Systemic Development > Psycho-Acoustics
Any Inguz Users Out There?
miklorsmith:
I'm embarking on Inguz room correction. I have (I think) all the software installed and running, the microphone preamp, and a microphone from my TacT 2.2XP (to be upgraded). I also still need a microphone stand but can limp along without it.
Any tips on measurement or quirks about the "system"? I'd love to hear experiences that could help me sidestep some of the learning curve.
Thanks in advance,
Mike
inguz:
Documentation about the measurement process is slightly in flux right now; the current version of of the software doesn't quite match the documentation, and the old doc is very crufty.
The old doc:
http://inguzaudio.com/Tools/
There's a new pass at the documentation -- work in progress, still -- here:
http://inguzaudio.com/measurement/
Here's a good place to ask questions if you get stuck any.
-Hugh
Carlman:
Welcome to AN, Hugh! It'd be nice to have an intro to your theory(ies) for applying room correction. Some of us have never heard of Inguz. How is it different than TacT or Behringer?
Thanks,
Carl
miklorsmith:
I'll start the discussion - my TacT preamp cost $6,000 and I'm getting going with Inguz for around $200! This setup is also transparent to the rest of your rig, i.e. you can try it out with all your existing electronics in place. Further, if your buddy has the stuff, he could bring it over to your house, conduct measurements, and leave with the stuff. Voila - room correction and it didn't cost a dime.
Hugh has been very proactive about adding to the software, all of which is free. He's even talking about doing a parametric option for the EQ plugin. This is VERY cool and seriously, a 5-band parametric, SPL meter, and test tone files would be all most folks would need.
Hugh has been EXTREMELY helpful with me over a product that he's not even selling. How's this - Some of the best customer service I've ever had.
I think I have enough to get off the ground, though I need a better microphone and stand to do it "right". I'll wade into the waters this w/e and see how much trouble I get in. :D
inguz:
Good to be here! Let's try some of my philosophical background on room correction. Coffee: check.
I'm not really a longstanding DSP expert. In fact I'm approximately a cheapskate audiophile: I want a truly great-sounding system for as little outlay as possible. Having heard a few really excellent systems (eg. Wadia SACD to 300lb Pass amps to esoteric ceramic-driver dynamic speakers; SET to Edgar-style Tractix horns), it's quite apparent that this game is endlessly subtle and very few systems can approach "live" reproduction (or however you define the ultimate goal).
Broadly, I'd like to see this as an engineering problem. So I'm skeptical of the tweaks brigade, and even more so the Mystical Dimensions of Horns stuff. But on the other hand, experience is undeniable: damping plates and isolation make many components sound better, and hooking up the humungous Virtual Dynamics Genesis power cable to my Transporter made an audible difference. So with a scientific attitude that everything is measureable, I'm not convinced we are measuring all the right things.
Back to room correction. It seems this all comes down to the relationship between the loudspeaker and the room. I don't have enough experience here yet to be definitive, but the results are dramatically different between speakers with different dispersion patterns -- I have to try corner-horns some time soon -- and removing the influence of the room on the direct and reverberant sound is just about always positive.
In a traditional stereo setup, the "sweet spot" is fairly diffuse: over a wide area, you have some experience of stereo positioning, although the hearing-into-the-recording-ambience experience only has its best effect at one narrow point. If you can add absorbtion panels, and maybe diffusers, to stop slap- and flutter-type echo in the room (what REG calls "room roar"), there's a broad improvement for all listening positions.
Digital correction on the other hand usually has the opposite effect. The sweet spot narrows, but the experience in the right spot can be astonishing. So I'd really advocate, in this order,
- Put some absorption in the room, especially at first-reflection points.
- Get a decent pair of loudspeakers.
- Add digital room correction.
Digital correction actually helps with #2 as well. I had an old pair of Linn Keosa speakers for a long while, and while they're nothing special, adding room correction made a huge difference -- not because it corrected the reverberant room very much, but because it drastically cleaned up the direct sound.
Where was I? Ah. Squeezebox.
I love the digital network player. The SB's user interface is a bit clunky (need spectacles... hrrrmph...) but for absolute convenience it's a winner. And surely, if enough attention is paid to the DAC and analog stages, this is superior to any spinning plastic disk or S/PDIF transmission. It's my only source now - bye-bye turntable, tape deck, CD player, even preamp.
On the back end, SlimServer has enough hackability to intercept the digital bits before they hit the network, and run DSP filters in advance. Hence the InguzDSP filter-processor and EQ plugin. A PC has enough horsepower to process several streams simultaneously at very high resolution. So, if you can live with the limitations -- EQ changes only taking effect once the SB's buffer empties, and losing the ability to fast-forward and rewind within a track -- this is approaching the ideal system.
(Architecturally I think the right thing is to move the PC-architecture processing as close to the loudspeakers as possible. I put the DSP in the source. TacT and so on put the processor in the hi-fi rack via a S/PDIF loop. It ought to end up inside the loudspeaker itself, eventually).
In building EQ filters, I tried to take a "no compromise" approach. Process at 64-bit-float resolution, output at 24-bit depth, and run thousands-of-taps filters even for trivial equalization. I think this works OK, since I can't hear any ill effects.
The measurement and correction process:
- Measurement is really OK, even with cheap equipment. Using long sweeps gives enormous noise-immunity, so even moderately loud background noises and so on don't make much of a difference to the results. It's a hassle to buy or borrow measurement equipment, and I'm sure my neighbours will complain if I run sweeps too early on Sunday mornings. But you only really need to measure every few months.
- Building the room correction filters is a black art, and I don't do that. The two software systems I know about are Uli Bruggemann's "Acourate", and Denis Sbragion's (GPL) "DRC". I haven't used Uli's system to generate correction filters, so I can't compare the two, but I expect they have enough subtle differences that you would make a personal-preference choice for one or the other. DRC is insanely configurable, but in places the documentation reads like a PhD.
- I think my measurement method (record both the unprocessed sweep and the microphone input, to a stereo file, for each loudspeaker) is more robust than deconvolution against a pre-generated inverse sweep. But as I may have mentioned, I'm not really a DSP expert. So if you want to use this method, I have documentation about how to set up the measurement, and my tools package lets you convert those measurements to impulse responses which Acourate can read. With DRC I went a step further, and my tools actually run the DRC part for you; it's still a pretty geeky business to get things configured properly, but much of the grunt-work is automated.
- Just Do It. The only time I've been disappointed with the results was when a bug in my DSP messed up the sound.
H
Navigation
[0] Message Index
[#] Next page
Go to full version