Difference Between Avia and DVE - Subwoofer Calibration (1 Viewer)

[Rick Cohen]

Does anyone know how the THX Optimizer test tones compare with these more complicate disks?

Perhaps Ed if you are in a mood to do a bit more testing you could provide some insights regarding the THX Optimizer tones.

RICK

 

[Jeremy Anderson]

Okay, I asked a friend of mine who does audio production, and he told me I've been analyzing the digital data incorrectly. I had PEAK HOLD enabled, wasn't using C-weighting, and wasn't averaging (which he says I should do for pink noise). The following numbers should accurately reflect the behavior of a SPL meter.

So using his method, here are the total power numbers:

Avia:
Front left, -20.15 (or 84.85dBC)
Front left sub tone, -22.39 (or 82.61dBC)
Therefore, with Avia you should be setting each channel to 85dB on your meter and the subwoofer to between 82-83dB. In other words, these are accurate tones if you account for C-weighting (which has been common knowledge).

DVE:
Front left, -31.06 (or 73.94dBC)
LFE tone, -35.11 (or 79.89dBC, since LFE's max value is 115dB, not 105dB like the other channels)

Here's the thing of note: According to this test, DVE's LFE tone should theoretically read 2dB lower than Avia's sub tone on a meter at reference level. See how much of a difference Edward got between DVE and Avia above? That's right... 2dB! Coincidence? I think not.

In short, Edward's correct. DVE does not compensate for the 10dB boost that the LFE channel gets. I think we can pretty much lay this debate to rest. DVE is apparently just WRONG. I have no idea why my system reads normally, but the digital data doesn't lie. Now, I'm going to go check everything I can think of on my gear until I figure out why my system reads so strangely.

 

[Jeremy Anderson]

Being the anal guy that I am, I just swapped the leads to my subwoofer (hence reversing phase). I then ran my 1/6 octave tones and re-equalized for flat response. Here are the numbers I get now at reference level:

Avia: 85dB speakers, 83dB sub.

DVE: 74dB speakers, 81dB sub.

So apparently there was some strange phase issue with my sub that was causing my system to read incorrectly. With my EQ set for flat response, DVE is 2dB lower than Avia... which matches both Edward's results and the corrected digital analysis of DVE's tone.

I think that pretty much seals it. Final verdict: DVE's sub tone is encoded 8dBC too hot.

 

[Brian L]

Ed/Jeremy,

To start with, great work on this. I was back there with you in March when this all came to light, but had no way to test as thoroughly as you guys did. Glad to see that first off, I was not on crack, and more importantly, my room and gear are not telling lies!

On a side note, do either of you guys remember any of the threads pertaining to universal players that have the .1 channel output about 10 dB too low? Lots of reports on that from users with many different universal players. My Pioneer 45a was one of the 1st.

So here's me theory on the DVE LFE channel problem.

Joe Kane used universal player decoder design as the target player type for DVE. As such, he is subconsciously throwing all those users a bone by authoring a set of test tones that, when played through the decoders of those players, will work fine.

Oh all right, its not likely. But, I can say with some certainty that any user that has such a player can rest assured that DVE will be the ticket to balance the 5.1 outputs, LFE channel included.

BGL

 

[Philip>L]

All of the files on my disc are from August 14, 2003 also.


I can get my Pioneer 563A univeral player to play the LFE test tones on DVE or the internal test tones anywhere near as close to the same SPL as the mains. When the mains are turned as low as they'll go and the LFE is turned up as high as it will go, the level is still 20db less in the sub than the mains.

But when I process them through my Yamaha receiver I have no problem at all. Well, except for the problem with the test tone in the first place...

 

[Edward J M]

More important to the lay user without the capability to analyze and subsequently EQ for a flat response, what are your Avia and DVE readings with the EQ disabled?

If I was a bettin' man....my guess for DVE would be 86-87 dB, or 12-13 dB hot.

Regardless, that's a wrap, guys. A huge thanks to Jeremy for gutting it out and correcting the methods for his digital analysis of the DVE signal. No need to swap discs now, and our data sets and results concur 100%.

Final Summary:

The DVE subwoofer calibration LFE tone is encoded exactly 10 dB too hot, and does not account for the fact that all pre/pro's boost the LFE channel 10 dB.

Due to its extra-wide bandwidth which contains strong content to below 20 Hz, the DVE tone will read about 8 dB too hot on a C-weighted SPL meter. If you have an unweighted mic, it will read 10 dB too hot. This is OUTSIDE with no room gain or boundary effects considered.

For in-room calibration, anything ABOVE 8 dB hot (with a C-weighted meter) or 10 dB hot (with an unweighted mic) can be attributed to room gain.

If you have a sub that can dig strong to below 20 Hz, and your room is small with lots of room gain below 30 Hz, DVE can/will read as much as 12-16 dB high with a C-weighted meter.

DVE is now firmly on my "not recommended" list for subwoofer calibration. DVE can still be used for surround channel calibration, but what's the point if you need to look elsewhere to calibrate the sub? Also, DVE is a navigation nightmare. IMHO, DVE is in need of a wholesale revision; after such a long gestation period, what a disappointment.

Because it is encoded at the proper level, and because it contains a sharply truncated 35-80 Hz bandwidth which will not be affected by room gain, Avia remains my recommendation for basic subwoofer and HT calibration.

Because Avia uses RB, I still recommend checking the sub level in all 5 channels. Select the channel which seems most representative of the average, and use that channel for final sub calibration. With Avia, go with 83-84 dB for a flat sub calibration, and 86-87 dB if you like things a little on the hot side.

Regards,

Ed

 

[John Tami]

Tim....

If you'd read all through this thread you'd see that the S&V HTT Disk is NOT recommended (By Ovation) for Sub Cal. Period. (we had this discussion at The Spot couple months ago..Ovation was contacted and that was their answer...)

And maybe mute at this point, but the DVE disc is only $16.76 (something like that) at deepdiscountdvd.com


Ed...I only have those two discs...I really don't feel like buying a third Cal Disc (Avia) for just the Sub. I have a 20x15 room, slight cathedral ceiling, open corner. No way I can use the DVE with correction values?

Maybe my next receiver will have sub tones? (August)

What BS...anyone talk to DVE yet?

 

[Jeremy Anderson]

If I turn off the EQ and set master volume so Avia reads 83dB, DVE reads 86dB now. Go figure.

 

[Edward J M]

Sure - an educated guess in that size room would be "around" 12-13 dB hot as read on the c-weighted meter. So look for 87-88 dB, depending on how deep your sub digs, and how much room gain you're experiencing.

An even better way might be to burn some test tones in the 200-20 Hz bandwidth and run a sweep with the mains and the sub and plot the FR with the RS meter correction factors. Then you'll know for sure what the sub is doing in-room.

Ed

 

[John Tami]

Should have mentioned...20-39 PCi. Thanks!

 

[MuneebM]

This is a very interesting thread, and I was also one of the first to discover the strange anomaly with my copy of DVE compared to Avia. In my set up, DVE's LFE test tone is 12-13 dB hotter than the surround channel tones, i.e. 87-88 dB vs. 75 dB.

I still prefer to use DVE for my 6.1 speaker level calibrations, but then switch to Avia for subwoofer calibration. However, what bothers me about the Avia test tone for the subwoofer is the fact that the needle on my Rat Shack SPL meter sways plus/minus 2-3 dB with the Avia test tone, whereas it's rather stable with the DVE LFE test tone. How do folks on this board compensate for the swings? Do you take the measurement based on the middle of the low and high swing?

http://207.107.210.136/sigserv/pl/index.pl?p=131
My DVD Collection

 

[Edward J M]

Set the meter to slow, and also to the next higher dB scale.

If you are using Avia, set the meter to the 90 dB scale and read in the negative section of the display. So -5 would be 85 dB if you are using the 90 dB scale.

Using the next higher scale creates much less fluctutation in the needle since the scale is always compressed below the 0 mark.

 

[Ilkka R]

Do you have your RS meter set to SLOW setting? It should tame those swings. Take the average if you already have it at slow.

Ed already said it.

 

[MuneebM]

Yes, I definitely always have it set to SLOW, in fact, I've never even used the FAST setting on my meter.

As for the subwoofer reading, it would be impossible to get a reading of 82-83 dB with the meter set at 90 because it's hard to tell when the needle is at -7 or -8. I set my subwoofer to a level when the meter reads 82-83 dB to compensate for the C-weighting.

 

[Ilkka R]

Yes, 82dB is the right level if you don't want it "hot". And I agree, it's very difficult to read it if using 90dB range. I use 80dB range.

 

[Brian L]

I would agree that the RS Meter, even set to slow, makes proper interpretation hard with certain test discs.

The most stable readings I have ever seen are with pure sine waves. Any tone that has a warble built in will be hard to correctly interpret with a RS meter.

I also have had reasonably steady readings using the Infinity RABOS Kit (Meter and CD). Either the meter is more heavily filtered or the tones have less warble, but they work well together.

But for absolute, dead nuts steady tones, you need sin waves.

BGL

 

[Ilkka R]

Actually the RS meter is very stable, but those test signals aren't. I've measured (with SpectraPRO) the levels of Avia test signals (subwoofer level) and they fluctuate atleast 2-3dB. So they are not stable. Also what makes accurate reading difficult is that the signal move from sub to other speaker too quickly. If your sub is calibrated very cold or very hot (more that 3-4dB) the needle is still moving when signal goes already to speaker. That causes errors.

Sinewaves are not good with calibrating unless you have a dead flat FR. So not suggested.

 

[Chase_Smith]

Hey guys. I hate to bump an ancient thread, but it has been my all-time favorite resource when it comes to subwoofer calibration.

My question with Avia has to do with choosing the channel to set the subwoofer with.

Each surround channel interacts differently, so I get different level readings depending on the channel.

Also, each speaker already contributes SPL to the subwoofer reading, so I was wondering if you considered my idea to have any merit:

My subwoofer is corner-loaded next to my front-left main channel. So, to try and isolate the subwoofer, I unhooked my front-left speaker so that I was only getting a reading from the subwoofer during it's portion of the test tone.

My speakers are calibrated to 85db when my receiver is set to 00, so I calibrated the subwoofer by setting the receiver to -3 on the dial, so that "0" on the SPL meter would be 83db, and the center point for calibration. I calibrated the sub level so that the needle bounced back and forth equally on both sides.

Sound right, or does the main channel information have importance in the calibration?

 

[Edward J M]

Avia is a Dolby Digital format, and relies on redirected bass in order to calibrate the subwoofer level. This requires the speakers to be set to small, and the speaker channel in question will then contribute bass to the subwoofer tone.

The amount of bass each speaker contributes to the sub tone is highly dependent on its room location, its inherent bass capabilities, its proximity to the subwoofer, the phase setting of the subwoofer, and the crossover frequency selected.

The best way to calibrate the subwoofer with Avia is to run the sub test tone for each speaker channel and note the results. Pick the channel that seems most representative of the average, and use it for final sub calibration. This method will tend to eliminate over or under calibration by disregarding the highest and the lowest values and using the average.

Using the left main for subwoofer calibration will likely result in undercalibration because that speaker is very close to the sub and you'll be getting good acoustic coupling. Try the center channel and the right main and note the results.

Avia will "get you close", but final calibration of the subwoofer is a bit subjective. If the sub seems a little too hot or too cool, then adjust accordingly - but I would't stray more than 3-4 dB in either direction.

Also, a good way to evaluate the effect of the phase control when using Avia is to run the sub calibration tone for the speaker closest to the subwoofer and try the phase at 0 and then at 180. Observes what happens on the SPL meter. Usually one setting will be considerably higher than the other, indicating the subwoofer and the nearest speaker are the most "in phase" at that phase setting. This may not be completely desirable from a smooth FR standpoint, but it can help you understand the effect of the phase control, since Avia does rely on redirected bass.

If you really want to see what is happening with your system, plotting the FR is more revealing than using a calibration DVD (even on as good as Avia). You could be calibrating to a peak or a null in the response, and this can affect the calibration level, obviously.

DVE can be used for sub calibration if you account for the 10 dB error in the subwoofer tone, and also account for room gain (if you have any) since DVE gets to 15 Hz. DVE doesn't use redirected bass - it uses the LFE channel - so it might appeal more to what you were trying to accomplish above. DVE also has 6.1 capability for the rear surrounds, whereas Avia doesn't.

 

[Chase_Smith]

Do you think using the Center Channel tone would be the best bet to get a good average for the system?

Also, can you answer my question about playing the subwoofer test alone, without the corresponding speaker being hooked up?

What I mean is, say I were to unplug all my speakers, and run through the test tones. The level would read the same for all channels, right? I know the speakers interact differently, but in theory the subwoofer should be outputting the same signal, right?