1st results and a reverb problem.. Quite a challenge!

[alexp]

Hi,

This is a fantastic forum - it made me think. A lot.

Well, I present you: a challenge. The main problem: reverb/echo. Not extremely bad, but definitely the key thing to improve. 38Hz peak (with no Audyssey on) would be second on the list.

Otherwise, the frequency response is fairly even, stage is there and stereo impression is very good (7.5/10 on a subjective scale).

Pictures say more than a 1000 words, so see the attachments first.

Att. 1 - the room. Floor: hard wood, walls and ceiling: dry wall. The slopes are what makes the room interesting.
Since the pix is crowded as it is, I didn't include any small objects such as CD stands (1m tall, next to the sub) nor the projector (above the sofa).

Att. 2 - low freq results without EQ

Att. 3 - REVERB - without EQ

Att. 4 - full freq without EQ

Att. 5 - full freq with EQ (Audyssey on Onkyo). Of course it has no impact on reverb

I used a DT8850 SPL - which has the same mfg specs as Galaxy 140 (although I'm not using any calibration file for SPL). It's certified from 35Hz to 8kHz so I didn't post anything outside this range.
Sound came from HDMI out and SPL was hooked to line in on my on-board Realtek card.


AV: Onkyo 805, 2xB&W 602S2, 2xB&W 601S2


Let me know what you think - I'm thinking about some absorbing treatment behind sofa or behind front speakers..

Cheers!

[brucek]

A few comments about REW first.

The small floppy icon in the lower left corner of your graphs is designed to save REW graphs in an 800 bit wide jpg suitable for posting.

Any plot higher then ~200Hz will benefit from smoothing (1/3 octave) to remove comb filtering and reveal the underlying trend.

It's hard to get too serious about any measurement at the extremes when you don't use a calibration file for your mic. You should attempt to get a hold of a Galaxy CM-140 and redo your measurements. We supply the cal file on our download page.

RT60 is really only valid in very large rooms and not suitable for home use.

Always start with a sub only measurement from 0-200Hz to get a feel for what the sub alone is adding to the mix. Then positioning or equalizing can be accomplished on the sub first. Once that's complete, you can add the mains and do another identical measurement from 0-200Hz and adjust the subs phase control to smooth the crossover region as good as possible.

Then do your 0-5KHz measure with smoothing and you're ready to evaluate. Anyway, I guess it's kind of obvious that you have a peak at 40Hz that needs addressing.

brucek

[alexp]

brucek wrote:

A few comments about REW first.

The small floppy icon in the lower left corner of your graphs is designed to save REW graphs in an 800 bit wide jpg suitable for posting.

Any plot higher then ~200Hz will benefit from smoothing (1/3 octave) to remove comb filtering and reveal the underlying trend.

It's hard to get too serious about any measurement at the extremes when you don't use a calibration file for your mic. You should attempt to get a hold of a Galaxy CM-140 and redo your measurements. We supply the cal file on our download page.

RT60 is really only valid in very large rooms and not suitable for home use.

Always start with a sub only measurement from 0-200Hz to get a feel for what the sub alone is adding to the mix. Then positioning or equalizing can be accomplished on the sub first. Once that's complete, you can add the mains and do another identical measurement from 0-200Hz and adjust the subs phase control to smooth the crossover region as good as possible.

Then do your 0-5KHz measure with smoothing and you're ready to evaluate. Anyway, I guess it's kind of obvious that you have a peak at 40Hz that needs addressing.

brucek

Ad. icon - got it

Ad. smoothing - see attached - full scale smoothed (1st without EQ & sub, second with EQ & sub)

Ad. SPL - no can do - not available in Poland (I tried reaching out to the apparent 'overseas' dealer in LA but he never replied)
My SPL has the same (or very similar) characteristics as Galaxy 140, so I'll check out results after applying the calibration file.

Nevertheless - I don't mind the extremes as my main speakers don't go below 35 and I'm listening to stereo without the sub (which is disabled in Pure Audio mode, and listening in any other mode significantly limits the sound stage).

Ad. RT60 - hmm, results looked very real, I'm surprised to hear that the function doesn't work at home... Do you know any other reliable way to measure reverb?

Ad. Sub measurement - I'll try it, although as mentioned earlier - so far I'm not using the sub in stereo.

Ad. peak at 40Hz Yep, that's the one. What do you recommend?

[brucek]

results looked very real, I'm surprised to hear that the function doesn't work at home... Do you know any other reliable way to measure reverb?

You'll always get results, it's just that they mean very little in a small room (i.e. smaller than a gym or a church).
For low frequencies < 200Hz, we use waterfall and decay plots to see what the room is doing. Above 200Hz we use ETC plots to examine reflections. They are easily identified and their distance revealed with the REW measurement tool.

peak at 40Hz Yep, that's the one. What do you recommend?

Looks like your EQ is already taking care of that quite well, although there are several deep dips in the after EQ graph..

brucek