The Dolby Atmos, DTS-X, and Auro-3D Discussion Thread

[AcuDefTechGuy]

Atmos is supposed to make a bird fly around your room three-dimensionally by triangulating the sound of that bird by multiple speakers simultaneously at different levels so as to "place" that bird within the prism of your room.

Like we’ve said many times, it depends on the sound mixers.

In “Snow White and the Huntsman DTSX”, there is a scene where the ravens fiercely fly around the evil queen in the chamber/room. You can hear them fly around you. Then they fly up to the ceiling.

I don’t know about 3D-pin-point accuracy, but it’s pretty close.

 

[AcuDefTechGuy]

I'm the opposite.

DSU seems accurate and appropriate. DTS NeuralX is overtly aggressive and puts sound in the wrong channels at times.

I don’t know about “more accurate”, but DSU is less aggressive than NeuralX.

Some people think NeuralX is too aggressive, and some people think DSU is too conservative, not “more accurate”.

DSU might work better for some movies and NeuralX works better for some movies.

Comes down to what you think is more accurate or preferable.

I think NeuralX does overhead rain, thunder, aircrafts, and music better than DSU.

I have one example DTS NeuralX scene to compare. In HBO’s Generation Kill, the final 5 minutes. It recaps the video footage of the war. As the men gather around to watch the video in horror and disgust, Johnny Cash’s “The Man Comes Around” prominently plays overhead. It just seems so appropriate to me how that music plays overhead with full authority with NeuralX.

When I watch that same scene with DSU, the music doesn’t play from overhead - it plays from the bed speakers and sounds like some elevator music in the background.

 

[VonMagnum]

I'm the opposite.

DSU seems accurate and appropriate. DTS NeuralX is overtly aggressive and puts sound in the wrong channels at times.

You can't get much more 'wrong' than putting a helicopter at ground level when it's supposed to be overhead, IMO. That's what DSU does as it only puts 'safe' ambient sounds overhead (e.g. crickets). The old Dolby assumed surround speakers above your ears. The new one pushes BED sounds (ear level). DSU is 10x more inaccurate for that reason alone. It doesn't even play the old movies correctly for the same reason. Play Top Gun in 6.1 on an old Dolby guideline system and the jets are above your head. Play it with DSU and they're at ear level. Play it in Neural X and they're not only overhead, but move front to back overhead.

It's another reason I suspect that even with Atmos at home, the speakers should ALL be a couple of feet above ear level, especially with multiple rows involved. Of course, that's not always practical, particularly with rooms with 7 or 8 foot ceilings and with one row, it doesn't make much difference.

 

[Erod]

You can't get much more 'wrong' than putting a helicopter at ground level when it's supposed to be overhead, IMO. That's what DSU does as it only puts 'safe' ambient sounds overhead (e.g. crickets). The old Dolby assumed surround speakers above your ears. The new one pushes BED sounds (ear level). DSU is 10x more inaccurate for that reason alone. It doesn't even play the old movies correctly for the same reason. Play Top Gun in 6.1 on an old Dolby guideline system and the jets are above your head. Play it with DSU and they're at ear level. Play it in Neural X and they're not only overhead, but move front to back overhead.

It's another reason I suspect that even with Atmos at home, the speakers should ALL be a couple of feet above ear level, especially with multiple rows involved. Of course, that's not always practical, particularly with rooms with 7 or 8 foot ceilings and with one row, it doesn't make much difference.

On multiple occassions, I've heard voices on screen put in the height channels with Neural X. I think that's why most people use DSU.

But this is hardly worthy of a big debate. Everybody should choose what they like, absolutely.

Also, with the new ARC Genesis, my height channels are much more active now, so that gave DSU some more punch.

 

[VonMagnum]

On multiple occassions, I've heard voices on screen put in the height channels with Neural X. I think that's why most people use DSU.

I've never noticed that, but then I'm already feeding some of the lower channels to the heights to create a phantom image in the middle so the voices come from my screen instead of below it ("dialog lift" effect). I also wasn't aware "most people" use DSU. I have no idea what people use. The AVRs typically default to the signal they get so if you're watching a DTS BD (about 75% of them),you get Neural X by default and if you're watching Dolby you get DSU by default. I kind of doubt the average person would bother to change them, really, but then I don't think the average person even uses an AVR these days. They tend to buy soundbars gauging by the electronics shows and Best Buy displays. I suppose they're better than the TV speaker. Actually, I don't think they're all as bad as some like to make them out to be.

I actually set up a soundbar with sub and rear channel speakers (5.1 system as the soundbar had three channels in it) at my mother's house on her smaller TV (I set up an AVR system in her larger system). It's not really great, but for the price it cost, it's kind of phenomenal in a way. The bar is the exact width of the TV so panned dialog lines right up. It has a center channel so dialog is centered from either side of the couch (neither sits directly in front of the set) and surround works surprisingly well. The subwoofer has limited range, but plays loud down to about 45Hz and then starts to drop off (since it plays up to like 200Hz I put it dead center behind her couch and it shakes the couch, which is kind of cool). The newer version of the same sound bar has Atmos reflection speakers on the top. I kind of wish I'd waited. I'd be curious how that would sound in her room which has the right sort of dimensions for it.

But this is hardly worthy of a big debate. Everybody should choose what they like, absolutely.

I kind of like Auro-3Ds upmixer for a more traditional 5.1 type mix (IMO, it's even lighter on the overhead use than DSU). Some DTS 5.1/6.1 CDs sound pretty good in Auro upmixed. Neural X tends to put instruments up in the air a lot (neat for soundtracks, not so much for live bands or traditional albums). DSU does better, but I was putting it into straight 5.1 (Auto) for the best effect, but found Auro didn't screw it up and added a bit of ambience.

Also, with the new ARC Genesis, my height channels are much more active now, so that gave DSU some more punch.

What does a room correction system have to do with how active the height channels are? Or do you mean they're just that much more even with the bed channels in terms of even output?

 

[Erod]

I've never noticed that, but then I'm already feeding some of the lower channels to the heights to create a phantom image in the middle so the voices come from my screen instead of below it ("dialog lift" effect). I also wasn't aware "most people" use DSU. I have no idea what people use. The AVRs typically default to the signal they get so if you're watching a DTS BD (about 75% of them),you get Neural X by default and if you're watching Dolby you get DSU by default. I kind of doubt the average person would bother to change them, really, but then I don't think the average person even uses an AVR these days. They tend to buy soundbars gauging by the electronics shows and Best Buy displays. I suppose they're better than the TV speaker. Actually, I don't think they're all as bad as some like to make them out to be.

I actually set up a soundbar with sub and rear channel speakers (5.1 system as the soundbar had three channels in it) at my mother's house on her smaller TV (I set up an AVR system in her larger system). It's not really great, but for the price it cost, it's kind of phenomenal in a way. The bar is the exact width of the TV so panned dialog lines right up. It has a center channel so dialog is centered from either side of the couch (neither sits directly in front of the set) and surround works surprisingly well. The subwoofer has limited range, but plays loud down to about 45Hz and then starts to drop off (since it plays up to like 200Hz I put it dead center behind her couch and it shakes the couch, which is kind of cool). The newer version of the same sound bar has Atmos reflection speakers on the top. I kind of wish I'd waited. I'd be curious how that would sound in her room which has the right sort of dimensions for it.



I kind of like Auro-3Ds upmixer for a more traditional 5.1 type mix (IMO, it's even lighter on the overhead use than DSU). Some DTS 5.1/6.1 CDs sound pretty good in Auro upmixed. Neural X tends to put instruments up in the air a lot (neat for soundtracks, not so much for live bands or traditional albums). DSU does better, but I was putting it into straight 5.1 (Auto) for the best effect, but found Auro didn't screw it up and added a bit of ambience.



What does a room correction system have to do with how active the height channels are? Or do you mean they're just that much more even with the bed channels in terms of even output?

The new ARC Genesis update significantly increased dynamics, especially in the height channels. As a result, DSU is not nearly as subtle on Anthem products as before. Heights are both aggressive on discrete effects like Neural X with the ambiance still intact from DSU before.

It's like DSU plus Neural X simultaneously.

 

[Bookmark]

Voices should appear in the Height/Overhead channels, when it is called for, up-mix or native. Example, in the updated Flatliners "intro" with Miss Page or Strange Days when someone is experiencing the memory rig. Or simply films with locational PA system like Harry Potters or Bourne at the train stations or Denzel in the Equalizer at work in the home depot. The list goes on. General on screen dialog should not really be present in the overheads/heights unless it is additional reverb or echoes. Dolby's recent restrictions on up-mixing non Atmos Dolby content is not something however, to be commended or admired.

Dialog lift is something else entirely and there to compensate for limited/compromised centre placement.

Dts Virtual X is for older layouts with 5.1 or 7.1 systems to simulate overhead/height speakers in their absence. Dts Neural X is there for when they are present. Dolby AFAIK do not have this option as yet. Both DSU and Neural X will make use of all the available speakers, however the results can vary between content. Don't expect miracles from old 30s,40s mono films. I was however pleased with up-mix results my D Day tribute selection of 40s,50s and 60s war films, however newer films like Saving Private Ryan and Clint's Iwo Jima double still work much better.

All EQ systems have their Pros and Cons, their advocates and their detractors. Some even avoid it completely. Personally I would be concerned if the older ARC EQ was significantly compressing the dynamics, as implied, this is not something I have seen reported before.

 

[Erod]

Voices should appear in the Height/Overhead channels, when it is called for, up-mix or native. Example, in the updated Flatliners "intro" with Miss Page or Strange Days when someone is experiencing the memory rig. Or simply films with locational PA system like Harry Potters or Bourne at the train stations or Denzel in the Equalizer at work in the home depot. The list goes on. General on screen dialog should not really be present in the overheads/heights unless it is additional reverb or echoes. Dolby's recent restrictions on up-mixing non Atmos Dolby content is not something however, to be commended or admired.

Dialog lift is something else entirely and there to compensate for limited/compromised centre placement.

Dts Virtual X is for older layouts with 5.1 or 7.1 systems to simulate overhead/height speakers in their absence. Dts Neural X is there for when they are present. Dolby AFAIK do not have this option as yet. Both DSU and Neural X will make use of all the available speakers, however the results can vary between content. Don't expect miracles from old 30s,40s mono films. I was however pleased with up-mix results my D Day tribute selection of 40s,50s and 60s war films, however newer films like Saving Private Ryan and Clint's Iwo Jima double still work much better.

All EQ systems have their Pros and Cons, their advocates and their detractors. Some even avoid it completely. Personally I would be concerned if the older ARC EQ was significantly compressing the dynamics, as implied, this is not something I have seen reported before.

These aren't the kind of voices I'm referring to. Yes, voices meant for the surround track belong there, but Neural X would occasionally take on screen sound and place them in odd places here and there. Nothing overly egregious, but it seemed confused at times.

To my ears, DSU is more about extending the soundstage to the heights with some discrete effects, while Neural X is just the opposite, more about discrete effects and very little soundstage extension up top. From that standpoint, it's a matter of taste.

However, Anthem's new update to ARC (called Genesis) has applied more dynamics to the surround and height channels than before. It's been like getting a brand new preamp/receiver for the Anthem crowd.

As a result, DSU in the Anthem acts more like a combination of DSU and Neural X. More spaciousness and more prominent discrete effects. Part of that is just volume balance, but there's something else going on in the algorithms that has made is quite different in a good way.

 

[panteragstk]

I tend to agree. These days I either use 1. NeuralX, 2. Atmos, or 3. DTSX.

I don’t use DSU.

Interesting. I've tried both and I tend to use DSU with Dolby content and Neural X with DTS content. I'll have to mix it up and see what I get.

@VonMagnum mentioned top gun and I have a copy of it that I was disappointing with how the jets didn't sound overhead when using DSU. I thought it was just a crap mix (top gun apparently sounds better with the regular DTS mix than the DTS-MA mix) so I just figured it was that. I'll have to try the DTS upmixer with it again.

 

[VonMagnum]

V
Dts Virtual X is for older layouts with 5.1 or 7.1 systems to simulate overhead/height speakers in their absence. Dts Neural X is there for when they are present. Dolby AFAIK do not have this option as yet.

I've read the newest AVRs either have or are supposed to get Dolby's virtual speaker simulator. It's listed on the newest models from D&M, but I think it's listed as firmware update. It'd be kind of a shame for older models to have Virtual X but no Dolby version as it becomes available. It's like being sold half a car for people that use it.

I seem to recall Virtual X still having combination settings (e.g. Virtual X + Neural X versus Virtual X + DTS:X for the same movie) to control discrete X versus just a neural upmix into virtual speakers. It does have a multi-in + Virtual:X option, but when I tried it with a few Dolby 7.1 signals (set KODI to 7.1 multi-channel output), it didn't sound as good as with a DTS one for whatever reason. It could have been just that soundtrack or it could be Virtual X works better with DTS encoding by some technical nuance of which I'm unaware.

Both DSU and Neural X will make use of all the available speakers, however the results can vary between content.

DSU will not use front wides or any other "special" designated speakers to my knowledge. It is limited purely to the home version base 7 bed speakers + 2 overhead channels which are Left/Right overheads using a combined array of whatever overheads you use (i.e. whether 2 overheads or 10 overheads, you get them all playing either the left or right channel overhead channel). In other words, if you have a 34 channel Trinnov Atmos setup, you will still only get 7 bed channels + 2 overhead (using all 10 as an array of 2-channels) unless you engage Trinnov's compensation mode (forget its name offhand that simulates any setup using as many speakers as possible).

Neural X can technically use any available channels, BUT is currently limited to 11-channels plus subwoofer channel and prioritizes the more common channels over the less common (e.g. four overheads instead of two plus front wides). DTS:X Pro (if it ever gets released this year, which is half over already) is supposed to remove that limitation and allow something like a Trinnov to use ALL available channels with Neural X at the same time, turning a 7.1.4 DTS:X movie into up to 32.2. This will give DTS:X a HUGE advantage over Dolby Atmos, which despite all its 'object' claims is still limited to whatever the mix specifies from everything I've read. If the mix doesn't call for Lc/Rc, you don't get them even if an object is present there. It will just pan L/R instead. This was made clear on AVS by a Trinnov owner. He said the majority of home Atmos tracks don't specify Lc/Rc (which he uses between left and center and right and center) so they just sit there silent. He doesn't have a lot of the more unusual surround speakers available to Trinnov so who knows how many tracks will use them for objects?

My point is that Atmos was sold under the umbrella of objects are objects and so no matter how many Atmos speakers you add, the system will use them if an object comes near that speaker in the mix. They neglected to say they will work IF and only IF the film mixing guy includes those speakers in the mix from the start. That's a whole different ball game to Disney just locking objects in 7.1.4 by using locked objects as channels and mixing as if it were a channel system. I was surprised to read that. Why on EARTH would you NOT use the whole 34-speaker setup for every single movie and just move your sound objects where you want them and let the system determine which speakers play as the object moves? I thought that was the entire point of Atmos. So I was surprised to read that Lc/Rc are just silent on many movies for this guy. If an object moves between L/R it should pass through Lc/Rc as a matter of course.....

Again, I don't know if it's just those speakers that are uncommon in Atmos soundtracks or whether some of the side/rear speakers at bed level are also not used in some soundtracks other than Disney. IMO, they should ALL be used based on the object behavior a a matter of course and I do not know why Disney is allowed to lock objects for the sole purpose of defeating the object system and turning Atmos into an 11-channel limited channel system. I do no know why any film mixer (even if their home mixing Atmos studio doesn't have all 34 speakers) would leave any speakers out of a mix when it should be a matter of defining a path for an object in 3D space and letting the system decide which and how many speakers to use. Why limit it? Atmos downmixes to fewer channels just fine and that screws over the high-end Trinnov owners who actually have 34 speakers connected.

I've never read ANY reason WHY Disney uses those limited 7.1.4 print-through soundtracks (Dolby SadMouse) or why Dolby hasn't done something in their license to use Atmos to prevent it. It seems that the reason must be rather nefarious. Filmmixer at AVS said he found out the reason and was surprised at what it was (and specified it had nothing to do with saving money on making a streaming mix that uses less data or whatever) and yet would not say what the actual reason was as he said that was privileged information that Disney does not want anyone to know, apparently. Well color me stupid. WTF gives Disney the right to screw their soundtracks and then not even answer for it or even let someone else state the reason?

I have no real respect for Disney anymore. They charge more because they CAN (whether amusement park tickets or content). I think Walt would be rolling over in his grave if he knew the sheer levels of GREED that Disney operates at these days. Many Americans cannot afford to go to their theme parks in our own country as it's over $100 per person per day unless you buy a ridiculous number of days per person (which gets the total up). For a family of 4 paying out $400 for ONE day plus food plus gas plus getting to one of only two parks in the country that are not near most other states (e.g. Florida is at an extreme Southern location and California at the extreme West) is out of reach for a family trying to make ends meet.

All EQ systems have their Pros and Cons, their advocates and their detractors. Some even avoid it completely. Personally I would be concerned if the older ARC EQ was significantly compressing the dynamics, as implied, this is not something I have seen reported before.

Yeah, that sounded strange to me too. I have gotten significantly different results from Audyssey just by changing the toe-in on the main speakers slightly (and who knows with so many mic readings what effect having the mic in a slightly different position for each mic could make, especially at higher frequencies that are extremely directional (i.e. I went from shrill sibilance to dead even just by doing it over again). I'm thinking he got a better result when he redid it. Nothing in their release suggests more dynamics.

 

[AcuDefTechGuy]

Interesting. I've tried both and I tend to use DSU with Dolby content and Neural X with DTS content. I'll have to mix it up and see what I get.

Sometimes they sound about the same, sometimes NeuralX might be a bit too aggressive.

But I think overall, I prefer NeuralX over DSU for most things (that are not Atmos/DTSX).

This is the only reason (for me) why it would suck if Dolby restricts the use of NeuralX on all Dolby contents - it's because I want to use NeuralX on everything.

 

[Erod]

Yeah, that sounded strange to me too. I have gotten significantly different results from Audyssey just by changing the toe-in on the main speakers slightly (and who knows with so many mic readings what effect having the mic in a slightly different position for each mic could make, especially at higher frequencies that are extremely directional (i.e. I went from shrill sibilance to dead even just by doing it over again). I'm thinking he got a better result when he redid it. Nothing in their release suggests more dynamics.

I can tell you that every Anthem owner to a person I've read (hundreds of them) has noticed a significant improvement in dynamics with Genesis versus ARC-2. It's not even comparable. ARC-2 is very nicely balanced, but conservative. Genesis turned the surrounds and heights loose.

 

[AcuDefTechGuy]

ARC-2 is very nicely balanced, but conservative. Genesis turned the surrounds and heights loose.

I thought most people wanted a more "balanced" conservative approach?

Now Genesis makes the surrounds more aggressive?

 

[VonMagnum]

Interesting. I've tried both and I tend to use DSU with Dolby content and Neural X with DTS content. I'll have to mix it up and see what I get.

Some DSU stuff sounded close or hard to tell from Neural X (I'm thinking the older demos, etc. that I could quickly compare between the two), but when it comes to real movies, there's a big difference, IMO.

@VonMagnum mentioned top gun and I have a copy of it that I was disappointing with how the jets didn't sound overhead when using DSU. I thought it was just a crap mix (top gun apparently sounds better with the regular DTS mix than the DTS-MA mix) so I just figured it was that. I'll have to try the DTS upmixer with it again.

I have the BD. Honestly, I don't usually bother with the DTS-HD type signals unless 'X' is on there as I can't really hear a difference encoding to regular DTS with its already pretty darn high bitrate default in Handbrake. Even Dolby Digital at 640kbps sounds fine to my ears. I know a LOT of people on some sites get upset by that notion that lossy compression is actually pretty darn good and very hard to tell from lossless (people just seem to like 'knowing' that it's lossless regardless of whether they can truly hear a difference and often base their assumptions on some already lousy sounding track or more often than not slight volume differences).

In other words, I don't think Top Gun is going to sound better using DTS-ES 6.1 over DTS-MA 6.1. But then there might be a version out there using DTS-MA 5.1 that's a different mix altogether or something. iTunes only has DD 5.1 for Top Gun, for example. All I know is the DTS-ES 6.1 mix sounds amazing here in Neural X. The planes fly overhead front to back, which is an improvement over the old 6.1 mix using just 6.1 speakers with the surrounds at the 2/3 height recommendation Dolby used to recommend before Atmos as that would be overhead, but it couldn't do the front/back overhead as the rear only had the one speaker at bed level here in the back (if it had been 2/3 up, it might have come close). But that was one of the movies I was highly concerned about before switching and people at AVS seemed to have this "Atmos is always better" attitude (implying DSU) which I found to be bizarre given DSU doesn't put much in the overheads.

There's actually a review of Top Gun with the different upmixers (think it was a shootoff article between Auro, DSU and Neural X as part of a larger article of Auro-3D, Atmos and X). The guy unplugged the bed speakers and listened to various fighter jet scenes with each of them and noticed DSU had almost nothing of the jets in the overheads. Auro had "some" in the overheads, but only Neural X not only had a LOT in the overheads, but correctly moved front to back across the room overhead as well. My test was simpler. I just tried them with a few scenes with each mode. DSU was barely above ear level (about where my surrounds normally sound at bed level). Auro-3D was a little better, but still pretty low. Neural X just sounded AWESOME. Better than the old 6.1 with true speakers.

Throw in 3D and it's even cooler (although sadly the 3D version of Top Gun is very grainy and the colors are shifted; it might be more accurate to the original film as they did a rescan), but at a certain level grain can be annoying (Ghostbusters has similar amounts, IMO), but it's more noticeable in 3D since both eyes get it and it sticks the 3D objects rather than sitting in the background as a true 3D camera does (it actually has more grain yet, but it's sitting in the background not on objects sticking out of the screen so you notice it less although it can cause 'shimmery' effects at times as the grain is different for each eye with actual 3D cameras).

 

[VonMagnum]

I can tell you that every Anthem owner to a person I've read (hundreds of them) has noticed a significant improvement in dynamics with Genesis versus ARC-2. It's not even comparable. ARC-2 is very nicely balanced, but conservative. Genesis turned the surrounds and heights loose.

That's just odd. It almost sounds like Anthem was SQUASHING dynamics before. You can't create more dynamics out of nothing but you can compress them. How did it sound with ARC turned off before?

 

[Erod]

I thought most people wanted a more "balanced" conservative approach?

Now Genesis makes the surrounds more aggressive?

Yes, in terms of separation. Surrounds seem more involved.

It also greatly improved bass response overall (which a bunch of new controls that I don't even use so far), and it improved the center channel dialog. Folks are loving the bass response.

They did a lot with the algorithms. Honestly, it's like a brand new preamp now.

 

[Erod]

That's just odd. It almost sounds like Anthem was SQUASHING dynamics before. You can't create more dynamics out of nothing but you can compress them. How did it sound with ARC turned off before?

Before, turning if off didn't sound quite as good to me. And I only corrected up to 500 Hz.

Now, I correct up to 5 kHz because the dynamics aren't negatively impacted at all. Bass response sounds so much cleaner and more punchy, and the surrounds came alive. They're more prominent now, and not just from volume level.

We're all trying to figure out exactly what they did, and they're even releasing beta versions of future firmware updates in the works for Genesis that we can play around with if we want to.

 

[AcuDefTechGuy]

Yes, in terms of separation. Surrounds seem more involved.

It also greatly improved bass response overall (which a bunch of new controls that I don't even use so far), and it improved the center channel dialog. Folks are loving the bass response.

They did a lot with the algorithms. Honestly, it's like a brand new preamp now.

You couldn't just increase the sub level or increase the Parametric EQ for the subs by a few decibels for better bass response?

Also increase the Center speaker levels a few decibels for better dialogue?

 

[AcuDefTechGuy]

We're all trying to figure out exactly what they did

You mean you're still trying to figure out exactly what and how they Equalized (frequencies and amplitudes).

For example, on my Yamaha, to get "much better bass response", I don't increase the subwoofer levels at all. In fact, I actually DECREASE the subwoofer levels. Instead, I increased the Parametric EQ of the subwoofers from about 40Hz-80Hz by 2dB.

For "Much cleaner and clearer" or crystal clear dialogue, I just increase the Center Speaker level by 3dB.

It's amazing what a few simple adjustments can do for your system!

 

[Erod]

You couldn't just increase the sub level or increase the Parametric EQ for the subs by a few decibels for better bass response?

Also increase the Center speaker levels a few decibels for better dialogue?

Far more to it than that. Louder doesn't mean better separation.

A lot of folks are wondering why they did this with it being aboutt time to release the next line of Anthem receivers and preamps. Perhaps it's because with HDMI 2.1 so slow to "matter", they might want to sit tight a bit longer and just provide some of the next gen improvements.

Genesis is getting rave reviews. A lot of folks are calling it a true like-for-like comparison with Dirac.