Using a High WAF Design to Block the Back Wave

Hi Bernard!

See the area within the elipse I've drawn on your plot?

That reflection has been sharply reduced by your traps. Timing looks perfect for a reflection from the front wall, too. Ideally, you want a smooth, dense set of reflections which diminish in volume over the first 40mS, or so. Reducing that early spike is a step in a good direction, acoustically speaking. To find the problematic spikes, one must use an unsmoothed response.

Best primer for audiophiles on this is Mellor/Hedbeck's paper I've linked to, before: http://blog.acousticfrontiers.com/w...-standards-for-stereo-listening-rooms-pu.html, as well as a follow-up by Nyal here.

Ideally, we'd also look at a smoothed response by octave, as well as left-only/right-only comparison. Unfortunately, Omnimic doesn't smooth, yet. I'm trying to get Bill Waslo to take this on, as it's really a very simple transform for the software. Until then, only programs like REW, ARTA, etc. will do.

Ken, thanks for your response. I did a search on getting Fourier coefficients using Excel (for the smoothing I mentioned above), and came up with this, which I have not yet tried:
http://www.ehow.com/how_4670778_fourier-transform-fft-microsoft-excel.html

I think your investment in test kit is great, Bernard. It is interesting to know what is going on, is it not?

BTW: re the tubes - RCAs followed the Amperex closely - not shown in the plots - much like the Shuguangs followed the GEs. Which is ironic, since the chinky tubes are an architectural copy of the RCAs!!!

If you have some 6550/KT88s you can roll, I'd be interested to see the plots. Be carefull not to nudge the mic or change anything in the room while taking measurements/rolling tubes. Paranoia is the best policy!

Bernard said:

Ken, thanks for your response. I did a search on getting Fourier coefficients using Excel (for the smoothing I mentioned above), and came up with this, which I have not yet tried:
http://www.ehow.com/how_4670778_fourier-transform-fft-microsoft-excel.html

Click to expand...

Owee! (Geez, I sound like Todd!)

Problem is, we not only want to smooth, but then have the program overlay the various by-octave plots so that we may easily see any spectral variance - we want reflections to be as full frequency as possible. In some utopian world, we'd then tailor the absorption/diffusion to achieve spectral balance. If Waslo doesn't do something, soon, I'll have to dredge up all the bits and pieces I've purchased for REW and learn to use it. Argh!

Be aware, Bernard, that you should be measuring one speaker at a time - left only, right only. What I did was to select a spike, then have a large-ish audio friend, aka The Human Absorber, move around the room until the spike was attenuated. We then moved an absorber around that general location until I got the best response. In lieu of THA, just move a throw pillow or something with some absorption around the front/rear/sides of the mic until you find the general direction from which the reflection originates and triangulate from there. Ironically, for me, the first reflection points weren't problematic(!). The contralateral points (reflection on left wall from right main and vice versa) were. I already had the rear wave absorbed.

Have fun!

Justin, it IS interesting to see what's going on.

I'm using SED Winged C 6550's in my ARC`s. I can`t remember what was in them when I got the monoblocks, but I still have them; the SED`s were much better. I`ll do the comparison (just for you, dear friend) one of these days. And don`t worry, I won`t change a thing. I may not have enough tubes to do both channels, so may do only one, which is probably best, with the mic firing straight at the speaker.

Ken, you`re asking for the impossible - audio friends are few and far between, but not only that, you want the guy to be large. Well, I do have smallish audio friend who may not object to my duct taping pillows to his person

I do realize that in order to do tweaking I should be measuring one speaker at a time, but I did both as I wanted to know the overall effect of the pair of traps. Unfortunately I cannot place absorbers all over the place as I do have wife who has to live with the visual; living with me is bad enough. And yes, we ARE beginning to sound like Todd, not that that`s a bad thing !

The thing to remember is all this stuff is just a guide. Subjectivity always rules, as it is you who is doing the listening, not the mic. Hence me deliberately committing the sin of rolling off the top end with KLM5.

This was my own idea - including the target frequency. 12KHz is quite high subjectively - try a signal generator and listen to it. However, subjectively it calmed down a hyper-real presentation into a more humane one. Being honest I have never, in my entire life, heard anything like a flat KLM5 for resolution - ever - anywhere. However, there is such a thing as too much - hence the calming action.

Also, cross-comparison between our plots is meaningless. Different test kit will yield different results (though it ought not).

The only meaningful technical comparison would be with the same test kit/ancillaries in the same room.

Justin, as someone who can alter the FR of his system at will, IME the hardest thing to get right is the 6-20kHz region. I'm constantly tweaking it to varying levels of dissatisfaction. To make matters worse and because there are no studio standards, FR varies from recording to recording, so that what sounds terrific with albums A-M may sound flat and lifeless for albums N-Z. There are no ideals and, yes, it's all subjective.

User211 said:

The thing to remember is all this stuff is just a guide. Subjectivity always rules, as it is you who is doing the listening, not the mic. Hence me deliberately committing the sin of rolling off the top end with KLM5.

Click to expand...

Remember that you're talking to someone who loves the sound of the voluptuously sweet Koetsu ! It may not be accurate/analytical, but I love it.

RUR said:

Justin, as someone who can alter the FR of his system at will, IME the hardest thing to get right is the 6-20kHz region.

Click to expand...

But it's highly unlikely that you can hear up to 20kHz.

User211 said:

Here's an interesting plot showing the effect of different 211 tube types on FR. This is the area of the FR that showed the biggest difference. So tube rolling does make a difference - at least measurably.

Test mic was the Behringer - pretty good mic I think. Beats an iPhone anyway! M-Audio external USB soundcard/preamp.

Coloured coding on the left should be obvious. Best save the pic to a file, then view in a JPG viewer to zoom.

Click to expand...

As high as 4db in variation with tube rolling? Yikes. I'll stick to solid state. I don't need any more variables in the equation. haha

Bernard said:

But it's highly unlikely that you can hear up to 20kHz.

Click to expand...

At my age, certainly not. I was reading a couple of VLF hearing acuity studies recently and we, more mature folks, appear to have better hearing @ VLF than young'uns. It all balances out!

RUR said:

At my age, certainly not. I was reading a couple of VLF hearing acuity studies recently and we, more mature folks, appear to have better hearing @ VLF than young'uns. It all balances out!

Click to expand...

Whatever you guys do, don't download the audiology test app for your phone... good lord, test results I just didn't need to know.

Justin, your HF roll-off preference is but an outward projection of most of our inward response curves.

Bernard, there are a couple of other measurements you could/should do. Try measuring FR for bass only - 5-200Hz with the resolution set to, say, 24ppo. That's about the minimum required resolution (for bass) in order to see what's really going on. Also try measuring bass decay.

/Ken

Ken, what is 24ppo, and how do you set it? I see nothing in the help section.

Hocky said:

As high as 4db in variation with tube rolling? Yikes. I'll stick to solid state. I don't need any more variables in the equation. haha

Click to expand...

Yeah, but Justin can tailor the sound to what is most pleasing to him, which is what it's all about. Can't do that with solid state.

Bernard said:

Ken, what is 24ppo, and how do you set it? I see nothing in the help section.

Click to expand...

Bernard, it's simply an easier to type name for 1/24th octave resolution (24 Parts Per Octave). On the Frequency Resolution tab, look in the upper right. You'll see a tiny window which is default to 1/6th Octave. Simply change that to 1/24th Octave using the little drop-down arrow. For bass, lower resolutions often don't show enough detail to be useful. For full FR measurements, 1/3 or 1/6 is fine.

Bernard said:

Yeah, but Justin can tailor the sound to what is most pleasing to him, which is what it's all about. Can't do that with solid state.

Click to expand...

No, but you can do that with EQ, too, where you can legitimately make the changes that you want.

Hocky said:

No, but you can do that with EQ, too, where you can legitimately make the changes that you want.

Click to expand...

Command central.:music:

Hocky said:

No, but you can do that with EQ, too, where you can legitimately make the changes that you want.

Click to expand...

Are you saying that Justin is illegitimate?