Jumped into REW a few weeks back and after figuring it out was able to get a baseline. The numbers weren't what I was expecting (a bit disappointing) but I was optimistic that I could use the data to find better locations for the subs etc.
Fast forward to today -I was planning on testing again with REW and trying various sub locations. Before I moved subs around I tested my setup as I left it to confirm my previous measurements. To my dismay the measurements came back completely different. Nothing has changed since my original measurements, but the REW measurements are night and day apart. I do two sweeps for every measurement but bizarrely even measurements taken back to back show noticeably different results. Am wondering what I might be doing wrong...
See below for examples:
SPL: The robust lines were the new tests that were taken back to back, the weaker line was the original baseline. The newer plots seem more in line with what my ears would guess, but there's no way I'm putting out that much output below 10hz.
Waterfalls
New waterfall plots: Notice the differences here, these tests were literally taken less than two minutes apart. I even shut my fridge and lights off to make sure there was no sound pollution... yet the graphs are still noticeably different.
Original baseline
Spectrograms: I'm no expert but my new spectrogram is the worst showing I've ever seen. It's a large room that's carpeted, I can't fathom it's actually that bad.
Original baseline:
New:
Quick details on setup:
Room: Roughly 4-5k cubic feet due to open layout and vaulted ceilings
Mic: RadioShack meter, the older digital style -there's literally a pic of it in the REW help guide.
Sound card: UCA202
Subs: 2x PSA XS30se's
Note on REW process: Sound card was calibrated and returned normal results. Levels checked out before every measurement. Mic was on a tripod just over the main listening position. I work the laptop in another room so as not to disrupt the test in any way.
Any suggestions? Sorry for writing a book. Thanks in advance!
Check the cal files. I would re-do the loopback to verify that it is flat. See if the mic calibration is jacking up the low end like crazy by doing a sweep with the mic cal file loaded, with the loopback still connected.
The only other thing I can think of is clipping. I've gotten beautiful flat sub responses before when I put too much input on the mic. My interface wouldn't even let the signal clip, IT WOULD COMPRESS it when it got too loud on the input. It was a real blast to track down.
IMHO it's ok if you are in the same room with the sweep going. There won't be a noticeable diff as long as you aren't in the main part of the soundfield.
Check the cal files. I would re-do the loopback to verify that it is flat. See if the mic calibration is jacking up the low end like crazy by doing a sweep with the mic cal file loaded, with the loopback still connected.
The only other thing I can think of is clipping. I've gotten beautiful flat sub responses before when I put too much input on the mic. My interface wouldn't even let the signal clip, IT WOULD COMPRESS it when it got too loud on the input. It was a real blast to track down.
I confirmed the loop back was mostly flat, but can check again and post results. I didn't think about doing a sweep with the loopback still connected, thanks for the suggestion! As far as the clipping goes, are you suggesting the levels output from the speakers are higher than what the mic can interpret? What's the best way to determine if that's the case?
While my system is the first experience I have with any of this, I'm fairly confident it doesn't sound like the gold trace. I feel like I have strong bass throughout the range much like the other plots suggest. Of course I'm testing with REW to validate my ears and further optimize. Like many here I enjoy optimizing, I'm not doing this because I feel I have to -I genuinely enjoy it. If REW can start giving me consistent results I'll really be enjoying it.
IMO, REW and any other like program or measurements of any type are absolutely and completely useless. Does it sound good? If yes, enjoy your system. If it sounds bad, adjust, or exchange your equipment until you're happy. It's really that simple.
IMO, REW and any other like program or measurements of any type are absolutely and completely useless. Does it sound good? If yes, enjoy your system. If it sounds bad, adjust, or exchange your equipment until you're happy. It's really that simple.
Not to disrespect your opinion but if his system sounds like the gold trace on the frequency response graph, I would bet it doesn't sound very good.
REW is a tool that helps us tweak our sound systems to sound as good as they can, as fast as they can. It also helps us to literally see what we are hearing and IMO there is no better tool for training your ear or diagnosing problems in your system. Is it easier to build a wall with a tape measure or by eye?
You can use the "Start Delay" option to get out of the way, if you like.
Yes, REW can be flakey. There have been times when I've spent more time setting up than actually generating results. Play with all the cables and connections, check the battery in the meter, reboot the computer, make sure the volume is at 100%, try changing the input and output devices and increasing the buffers to 128k, and wave a live chicken over your head three times.
Yea REW being flaky crossed my mind... The one suggestion which confuses me is the volume at full. I actually need my computer volume (USB sound card volume if I'm being specific) set to 2 (out of 100) for the levels to be ok during the sound card calibration. Is that unusual? What's more bizarre is if I use the right output (or input I can't remember) the volume is so low I can't get the levels high enough even if I turn the computer volume up to 100... Using the left input/output resolve this issue but again I have to set the volume at 2 to be in check. Any suggestions?
Don't have time tonight to set the mic/tripod etc up, but I ran another sound check and did a sweep while the loop back was configured. See below for results. I think the calibration looks somewhat unusual, but it's fairly flat outside of the extremes. The sweep after the calibration file is in place isn't as flat as I'd like it to be though...
You guys crack me up btw, it's quite easy to route cables over to another room and sit in the comforts of a desk while running sweeps. No need to hide behind my couch with a laptop while testing. I'm pretty sure I'm not clipping the mic but I can test that when I have time to set everything up again. If anyone has any other suggestions I'm all ears!
Something ain't right. You're hot 10dB at 10Hz on your loopback. Are you using an interface's digital out to your AVR and then out of the L or R back into the front of the interface? Have you forgot to clear your mic's cal file while you're doing the loopback?
Run it next time on 5dB scale, all the way down to 2Hz, clear the mic cal file, clear the soundcard/interface cal file and let's see what the rolloff is of your AVR and soundcard/interface. Then just make a new cal file. Then with the soundcard/interface's cal file loaded, run it again and see if it is ruler flat.
To make a cal file you will have to use your left or right out of your AVR setting it for fullrange. You can't use your LFE out because you need it to go to 20k. After you make a new cal file that is ruler flat from 2 to 20k, you can then loop out of your AVR's LFE and see if the rolloff is about the same as the left and right out. There will be some variation but most likely it will be acceptable.
You have to get your cal file as ruler flat as possible in the whole band so that you can trust your measurements.
Are you using an interface or usb mic or soundcard? You need to monitor levels on whatever hardware it is as well as in the software that nothing clips but that your levels are robust enough to trust.
REW is not flaky. I've been measuring with mics and with line level signals for years and it never lets me down. There is just a learning curve involved that you have to get over first.
As an added bonus see below for my frequency sweeps. Lower sweep is post Audyssey, higher plot is 3.5 db hotter in the sub (as set in the receiver). System is now currently set just 1db hotter than Audyssey intended (currently -5.5 in the receiver) and it it does a nice job of reducing the dip between 10-20hz while remaining flatter throughout the range than the 3db hot chart. Sorry don't have a plot of the 1db hot setup as I ran out of alone time in the house and couldn't get a clean (quiet) sweep so I didn't bother saving the result. Passed the goosebumps test so I'm happy!
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Related Threads
?
?
?
?
?
AVS Forum
34M posts
1.5M members
Since 1999
A forum community dedicated to home theater owners and enthusiasts. Come join the discussion about home audio/video, TVs, projectors, screens, receivers, speakers, projects, DIY’s, product reviews, accessories, classifieds, and more!