[BobearQSI]

I've searched through the forums here and quite a few online guides, but I can't seem to find an answer to my specific question.


I have the AIX Blu-ray calibration disc I use to calibrate speaker levels using an SPL meter. I use this because the primary application of my theater is Blu-ray movies, and it gives different measurements compared to the amp's built-in pink noise, using either of the methods from my subwoofer question below.


So, my question is this: When the AIX test tones are playing through the front/side/rear speakers, I notice that the subwoofer is also making a noticeable amount of noise (my crossover is set at 80Hz). Should I be turning the powered subwoofer off when measuring levels of the satellites, or leave it on?


Thanks.

 

[BoogerBrains]


Speaker SPL calibration should be performed on each speaker separately. There should be no sound coming through the sub when any of the other speakers are being leveled.

 

[craig john]



Quote:
Originally Posted by BobearQSI  /t/1519911/speaker-level-calibration-using-test-disc#post_24410126


I've searched through the forums here and quite a few online guides, but I can't seem to find an answer to my specific question.


I have the AIX Blu-ray calibration disc I use to calibrate speaker levels using an SPL meter. I use this because the primary application of my theater is Blu-ray movies, and it gives different measurements compared to the amp's built-in pink noise, using either of the methods from my subwoofer question below.


So, my question is this: When the AIX test tones are playing through the front/side/rear speakers, I notice that the subwoofer is also making a noticeable amount of noise (my crossover is set at 80Hz). Should I be turning the powered subwoofer off when measuring levels of the satellites, or leave it on?


Thanks.

Does your receiver have Audyssey?

 

[BobearQSI]

Yes, however, I don't use it for levels. The reason is, I get different levels from Audyssey, the AVR's internal pink noise, and the Blu-ray test disc pink noise (all within 3db, though). I'm guessing the AVR's pink noise generator isn't going through the EQ processing, which may be why I see this difference.


So I'm using the test disc after running Audyssey, because this provides a full path from my most common usage of the system: Disc->Player->LPCM->AVR->Audyssey->Speakers.

 

[mogorf]



Quote:
Originally Posted by BobearQSI  /t/1519911/speaker-level-calibration-using-test-disc#post_24443746


Yes, however, I don't use it for levels. The reason is, I get different levels from Audyssey, the AVR's internal pink noise, and the Blu-ray test disc pink noise (all within 3db, though). I'm guessing the AVR's pink noise generator isn't going through the EQ processing, which may be why I see this difference.


So I'm using the test disc after running Audyssey, because this provides a full path from my most common usage of the system: Disc->Player->LPCM->AVR->Audyssey->Speakers.

I fear some form of over thinking here. As long as you are within 3 dB no matter how you measure, you are good to go.


BTW, how is your issue of sound coming thru the sub? That should be solved, indeed.

 

[craig john]



Quote:
Originally Posted by BobearQSI  /t/1519911/speaker-level-calibration-using-test-disc#post_24443746


Yes, however, I don't use it for levels. The reason is, I get different levels from Audyssey, the AVR's internal pink noise, and the Blu-ray test disc pink noise (all within 3db, though). I'm guessing the AVR's pink noise generator isn't going through the EQ processing, which may be why I see this difference.


So I'm using the test disc after running Audyssey, because this provides a full path from my most common usage of the system: Disc->Player->LPCM->AVR->Audyssey->Speakers.

You've nailed the reason I asked the question. The levels are different when using the external test disc because they are going through the Audyssey EQ filters while the internal test tones do not, even after running Audyssey. Your process using the external test tones is the *right* way to do it. However, you should shut the sub off while running the tones through the speakers.


Craig

 

[mogorf]



Quote:
Originally Posted by craig john  /t/1519911/speaker-level-calibration-using-test-disc#post_24444596


You've nailed the reason I asked the question. The levels are different when using the external test disc because they are going through the Audyssey EQ filters while the internal test tones do not, even after running Audyssey. Your process using the external test tones is the *right* way to do it. However, you should shut the sub off while running the tones through the speakers.


Craig

Craig, actually the internal test tones, despite the fact that the test tone block is after the Audyssey block (as Chris K. once explained), thus lack frequency response equalization in the 500 Hz - 2 KHz pink noise range, yet do run through with Audyssey's level adjustments engaged. AFAIU.

 

[craig john]



Quote:
Originally Posted by mogorf  /t/1519911/speaker-level-calibration-using-test-disc#post_24444637


Craig, actually the internal test tones, despite the fact that the test tone block is after the Audyssey block (as Chris K. once explained), thus lack frequency response equalization in the 500 Hz - 2 KHz pink noise range, yet do run through with Audyssey's level adjustments engaged. AFAIU.

Yes, the level adjustments, (aka the level "trims"), are engaged, but the boosts and cuts of the EQ filters are not. If Audyssey has implemented a wide cut of 6 to 9 dB in the range of the test tones, the average level read by the SPL meter will be quite different with the Audyssey filters engaged or not engaged.

 

[mogorf]



Quote:
Originally Posted by craig john  /t/1519911/speaker-level-calibration-using-test-disc#post_24444695


Yes, the level adjustments, (aka the level "trims"), are engaged, but the boosts and cuts of the EQ filters are not. If Audyssey has implemented a wide cut of 6 to 9 dB in the range of the test tones, the average level read by the SPL meter will be quite different with the Audyssey filters engaged or not engaged.

Thanks for your feedback Craig, yet, in my u'standing the boosts and cuts are most typical in the low frequency end of the spectrum (below Schroeder, mostly below 300 Hz-ish), while the range of the pink noise test signal (500 Hz - 2 kHz) is not "endangered" by such phenomena. The average level read of the SPL meter in this regard should/will remain intact whether Audyssey filters are engaged or not. Again AFAIU.


Nice to talk to you!

 

[craig john]



Quote:
Originally Posted by mogorf  /t/1519911/speaker-level-calibration-using-test-disc#post_24444748


Thanks for your feedback Craig, yet, in my u'standing the boosts and cuts are most typical in the low frequency end of the spectrum (below Schroeder, mostly below 300 Hz-ish), while the range of the pink noise test signal (500 Hz - 2 kHz) is not "endangered" by such phenomena. The average level read of the SPL meter in this regard should/will remain intact whether Audyssey filters are engaged or not. Again AFAIU.


Nice to talk to you!

Here is a post I made about this. I tested 2 different test discs and the internal test tones. The results and explanation are in this post:
http://www.avsforum.com/t/1165099/official-jtr-speaker-thread/15870#post_24320284


And you are right... the biggest differences are in the subwoofer range.

 

[BobearQSI]



Quote:
Originally Posted by mogorf  /t/1519911/speaker-level-calibration-using-test-disc#post_24444176


I fear some form of over thinking here. As long as you are within 3 dB no matter how you measure, you are good to go.


BTW, how is your issue of sound coming thru the sub? That should be solved, indeed.

Because, why not get it within 0.5db when I have a meter and it only takes 10 minutes. I probably wouldn't go get a meter for this, but I got one for Christmas a few years back.


The test disc plays full range pink noise which is why the sub was engaging, so I just need to hit the power switch on my sub when measuring.

 

[mogorf]



Quote:
Originally Posted by BobearQSI  /t/1519911/speaker-level-calibration-using-test-disc#post_24445675


Because, why not get it within 0.5db when I have a meter and it only takes 10 minutes. I probably wouldn't go get a meter for this, but I got one for Christmas a few years back.


The test disc plays full range pink noise which is why the sub was engaging, so I just need to hit the power switch on my sub when measuring.

The meter has an accuracy of +/- 2 dB, just the same as the Audyssey commercial mic. As regards the test signal, it should be band limited (0.5 - 2 kHz) pink noise signal instead of full band. For the sub: 40 - 80 Hz band limited, with calibration data for correction of the SLP at low frequencies. Hitting the power switch on the sub won't solve the problem. Hope this helps.

 

[BobearQSI]



Quote:
Originally Posted by mogorf  /t/1519911/speaker-level-calibration-using-test-disc#post_24445846


The meter has an accuracy of +/- 2 dB, just the same as the Audyssey commercial mic. As regards the test signal, it should be band limited (0.5 - 2 kHz) pink noise signal instead of full band. For the sub: 40 - 80 Hz band limited, with calibration data for correction of the SLP at low frequencies. Hitting the power switch on the sub won't solve the problem. Hope this helps.

Thanks, this is all good info and eye opening. My meter's specification claims +/- 1.5 dB accuracy, although I doubt 0.5 dB of improvement would even be noticeable. I'm seeing up to 3 dB difference, but it sounds like the AIX test disc doesn't have proper signals. If I had to guess just by listening, the receiver tones sound like they are correctly band limited and the AIX disc does not.


I do have the Spears and Munsil HD Benchmark 2nd edition which has properly band limited signals, I might try those and compare to the receiver's internal tones, and see if the difference goes away.

 

[jdlynch]

I use a Marantz 8801 with Audyssey XT32. Is it being suggested here that I should go back, after the Audyssey set-up, and check my trim settings with Audyssey room correction engaged? I already own an AIX test disc and an SPL meter.

 

[BoogerBrains]


The following is a procedure I have used with success many times. Maybe it will work for you:

 

The fun part

1. Find the best location possible for the speakers - for the mains this is fairly straight forward. for the surround and sub(s) you may want to test them a bit.

 

The Really fun part

2 Treat the environment directly surrounding the speaker (behind and below usually) with sound absorption / isolation.

 

The really really fun part

3. Add processing to the speakers (EQ normally. Sometimes the automated EQs work, but ALWAYS test against your ear with familiar music or program material. In the end, always go back to your ears). You want processing to be in the signal chain during the speaker leveling. If you level the speakers dry and there is a significant curve, you may defeat all your hard work once you are playing back music or a movie through the processing.

 

The exciting part

4. Level the speakers:

    Play band-pass limited pink noise through the system. I have found one octave (500-1k or something like that) works well for music and two octaves (500-2K) for movies. (if you don't know what you have, that is not going to be a big deal. In a crunch I have also tuned systems with full range pink noise, but this is not ideal). Start with playing the band-passed pink noise through the FL, FR, C speakers one at a time. Position the SPL meter at the listening location and at the height of the ear. If you are using a directional SPL meter I have found pointing the meter directly at the speaker you are testing works well. If you are using an omni mic you can just have it point to the screen during the entire test. Ideally you would mount it on a tripod. The Dolby Digital and DTS theater spec is 85 dB SPLC for each of the front speakers (this assumes the pink noise is playing at -18dB FS - this will provide an individual speaker peak of 103 dB SPLC at 0dB FS). This maybe really loud for normal listening - be careful not to blow your speakers either (you could start at 75 or 80 dB and work your way up or down to a comfortable listening level or your normal listening level.)

    Surrounds are to be levelled at 82 dB SPLC (again through band pass limited pink noise), but I tend to goose them up a bit to 85 (that's just my preference).

    The sub is a bit trickier. You HAVE TO band-pass limit this pink noise signal - most likely the disc already is and so is any tone generated by the processor.The tricky part is this: There is a 10 dB gain boost as part of the ".1" channel in a dolby or DTS environment (Sony's and Meyers's may also have this, but I have never set up either of them) - this means that you tune your sub to 95 dBC instead of 85 dBC IF THERE IS NOT PROCESSING. Most consumer and home theater processors do this automatically, so I would start with the 85 dB SPLC level and work your way up if you want more. 

 

The Successful part

Once this is all done, play excerpts of lots of different material with which you are familiar. You may find yourself thinking the sub is too loud or something like "the left surround speaker is not loud enough" - trust your ears. If it does not sound like how you want it, adjust it - you can always retune the system.

 

I'll see if I can scrounge up the Dolby and DTS specs as a reference.

 

[BobearQSI]



Quote:
Originally Posted by jdlynch  /t/1519911/speaker-level-calibration-using-test-disc#post_24463628


I use a Marantz 8801 with Audyssey XT32. Is it being suggested here that I should go back, after the Audyssey set-up, and check my trim settings with Audyssey room correction engaged? I already own an AIX test disc and an SPL meter.

I have reason to believe the AIX test disc does not play band limited pink noise. My sub crossover is at 100Hz, and I hear it pretty well when playing the AIX tones, which means the test tones are likely not band limited at 500Hz.


And as others have pointed out, meters' accuracy is around 2 dB, so fine tuning probably won't help much. You could, however, at least measure with a test disc and meter to make sure the levels are within 2-3 dB, just to verify the Audyssey worked correctly. But you should use a disc with proper band limited pink noise to do this, not the AIX disc.


And I second the notion to use your ears. I personally calibrate my sub and surrounds to 85 dB and the fronts to 82 dB, because that's how I like it when the volume is set to 0db. My system's not that high end and I think the front spearkers which carry more of the sound are too harsh at 85 dB.


I just watched Hunger Games: Catching Fire this weekend in 7.1 and it makes much heavier use of the surrounds than most stuff. It's one of my favorite mixes. Another of my all time favorites for the sound is the Drag Me To Hell Blu-ray

 

[BoogerBrains]


The band limitation is a bummer, but won't prevent you from setting up a good system. And it sounds like you have a good grasp of what you are working with as well.

 

The 2-3dB or the +/-1.5 dB as mentioned in the string is not really a problem because you are adjusting the levels to match the other speakers in your system. The 2-3 dB or +/- 1.5 dB of accuracy usually refers to the accuracy of the meter when compared to the physical reference. But, the meter should give a very consistent reading between uses, which is the most important part of setting up speaker levels to match each other - not the actual reference level.

One strange thing, which has been mentioned, is that pink noise and white noise "huff" or "breathe". This will cause a fluctuating dB level. I have seen anywhere from 1 to 4 dB of fluctuation (depending on the band limiting). 

One way to mitigate this is to find a SPL level with pink noise and then change to a single (or multiphonic) tone(s). Then use the level of the tone as a reference for the remainder of your system leveling. You will always need to go back and verify the speakers sound even with pink noise and program material. One caveat to this is that if you use a tone, and that tone matches a prominent room mode, your readings will not be reliable (note that if the tone does NOT match a room mode, the readings will be very reliable). If the tone matches a room mode, the tone will be louder or softer (sometimes by +/- 6 dB) YIKES! So, if you use the single tone method you will need to use many tones which are not harmonically related - your source test disc may or may not have the tones needed. You can always try it and hear if you like the results.

I think some of the fun of setting up all these systems is that no two are the same. We (as audio peoples) follow guidelines, know how to use the equipment, but then ultimately need use our ears and play with positions and equipment settings to get a good sound. (this happens with systems from the home theater to movie theaters, to stadium systems.)

 

[craig john]



Quote:
Originally Posted by jdlynch  /t/1519911/speaker-level-calibration-using-test-disc#post_24463628


I use a Marantz 8801 with Audyssey XT32. Is it being suggested here that I should go back, after the Audyssey set-up, and check my trim settings with Audyssey room correction engaged? I already own an AIX test disc and an SPL meter.

I'm certainly not suggesting that. I'm suggesting you simply use Audyssey's settings. I know of no reason Audyssey would have gotten them wrong. The goal of an accurate calibration is to set the levels of all the channels to be exactly the same at the primary listening position. Audyssey uses the same mic, in the same position, to measure the response, and then it uses the same correction algorithm on all the channels. So, if the goal is to get an accurate calibration, Audyssey has checked all the boxes.


However, what often happens is that, after an Audyssey run, guys pull out their SPL meter and play the receiver's test tones to measure the SPL's of the various channels, checking to ensure Audyssey got the levels right. When they see that the SPL meter readings are different, they assume Audyssey got it wrong, and their SPL meter is correct. But what they haven't factored in is that the receiver's internal test tones are not exposed to the FR changes Audyssey has made with it's EQ filters, so they are *bound* to measure differently. That doesn't mean they are "correct" when Audyssey is re-engaged. In fact, they are actually wrong.


Another factor is that Audyssey uses frequency sweeps and looks at the SPL of individual frequencies within the sweep inside a very short time window. That is a completely different test signal than the bandwidth limited "noise" tone the receiver uses. It's also a very different way to examine the SPL. Audyssey "calculates" the average SPL of all the frequencies within the bandwidth it examines. It is impossible to "calculate" the average levels the Audyssey-way using a "noise" signal and a simple SPL meter. Even using external test tones, and making sure the Audyssey filters are engaged, is not going to perfectly duplicate the Audyssey "paradigm." It will be a better approximation, but it won't be the exact same result.


So I'm suggesting... don't even bother. Just leave the levels set where they are. There are other things to tweak after an Audyssey calibration and EQ, (crossovers, subwoofer distance settings, etc.), but the relative channel trims... are not one of them... in spite of what your SPL meter is telling you... whether you're using the internal test tones, or some external test tones, or whatever...


Craig

 

[BoogerBrains]


Good info Craig. I have not used Audyssey, and given the way it works from how you describe, the conventional methods of speaker levelling may be counterproductive - as you mention. Thanks for the good info.

 

[jdlynch]

Thank you Craig for the explanation.