Level calibration signal volume varies

[DYohn]

Thosde are very interesting readings. Were they recorded off the line-level outputs? Was Dolby Volume enabled?

[markus]

Sorry, had to delete the first post because the graph was not correct. The volume relation doesn't change with different volume settings but the subwoofer calibration signal is 15dB louder than that of other channels. Here's a new measurement (measured the line outs):



Blue/yellow = subwoofer/left channel @ volume 60
Green/orange = subwoofer/left channel @ volume 20

Best, Markus

[DYohn]

OK, that makes more sense.

[markus]

Does it?

[Topend]

Let me get this right, there is a 15db difference, so it would have to be corrected by having the sub db adjusted to -7.5db and the other speakers to +7.5db. not counting room acoustics.

[marius]

i would have guessed about 9db difference based on my measurements (ratshack meter), but 15db doesn't surprise me.

[simarddominic]

I made several tests and I am rather perplexed…

Since EmoQ gives me eccentric results for the distances, the levels and the xovers, I put the real distances, the cut-off frequencies according to frequencys response of my speakers and tried to adjust the levels of each speakers.

The problem is that it's significaly different from a test tones to another.

From the internal test tone over the " limited pink noise 5.1" in dolby digital EX test tone of DVE to the " limited pink noise 5.1 DTS-ES always on the DVD DVE" to the many test tone included in the various bd and dvd films I have.

And don't tell me " master volume interfer with this" because I always start to have a 75db reading for the left front speaker at 0db on UMC-1 speaker level menu and after, I don't touch the master volume in the process.

I would like to use DVE Blu-Ray but I will have it only this week…

The results are rather different that when, for example I apply the results of the internal test tone, my left speaker have an higher outpout level than right in stereo listening, enough that I ear it easy.

Same thing with the left and right surround speakers…only the center and the subwoofer remains constant.

I suspect that dependently wich decoding carried out by the UMC-1 (Dolby DIGITAL, Dolby True-HD, DTS-HD Multi-PCM etc) the level of each speaker changes and it's the case, it's a bug that we should tell to Emotiva tech team because the levels must remain balanced independently of the signal processing.

I would like that someone conter check that...

Balance your speaker at 75db with a sonometer or with REW, with for exemple, the internal test tone and remakes the process with several other test tones encoded in different formats and let us know here if your results are the same than me.

Thanks !

PS. Sorry for my bad english...I'm a french canadian ;D

[markus]

Some more observations on level calibration.

The UMC-1's calibration signal is white noise. Left RCA out:



Pink noise would produce a graph that is flat. Just for the records.


After calibrating mains and sub with a SPL meter set to C-weighting, an in-room measurement at the listening position with an RTA shows something like this for subwoofer and left speaker:




The 1kHz band and the 63Hz band should show the same level. Instead we see that the subwoofer is a whopping 15dB too loud.

Confirmation with pink noise - subwoofer and left speaker:




Again, compare the 63Hz and 1kHz band. Same result, subwoofer is 15dB too loud.

Most home theaters are set up with the subwoofer too hot but Emotiva, please, calibration should give us reference. From there everybody can adjust to his preference.

Hope this gets fixed in one of the upcoming firmware releases.

Best, Markus

[mysterymachine]

Hmmm I still haven't wrapped my brain around this issue. I ran the test tones in the UMC, and then the test tones on a disc and they were even. The subwoofer tone was not more hot on the disc than the internal tones which is what it seems you are saying.

The only thing I saw was a test on a disc that did the full frequency sweep for all channels at once to check for room nulls etc - once that test tone went below the crossovers the bass went through the roof pegging the needle on my meter. I wasn't sure if that was a problem with the test or I just didn't understand that test. I think essentially since all 5 of the other speakers were rerouting to the subwoofer and it was sending the signal sky high. If I did a single channel crossover test it seemed fine. I am probably just misunderstanding that test.. all the others were fine.

I have my UMC set for -5 for LFE in the input adjustment for all sources except analog and 0 for analog (and +5 LFE in the player with the analog out) so there might be some issue with not applying the +10 to analog LFE.

[markus]

What kind of test signal is on that disc? Is it the LFE the same level as other channels? Then it would be 10dB too loud (UMC applies LFE +10dB boost).

[mysterymachine]

I'm not sure, it has lots of test signals in various codecs and khz. Its the AIX sampler/calibration blu-ray that came with my Oppo BDP-83. Dan and Lonnie are watching it in that first demo video they did of the UMC.

For levels I think I used the standard dolby digital track as the MPCM tracks were not always working right with the Oppo+UMC.

I was on a track meant to test crossovers and such when my needle pegged in the bass freq (and my whole house felt like it was shaking)

[wizardofoz]

markus...What software are you using for your graphs... can you post a link to it please

[markus]

Graphs are from RTA of Signal Scope Pro:
www.faberacoustical.com/products/signalscope_pro/octave_analyzer/