Home Theater Forum and Systems banner

1 - 3 of 3 Posts

1 Posts
Discussion Starter · #1 ·
Hello all. This is my first post. I hope someone can help point me in the right direction. I've searched most of the forum, and I've learned quite a bit. But I still don't have all the answers. Basically my question is, why do I need REW? I mean I understand all the things it can do, but how can I use it to make my system sound better? I don't even know what I want, other than that - making my system sound better. How can REW help me achieve that?.

I have already calibrated my mic and my soundcard. I've run a few tests. Now what? I have some cool looking graphs. What's next? Let me first describe my equipment, room, and current sound. Our home has an upstairs gameroom where we have a full htpc 5.1 setup. My htpc is running Win 7 64 bit. I use J. River as my main front end. The mobo has a fiber connection which I am feeding into a Krell Showcase preamp. 90% of the time I use this room to listen to music. My files are everything from 16/44.1 all the way up to 24/192 FLAC 2 channel, with a few 5.1 rips like the Immersion Pink Floyd rips and several SACD DSD rips as well. Anyway, the Krell's DAC's handle up to 24/192, but soon I will be adding a new DAC that can decode DSD natively (when prices drop to something more affordable). My amplifiers are Denon, and my speakers are B&W, while my subwoofer is an M&K dual 12" with 400 watts. I purchased a Dayton MME-6 calibrated microphone and a Tascam US-144mkII to use for the calibration, as well as a Radio Shack digital SPL meter. So far I have managed to calibrate the Tascam. It was difficult because many times a pop up message kept alerting me that the level was too low, or that the signal was clipping, or the signal was distorted and could not be used. That went on for 2 days. Finally today, using the loopback method, I got a nice measurement. I confirmed it was a good measurement by running it a few more times. The Dayton mic already had it's own calibration graph. So now with the hardware ready to perform the tests I was ready to start measureing. But what exactly do I measure?

I guess now I must explain what I'm after. My gameroom sounds great, but I know I can improve on it. I have never used, nor do I own and eq. I am pretty sure I will need this at a minimum, along with absorption panels, to improve my rooms acoustics. My Krell has a 3 band frequency parametric eq built in that I can start at. My goal, I guess, is to take a frequency response reading and see where are the points that need boosting or cutting. I'm assuming I would first want to attempt to correct the rooms acoustics with panels and traps before using the eq. Once I've achieved a nice and flat response, I'm thinking I could use the DSP settings, as well as any additional small scale eq'ing, in J. River. I understand that J. River also has the capability to take the calibration files that I've made in REW and import them into it's own room convolution software. I'd like to be able to do this. Basically, I'd like to use digital room correction to achieve the best sound possible from my system and my room. I just don't know what to do next.

When I performed the measurement in REW I had some really cool looking graphs and charts that impressed my wife and kids, but left me looking like a deer in headlights. I just didn't know what to do next. I don't even know how to see my rooms frequency response curve. I saw the SPL graph. Is this my room's and equipment frequency response? I smoothed it out and it made more sense to me but I don't know if I'm looking at the right chart. I'm kinda stuck at this point. Where do I go next? What measurements does one take to get a better view of his rooms acoustics? And once I have that view how do I apply it to J. River so that it can do it's magic and digitally "fix" my room?

I understand these are a lot of questions. I just need to be pointed into the right direction, then I can figure it out. Any help would be greatly appreciated.

Thanks in advance.


239 Posts
What are your listening impressions of system as is?

What does it do well? And where does is disappoint?

For best sound stage speaker placement is fundamental. Speakers should be at least two feat in front of wall, and at least three feat from each side wall. This allows for each speaker to have same bass characteristic. First reflections from side walls arrive a crucial few milliseconds later then when speakers are jammed into corners, relative to the direct sound. This gives brain critical time for extracting directional information from direct sound, which is mostly contained in 500Hz-4kHz range.

With subwoofer great to start incorner, where max room gain happens, and making most equalization cuts rather than boosts.

If room is sparsely furnished and without carpet and friendly ceiling, then absorptive panels may be beneficial. Effective bass traps take up lots of real estate, and can't just be placed anywhere.

Surprisingly easy test for need of panels is clapping your hands from listening position. If sound hangs, or reverberation sound comes from a particular direction, room may be too lively.

So, for stereo start with speaker placement to get good starting sound stage. Take measurements of each speaker individually and check for gross differences in frequency response (FR). Apply at least 1/12 octave smoothing, but likely 1/6 octave or even 1/3 octave to get flavor of the situation. Take several measurements with slightly different microphone placement, as in just 2-3" inch shifts. Apply enough smoothing so plots from these similar locations look very similar. Likewise try a couple of measurements with microphone stationary, and just moving a speaker, first just inches, and then about a foot. This will give you a sense of how sensitive measurement is, and how sensitive sound stage is to speaker and listener positioning.

Once gains from placement are tapped out EQ mains to desired flatness without subwoofer, and without crossover for subwoofer. Likewise EQ subwoofer without crossover so it is flat to at least an octave above desired crossover frequency. Example: if crossover is intended at 80Hz, EQ subwoofer flat up to >160Hz, and EQ mains flat down to 40Hz. This way when crossover is engaged it will be much easier to get delay/phasing and final EQ in crossover region set.

Convolution based EQ and room correction used well are extremely effective.

With htpc and JR MC convolution engine it is suppose to be fairly straight forward to apply filters generated based on room measurements (room's impulse response (IR)), using REW. Filters for full DRC may be created from room IR. Sourceforge DRC, even with Correction Designer gui requires much care and repeated attempts with variations of the many user variables. Audiolense and Acourate are friendlier, but come with up front price. Convolution configuration file is scary at first.

There are a few experienced JR MC convolutionists poking around this forum, but plenty on JR MC forum.

To get good results with convolution it is necessary make measurements through convolution engine. This is still somewhat weak link with REW. Sweep output needs to be input into convolution engine. Supposedly using two sound cards it is possible to select output of one card (measurement card) from within MC as input for multi channel processing and output by fiber to Krell. Testing the set up with deep narrow EQ band and being able to apply it separately to each speaker is good way to verify that rig is doing what you are asking of it.

Do some measurements and share them here! Personally I find that saving sets of REW results as .mdat file for posting gives outside observer most info in least space. IMHO, anyone on this site that is interested in what others are doing will have REW running.

So, follow REW calibration procedures for sound card and for applying microphone calibration file.

For big spaces, lively spaces I recommend at least 256k length measurements, and working at 44.1kHz or 48kHz sample rates. And, REW beta has nice distortion measurement feature and ASIO sound card driver support.

And oh yes, what are basic dimensions of listening space, and which B & W do you have for mains?



32 Posts
This is an excellent walkthrough of what you can do with REW and JRiver to improve the sound of your two-channel system, from a blogger at Computer Audiophile.

Personally, I'd start with:
1. Optimizing sub placement.
2. Dialing in sub/mains integration.
3. DRC of sub/mains using JRiver convolution engine.

1 - 3 of 3 Posts