Attention all Denon 3803 Owners - Major Bass Management Flaw (1 Viewer)

[Edward J M]

Thanks Mike. I'm not ready to throw in the towel just yet. The AVR is otherwise functioning perfectly in all respects and provides absolutely killer sound in all formats.

I have a friend with a new 3803 and I will be running the same test on his system shortly. If his 3803 acts differently, I'll have my answer. If it acts the same, I'll have some head scratching to do.

Not all of this really adds up just yet. For example, despite the fact that Avia is Dolby Digital, is it possible this fixed 8 dB LFE deficit has anything to do with the fact that most DD and DTS soundtracks have the LFE channel recorded 10 dB louder than the rest of the soundtrack? You would think Avia accounted for this when mastering the disc, but what if they didn't?

Also, if I was truly operating at an 8 dB LFE deficit at all times (as Avia indicates) regardless of the sub level setting, why is my bass excellent at my current setting of Mains -3 and sub 0 with sub calibrated 3 dB hot? And why is it equally bad at sub setting -10 and the sub calibrated 3 dB hot? You would expect terrible bass at any setting, but it simply isn't true.

At the current setting of mains -3 and sub 0 and 3 dB hot, the bass is fabulous in all respects - tone, texture, detail, power, dynamic range, and consistency - all excellent. I'm hitting bass peaks anywhere from 108-112 dB at the couch at or near 00 Master Volume, and the bass seems perfectly balanced with the other speaker volumes. Adding 8 dB to the LFE track would simply overwhelm the bass response - so this just doesn't add up either.

Case in point, the sock explosion in Monster's Inc. gives me 112 dB at Master Volume -5. If true Reference Level (according to Avia) is +5 on my Master Volume control, the sock explosion should give me 122 dB at Reference Level, which seems just about right. That sure doesn't seem like an 8 dB LFE deficit to me does it? Of course I won't try that since it would likely damage the sub, but it still adds up upon playback at the SPL meter.

Regarding the internal test tones, and my speakers, and their respective distances, here is the breakdown:

LF: -3 Polk RT800i 11 feet
C: -4.5 Polk CS400i 10 feet
RF: -3 Polk RT800i 11 feet
RS: -3.5 Polk f/x 1000 6 feet
RSB: -3.5 Polk CS245i 6 feet
LSB: -1.5 Polk CS 245i 6 feet
LS: -1.5 Polk f/x 1000 6 feet

These results didn't seem that unusual to me, but I agree the settings are on the cool side for 75 dB, which would mean that the test tones from my 3803 are on the hot side. Especially since Avia at the same volume only bumped up 5 dB to 80 dB instead of the expected 10 dB jump to 85 dB. The only thing I concluded here is that the 3803 test tones are not 30 dB below Reference Level, they are more like 25 dB below Reference Level. That's assuming Avia at 85 dB is 20 dB below Reference Level as they claim.

When I test the other local 3803, I'll report back with more data. Just wanted to give you something to chew on.

Regards,

Ed

 

[Jonathan M]

Ed,

I do not own the 3803, nor have I heard it or played with it, so this post is somewhat off-topic - I apologize for this in advance.

However I believe there is a great deal of doubt surrounding bass management with a great many (if not ALL) surround processors. Many people do not KNOW for themselves what is actually going on with their processors, and what should be going on. Case in point is the recent Outlaw debarcle with the discovery of a bug in it's processing some 2 years after it's inception. Surely if people were more knowledgable as to what should be happening, this would have been picked up much, much earlier. I would not be surprised in the least to find out that the 3803 is doing something in a different way than one would expect (whether it is a bug or not is a different issue).

There is very little information available with regards to the f3, slope, level, and alignment of the filters used in the bass management of today's processors. Most of the information that I have been able to find is based on what is easily found in the manuals etc. This is airy-fairy to say the least. (ie they'll say something like "The bass signal below 80Hz is routed to the subwoofer or mains").

I am currently developing a system to test the bass management of the receiver of my choice, the NAD T752. It basically involves a bunch of MLS signals on a Dolby Digital WAV CD, and a DTS CD. You play the test signals and record them (via the pre-outs) via the line-in on a computer. A simple FFT conversion then allows you to get the frequency response of each channel, from which the f3, alignment, relative levels, and slope info can be easily observed. The advantage of this method is that it allows you to compare what is happening in ANY of the bass management modes without the vagarities of speakers being involved. You can compare channel levels, bass levels, phase issues, and everything. It is also relatively easy to do the tests.

I am currently evaluating how successful this method will be, and will post a fresh thread once I've done some tests. If you (or anyone else) wishes to test out the 3803's response, feel free to contact me and I'll hook you up with what's needed. Also, if anyone has any ideas as to what could be included, feel free to contact me.

My suggestions to you (Back on topic now is to test the other local machine in the same way and see if it responds the same. My guess is it will, as corrupted firmware data would likely cause far more problems than the one's you are seeing.

Cheers,
Jonathan

 

[Mike Up]

Jonathan M,

see if it responds the same. My guess is it will,

Just out of curiosity, why do you think it will perform the same when a couple of us already confirmed ours do not perform that way?

Seeing that mine does not, I surely wouldn't expect him to be that unlucky in finding another defective unit.:frowning:

 

[Edward J M]

Jonathan:

That is a gracious offer and I may take you up on it pending the results of testing the other local 3803. It sounds like an interesting project and I wish you success.

Mike:

One variable is that you have not used the Avia disc, but instead used two other discs. Given the overall excellence of the big name calibration discs on the market, we should be able to eliminate that as a variable, but I still want to use the same Avia disc on that other 3803 before I draw any firm conclusions. You will note further up this thread, a 2803 owner experienced the same results with the Avia disc. I agree, unless the problem is endemic, my chances of finding a second corrupt 3803 should be nil.

I have also emailed Avia (Ovation) with my results and asked their opinion.

To anyone still reading this thread - if you have the Avia disc and a 3803 (or even a Denon "XX03" AVR), please run the 6 channel (i.e., 5 channel low passed bass plus LFE channel) panning test in Chapter 7.4. Avia claims all channels should be the same SPL when the speaks are set to Small. Make sure your LFE Level Control (available only in DD or DTS playback) is set to 0.

Regards,

Ed

 

[EricSm]

I just ran a quick 6 six channel pan on the Avia disk. And while the test wasn't exactly scientific, it gives a clear impression of what was going on with my 1803.

I have my system calibrated to 75 db across the board.

Run #1:

Sub Level: 0db

LF High Passed Bass Level: 75 dB
C High Passed Bass Level: 75 dB
RF High Passed Bass Level: 73 dB
RS High Passed Bass Level: 72 dB
LS High Passed Bass Level: 72 dB
LFE Channel Bass Level: 69 dB

The lower bass levels in the rear channels probably can be accounted for because I'm running satellite speakers back there. I set the receiver's crossover setting at 120hz to truly try and have as much bass as possible directed to the sub. I find it interesting that my right channel reading was 2db lower than the left, just as Ed's was.

Is it possible that the lower LFE channel can be accounted for because the sound isn't being produced simulatenously from, say, a left speaker and the sub? I know in theory, the bass should be directed to the sub, but if the signal used for the pan isn't low enough there probably would be some crossover exchange and a certain amount of the signal coming through the speakers in addition to the sub. That would also account for the slight drop off in my rears which simply aren't capable of producing bass.

Not to muddy the waters here, but as I was preparing to recalibrate the sub by lowering the sub level at the receiver, I noticed something interesting. If I simply brought the db level down at the receiver for the sub and did the subwoofer level, left-front test the measured db reading between having the sub level at 0 vs -10 was -3 db. That just doesn't make sense. Lowering the front left channel -10 db makes a -10 db level reading difference. With the front left channel lowered -10 db I decided to see how that effected the sub woofer's readout(with sub at 0db) and it lowered the overall volume -4db then it otherwise would be with my front left at the 0db setting. That would seem to suggest what I had suspected: the crossover interlapping between the sub and left channel is boosting the read out strength for the subwoofer. That's fine for those frequencies, but if I have my crossover at 120hz and get a low boom at 35hz it's going to be significantly lower than perhaps it should be. On the same token, if I try to compensate for that by boosting the sub levels the upper bass frequencies well be significantly too strong.

In that six channel sweep the only time the subwoofer is completely isolated the db reading drops significantly. I've gotten a little off track it seems, but it certainly seems related.

-eric

 

[Edward J M]

Eric - thanks for checking your unit.

Your point on the 6 channel panning test is interesting - are you actually hearing sound from the respective high passed speaker when it pans on the 5 surround channels? I have not tried this yet. I would suspect not, but it's worth double checking.

Your results when you adjust from sub level 0 to sub level -10 are interesting and show that this level control is not necessarily linear and does not march in lockstep with an absolute dB scale.

Regardless, in your future posts, I would avoid assigning a dB value when you describe any sub level or speaker level settings on your Denon, because they are only relative settings, and not absolutes.

I also ran the LF low-pass and RF low-pass and LFE reverse sweeps last night. These are neat because they go right down to 20 Hz and the SVS really rumbles the couch way down there. The LF and RF sweeps on this test DO include the high passed speaker (since they start at 200 Hz) and I found that the LF and RF reverse sweeps are on average several dB louder than the LFE reverse sweep until below about 50 Hz (this makes sense since my filter point is 80 Hz and the mains will still contribute below that), and then the LFE sweep starts gaining ground and by 40 Hz and lower it almost equals the levels in the LF and RF reverse sweeps. This was very interesting, because it somewhat contradicts the 6 channel panning test.

I'd be interested to see if you - using a 120 Hz high pass - obtain different readings on these reverse sweeps. I wonder if your LFE reverse sweep will more quickly equalize with your RF and LF reverse sweeps because you are using a higher filter point and the mains are dropping out of the equation much higher up in the frequency band. Try and see what happens.

Regards,

Ed

 

[EricSm]

then the LFE sweep starts gaining ground and by 40 Hz and lower it almost equals the levels in the LF and RF reverse sweeps.

room acoustics and waves would definetly be a factor here. Having done those sweeps I can confirm that on average the dedicated LFE sweep is definetly a few -db lower than, say, a front left channel sweep. On a side note, when I do the center channel sweep there is a monster boost(like 10 db gain) around the 130hz-100hz region(crossover at 100hz). I wonder if this would be intentional on the part of polk(csi40 center) to make for more clear articulation in dialogue or a room acoustic issue. It gets so loud I almost always reach for the remote before realizing that would sort of defeat the relevance of the sweep!

It's wierd. I still don't have a clear perception of what is going on with the bass level control at the receiver. The AVIA tests show that the control itself makes little difference, but I swear in movies that there is a world of difference between calibrating my sub with the level -5 and having it calibrated at 0.

 

[Edward J M]

The AVIA tests show that the control itself makes little difference, but I swear in movies that there is a world of difference between calibrating my sub with the level -5 and having it calibrated at 0.

Yes, let's not loose sight of why this thread was started. You are absolutely correct. With the mains set to -3, the bass quality at sub setting -10 calibrated 3 dB hot is rotten, and the bass quality at sub setting 0 calibrated 3 dB hot is fabulous.

I may never be able to put my finger on this with any absolute certainty. Yes, Avia is showing a fixed LFE channel deficit, but said deficit never changes, regardless of the sub setting. That sure doesn't jive with what I'm hearing and feeling on DVD playback.

I guess the important thing that has come out of this thread thus far, is that the quality of the bass (at least in our AVRs) is highly dependent on the sub level setting in relation to the mains setting. If we (and possibly other "03" owners) can deal effectively with this "quirk", then that's really all that matters in the end.

I will be testing the other local 3803 on this Saturday. I'll report back then.

Regards,

Ed

 

[Mike Up]

I will be testing the other local 3803 on this Saturday

If you can, why not try other makes also, as Sony ES, Yamaha, HK, Marantz Etc. I wouldn't try the Pioneer Elite as it uses MCACC and that modifies all bass management.

Have a good one.

 

[phyllis]

I checked my denon 5803 to see if there was any flaws and to my hearing and to avia tests,there was no differences that i could tell.If the sub level was at -5 on receiver or +5 it brought on no changes.I know this is not a 3803 but just adding comments as an denon 03 owner.

 

[MatthewJ S]

If you can, why not try other makes also, as Sony ES, Yamaha, HK, Marantz Etc. I wouldn't try the Pioneer Elite as it uses MCACC and that modifies all bass management.
...........end quote............
Yeah ,I hope this problem isn't really just with the 3803!

 

[jeff_coil]

Edward
I will run my tests again in the same manner you did above and post my results. I was reading another thread and it appeared that someone else was running into some similar issues with Aiva and not a 3803, I don't recall what rcvr it was.
Jeff

 

[Larry Marq]

Edward,

Did you come to any conclusion on this situation?
Thanks, Larry

 

[Edward J M]

My sincere apologies to all for not updating this thread per my posted intentions.

Upstate NY was hit by an ice storm that rivaled the famous '91 ice storm in intensity and magnitude. Storm damage to my property was severe, and we've been digging out and repairing ever since.

I therefore missed the opportunity to test my friend's new 3803, but God willing and the river don't rise, I'll get over there this weekend and run it through the paces with my Avia disc and some DVDs and report back.

Regards,

Ed

 

[Edward J M]

Update:

I had a chance this past weekend to test my friend's 3803, and also to run more extensive tests on my own 3803 with respect to bass management issues.

Since there was a power surge during the aforementioned ice storm and power loss, and my 3803 was running at the time (I do use surge protectors), I decided to not take a chance and I reset the microprocessor, set all speakers to Small, the subwoofer to On, and the high/low pass filter point to 80 Hz.

I then recalibrated all five surround speakers to 85 dB using Avia and Master Volume 00. As expected, my individual speaker levels ended up in the low plus range in order to hit 85 dB all around. Satisfied that I was now calibrated to Reference Level on the surrounds, I bit the bullet and arbitrarily set the sub level to -4 and proceeded with the subwoofer calibration stage.

I made some very interesting observations about the Avia calibration DVD that bear mentioning up front. Since the Avia DVD operates in Dolby Digital 5.1 mode, both the surround speaker and the subwoofer contribute to the bass rumble tone when the user is calibrating the subwoofer level in relation to the surround speaker level.

To see just how much of an effect this phenomenon has on subwoofer calibration, I physically powered down my subwoofer and ran the speaker/sub calibration level test at Master Volume 00 for all five surround channel speakers. Using an 80 Hz filter point for the high/low pass, I was frankly astounded at how much bass emanates from the surround speaker in question when running this particular test. I use large, bass capable surround speakers in all channels, and this of course can exacerbate this phenomenon.

Here are my test results from this particular experiment. The first number is for the high pass pink noise, and the second number is for how much bass the surround speaker still cranks out on the low pass rumble tone with the subwoofer powered off.

Left Main: 85 dB / 82 dB
Center: 85 dB / 78 dB
Right Main: 85 dB / 79 dB
Right Surround: 85 dB / 78 dB
Left Surround: 85 dB / 78 dB

Obviously, room acoustics caused the left main to be 3 dB louder than the right main, and bear in mind I was innocently calibrating my sub off the left main earlier in the history of this problem. I temporarily powered the sub back up, adjusted the plate amp accordingly, and was no longer surprised to see it didn't take much contribution from the sub at all to reach 88 dB (i.e., "supposedly" 3 dB hot) at the SPL meter since the left main was already contributing 82 dB of bass for the rumble tone.

Next, I physically shut the sub back off and ran a reverse bass sweep for each surround channel. Again, I was impressed and amazed at how loud and how deep the surround speaker in question continued to play set to Small with an 80 Hz high pass. Clearly, the Denon was doing its job and ramping down the volume steadily as the frequency dropped, but the high pass filter rate can't be much steeper than 12 dB/octave because I was still getting significant contribution from the surround speaker in question (particularly from the mains) well down into the 40 Hz range. In fact, I was still getting something like 60-65 dB from the mains at 25 Hz! Anyone who thinks 80 Hz is too high for a filter point should run this test - it will change your mind!

Because the Avia disc allows the surround speakers to contribute so much bass to the rumble tone for the subwoofer, I no longer felt comfortable calibrating the sub from the Avia disc and decided to use the Denon internal test tones instead - just for the subwoofer. At Master Volume 00 (and with all speakers already calibrated to 85 dB with Avia), the Denon internal test tones generated exactly 81 dB for each surround speaker. Since the Avia tones are mastered at 20 dB below Reference Level, this tells me the Denon test tones are not at 30 dB below Reference Level as most people assume. This little bit of information might be of interest to anyone who calibrates their surround speaker levels with the AVR internal test tones.

Regardless, I left the Master Volume at 00 and left the sub level at the arbitrary -4 on the Denon, and adjusted the SVS plate amp until I was getting 85 dB (i.e., 4 dB hot) at the meter with the Denon discrete subwoofer test tone. Since this tone does not play at all through the surround speakers, it is IMHO a truer measure of how hot the sub is running as compared to Avia, which masks the sub level with contribution from the surround speaker due to the Dolby Digital encoding.

Anyway, I then proceeded to the six channel bass panning test, which compares the low passed bass level to the LFE channel level. For this test, I did just the opposite and disconnected every surround speaker, and just ran the sub. It still showed an 8-10 dB deficit for the LFE channel as compared to the low passed surround bass and my friend's Denon 3803 exhibited the exact same behavior. Since Avia claims they should all be the same level, and since I specifically disconnected the surround speakers to eliminate them as a contributing source of bass volume on this test, I cannot explain this and I do not know for absolute certainty if it is a flaw of the Avia disc, or a flaw of the Denon. Based on the results of the next test, I suspect the Avia disc is at fault, but I'd like others who have this disc to run the same tests if they are inclined.

I then did a reverse sweep of all the surround channels and the LFE channel with just the sub running and all the surrounds still disconnected. The SPL levels for the LFE channel and the low passed bass for this particular run of tests were nearly identical all the way down to 20 Hz (as they should be), which leads me to believe the six channel panning test in Avia is flawed, but again I cannot be positive of this.

Now came the acid test. With all the speaker levels calibrated to Avia Reference Level at Master Volume 00, and well into the plus range on the Denon, and with the sub set at -4 and calibrated 4 dB hot at the plate amp, it was time to spin some DVDs and see if my problem still existed. The short answer - the problem is gone. I don't know if it was because I was calibrating the sub off the Avia left main channel (which contributed a bunch of bass on the rumble tone), or if it was because I reset the microprocessor again, or possibly both, but the bass is excellent in all respects and the Denon sub level is still far lower than the Denon mains level setting.

A happy ending for me and I learned a lot in the process - thanks for hanging with me and if you have also run your own tests and would like to share the results, please do.

Regards,

Ed

 

[Jamey F]

If the problem is gone, perhaps it didn't exist. If there is a chance it didn't exist, I would suggest editting your title.

I'm sure I will have many questions as to how to setup my 3803 when my studio 100's get here.

 

[Edward J M]

If the problem is gone, perhaps it didn't exist. If there is a chance it didn't exist, I would suggest editting your title.

I don't think the title can be edited - only the post content, which has already been done as this thread progressed. I don't have sufficient information about that particular test on the Avia disc, or from other Denon 3803 owners who also own the Avia disc, to say exactly where the problem existed. I inquired at Avia about this test, and they never replied to my email. Regardless, prior to resetting the processor and subsequently recalibrating, the problem definitely existed. At a minimum, you are now armed with the knowledge from my own experience, and you can be on the look-out for it in your own 3803.

Good luck with your 3803 - aside from this problem which has cost me a huge amount of time that could have otherwise been spent more enjoyably, I really like the 3803.

Regards,

Ed

 

[Mike Up]

Edward,

Glad everything worked out. I have the S&V Home Theater tune up DVD and didn't care for the disc's test tones. The subwoofer test tone was about 4db hotter than my other discs plus it used a different bandwidth. I felt it to be the most inaccurate although I don't have test equipment to measure that. Avia and S&V HT tune up disc are both from Ovation software and are very similar except Avia has more tests.

For what it's worth, the receiver's internal subwoofer tone is 3db low compared to Video Essentials and my other test disc.

It's for this reason I believe in calibrating the subwoofer by ear after an initial test tone calibration is made. These subwoofer test tone level inconsistencies from differing test discs and the receiver's internal subwoofer tone, makes me wonder if level calibrating the sub is all that it's made out to be. Just using a different bandwidth can lead to different level results.:frowning:

Hope everything keeps working good for you, ENJOY!

 

[Jamey F]

Anyone care to make it easy on fella? Could you list exactly what you are doing with Avia to calibrate your system. I have a digital RS meter. I haven't opened the box for my Avia. This could explain why I don't understand some of what's going on. I used the test tones of the 3803 to calibrate my system the first time. I'm expecting studio 100's soon so I'll have to do it all again when I get them. I wanted to see the procedure you guys are following to check your settings.

 

[Bill Kane]

Jamey,

The aforegoing thread is a graduate level course in subwoofer bass management!

Spend some time reviewing the Avia disk and this Beginners Primer selection to familiarize yourself.

And of course there's a gold mine of archival stuff here in Search.

It's fun to calibrate: my RS meter is permanently attached to the tripod ready to use! I find that little changes in placement, or reading others' little tweaks, inspires me to fire it up.

bill