[Philm63]

Posted this question last month in the subwoofer forum and one suggestion was to post it here to see if this was something that could be solved in REW. Below are a couple of shots of 3 scans - just the sub; just the mains; and combined. The first with no smoothing, the second with 1/6 smoothing for clarity.

My question was (and still is...) why should there be a difference in the recorded SPL between the scans of the sub with and without the mains? Granted it's only a few dBs, but still, why are they coming in different? Is there actually enough energy coming from the mains at the lower frequencies (below the 80 Hz XO point) to effectively interfere with and have a negative impact on the measured output?

I've messed around with this for a bit now - did a bunch of scans this way and that way - played with phase tuning - level matching - and am still scratching my head. I know a few dBs is not really a big deal, but it's just that it's still as of yet unexplained, and the engineer in me can't let it go.

The measurements are with one sub and the LR mains. AVR volume is pegged at -30 for all measurements, Audyssey is ON (post calibration), Dynamic Volume and Dynamic EQ are both OFF for all measurements (yes; I double-checked each time - not gonna let THAT one fool me again!)

In the AVR; mains are set at 80 Hz, sub at 120 Hz - LFE only (no LFE+Mains in this house!) Large open living room (if this matters); >5000 cu ft. Using REW with HDMI through a Denon 4520 into Chane A5rx-c's and a PSA S3000i sub. (again, if this matters) UMIK1 used at the MLP for all measurements, calibrated (90 degrees), did not move during measurements. Noise floor was around 40-45 dB - AC off, wife/baby/dog in other room, doors closed, sun was shining. (again, if this matters)

Lou, aka Lumen (formally known as "Bluerockinlou") suggested there may be something regarding setting the levels using "Check Levels" in REW - I don't fully understand how this works, or if I may have inadvertently done something that gave rise to this being a contributing factor...

 

[jtalden]

Just Guessing:
The level of the SW alone appears to be 5-6dB higher than the other 2 measurements. 6db is the expected difference in level if for instance the SW is driven with a R+L rather than R or L alone. From your post it does not clearly indicate what channels were fed the signal and what mode the AVR was in. It also sounds like you are testing R+L mains together. This then does not lead me to any specific cause, but my thought is that somehow the signal routing/processing is the related to the issue.

As a test you may want to try each channel separately: L, SW, L+SW then R, SW, R+SW then L+R, SW, L+R+SW. Use stereo mode and feed the signal for each series to the correct analog inputs for each of the 3 series; L, R, L+R. Stay away from mono mode or any matrix mode. Do not send the SW signal to the LFE multichannel input input unless you adjust the level in the amp accordingly (-10dB for R and L and -4dB for L+R ?Maybe?). I am not really sure as I have never used the LFE input for measurements.

This is a good series anyway even if it does not locate the problem. It can help to highlight how well the phantom center is operating in your setup if you use stereo mode for music.

I hope this somehow helps identify the problem.

 

[Philm63]

It does add insight, thanks. I do run the AVR in Stereo mode for all such measurements which gives me a sub output during LR measurements, if I want it; otherwise I can just turn the sub OFF during measurements of LR.

But your point gives me pause; when I measure just the sub, I select Channel 4 in REW and it outputs only to the sub channel in the AVR. When I measure the mains, I select Channel 1 on the primary output, and 2 on the secondary output, and because the AVR is in Stereo mode it gives me also the sub output.

Could it be that, because the sub is playing during LR measurements only because the AVR is in Stereo mode, not because REW is outputting on Channel 4 (for LR it is outputting on Channels 1 and 2 only), that the level of the sub's output is different from that which would exist if I were outputting from Channel 4?

An experiment is in order, me thinks. Run a sweep using Channels 1 and 2 as if to measure just my mains, leave the AVR in Stereo mode, and unplug the mains so only the sub will play. Then run a scan of Channel 4 and compare the output levels. I'll report my results as soon as I can get some scan time (hopefully today).

 

[jtalden]

Yes, channel 4 in HDMI is the LFE. If it is much easier for you to use that LFE input rather than analog methods that is fine. Just adjust the AVR level accordingly. From your chart that looks like it is probably going to be -4dB for your current measurement scheme.

I would do loopback measurements of the Preouts and confirm the needed difference in levels just to be sure. That's probably not necessary however. Other here that use HDMI have probably have sorted through all this and can indicate exactly what settings are easiest to use.

 

[JohnM]

Implicit in some of the comments above, but just to be explicit: the LFE channel gets 10 dB gain applied in the processor.