I wasn't clear about this, I'm sorry. I have a +26 dB EQ peak at 20 Hz, but also -29 dB general attenuation. The *result* is -3 dB at 20 Hz.
Obviously large vs small room acoustics are very different. I agree that a "dead" room sounds unnatural and should be avoided in typical listening rooms. It would be helpful to measure your rooms response at all frequencies to see what is actually going on in your room. Bass traps / panels with a mix of reflection and absorption is the key to getting good sound in a small room. Diffusion can also help make a small room also sound larger than it really is. The end result is a tighter, crisper sounding room. A quick experiment is to play some music through your system and record it at the listening position with a good stereo microphone recorder of some type like a Zoom H2 for example. When you listen to the recording you get a good evaluation of how your "room" actually sounds and how it is affecting your music. Properly treated rooms will have a closer representation to the actually source, which should be the goal in all listening rooms, right?
Concerning OB, I don't have a ton of first hand experience. However I am currently testing some 25" in planar/ dipole midrange elements. It's my understanding that OB's goal is to not have colorations of an enclosure, produced by a back wave reflection. My concern would be the same reflection OB design's try to eliminate would be introduced via the acoustic element of the room and the back wave to the listener. In the case of Stig Erik's room, he already stated that he has treated the whole front wall to about 250hz. I am curious how the rest of the room is treated? I'd also be curious to see some waterfall graphs of frequency sweeps.
The alternative to a damped front wall reflection is an undamped frontwall reflection... a damped frontwall will still reflect something, but its possible to get the reflection more than 30 dB down over a large frequency range using the right materials.
My room is damped on the walls and ceciling in the entire 1/3 of the front end. The rest is untreated, but there's furniture giving some diffusion.
Will look into doing a waterfall measurement, and maybe an ETC measurement.
My room is damped on the walls and ceciling in the entire 1/3 of the front end. The rest is untreated, but there's furniture giving some diffusion.
Will look into doing a waterfall measurement, and maybe an ETC measurement.
Decay in listening position, no smoothing:
With smoothing:
ETC:
There's just a single reflection that's not below -20 dB, and that's from the floor. I have a very thick rug there, but it doesn really help...
An externally hosted image should be here but it was not working when we last tested it.
With smoothing:
An externally hosted image should be here but it was not working when we last tested it.
ETC:
There's just a single reflection that's not below -20 dB, and that's from the floor. I have a very thick rug there, but it doesn really help...
An externally hosted image should be here but it was not working when we last tested it.
Roughly calculated, the EQ is effectively equal to a low pass filter with a slope slightly bigger than 12dB/oct. (Correct me if I'm wrong) Anyway, no matter what the 'slope' is, the overall shape of EQ curve is inevitably a low pass.
Such filter has a 'delayed' phase response in general. That maybe one of the causes to 'lack of slam'. (again, correct me if I'm wrong, or someone can explain it much better) You may reduce such delay by pushing the 'knee' of the filter lower, or lower the filter Q (make the knee rounder/flatter).
I'm glad you got a good result from H baffle. I believe the improvement you got is by the raised efficiency (at least partially). However the major acoustic loss of dipole and low Q rolloff are still there. It's good to get the slam 'back' by the more efficient H baffle, but I assure you, it can be better than that (if not 'much better'). With the woofers you're using, your OB can eat any boxed speakers in EVERY respect hands down (except subsonic range and sheer SPL).
Sure, that's a very good scientific spirit. So I keep repeating myself "it's such a pity I can not explain it well enough". And I don't have sophisticated measuring tools to prove it.
But still, I'd like to suggest, the following articles are great, you might like to read them (or, I bet you've already have.)
http://www.passdiy.com/pdf/cs-amps-speakers.pdf
Jordan JX92S OB with a Goldwood GW-1858 Woofer in an H Frame Project
PC-based XO.
All EQ and filtering is made with the PLParEQ from Refined Audiometrics Lab. Its the most transparent sounding EQ I've heard. The only drawbacks are very high CPU utilization and high latency. The latency is so long that its difficult to measure the output after EQ and XO properly. I guess its a second or maybe even more.
All EQ and filtering is made with the PLParEQ from Refined Audiometrics Lab. Its the most transparent sounding EQ I've heard. The only drawbacks are very high CPU utilization and high latency. The latency is so long that its difficult to measure the output after EQ and XO properly. I guess its a second or maybe even more.
Square wave?
I bet you could measure a good square wave even when most passive cross speakers give a result that is not recognizable.
This idea begs an el-cheapo clone ... I'm seriously thinking about B&G NEO3, some Zaph ZA14's (or Vifa P13's, now on special) and Seas L22RNX's. DCX2496 should cover those well enough. No way though I can run 21" woofers and stay married though 
But if I get enough feedback about GR-research servo 12" ... supposedly two of those per side should be enough for everyone but extreme bass freaks.
Now to figure how to sneak through speakers on chains into the living room
But if I get enough feedback about GR-research servo 12" ... supposedly two of those per side should be enough for everyone but extreme bass freaks.
Now to figure how to sneak through speakers on chains into the living room
This idea begs an el-cheapo clone ... I'm seriously thinking about B&G NEO3, some Zaph ZA14's (or Vifa P13's, now on special) and Seas L22RNX's. DCX2496 should cover those well enough. No way though I can run 21" woofers and stay married though
But if I get enough feedback about GR-research servo 12" ... supposedly two of those per side should be enough for everyone but extreme bass freaks.
Now to figure how to sneak through speakers on chains into the living room
Tell her that it's all a part of your new weight lifting workout.
Bench Press Chains
In the form of a firewire audio interface, usually. The mic preamps could be reused for vinyl though standalone phono preamps aren't hard to get. Echo, Emu, Focusrite, M-Audio, Presonus, RME, and Roland/Eidirol are just a few of the companies in the space. Get into the rackmount gear and my experience is the quality's tough to match in the hi-fi market for anywhere near the same price.
I'm curious what's Erik's using though.
ASIO yes. I have an RME Fireface 800 external interface, which is connected to a DAD AX-24 8-channel DAC.
Analog sources and vinyl is easy. I just plug then into any analog input of the RME Fireface and route them through the XO. For the vinyl turntable I use a microphone input on the Fireface to get enough gain, and run the RIAA correction as a part of the digital XO.
Nice, i didnt know this EQ
Did you try Algorithmix too?
Algorithmix LinearPhase PEQ Red --- DirectX/VST 10-band precise linear-phase parametric Equalizer featuring unique shelving filters for mastering of complex mixes without changing their sonic character