The Problem With Dolby Atmos

[goat76]

Just for interest and understanding, can you point out some specific recordings where you find the mix to be gimmicky or otherwise unsuitable? This is not to challenge you at all, I am just curious. Thanks.

I already gave an example of a track by the band Yes. You can find it at the end of post #196.
I can easily find more examples, but it's past 2 a.m where I live, so that have to wait until tomorrow.

I was supposed to listen to your examples you mentioned in your previous post, but I forgot about it. Will listen to those mixes as well.

 

[BJL]

Yes, and the Atmos mix has been widely criticized for being too "front centric", specially in comparison to the older, more inventive 5.1 mix.
But each to their own, it's great to have many options. What a incredible time the last decade+ has been for the surround music revival!

This must be a joke. Have you heard the two mixes? The 5.1 mix (which is included on the Atmos blu ray) is also front centric, even more so than the Atmos mix. There are no surround back speakers (5.1 not 7.1) unless someone used an upmixer. The Atmos mix uses all 15 plus LFE speaker locations creating a tremendous sense of space. So yes, it is oriented toward the front, but the other speaker positions are used to create space. Whomever says that the 5.1 mix is more inventive must have goo in their ears, or they are trolling.

I can see this objectively by observing the levels in all 15 output channels.

 

[BJL]

I already gave an example of a track by the band Yes. You can find it at the end of post #196.
I can easily find more examples, but it's past 2 a.m where I live, so that have to wait until tomorrow.

I was supposed to listen to your examples you mentioned in your previous post, but I forgot about it. Will listen to those mixes as well.

Oh sorry I didn't notice. I will look for a blu ray of the Yes album and listen to it, I was planning on getting those anyway.

 

[Sal1950]

I can see this objectively by observing the levels in all 15 output channels.

Forget about your 15 object based channel levels and actually listen to the Atmos vs 5.1 mix to find which sounds more active in the bed channels.

The only problem is that with more channels available, it's obviously even harder for many mixing engineers not to use them in a gimmicky way.

There's the problem, what you may find gimmicky, many others find inventive, exciting presentations.
I wonder how you feel about albums such as Yello - Point, Booka Shade - Dear Future Self, Steven Wilson - The Raven that Refused to Sing ?

 

[BJL]

Forget about your 15 object based channel levels and actually listen to the Atmos vs 5.1 mix to find which sounds more active in the bed channels.

I thought that is what I was trying to say, politely, but I can be blunt; compared to the Atmos mix, the old 5.1 mix is flat, boring, and unimaginative. Does that work better for you?

I wonder how you feel about albums such as Yello - Point, Booka Shade - Dear Future Self, Steven Wilson - The Raven that Refused to Sing ?

The OP was talking about rock albums, otherwise I would have mentioned them. Point Yello and Touch Yello are superb Atmos mixes and I personally love the music. I have every Atmos blu ray that Booka Shade has released and they are terrific. And then, going into current electronic music, there are the two difficult to acquire Max Cooper blu rays. Steven Wilson? I don't particularly care for his music, though I appreciate some of his engineering, but that is purely personal, I just don't like any of it, though I've tried.

Edit: I am listening on a 9.1.6 system. I'm not sure if some of the recordings would have quite the same impact without the front wides (meaning the Knopfler/Dire Straits.)

 

[Sal1950]

The OP was talking about rock albums, otherwise I would have mentioned them. Point Yello and Touch Yello are superb Atmos mixes and I personally love the music. I have every Atmos blu ray that Booka Shade has released and they are terrific.

IMHO it's all Rock, just a sub-gender. Well then if you don't consider Boris and Booka a gimmicky use of 5.1 & Atmos I guess we're not so far apart.
All their works are considered Demo Quality surround mixing.

I thought that is what I was trying to say, politely, but I can be blunt; compared to the Atmos mix, the old 5.1 mix is flat, boring, and unimaginative. Does that work better for you?

Maybe then it's you who has goo in his ears. LOL

Steven Wilson? I don't particularly care for his music, though I appreciate some of his engineering, but that is purely personal, I just don't like any of it, though I've tried.

Are you referring to his music, or his outsourced mixing projects?
For the most part I can't name anyone I'd like to see doing the remixing of classic rock more than him.

His personal music, I loved most all his earlier more prog rock work, both solo and with Porcupine Tree, Blackfield, etc.
OTOH the last 3 or 4 albums, not so much, though 2025s The Overview was a little better.
Like David Bowie he's constantly trying to reinvent himself, not always to popular acclaim.

 

[goat76]

For the type of mix you describe as ideal for a rock band, Mark Knopfler's albums Sailing to Philadelphia and, especially One Deep River are great examples, and Dire Straits Brothers in Arms (Atmos mixes co-produced by Mark Knopfler and mixed by band mate Guy Fletcher.) These three albums make extensive use of the L/R Wide positions, supported by Top Front and Top Middle, with little or no signal from Surround Back and Top Rear. The Dire Straits album does use the Surrounds, primarily for ambience. (NB my comment based on the blu ray releases.) The audio presentation, especially the Knopfler solo albums, is to place you in a club or the studio audience. My recollection is that Steve Wilson's Atmos remixes of the Grateful Dead Europe '72 and Blues for Allah also emphasize the front and my extensive use of the Front Wide positions.

In my opinion, this should not be dogma.

The Atmos mix of the album Sailing to Philadelphia sounds pretty good; the surround channels are used just enough to add envelopment without “giving themselves away” as individual sound sources. It’s a very fine line for a mixing engineer to balance the energy between the front channels and the surround channels, not making the mix tip over too much to the channels behind the listener. If anything, they may even have been a bit too conservative on that aspect for Sailing to Philadelphia, but it’s better to have a little margin, as this is very sensitive for a multichannel mix.

For the One Deep River Atmos mix, they dared to let it be a little bit less front-focused, making it sound large and room-filling without anything obvious tipping over behind me at the listening position. None of Knopfler’s mixes use any gimmicks like panning any instruments to the surround channels, and no strange decisions to pan, for example, any single tom drums to any specific surround channel or similar things like that.
It seems like both these mixes were done with a speaker system balanced similarly to mine, which I suspect is not always the case with some other Atmos mixes, with way too much energy in the surround channels.

When listening to the Atmos version of the first track on Brothers In Arms, the mix has a similar balance as the album One Deep River. But I must mention that even the original 2-channel stereo mix of this song has an “airy” quality, which made me have to check if there was any sound coming from the surround channels.

All the above mixes use the surround channels to widen the main instruments up to 90 degrees to the side of the listening position, but never seem to pass that point, and tip over behind me.

Staying with rock, consider Eddie Kramer's magnificent Atmos reconstruction of Jimi Hendrix Axis:Bold As Love. Kramer was the original engineer, and recreated the album from the original session tapes, based on what he believed Hendrix would have wanted. Kramer does incorporate extensive movement and panning to artistic effect, what some might call "gimmicks."

I like the Atmos mix of Jimi Hendrix’s Axis: Bold As Love. There are some artistic playing with panning in the mixes, but as none of these sounds has a static position to any of the surround channels, I don’t feel they are distracting in a “ what the f..” is that instrument doing over there?

A poorly done mix, is a poorly done mix, whether it be Atmos or quad or stereo or mono. Consider that in the 1960s there were powerful mono mixes (Motown and Stax, as examples.) There were also sloppy and haphazard mono mixes. This isn't a problem of Atmos, it is a problem of mediocrity. I suspect that many of the "spatial audio" mixes being churned out are automated and/or not done with thought or care. We can't realistically expect that every mix in any given format is going to be a work of art, there are differences in skill between engineers.

Consider as an example, Teo Macero's extraordinary mixes from Miles Davis electric band period (stereo) and compare with the mediocre mixes on the outtakes now incorporated with many releases. Then you will understand why Macero generally opposed releasing outtakes. They were rejected for good reasons.

While poor mixes can be found in everything from Atmos to stereo, I find a much larger portion of that can be found in Atmos mixes. The mixing engineers knows the stereo format inside and out, but when it comes to Atmos ( and probably any multichannel formats), it seems like many of them are at the stage of taking baby steps.

I’m not as impressed with Steven Wilson’s Atmos mixes as many people here, and often find them sounding somewhat static and stale, as if he doesn’t have a sense for making a surround mix to sound “organic” and lively, which is done by panning the same sound object to different positions in the sound field and then using delays to create some movement in the mixes, like in the Hendrix Atmos mix mentioned above. What Steven mostly does instead of that is to pan single objects to different positions in the sound field, which often makes them sound unnaturally isolated in their static positions, and that is my main problem, especially when that sound object is positioned to a point behind the listening position. I hope this explanation goes through to everyone not used to mixing music themselves.

 

[Sal1950]

What Steven mostly does instead of that is to pan single objects to different positions in the sound field, which often makes them sound unnaturally isolated in their static positions, and that is my main problem, especially when that sound object is positioned to a point behind the listening position. I hope this explanation goes through to everyone not used to mixing music themselves.

Fully understand your preferences here. I and a large portion of the Atmos/5.1 popular music listening audience disagree. A good immersive mix should set me in the middle of a performance with lead vocals mostly upfront but the rest positioned all around me, maybe even behind. If you sat in the middle of the band, that's what you'd hear. That's why it's called immersive. LOL
Mark Waldrep of AIX used to do multi-positioned releases with On-Stage, Audience, and Headphone mixes in an effort to please everyone, a good option for all I guess. I still maintain if you don't enjoy immersive type mixing, using the stereo track with the upmixer of choice is your best option. With the wide variety and tweaking options available today they can create a presentation to most everyone's liking.

 

[dlaloum]

Forget about your 15 object based channel levels and actually listen to the Atmos vs 5.1 mix to find which sounds more active in the bed channels.


There's the problem, what you may find gimmicky, many others find inventive, exciting presentations.
I wonder how you feel about albums such as Yello - Point, Booka Shade - Dear Future Self, Steven Wilson - The Raven that Refused to Sing ?

There is the seperation here between those who look at high fidelity as the reproduction of a live acoustic performance, vs those who focus on recordings where the mastering/engineering itself is a part of the end result (rather than being merely a step in the chain of delivery)...

I like both.

But I find that reproduction of live acoustic performances are a greater test of a systems resolution than synthesized effects/environments - as impressive as they are. There is no (and can be no) baseline for measurement of the accuracy of reproduction of an artificial sound environment - the very question is nonsense really, as it is an intentionally artificial construct.
Whereas live acoustics are something that those with experience with those instruments and environments, can quickly gauge whether the end result is authentic sounding or not.

A set of speakers and environment that can provide excellent results with live acoustic performances, will always be a great starting point for multi channel artificial recordings as well.... the opposite is sadly not true.

Hence I find authentic reproduction of guitar, violin, piano, and voice (unamplified) - to be a far better gauge of a systems capabilities.

There were a number of efforts to develop recording methods and encoding methods to bring a high fidelity approach to the multi channel surround environment eg: ambisonics
But this has become a niche technology ...

 

[Sal1950]

There is the seperation here between those who look at high fidelity as the reproduction of a live acoustic performance, vs those who focus on recordings where the mastering/engineering itself is a part of the end result (rather than being merely a step in the chain of delivery)...

I like both.

Good post. The big bugaboo being that the number of rock/pop recordings done in a perfectionist manner we could count on may be one hand. The other is there really isn't any absolute as to what a Fender Stratocaster sounds like, vs a Stradivarius violin. Bottom line is we shall forever live in 2 different worlds.
That said High-Fidelity can be very closely defined with todays level of measurement detail, only in the realm of the speaker/room are we still left with quite imperfect products. Casting aside some of the very poor performing speakers out there, the TOTL models variances will then be determined by personal preferences since none are perfect but very many are very damn good.

But I find that reproduction of live acoustic performances are a greater test of a systems resolution than synthesized effects/environments - as impressive as they are. There is no (and can be no) baseline for measurement of the accuracy of reproduction of an artificial sound environment - the very question is nonsense really, as it is an intentionally artificial construct.
Whereas live acoustics are something that those with experience with those instruments and environments, can quickly gauge whether the end result is authentic sounding or not.

I'd call it more the ability to deliver flat FR at the MLP than resolution but I get what your saying. But again, since none are perfect, one may sound just right with a violin but wrong with a cello, or again with a double bass vs viola.
Speaking honestly, much of that means nothing to me. I've never attended a real live fully acoustic symphony in my life, it's not the kind of music I'd go too far out of my way to hear.

A set of speakers and environment that can provide excellent results with live acoustic performances, will always be a great starting point for multi channel artificial recordings as well.... the opposite is sadly not true.

But again that's only one aspect of being able to reproduce any type of music, things like dynamic range, bass extension, lack of distortion are also very important.
Much of the classical world has drooled over speakers like LS3/5A, the original Quads, etc that may make pretty violin sounds but most rockers would send into a shower of sparks on the second day. So in the end what good was that natural sounding violin when instead David Gilmours Black Strat sounds just perfect on JBL's?
The real bottom line is not many can have it all, most of us all have to make some compromises depending on our needs and preferences.

But how multich or 2ch became any part of this I have no idea, the same principles apply equally to both.

 

[napfkuchen]

Ours is a small segmented community of special interests and tastes that has little connection to the John Doe soundbar and earplugs world.

Thanks to all AirPod users!

Fully understand your preferences here. I and a large portion of the Atmos/5.1 popular music listening audience disagree. A good immersive mix should set me in the middle of a performance with lead vocals mostly upfront but the rest positioned all around me, maybe even behind. If you sat in the middle of the band, that's what you'd hear. That's why it's called immersive. LOL

A typical concert setup doesn't require full-channel 3D audio, which is why I enjoy listening to music that transcends those boundaries.

That way, you very quickly cross the boundaries into other music genres, too. Without that "3D nonsense" I wouldn't have paid much attention to R&B, jazz or experimental electronic music until now.

 

[BJL]

The Atmos mix of the album Sailing to Philadelphia sounds pretty good; the surround channels are used just enough to add envelopment without “giving themselves away” as individual sound sources. It’s a very fine line for a mixing engineer to balance the energy between the front channels and the surround channels, not making the mix tip over too much to the channels behind the listener. If anything, they may even have been a bit too conservative on that aspect for Sailing to Philadelphia, but it’s better to have a little margin, as this is very sensitive for a multichannel mix.

For the One Deep River Atmos mix, they dared to let it be a little bit less front-focused, making it sound large and room-filling without anything obvious tipping over behind me at the listening position. None of Knopfler’s mixes use any gimmicks like panning any instruments to the surround channels, and no strange decisions to pan, for example, any single tom drums to any specific surround channel or similar things like that.
It seems like both these mixes were done with a speaker system balanced similarly to mine, which I suspect is not always the case with some other Atmos mixes, with way too much energy in the surround channels.

When listening to the Atmos version of the first track on Brothers In Arms, the mix has a similar balance as the album One Deep River. But I must mention that even the original 2-channel stereo mix of this song has an “airy” quality, which made me have to check if there was any sound coming from the surround channels.

All the above mixes use the surround channels to widen the main instruments up to 90 degrees to the side of the listening position, but never seem to pass that point, and tip over behind me.

Thanks for your thoughtful comments on your listening. It is also worth mentioning that the Atmos mixes for Knopfler's albums (and Brothers in Arms) were done by Guy Fletcher, played keyboards for Dire Straits (starting with Brothers), on Knopfler's solo albums and whom Knopfler acknowledged as long term musical collaborator. Mark Knopfler co-produced all three. Thus these mixes should be considered authoritative as to the intent of the artist.

Edit: also interesting that you mention the stereo mix of Brothers. In his liner notes, Fletcher states that he wanted to retain the feeling of the original stereo mix in the Atmos version.

 

[BJL]

I'm curious if @Sal1950 and @goat76 have listened to the Atmos mix of Talking Heads More Songs etc. Mix by Jerry Harrison & engineer Eric Thorngren. The approach to Atmos varies from song to song. Some are firmly anchored to the front (@goat76) while others use somewhat unconventional panning to bring out processed guitar and similar abstract sounds, (@Sal1950 ). I happen to have just listened to the album straight through, and because of this recent back and forth on Atmos approaches, and because you both had interesting insights, I wondered if you have listened to this and had any comments. Apologies if you dislike Talking Heads.

Most of the songs seem to be 9.x.4 (no top middle) with one or two that do not use the front wide position.

 

[goat76]

I'm curious if @Sal1950 and @goat76 have listened to the Atmos mix of Talking Heads More Songs etc. Mix by Jerry Harrison & engineer Eric Thorngren. The approach to Atmos varies from song to song. Some are firmly anchored to the front (@goat76) while others use somewhat unconventional panning to bring out processed guitar and similar abstract sounds, (@Sal1950 ). I happen to have just listened to the album straight through, and because of this recent back and forth on Atmos approaches, and because you both had interesting insights, I wondered if you have listened to this and had any comments. Apologies if you dislike Talking Heads.

Most of the songs seem to be 9.x.4 (no top middle) with one or two that do not use the front wide position.

I have to get back to you on this as I haven’t had time to listen to it yet.

One thing I must mention is that I only listen to the Atmos mixes through a 5.1-system, so it all depends on how well the Atmos down-mixer works. As I have never had the opportunity to compared the same mix in full Atmos vs the 5.1 version I get to hear, it’s a bit of a risk that I miss something you hear. It’s a really good 5.1 system though with great speakers all around.

And speaking of loudspeakers configuration for surround music, have you ever compared what goes missing if you set your receiver to say a 7.1 system compared to your full Atmos system, and how much are the ceiling speakers add to the listening experience when it comes to Atmos music mixes?

 

[Sal1950]

I have to get back to you on this as I haven’t had time to listen to it yet.

Same here, maybe later tonight

As I have never had the opportunity to compared the same mix in full Atmos vs the 5.1 version I get to hear, it’s a bit of a risk that I miss something you hear. It’s a really good 5.1 system though with great speakers all around.

You have to be careful on that. Even on the newly released BluRays,. the 5.1 and Atmos mixes can be completely different, sometimes even using different engineers with the 5.1 mix being done 10 years ago.
And then you can't expect a downsampled Atmos mix on a 5.1 system, to give the same presentation as when listening to it being decoded by true Atmos decoding software and played on a real Atmos speaker spread.
You just can't know with these things.

 

[BJL]

I have to get back to you on this as I haven’t had time to listen to it yet.

One thing I must mention is that I only listen to the Atmos mixes through a 5.1-system, so it all depends on how well the Atmos down-mixer works. As I have never had the opportunity to compared the same mix in full Atmos vs the 5.1 version I get to hear, it’s a bit of a risk that I miss something you hear. It’s a really good 5.1 system though with great speakers all around.

And speaking of loudspeakers configuration for surround music, have you ever compared what goes missing if you set your receiver to say a 7.1 system compared to your full Atmos system, and how much are the ceiling speakers add to the listening experience when it comes to Atmos music mixes?

Just for fun, later tonight I will turn off my AVP's Atmos decoder and re-listen to the T-Heads album. That will give me the underlying TrueHD 7.1 mix.

I have compared the 7.1 to Atmos with a couple of other recordings, including one of the Knopfler's albums. Even more than the height speakers, it is the absence of the front wides that I notice most immediately (for those albums that take advantage of that position, that is), but the height speakers can, and often do, make a big difference. As you probably noticed, there are wide variations in quality. The best albums are effective at using the heights to create a sense of space (e.g. Anette Askvik - Liberty, a 7.1.4 mix) or imaginative placement of sounds and sense of space (e.g. Touch Yello). If you can do it, I would start by adding wides, and then at least one pair of heights.

By the way, I always hear references to "downmixing Atmos to 5.1 (or 7.1). I am definitely not knowledgable on the technical underpinnings, but that seems to me to be a misnomer. The underlying file in Atmos is 7.1 TrueHD plus metadata for decoding the positional information in capable systems. Therefore I think that it is more correct to refer to Atmos as an upmixer, as the "downmixed" file is always there. Maybe someone smarter than me knows the answer, but my processor refers to an "Atmos Upmixer" which I can disable if I want to listen to the underlying 7.1 file.

 

[Sal1950]

By the way, I always hear references to "downmixing Atmos to 5.1 (or 7.1). I am definitely not knowledgable on the technical underpinnings, but that seems to me to be a misnomer. The underlying file in Atmos is 7.1 TrueHD plus metadata for decoding the positional information in capable systems. Therefore I think that it is more correct to refer to Atmos as an upmixer, as the "downmixed" file is always there. Maybe someone smarter than me knows the answer, but my processor refers to an "Atmos Upmixer" which I can disable if I want to listen to the underlying 7.1 file.

Boy you really want to confuse the hell out of the commonly used terms. LOL
As far as a "Atmos Upmixer" ? what processor are you using, never heard that one before. ?
I'm mostly aware of the upmixing done to overheads from non-Atmos sources as Multi In + DSurround, or + Neurall X, or + Auro?

 

[BJL]

Boy you really want to confuse the hell out of the commonly used terms. LOL
As far as a "Atmos Upmixer" ? what processor are you using, never heard that one before. ?
I'm mostly aware of the upmixing done to overheads from non-Atmos sources as Multi In + DSurround, or + Neurall X, or + Auro?

Tide16 and yes it was confusing until I had some communication with the miniDSP DevTeam and gave it some thought. One of the options available for the Tide's Dolby processor is captioned "Atmos Upmixing" which I took to refer to the Dolby Upmixer (which will process LPCM sources to upmix to the wides and heights.) But no, it means that enabled everything is processed through the Atmos decoder, including encoded TrueHD files. Then, I thought about, and realized that makes perfect sense. An Atmos file (at least for consumer products) is a 7.1 Dolby TrueHD file with metadata. If your processor ignores the metadata (i.e. no Atmos decoding or no supported speaker configuration), it will pass through as a 7.1 channel file.

And actually, from the hobbyist standpoint it is quite handy to be able to turn off Atmos processing, and review the 7.1 file.

I suggested a clarification to the Tide16 user manual . . .

 

[Sal1950]

Then, I thought about, and realized that makes perfect sense. An Atmos file (at least for consumer products) is a 7.1 Dolby TrueHD file with metadata. If your processor ignores the metadata (i.e. no Atmos decoding or no supported speaker configuration), it will pass through as a 7.1 channel file.

And actually, from the hobbyist standpoint it is quite handy to be able to turn off Atmos processing, and review the 7.1 file.

LOL ??? All the Atmos capable processors I've owned Marantz and Denon do automagically recognize Atmos meta data and turn it on.
If for some reason you want something else, a push of the Mode button will let you play the bed alone or just about any other codec you want to screw with.
To bad IME that's not true for Auro 3D, there doesn't seem to be anyway to auto flag it.

I suggested a clarification to the Tide16 user manual . . .

Good idea

 

[BJL]

LOL ??? All the Atmos capable processors I've owned Marantz and Denon do automagically recognize Atmos meta data and turn it on.
If for some reason you want something else, a push of the Mode button will let you play the bed alone or just about any other codec you want to screw with.

I had a processor like that. Interesting question, when you change the mode is it playing the bed, or is it playing back the entire TrueHD file. The Tide is playing the entire 7.1 file. There is no way (that I've found) to disable the Atmos feature leaving the bed channels only. I'm not sure if I'd want to, but so far as I know, it can't be done on the Tide.

It will detect and automatically decode. This is an option to disable automatic detection of Atmos material, to limit the Dolby Upmixer to the bed channels, and/or play back the 7 (or occasionally 5) channel TrueHD file without processing.