do some EQing on your sub to fix peaks all auto-EQs I've seen will set your sub too low. Make sure you fix room modes first! The second thing is that I do not fully understand how to setup subs on RoomEQ for systems that have different crossover points for each speaker (as Audessey does). RoomEQ wants you to fit the EQ to a specific crossover point and slope - something that might not be constant in my system.

[BoomieMCT]

Anthony wrote:

The FR and waterfall plots were virtually identical (well within the measurement to measurement error) with the panel removed.

Yeah, when Anthony says "virtually identical" he means we looked and looked and couldn't find the difference. Those features are very neat and I'm jealous that I can't use them (as Mac's won't run the newer versions).

[Wayne A. Pflughaupt]

BoomieMCT wrote:

The second thing is that I do not fully understand how to setup subs on RoomEQ for systems that have different crossover points for each speaker (as Audessey does). RoomEQ wants you to fit the EQ to a specific crossover point and slope - something that might not be constant in my system.

Assuming you have only one sub, its crossover point isn’t going to change. The only difference that would show up in full-range measurements would be where some of the speakers’ response would overlap the sub’s (or perhaps underlap), if their filtering was at a different frequency from the sub.

Regards,
Wayne

[BoomieMCT]

Wayne A. Pflughaupt wrote:

Assuming you have only one sub, its crossover point isn’t going to change. The only difference that would show up in full-range measurements would be where some of the speakers’ response would overlap the sub’s (or perhaps underlap), if their filtering was at a different frequency from the sub.

Regards,
Wayne

I buy that - especially if the slope of the sub's crossover is less than the slope of the speakers crossover (as is often the case). Does anyone know the slopes for speakers / subs on a Denon AVR-2807?

I think we used a fourth order slope at 90 Hz to set up the sub's EQ. I'll check the setting to see where it is setting the sub at (I'll have to guess at the slope). Either way, the final result with everything together has a pretty flat FR.

[thxgoon]

BoomieMCT wrote:

Yeah, when Anthony says "virtually identical" he means we looked and looked and couldn't find the difference. Those features are very neat and I'm jealous that I can't use them (as Mac's won't run the newer versions).

The waterfall plot feature should be included in the latest mac release. It is on the one I downloaded anyway. Stability has been an issue though as I think mac's java implementation could use quite a bit of work. I end up restarting into windows and it works great... the on board sound card actually measures really flat.

Otherwise good job! I agree REW is an absolutely awesome resource for all of us and John does not get enough thanks so here's a big one from me. Thanks!

[BoomieMCT]

thxgoon wrote:

the on board sound card actually measures really flat.

Which Mac do you have? I'll have to look up which one my wife has but it has almost no response above 12kHz or below 40Hz. This has been noticed in more than using RoomEQ. When my wife's band is practicing they usually run a line from the PAs to the Mac, record using Garageband and then play back through the same mixer. Since all the equipment is the same it *should* sound the same. It doesn't. The dynamics are completely flat and it is missing the same highs and lows as noticed on RoomEQ.

On other forums I've heard similar stories from other Mac owners.

[thxgoon]

I have a MBP 2.2 that I just got a few months ago. I'll have to see if I can dig up a pic of the calibration file when I get home. I used to use my roommates Toshiba laptop and the built in sound card had pretty bad rolloff for the lows so having bad FR is not primarily a mac thing. Can't remember how the highs looked on it though. Sorry to hear the computer has been a disapointment.

For recordings I always use external hardware. If you're looking for professional quality results you're probably not going to find it with a built in sound card using a 1/8th" mini jack. You might also check in Garageband and make sure that all of the Compressors are turned off. IIRC when I used it a compressor was added to the master track by default. Also, check and make sure that you are not clipping levels on the computer coming in or going out. You can check that in the audio control panel. Might be a solution to the problem of dynamics...

[BoomieMCT]

Ah, we were using a Mac Powerbook G4. I guess your MacBook must have a better soundcard - the one in the G4 is absolute trash. For the recordings it wasn't clipping, it was just bad (like a really poor MP3). We also could not find any compression settings in Garageband. Maybe we were using an older copy.

Someday (probably when I get a laptop) I'll get an nice external soundcard. In the past it wasn't needed. The soundcard on my homebuilt PC in my office / studio (which uses a 1/8" stereo plug) actually got pretty good results. I would use the same PC to set up my HT but it isn't easily portable.

[thxgoon]

Got home and took a pic of the sound card calibration if you're interested. The dashed black line is the calibration file and I assume that the high end roll off has to do with reaching the sampling limit though I could be wrong...

[brucek]

and I assume that the high end roll off has to do with reaching the sampling limit though I could be wrong

It's the REW target line for filtering............ overlay lines are turned on and off with the checkboxes at the bottom of the graph screen...

[thxgoon]

Here's a pic without the Blue REW target response. This is just the soundcard.

[DrWho]

Yea, the HF rolloff is the anti-aliasing filter on your soundcard...nothing to worry about. In fact, the cal looks pretty good.