Upmixing - where are we at? Have people compared upmixers?

[Sancus]

Auro3D's bass boost has never really bothered me because actual multi-channel music is rarely tonally identical to stereo recordings of the same material. It often has more bass, sometimes substantially so. And there are other differences. It's also already impossible to tune bass precisely since it vastly varies from recording to recording to an almost comical degree.

That said, if you really want to remove the bass boost properly, it should be possible with the newer Denons(X4700H) that offer a quick switch profile which include Audyssey settings. So you could make a curve for with-Auro3D and then one for everything else. Alternatively, you could use profiles saved in the app I suppose as well, though that's getting too tedious for me to bother with.

There is more and more Atmos music on Apple Music every week(including older albums), and the amount of pure-stereo I listen to continues to decrease, so I've honestly become less concerned about figuring out the perfect upmixing. In the long term it doesn't seem like it will matter.

 

[Chromatischism]

Auro3D's bass boost has never really bothered me because actual multi-channel music is rarely tonally identical to stereo recordings of the same material. It often has more bass, sometimes substantially so.

I have plenty of content with "lots of bass" and so that wouldn't change anything. My system is tuned accordingly. The problem would be if I heard another version of the same song(s) with a massive bass difference to them, it probably wouldn't be welcome. If it's caused by bass management, that's a shame because a multi-channel creator already knows the implications and should have adjusted the mix accordingly. I want to explore more multi-channel music but I'm not currently set up for that.

 

[Sancus]

I have plenty of content with "lots of bass" and so that wouldn't change anything. My system is tuned accordingly. The problem would be if I heard another version of the same song(s) with a massive bass difference to them, it probably wouldn't be welcome. If it's caused by bass management, that's a shame because a multi-channel creator already knows the implications and should have adjusted the mix accordingly. I want to explore more multi-channel music but I'm not currently set up for that.

Well, it's complicated. For one thing, the LFE allows you to include more bass than you can put into a stereo recording without hurting the SNR, because it is 10dB louder at the output stage than the other channels. To downmix LFE bass into other channels you have to lower them all by 10dB, which is why downmixing the LFE is generall not done.

As an example, the DVD-A 5.1 recording of Hotel California has low bass that does not exist in the stereo versions. And sounds much better for it, IMO. The kick drum and bass both sound *vastly* better on the multi-channel version.

Some of this is the difference between multi-channel recordings that treat mch as "stereo++"(which is basically what an upmixer does), and ones that actually utilize it to do more than that. Many multi-channel recordings use the additional channels to produce a horseshoe-shaped soundstage. To try to exactly duplicate the tonality of a stereo recording this way just doesn't make any sense. Using the additional channels gives you a lot of extra utility, you can include things that would simply get lost in a stereo recording, but don't in multi-channel because the spatial separation improves the ability to distinguish different sounds.

I suspect another reason is that with multi-channel mixes you can assume people are more likely to have systems that can reproduce low bass properly. There's no point to mixing 5.1 so that it sounds good on poor speakers, as is frequently done with standard stereo mixes. The same applies to Atmos where the object-based stream and the 2.0 bed channel mix are separate and can be treated separately. Engineers don't ALWAYS take advantage of this, but sometimes they do.

 

[Chromatischism]

What I meant was when each channel has bass content, it gets added to the sub channel by the bass manager. The more channels, the more bass. The mixer either needs to design the album assuming a subwoofer or not. I presume they would mix as if there would be one as that makes the most sense. Then they need to balance the bass levels so that you don't get too much bass when it's all said and done.

@audio2920 described the problem of bass build-up occurring in the context of high-channel count systems playing back an Atmos track, so I thought it could be happening with multi-channel music, too.

 

[Sancus]

What I meant was when each channel has bass content, it gets added to the sub channel by the bass manager. The more channels, the more bass. The mixer either needs to design the album assuming a subwoofer or not. I presume they would mix as if there would be one as that makes the most sense. Then they need to balance the bass levels so that you don't get too much bass when it's all said and done.

@audio2920 described the problem of bass build-up occurring in the context of high-channel count systems playing back an Atmos track, so I thought it could be happening with multi-channel music, too.

No, I was talking about intentional changes in the material. That bass summing effect is relatively small, like 2-3dB at most with 5 channels, and that's probably a worst case. The differences I'm referring to are much larger than that and also more complex than just boosting <100hz by a flat amount. eg the kick drum and the bass in that eagles track sound substantially different(better to me) than any of the stereo versions I've heard.

I would say that generally speaking, multi-channel music, when it IS significantly different, has better dynamic range, less of the bass high passed out, and better clarity, on top of the spatial benefits. There are even some cases where people have suggested that downmixing multi-channel versions to 2.0 will produce a better recording because the material is just that much better.

 

[Abe_W]

In one case testing DSU I had to turn center spread *off* because it was substantially reducing the prominence of a guitar line. So yeah, the idea that it's always universally better is definitely wrong. This is one reason I like Auro, I have literally never caught it doing anything like that.

DSU without center spread works well more often if you have a center that is identical and as close as possible to the same height as your L/R I think. But even then for me there are definite cases where it locks far too much to the center channel.

This is one reason I don't like DSU as my "standard" upmixer.

It is better to "disable" the center channel completely on DSU for music.

 

[dlaloum]

It is better to "disable" the center channel completely on DSU for music.

That is sad.

There is magic to be found in properly set up 3 channel stereo... and Dolby PLII, properly configured would achieve that - and maintain a sweet spot that one could truly walk around in... DSU doesn't seem to get there.

 

[Chromatischism]

It is better to "disable" the center channel completely on DSU for music.

Do you mean disable the center, and turn center spread on to make a center image?

 

[Abe_W]

Do you mean disable the center, and turn center spread on to make a center image?

I use a 5.4.4 for movies.
When upmixing music, it gets dropped to 4.4.2
The center channel is a fk up. Even if you manage to find a center channel speaker that isn't gimped or inherently flawed by design, the channel's existence confuses the hell out of 3D object based codecs for music upmixing.

It helps to have a quality over quantity approach. Get capable fronts+surrounds and prepare for sonic nirvana with 3D object based codes when you discard the cursed center channel sloppy Joe.

 

[Chromatischism]

I use a 5.4.4 for movies.
When upmixing music, it gets dropped to 4.4.2
The center channel is a fk up. Even if you manage to find a center channel speaker that isn't gimped or inherently flawed by design, the channel's existence confuses the hell out of 3D object based codecs for music upmixing.

It helps to have a quality over quantity approach. Get capable fronts+surrounds and prepare for sonic nirvana with 3D object based codes when you discard the cursed center channel sloppy Joe.

My center speaker is identical to the L&R.

Still, in my music testing, Center Spread sounded better ON vs off.

So you are suggesting to turn off the center, and enable center spread?

Unfortunately that will disable Audyssey and I don't think I could be bothered to run another calibration just for that. It's a shame Audyssey doesn't just let you delete a speaker.

 

[Abe_W]

My center speaker is identical to the L&R.

Still, in my music testing, Center Spread sounded better ON vs off.

So you are suggesting to turn off the center, and enable center spread?

Unfortunately that will disable Audyssey and I don't think I could be bothered to run another calibration just for that. It's a shame Audyssey doesn't just let you delete a speaker.

I have a Yamaha prepro, no audyssey. Yes, center spread on, center speaker disabled.

 

[KMO]

If you're suggesting the "centre spread" option does something when there's no centre speaker configured, I'd be fascinated to know what.

 

[Chromatischism]

I just tried it on my X4500H. When the center speaker is set to none, the center spread option disappears.

 

[Sancus]

It is better to "disable" the center channel completely on DSU for music.

I don't agree, but it doesn't really matter to me as I don't bother with anything other than Auro3D for music and DTS Neural X for tv/movies anymore.

 

[Abe_W]

I just tried it on my X4500H. When the center speaker is set to none, the center spread option disappears.

The center spread option is available after disabling the center channel on my Yamaha.

 

[ivo.f.doma]

The center spread option is available after disabling the center channel on my Yamaha.

And then what does the center spread function expand when the center channel is not there?

 

[Sancus]

There is no logical purpose for center spread with the center disabled. If it actually does anything audible -- and frankly I wouldn't believe it did without evidence -- then it's definitely a bug in Yamaha's implementation(not at all surprising). The entire point of "center spread" is to tell the DSU upmixer to put less sound into the center and "spread" it to the L/R more. Without a center there is nothing for "center spread" to do at all.

 

[Chromatischism]

There is no logical purpose for center spread with the center disabled. If it actually does anything audible -- and frankly I wouldn't believe it did without evidence -- then it's definitely a bug in Yamaha's implementation(not at all surprising). The entire point of "center spread" is to tell the DSU upmixer to put less sound into the center and "spread" it to the L/R more. Without a center there is nothing for "center spread" to do at all.

Well, I see how it could work, so it shouldn't be impossible.

 

[Chromatischism]

I just listened to Blade Runner by Hans Zimmer. It is my impression that Dolby Surround (with Center Spread) is the best of the 3 upmixers on this track.

Edit: on another day, I preferred Auro3D to keep more of the sound up front. This is going to be affected by the mood you're in, so I love having choices!

It's also interesting that even Stereo sounds like more than stereo on this track. On a good system, it sounds like other speakers are on.

 

[Chromatischism]

I just noticed that Dolby Surround puts a lot of signal to the center channel, more so than Auro3D. In order to level things out and make it sound even better, I reduced the center channel level by -6 and increased the overall master volume to compensate. I've noted that in my settings for anyone wishing to replicate:

Upmixing - where are we at? Have people compared upmixers?

Reliability is pretty important, as are functions.

www.audiosciencereview.com