Is DTS Neural-X sounding dull only on my Onkyo TX-NR7100?

[keks8430]

Has anyone else been looking at tonal differences between upmixing Stereo to 5.1 by Dolby Surround and DTS Neural-X?

I always found Neural-X more distant or muffled. Simply playing with center level, center spread or treble did not work out. Finally, I measured it.

Setup: SynRTA with noise signal for center only, measured the frequency response with Neural-X, set it as reference and measured again with Dolby Surround. I got the following curve:

I then created two Dirac-Live targets adding a) the difference and b) half the difference to my usual simple sloped target curve. For simplicity, I only considered the correction above 1 kHz. The L/R ratio was flat with a below 1dB difference.

While the full correction was somehow too much (difficult to say of what), the half correction was surprisingly pleasant.

I wonder if the measured difference and dullness of the Neural-X upmixer is systematic or something particular to the Onkyo?

 

[EWL5]

Depends on what speakers you have. The 7100 suffers from a "limp mode" when pushing to low impedance speakers!

 

[keks8430]

Depends on what speakers you have. The 7100 suffers from a "limp mode" when pushing to low impedance speakers!

Thx, but can definitively exclude any limp mode.
As said, the Neural-X dullness was observed when switching Stereo sources between Dolby Surround and Neural-X.

To my ears, Dolby Surround often sounds too close. Neural-X too distant, now just right with the half correction.
For sure a matter of taste, I think Neural-X does a better (stronger) job in creating ambiance.

 

[EWL5]

1) Does Dolby Digital/Atmos sound normal?
2) Does DTS or DTS HD-MA sound normal?
3) Have you run Dirac or any type of room correction?

 

[keks8430]

1) Does Dolby Digital/Atmos sound normal? - yes (5.1)
2) Does DTS or DTS HD-MA sound normal? - no idea, main source is sat with PCM/Dolby 2ch or 5.1
3) Have you run Dirac or any type of room correction? - yes, I mentioned Dirac Live above

The issue is with Stereo -> 5.1 only, used mostly for films.
Dolby Surround being more bright/direct but also "flat", Neural-X having more depth and with correction, sonically balanced.

 

[EWL5]

1) Does Dolby Digital/Atmos sound normal? - yes (5.1)
2) Does DTS or DTS HD-MA sound normal? - no idea, main source is sat with PCM/Dolby 2ch or 5.1
3) Have you run Dirac or any type of room correction? - yes, I mentioned Dirac Live above

The issue is with Stereo -> 5.1 only, used mostly for films.
Dolby Surround being more bright/direct but also "flat", Neural-X having more depth and with correction, sonically balanced.

I'm trying to isolate whether you have a DTS chips problem vs a stereo problem.

You could test a DVD w/a DTS track or a Blu-ray w/a DTS-HD MA track. Worth the trouble...

 

[keks8430]

Good point. While I use it for upmixing, the frequency measurement used an already given center channel.

The DTS and other demo clips favour effects, but I want to judge dialog volume/timbre vs "other".
Youtube Chromecast streaming is stereo only(?), will have to download and convert for playing on Vu+..

If possible, could you or anyone repeat the above measurement I did?
Links to noise files:
SynRTA_12th_center.wav
SynRTA_12th_front_right.wav
Can be done with laptop mic since it is a relative measurement, auto calibrating.
Would be interesting to see if the frequency response differs over components/types/brands.

 

[BJL]

Has anyone else been looking at tonal differences between upmixing Stereo to 5.1 by Dolby Surround and DTS Neural-X?

I always found Neural-X more distant or muffled. Simply playing with center level, center spread or treble did not work out. Finally, I measured it.

Setup: SynRTA with noise signal for center only, measured the frequency response with Neural-X, set it as reference and measured again with Dolby Surround. I got the following curve:

I then created two Dirac-Live targets adding a) the difference and b) half the difference to my usual simple sloped target curve. For simplicity, I only considered the correction above 1 kHz. The L/R ratio was flat with a below 1dB difference.

While the full correction was somehow too much (difficult to say of what), the half correction was surprisingly pleasant.

I wonder if the measured difference and dullness of the Neural-X upmixer is systematic or something particular to the Onkyo?

I mostly use dts Neural:X to place mono signals in the center channel (e.g. old rock and soul music) and occasionally with other music, where it is mixed bag, sometimes great, sometimes not good, depends on the source. dts tends to move sound as objects vs. dolby which is mainly ambience. For example, if you happen to have a 1970s recording that was SQ encoded, Neural:X is quite competent at bringing out the rear channels (many CDs of older recordings have the SQ encoding intact) and synthesizing a center channel consistent with the overall mix.

I have never found dts to be muffled or dull, and this includes with the Pioneer AVR I previously used, similar to the Onkyo I believe. I calibrate with Dirac using the target curve suggested by that software with no modifications. If your problem persists I suggest resetting the AVR to factory default and trying again, to clear out any software or settings issues. If that doesn't work there is a problem with you AVR, a defect of some sort and I would replace it.

 

[keks8430]

Ok, thx.
Resets already done, will present it to Onkyo Product Service.