[bfreedma]

A lot of people are saying that it is exactly like listening to speakers. If that's the case, I'm interested. I rarely listen to headphones for serious listening. I wouldn't be looking to make headphones sound better. My interest would be in having more control over room acoustics and distortion in a simulated speaker system to make simulated speakers sound better that real ones.

Same goals here. I’m hoping to get closer to the surround experience in a well calibrated system. If they can work out the system to the point where the control over simulated room acoustics/distortion allows for a better experience than my current speakers, I will not only be impressed, I will be as early an adopter as possible. The potential to simulate different spaces without dealing with the physical listening environment is also intriguing.

 

[jgazal]

A lot of people are saying that it is exactly like listening to speakers. If that's the case, I'm interested. I rarely listen to headphones for serious listening. I wouldn't be looking to make headphones sound better. My interest would be in having more control over room acoustics and distortion in a simulated speaker system to make simulated speakers sound better that real ones.

Me too!!!

 

[gregorio]

Most movies and a lot of multichannel music is recorded by pairing sets of speakers... for instance front left and rear left... and then balancing the levels to place the sound somewhere along the left wall. I call this "the four walls approach". They might pair the left and right mains, along with the center channel to define the front soundstage; or pair the two rears to place something along the back wall. But rarely do they pair diagonally across the room. I think the reason for this is that the wide variety of room acoustics make it difficult to predict if pairs would mesh properly that way. It's more foolproof to place along walls.
But Atmos uses the "sound field approach". There's no pairing of 2 channels. All of the speakers contain bits of the sound field in varying levels so the sound localizes in the middle of the room. It seems to me that this approach only works if all of your speakers are full range and the response and levels are precisely tuned to the room. If there is any problem with the room's acoustics, the sound field dissolves and it's just bland directionless sound coming at you from all directions.

Hmmm, not really. With movies there is no rear left (or right) point source, there's a rear left audio channel but no rear left speaker. Have a look next time you're in a cinema, you've got the three front speakers (behind the screen) and then a bunch of small speakers equally spaced all the way around the two side walls and rear wall. The rear left audio channel is proportionally split amongst the speakers along the left wall and the left side of the rear wall. Therefore, unlike the front left, the rear left is not a specific point but a "zone". Hence why you can't really have a diagonal pair (say front right + rear left), it's at at best a vague approximation and at worst, just a bit of a mess. This also explains why 7.1 was invented, the two relatively massive rear zones (left wall + left rear wall and right wall + right rear wall) are split into four smaller zones, providing somewhat more precision but we're still dealing with zones containing multiple speakers, not individual point sources. ... Atmos is effectively the exact same seven zones, the three point source front speakers and 4 side/rear zones. However, it additionally enables the use of "audio objects", specific sound objects which can be mapped to much smaller zones or even individual speakers (including height speakers) although that is controlled by the cinema processor rather than by the mix engineer/s.

All the above is for movies, all bets are off with regard to music mixing. Sometimes multi-channel music is mixed on film stages, effectively using the same "zone" approach to the rear channels, sometimes it's mixed in relatively small music studios to specific rear left and rear right speakers and sometimes it's mixed in small boutique film studios which have specific rear left and right speakers but are designed to somewhat emulate the multi-speaker setup in a cinema. With music it's effectively whatever anyone decides, there's no standard convention to follow or right/wrong way to do it. From an individual consumer's point of view, the positioning fidelity (and phase/freq fidelity) is like to be very variable, more so than with a stereo setup and additionally, there's of course their personal preferences. A consumer might get quite reasonable fidelity but just not "get it" or like the mixing decisions or might get relatively poor positional fidelity and like (or not) what they're hearing. For this reason, there's likely to be very conflicting opinions between different consumers and I'd personally take many of those opinions with a certain level of scepticism.

And putting the bass in one side of the rear when most people use bass management with a sub up front seems like it would sound like the bass was smeared across the room with the low part up front and the high part to one side in the rear.

In theory that wouldn't happen. The low freqs from the sub should sound non-directional and therefore you wouldn't perceive the low freqs coming from a different location to the higher freqs. In practice of course that might not be the case, depending on how an individual consumer's sub has been positioned and setup and also on the competence of whoever mixed the recording. Generally, with higher budget mixes (done somewhere like Abbey Road), I wouldn't expect the mixer's competence to be an issue but many/most music producers and engineers have relatively little experience of the practicalities of surround mixing and consumer reproduction, don't know how to mitigate the risks and often just "go for it" and hope for the best!

G

 

[bigshot]

I'm referring to the recent Atmos music mix for Sgt Pepper and Kraftwerk's Catalogue. Those were both mixed with object based sound and they've been folded down to 5.1. Even though I don't have an Atmos setup, these mixes have lots of instances of sound being placed in the middle of the room, or crossing diagonally from front to back. The mixes that create a sound field in the middle gets widely varying reviews online. I think that getting the front and rear to mesh properly in the middle isn't particularly common in home setups. I wasn't getting a good phantom center in the middle of the room until I replaced my rear speakers with full range, widely dispersing ones and recalibrated. The directionality of a speaker can affect it a lot. If a speaker is too directional, the phantom spot is very small. Wider dispersion seems to increase the size of the sweet spot.

The Lennon Imagine mix is the one that puts the bass in a single rear channel. The rest of the mix is all isolated in specific channels too except for the drums which are spread across the L&R mains. Whenever I hear an old quad mix that puts each thing in its own speaker, it sounds really thin. And since many people have smaller speakers in the rear than in front, putting the bass in the back seems like an invitation for big firebreaks in the bass response. I don't know if I'm going to like that mix based on what I've heard about it.

 

[gregorio]

The mixes that create a sound field in the middle gets widely varying reviews online. I think that getting the front and rear to mesh properly in the middle isn't particularly common in home setups.

It's not common in any setup! If you think about even a theoretically perfect cinema, then a sound panned to the centre of the room is going to be produced by pretty much every speaker in the cinema and there is going to be time/phase issues between all those speakers, regardless of where you're sitting. For this reason panning to a centre of the room position is generally avoided except in certain specific circumstances. If we're talking about a 5 point source mix (non-theatrical) done in a music studio with 5 speakers, then whether or not your front and rear speakers "mesh" depends on the distance relationship between your front and rear speakers relative to the distance between the front and rear speakers in the studio where the mix was created and your relative listening position.

Wider dispersion seems to increase the size of the sweet spot.

Indeed and that's the principle that cinemas use. A lot of smaller speakers dispersed around the walls, so the sweet spot isn't restricted to a couple of seats in the centre of the cinema. The downside is more vague positioning, zones rather than precise specific points, as I mentioned previously. Atmos gets around this to some extent with the use of audio objects. However, this is only a partial solution and you can't generally create a mix from audio objects alone but of course that audio object approach is largely lost when rendered in 5.1.

G

 

[bigshot]

Again, I'm talking about multichannel music in a home theater, not movies in a large cinema. When they mix music for Atmos, they can place things in the middle of the room. When it gets folded down into 5.1, those objects remain placed in the middle, just without a dedicated channel to precisely define them. If you can create a solid phantom center between front and back and right and left, a lot of that placement can come through. I have multichannel music recordings that have guitar solos that start against the front wall and slowly move towards the center of the room. There are many that have sound that crosses the room diagonally across the listening position. On Sgt Pepper, the calliope part of Mr Kite swirls around the room like a vortex. It takes open space, a room that isn't too large or too small, well calibrated levels and a fairly high volume level to pull off, but it is possible.

Most multichannel mixes are created with the "four walls" approach, where sound objects are placed somewhere between two paired speakers along a wall... Front Left paired with Rear Left, or Front Left paired with Front Right. But lately, there have been some music mixes that are trying to push into the middle of the room. Atmos mixes in particular. It's pretty amazing when it works right.

 

[gregorio]

[1] Again, I'm talking about multichannel music in a home theater, not movies in a large cinema.
[2] When they mix music for Atmos, they can place things in the middle of the room.
[3] When it gets folded down into 5.1, those objects remain placed in the middle, just without a dedicated channel to precisely define them.
[4] If you can create a solid phantom center between front and back and right and left, a lot of that placement can come through.

1. You actually said: "Most movies and a lot of multichannel music is recorded by pairing sets of speakers". Also, Atmos requires a bare minimum of a 9.1 system and all the commercial Atmos mixing facilities I know of have an Atmos setup consisting of at least 20 speakers. Music studios, even top class ones, simply don't have a control room of the size and dimensions necessary for a professional Atmos setup, so I assume that Atmos music mixes are actually made in theatrical dub stages.

2. Yes they/we can but no more so than with 5.1/7.1. There is no point source speaker in the middle of the room with an Atmos setup, it's the same phantom middle as with 7.1. When converting a 7.1 dub stage (or cinema) to Atmos none of the existing speakers which comprise the 7.1 system are changed or moved, it just requires the addition of two rows of ceiling speakers (to provide the dimension of height info) and a Dolby Atmos cinema processor, which provides the possibility, when a sound is dedicated to an audio object, to target smaller zones (or individual speakers in smaller setups depending on the Atmos processor) but obviously you can't target a middle speaker because there isn't one.

3. There is no dedicated middle channel with Atmos either.

4. Sure but again, creating that solid middle is very difficult and in commercial situations pretty much impossible to achieve. You could do it if you had a sweet spot of just one or two seats but then a dub stage or cinema with sweet spot of only a couple of seats is pretty useless. A larger, more diffuse sweet spot can be achieved but that's at the expense of the solidity of the middle, which is why the middle position is generally avoided, except in the case of certain sounds which are diffuse to start with, or are "passby's", moving sounds which pass through the middle position.

Your categorisations of a "four wall approach" and a "soundfield approach" don't exist in practise. In 5.1/7.1 we're always thinking in terms of both point sources and the entire soundfield. Atmos doesn't change this approach, it only affects it in so much as we can target smaller, more precise side/surround zones but that's only of real benefit to large spaces (cinemas and dub stages) where each side/surround zone comprises several/many individual speakers and in the fact that the soundfield is expanded to include the dimension of height but of course that's also irrelevant when folded down to 5.1/7.1.

G

 

[bigshot]

Sorry. I meant most multichannel music, not movies. And I was indicating that multichannel music mixes are beginning to become more sophisticated. Perhaps faster than people's home theaters are.

 

[gregorio]

[1] Sorry. I meant most multichannel music, not movies.
[2] And I was indicating that multichannel music mixes are beginning to become more sophisticated. Perhaps faster than people's home theaters are.

1. I realise that your main focus was obviously on multi-channel music. I was just pointing out that as far as Atmos music mixes are concerned, the lines are blurred between music and movies, as Atmos music mixes are presumably made in movie mix stages. The point I was trying to make is that even with 5.1/7.1 mixes, pretty much all movies and some/many music mixes are not recorded or mixed as "pairs of speakers" (your "four wall approach"). Certainly "pairs of speakers" are sometimes employed for some element/s but only as a part of the full soundfield, which will also include many mono elements (panned to different locations within the full soundfield) and some full 5.1/7.1 elements. With music mixes this can be less true than with movies, there are few conventions for multi-channel music mixing, most music producers are far less experienced with multi-channel music mixing, some are so used to stereo that they may approach surround mixing as a set of stereo soundfields (your "four wall approach") and some are often tempted to really experiment with all the 5.1/7.1 options, despite (or even ignorant of) the risks of translation issues when reproduced on consumer systems. Furthermore, older, traditional popular music genres (rock, etc.) evolved hand in hand with stereo and don't easily lend themselves to multi-channel.

2. Atmos is a relatively new format for movies and a very new format for music. As with all new music formats, music producers will experiment with the "new toys" the format allows and with little/no precedents or conventions, it's difficult to judge whether one is not going far enough and using the format to it's full advantage or going too far and being gimmicky. An obvious example of the latter is some of those early stereo pop music mixes which were hard-panned and employed "ping-ponged" elements. They sound rather ridiculous to modern ears accustomed to the results of decades of evolution of stereo mixing techniques and conventions but at the time probably didn't seem nearly so ridiculous. Even considering more modern stereo conventions and consumer expectations, I'd take a guess that on the relatively noisy music reproduction systems of the day and probably less widely spaced stereo speakers, that some/many of those extreme stereo mixes sounded a lot more sophisticated and less gimmicky than on modern stereo setups. So, even with stereo we see translation issues significantly affecting judgements of sophistication vs gimmicky and, with ever more speakers to deal with (5.1, then 7.1 and now Atmos), these issues are potentially increased. This is especially likely with Atmos, as the mixing system probably has 20-60 speakers and the consumer reproduction system probably 6 (5.1), 8 (7.1) or 10 (Atmos).

G

 

[bigshot]

To simplify what I was saying... Excluding movies, which are a totally different animal, I've noticed three mixing approaches to mixing multichannel music...

ISOLATED CHANNELS: Each sound element in the mix comes out of its own speaker. This was common in early quad mixes and it sounds thin

FOUR WALLS: Pairs of speakers along the walls are joined to create a soundstage along that wall. This is common in multichannel music mixes and it can sound quite good. (I'm included mono elements that are potted between two speakers in this.)

OBJECT BASED: Sound objects are placed within the room by balancing the sound across more than just two speakers. This is relatively new and I've found the most vivid examples in music mixed for Atmos. This kind of mix is hard for home stereos to reproduce, but when they can, the sound is astonishing. When they can't it turns into a sonic mush. It's kind of all or nothing.

Most mixes blend more than one of these techniques, but tend to favor one kind of mixing over the other. For instance, object based mixes almost always have four walls elements too, usually the rhythm section. And four walls often have small fills and featured instrumentation occupying isolated channels.

New mixing techniques are being developed for mixing rock music for multichannel. A lot of bands recorded to 24 tracks, which provides ample opportunity for creating three dimensional mixes that are very sophisticated. Elton John, The Doors, Steely Dan, Talking Heads, The Beatles, Yes, Dire Straits, Jethro Tull and King Crimson have all put out fantastic multichannel remixes of their back catalog that blow their original stereo mixes out of the water for clarity and balance. Jazz is being remixed for 5.1 too- Herbie Hancock, Dave Grusin, Lee Ritenour... I just got a 5.1 remix of Miles Davis's B itches Brew that completely replaces the original stereo mix. It's exciting that the art of mixing continues to advance and it can even shine new light on older music.

The only limitation is the implementation of the systems in homes. It can be complicated to set up. But home theater installers are pretty common now and they generally can get multichannel systems working properly. I see a lot of people online who are building home theaters into new houses they are having built. I think eventually, multichannel will be built into homes the same way that nooks for refrigerators and networking ports are.

 

[gregorio]

[1] OBJECT BASED: Sound objects are placed within the room by balancing the sound across more than just two speakers.
[2] This is relatively new and I've found the most vivid examples in music mixed for Atmos.
[2a] This kind of mix is hard for home stereos to reproduce, but when they can, the sound is astonishing. When they can't it turns into a sonic mush. It's kind of all or nothing.

1. No, this is incorrect. We've be able to do that ("sound objects are placed within the room by balancing sound across more than just two speaker") and have done so as a matter of routine for more than 25 years. It's a fundamental facility of 5.1 mixing. The "Object Based Sounds" offered by Atmos differ from 5.1 and 7.1 systems in that the sounds are not directly routed or recorded to the available audio channels (6 in the case of 5.1). Object based sounds are treated independently from the 5.1/7.1 audio channels and only routed to speakers on playback, as decided/determined by the Dolby processor. This only makes much of an audible difference in large venues (such as cinemas) where the side/rear zones are each comprised of multiple speakers and in the fact that the two rows of height info speakers can be addressed. When folding-down an Atmos mix to 5.1 or 7.1, all the object based data is simply recorded down to the 5.1/7.1 audio channels, obviously without any regard to the speaker layout/setup of any individual playback venue, which is the WHOLE advantage of object based sounds in the first place.

2. Object based sounds are quite new but what you are describing is not new. It may just be some expectation bias you're experiencing or it could be that the Atmos mixes you've heard were experimenting with the facilities of Atmos, including those facilities which were ALREADY AVAILABLE in 5.1/7.1 but were used sparingly due to the translation risks (the use of the middle pan position for example).
2a. Yes it is, that's the translation risks/issues I've been talking about, the same translation issues for 5.1/7.1 or Atmos. What you're talking about isn't due to Object Based sounds or Atmos itself, what you're talking about is most likely just more experimental mixes, which could have been done identically in 5.1.

[1] Most mixes blend more than one of these techniques, but tend to favor one kind of mixing over the other. For instance, object based mixes almost always have four walls elements too, usually the rhythm section. And four walls often have small fills and featured instrumentation occupying isolated channels.
2. New mixing techniques are being developed for mixing rock music for multichannel.
3. A lot of bands recorded to 24 tracks, which provides ample opportunity for creating three dimensional mixes that are very sophisticated.
3a. Elton John, The Doors, Steely Dan, Talking Heads, The Beatles, Yes, Dire Straits, Jethro Tull and King Crimson have all put out fantastic multichannel remixes of their back catalog that blow their original stereo mixes out of the water for clarity and balance.

1. A modern surround mix which didn't blend those techniques would be a strange mix indeed. A "tendency to favour one kind of mixing" may be dictated by the original material being remixed and where it isn't, it might be a somewhat erroreous artistic decision and/or simply inexperience. Generally though I dispute your categorisations, generally that isn't how music producers approach or mix surround material unless the material provides no alternative.

2. That's true to an extent. For example, the recent developments in "machine learning" tools and ambience synthesis provide more opportunities for separation, instrument placement and full use of the full surround soundfield. Of course though, separation depends on what's there to be separated in the first place...

3. It doesn't necessarily provide more or ample opportunity for creating sophisticated 3d mixes, it depends what's on those 24 tracks. For example, prior to the advent of 24 track recorders, the drumkit was typically recorded with two channels (stereo left/right) but with a 24 track recorder drumkits were typically recorded using 6-10 tracks. That doesn't provide more opportunity for a sophisticated 3d mix, unless you want to have say the snare drum or hihat positioned completely differently from the rest of the drumkit and that wouldn't work technically in most situations even if it were artistically desirable (which it typically isn't). Others of those 24 tracks are likely used for overdubs and probably some for effects. If your 24 tracks are filled with various synths, perc, backing vox and other instruments in addition to the standard rock band (drumkit, bass, rhythm and lead guitars and vox), then there's obviously a lot more opportunity for placement and filling the soundfield with instruments rather than just ambience or other effects. Furthermore, a standard rock band evolved from a mono consumer market and even when the market went to stereo, everything but the drumkit was still effectively a mono sound, recorded in mono. A limiting factor when trying to turn such an ensemble into a surround mix.
3a. In addition to the previous point, many of those earlier bands did not have 24 track recorders available at the time. Early Beatles only had two track recorders, Sgt. Pepper was their first use of a 4 track and 24 track didn't become available until some years after they split up. The only option in many of these cases is to separate what's there as much as practical, possibly re-arrange the music (maybe using discarded bounced tapes in the rare instances they're still available) and/or add full soundfield effects (for example surround reverbs, delays, etc.) and even new takes of additional instruments.

G

 

[bigshot]

I don't know what we're arguing about. We agree that object based mixing exists. Are we just arguing over it being new or not? Because I know it isn't new. I have pre-Atmos recordings that have sound object placement. No argument there. It's just not done as much because there isn't a lot of control over how it reproduces in most home 5.1 installations. I'm not disagreeing with you. I'm just trying to describe the types of approaches. Of course there are blends of approaches that work better or worse depending on the playback system. My main point was that mixes designed for Atmos like the Beatles Sgt Pepper fold down into 5.1 mixes that don't work properly on a lot of systems, though if a system has a reasonably strong phantom center ine middle of the room, a lot of the appeal of the effects of Atmos come though. Four walls approach is a lot more compatible with less than ideal home theater sound systems.

I promise I'll come back and parse more of your comment later. I really like hearing what you have to say and I learn a lot, but you don't say it in a way that makes it enjoyable to read or reply to. It becomes a chore to translate and reply to point by point replies that make the overall context unclear, so I tend to skip past it because it just takes too much effort. I really don't want to argue for arguing sake, and I don't have a dog in the fight for superiority. .I just like figuring things out and sharing information. It's the internet, so clarity is important. People don't read the internet the same way they read reference manuals.

I have a confession to make... I rarely read footnotes in books either. If they can't say it clearly in the body of the text, I figure it's probably tangential to the point.

 

[bigshot]

I don't have Atmos yet in my system. I probably can't with the room I am working with unfortunately. But as I understand it, there are Atmos encoded blu-rays that have discrete overhead channels. They aren't encoded and decoded from other channels like Dolby Pro Logic IIz. I think DTS has a format that is very similar. Blu-ray Atmos doesn't have as many channels as theatrical Atmos, which I think is capable of as many as 64, but in the average living room it should be able to make a significant improvement on imaging in the center of the sound field.

Also, as I understand it, multichannel playback equipment has a built in ability to fold down mixes with more channels than the system has speakers, and to scale up using DSPs if the recording has fewer channels than there are speakers. Again, nowhere near the precision of theatrical Atmos, but I have heard what is possible scaling stereo up to 5.1 using DSPs and it is a significant improvement.

I don't know exactly how scaling Atmos down works on a technical level, but I've heard the results on my own system with the two examples I have... Kraftwerk's Catalogue box set is a blu-ray with Atmos channels, and the Beatles Sgt Pepper is a mix that was created for theatrical Atmos, then folded down into 5.1 for release on blu-ray. That is two different ways of translating Atmos to 5.1, but the result is the same. Both of these discs have lots of examples of sound occupying the center of the room and crossing from front to back and from one corner to the opposite diagonal corner.

My rear channels are raised a bit to accommodate the layout of my room and when one of these effects happen, it is pretty vivid and pretty tightly produced. There is no dip out in the middle of the panning, and objects are clearly moving through the room, or standing well in front of the front speakers. I only have one other multichannel music recording that does this. It's an Elton John song where the guitar solos starts on the front wall and slowly walks towards the listener, ending when it is about 2/3rds of the way from the front to the listening position. That one was done for a 5.1 mix, not Atmos, so I know that it was possible before. But Atmos gives music producers the tools to do it in a controlled way if the home theater is equipped for Atmos blu-ray playback. And if they do it for Atmos blu-ray playback, it will be folded down into the 5.1 as well, just not as tightly controlled.

The problem is that these sorts of effects seem to fall apart if the 5.1 or 7.1 system isn't well implemented in the room. The Sgt Pepper mix was roundly criticized in internet forums for being "front weighted" and "flat sounding". On my system, it's the exact opposite. I hear depth and sounds within the room. But I can imagine that systems where the front speakers are full range and the rear speakers are small, and in systems where the front and rear speakers are so far apart that they don't reach the middle of the room, that would sound pretty much front weighted and flat sounding. It didn't help that they had a mastering error that attenuated the rear channel -6dB. Most home theater people are terrified of changing settings once they've run Audyssy or had an AV tech calibrate for them. I'm betting a big chunk of the listeners aren't hearing what was intended. That's probably the reason they decided not to go with Atmos for the White Album.

By the way, all of the original elements for the Beatles albums were vaulted. They had all the original 4 track elements that were merged down to sub mixes. They digitized them and rebuilt the whole album digitally from the ground up. They used George Martin's notes and reverse engineered the mixing decisions in the mono mix to create the Atmos mix. It was a huge project overseen by George Martin's son, Giles. He previously had taken the Beatles' TV appearances on film and kinescope and isolated the live vocals from the on-set playback and then re-composited the live vocals over the original instrumentation from the multitrack album masters of the song used to create the on-set playback. That was another monster of a project, I'm sure. Kind of like taking all the pieces of a jigsaw puzzle and putting them back together even though most of the pieces don't fit quite right!

 

[gregorio]

[1] I'm just trying to describe the types of approaches.
[2] I don't know what we're arguing about. We agree that object based mixing exists. Are we just arguing over it being new or not? ...
[3] My main point was that mixes designed for Atmos like the Beatles Sgt Pepper fold down into 5.1 mixes that don't work properly on a lot of systems, though if a system has a reasonably strong phantom center ine middle of the room, a lot of the appeal of the effects of Atmos come though.
[3a] Four walls approach is a lot more compatible with less than ideal home theater sound systems.

1. I realise that you're just trying to describe types of approaches. You've created explanations and conclusions of your impressions of listening to certain mixes on your system and used terminology to describe those conclusions. The problem/danger, is falling into the exact same trap the audiophile world so commonly falls into and for exactly the same reason. Audiophiles listen to mixes, get "impressions" and then create explanations and conclusions from those impressions using terminology. What usually results is confusion from misusing/misappropriating terminology and explanations/conclusions which are inaccurate (or just plain wrong) because there are other facts about the way things work and the way mixes are created which they either don't know about or are failing to consider. For example, it's perfectly possible to listen to a bunch of HRA recordings and gain the ACCURATE impression that HRA recordings are better than CD recordings and conclude that HRA is therefore superior to CD audio. We both know this conclusion is incorrect, this accurate impression can be caused for example by different masters and intrinsically there is no audible difference between HRA and CD audio. To a certain extent this is what's going on here, it's why we have a Sound Science forum and why I'm responding to your posts. For example: ...

2. We do agree that object based mixing exists but unfortunately, you are misusing/misappropriating the term, it has a specific and different meaning from the one you are using it for. What you are describing as "object based mixing" has existed since 5.1 was invented but "audio objects" and "object based mixing" has a completely different meaning and only exists in Dolby Atmos.

3. No, none of the effects of Atmos "come through". The only effects which can "come through" are those effects which can be created with 5.1. The only potential difference is that the 5.1 effects available might have been experimented with and employed more liberally than other 5.1 mixes.

4. The "four walls approach" you keep repeating doesn't really exist, it's just a term you've invented to explain your impressions and, it would be no more or less compatible with less than ideal home theatre setups.

[1] Blu-ray Atmos doesn't have as many channels as theatrical Atmos, [1a] which I think is capable of as many as 64,
[2] but in the average living room it should be able to make a significant improvement on imaging in the center of the sound field.
[3] Also, as I understand it, multichannel playback equipment has a built in ability to fold down mixes with more channels than the system has speakers, and to scale up using DSPs if the recording has fewer channels than there are speakers.
[3a] Again, nowhere near the precision of theatrical Atmos, but I have heard what is possible scaling stereo up to 5.1 using DSPs and it is a significant improvement.
[4] ... Both of these discs have lots of examples of sound occupying the center of the room and crossing from front to back and from one corner to the opposite diagonal corner.

1. Yes it does! Blu-ray Atmos has 8 channels (7.1), the same as theatrical Atmos. The problem here is that you are thinking in terms of the old paradigm, the paradigm that applied to all previous formats, which is audio channels routed to speakers. Dolby Atmos creates a new paradigm and you have to think about it differently or end-up making incorrect assumptions/conclusions! Atmos has the same number and function of audio channels as 7.1 but adds a new concept, the concept of "Audio Objects". Audio objects are certain sounds which can be designated as "audio objects" and are not recorded to any audio channel, they just exist as data/metadata. The decoder/processor then assigns these audio object directly to the appropriate speaker/s. In theatrical Atmos, the decoder/processor is in effect programmed with all the cinema's speakers and their position during installation and this is the real advantage of Atmos. If you have a small cinema with say 25 speakers or a large cinema with 120 speakers, the processor can intelligently work out for an audio object when and how to use a phantom position or point source position depending on what speakers the system has installed. So for example, let's take a sound panned to the centre position of a standard 2 channel stereo system. Obviously this centre position is a phantom centre but what happens if we add a centre speaker and a third audio channel? When mixing we now have the option of doing the the exact same thing, creating a phantom centre by recording the audio equally to both the left and right channels but we also have the different option of recording this sound to the third audio channel which is routed to the centre speaker. The advantage of using this second option is that we get an actual centre position rather than just a phantom centre but the disadvantage is that this mix is no longer compatible with a standard 2 channel stereo system, everything in that 3rd audio channel will simply be missing. Now let's apply the new Dolby Atmos paradigm to this example. We still have two audio channels but in addition we have the concept of an "audio object". When mixing we can still use the two audio channels as we would with a standard stereo system and pan our sound as a phantom centre but if we want, we can assign this sound as an "audio object" (which is not recorded to any audio channel/s) and pan it to the centre. What happens then is that the decoder/processor knows whether your specific system has an additional centre speaker or just a stereo pair and on playback routes the audio object (our sound) to the centre speaker if there is one or applies the panning law and routes it to the left and right speakers equally to create a phantom centre if we don't have that additional centre speaker. From this somewhat simplified example, I hope you can now see that if you've only got a stereo two speaker setup, the end result is exactly the same, regardless of whether you mix using the "audio object" paradigm of Atmos or just a standard stereo mixing. The real advantage of Atmos only makes a difference in cinemas, large spaces with loads of speakers. Of course, this explanation of Atmos ignores the height info, which is also an advantage/improvement over 5.1/7.1 but that's all folded down when reproduced on a 5.1/7.1 system anyway and I've ignored it to make the explanation simpler.
1a. Without looking up the specs again: Dolby Atmos has 8 audio channels, up to 64 audio objects and (if I remember correctly) can address a system of up to 128 speakers. But remember, the paradigm of audio channels and speakers is quite different with Atmos.

2. No, it won't make any difference to the imaging of the middle position of the soundfield.

3. That same fold-down ability (from say 5.1 to stereo) is exactly the same as what Atmos uses. Atmos does not use the DSP upscaling ability found in some AVRs though.
3a. I have a somewhat different opinion of consumer upscalers. Some of the early ones just applied a big surround reverb to the stereo and more recent ones are more sophisticated, they attempt to deconstruct the stereo mix and extract elements which are then placed around the soundfield with some surround reverb added. With some material this more recent approach works surprising well but other times not so well IMO. Some people though just like sound coming from all around them and prefer even poor upscaling to stereo. It also partly depends on the acoustics and quality of the stereo system, how well the stereo system can reproduce the depth cues. None of this is related to Atmos though, there is no upscaling, extracting elements of the mix or equivalent DSP effects. The problem of extracting elements to route to additional speakers is bypassed with Atmos by not having those elements "mixed" in the first place, in effect they're stored separately (outside the mix) and then added to the mix on playback by the decoder/processor. This obviously isn't the case with a mix that's been folded-down prior to distribution.

4. That's just a mixing decision which has nothing to do with it being originally mixed in Atmos. They could have done the exact same thing in 5.1 or 7.1. Your point here resembles the HRA/CD example above, HRA recordings can sound better than CD recording but because of somewhat different mastering choices, not because of the difference between Hires and CD.

[1] ...It's an Elton John song where the guitar solos starts on the front wall and slowly walks towards the listener, ending when it is about 2/3rds of the way from the front to the listening position. That one was done for a 5.1 mix, not Atmos, so I know that it was possible before. But Atmos gives music producers the tools to do it in a controlled way if the home theater is equipped for Atmos blu-ray playback.
[1a] And if they do it for Atmos blu-ray playback, it will be folded down into the 5.1 as well, just not as tightly controlled.
[2] The Sgt Pepper mix was roundly criticized in internet forums for being "front weighted" and "flat sounding". On my system, it's the exact opposite. I hear depth and sounds within the room. But I can imagine that systems where the front speakers are full range and the rear speakers are small, and in systems where the front and rear speakers are so far apart that they don't reach the middle of the room, that would sound pretty much front weighted and flat sounding. It didn't help that they had a mastering error that attenuated the rear channel -6dB.

1. No, Atmos does not provide the tools to accomplish that effect in a more controlled way, neither for a home theatre nor a cinema.
1a. The tightness of control you end-up with should be exactly the same as if it was just mixed in 5.1 in the first place, although it is possible under certain circumstances that the folded-down Atmos mix would be less "tightly controlled"! Again, your impressions are the result of mixing decisions, not the format (Atmos) itself. Most likely what you're hearing is the result of more effort, time and money being spent to create the Atmos mix.

2. If there was a mastering error with the rear channels being -6dB then the criticisms were justified, it should sound "front weighted and flat"! The fact it doesn't sound like that on your system indicates that your system is incorrectly calibrated, your rear speakers are too loud.

G

 

[bigshot]

I think we're talking on different wavelengths. I'm not talking about theory or mixing techniques in general. I'm talking specifically about recent trends in mixing for multichannel popular music for home audio- Blu-ray Audio, DVD-A and SACD. Two of the top engineers in this field right now are Steven Wilson and Giles Martin. Their styles are quite different because they have different techniques and goals for their mixes. Have either of you heard any of their work? Are you familiar with any multichannel music recordings, particularly in the pop/rock field? Have you heard one of the new Atmos music albums (Sgt Pepper, REM Automatic For The People, Kraftwerk Catalogue) on a good home Atmos system? I would love to hear about that. I've only heard Atmos on 5.1. And it appears that different 5.1 systems render Atmos radically differently. (I already explained why.)

I'm talking about something really specific, but the replies keep flying off into theoretical, historical and technical tangents. I'm trying to be really clear and give examples of what I'm talking about, but the replies don't seem to even relate to the recordings I'm discussing. It feels like when I asked for specific examples of binaural music recordings that showed the full potential of the technique and everyone kept talking about theory and no one could recommend any specific good examples for me to check out. Everyone knew all about binaural recording theory, but no one seemed to actually listen to it! Well, I listen to multichannel music. I have hundreds of albums in quad and 5.1. I'd love to be able to talk to someone here about it who also has a multichannel system.

I guess multichannel music is such a niche market, very few people keep up with it. It's a shame because 5.1 and Atmos are the biggest advances in sound quality since the introduction of stereo, and multichannel audio opens doors for whole new ways of constructing mixes.

Over a decade ago, Capitol records released a line of 5.1 DVD-As where they went into their back catalog and reworked mixes to show off the potential for bringing new life to old music. They released a bunch of discs called "Inside the Music" which were compilations based on different genres in the Capitol catalog... 60s pop vocals, jazz, new wave, country music, rock n roll, new age, etc. I've been picking these up for cheap because no one wants DVD-A any more, and I can feel the excitement that they felt when this was all new. Hearing Dean Martin sing "Ain't That A Kick In The Head" or Buddy Rich's band from the late 60s or George Jones in glorious surround is a revelation. There's some incredible sound embedded in old multitrack masters that just needs a little bit of modern remixing for 5.1 to bring out. Elton John already did this brilliantly with his albums. So has Jethro Tull, King Crimson and Yes. The Beatles albums are being remixed for surround now too. The multichannel remixes blow the old stereo albums out of the water. Things are happening. It's a small market share, but I think it's important because it shows the future isn't just earbuds and bluetooth boom boxes.

I get really excited about this stuff and everyone who hears these new remixes on my system are blown away. But it seems like most people aren't even aware of what is going on. They think multichannel is just for Star Wars and superhero movies. Kind of discouraging.