How are Dolby Atmos mixes created?

[mhardy6647]

You aren't a pro mixer. You aren't mixing in Atmos. And I'm not convinced you have a clue how many Atmos mixes are being dumped into streaming every week. Hence your posts are full of suppositions.

Here's an actual Atmos mixer. Does his mix process (after all the setup, which he can use in multiple mixes) sound like it happens 'in no time'?

One could - and it might even be interesting so to do - effect a Fermi estimate based of the number of tracks mixed to ATMOS to date, the number of mixers (i.e., people) doing the mixing, and the amount of time (total hours) that have elapsed since the first ATMOS-mixed track was uploaded.

 

[59Tweed]

Multi track for commercial product has been reality for 65 yrs since stereo pop music records hit the market. There are multitrack sessions on tape by the ton. There are industry towns like Nashville that have several recording studios in every neighborhood. Conversion to Atmos has a nominal requirement of 12 speakers and associated support equipment. An unused b room could be converted in a long weekend.
Given the present reality that school children are packing a multi track production rig in their smart phones, pro recording houses are stuck with the dilemma of competing with bedroom studios. Minimum spec licensed dolby atmos audio recording software is below $100 now. If they are fancy they could deploy an array of THX spec speakers ( calibrated ) ordered from Amazon for a modest amount.
The money is in recording and mastering in Atmos now. Content creaters are paid a higher percentage for product delivered to Apple. Thats "spatial" in Apple world. The financial incentive is there. Converting an existing control room to a wall and ceiling mounted playback system is not very challenging.
The short answer is there are lots of Atmos mixing rooms going on line presently. Ambitous content creators can afford an Atmos room.

 

[krabapple]

No one is saying multitrack recording is new!

What is new is the sheer volume of new and old consumer releases in multichannel format.

For new recordings by current artists, where the digital multitracks are ready to hand, this is 'simply' a matter of making a MCH mixdown as well as a stereo mixdown.

For 'legacy' recordings, it means 1) exhuming the old multitrack tapes/files, then 2) archiving them to digital if they are tapes , then 3) making a new MCH mix from the digital multis.

for 1, sometimes tapes/files/elements go missing and take detective work to find, depending on how diligent the record companies have been about storage
for 2, this along with 1 may or may not have been done already , at some time in the past.... or it might still need doing
for 3, you've also got to keep the familiar and perhaps beloved original stereo mix in mind


And all those things make me wonder about the flood of legacy remixes we are seeing...the how of it.

 

[rirelien]

And all those things make me wonder about the flood of legacy remixes we are seeing...the how of it.

Listening to an Atmos mix of Patricia Barber’s Modern Cool while responding: check out interviews with people like Steven Wilson, who has become well-known for doing immersive mixes of older well-known albums (not the Barber, though.)

 

[goat76]

No one is saying multitrack recording is new!

What is new is the sheer volume of new and old consumer releases in multichannel format.

For new recordings by current artists, where the digital multitracks are ready to hand, this is 'simply' a matter of making a MCH mixdown as well as a stereo mixdown.

For 'legacy' recordings, it means 1) exhuming the old multitrack tapes/files, then 2) archiving them to digital if they are tapes , then 3) making a new MCH mix from the digital multis.

for 1, sometimes tapes/files/elements go missing and take detective work to find, depending on how diligent the record companies have been about storage
for 2, this along with 1 may or may not have been done already , at some time in the past.... or it might still need doing
for 3, you've also got to keep the familiar and perhaps beloved original stereo mix in mind


And all those things make me wonder about the flood of legacy remixes we are seeing...the how of it.

1. I would say most of that job has already been done, the recording companies have been releasing stereo remixes over the years so many of them have already been digitized because of that.

2. And many of them have been digitized for archiving reasons as the companies fully well know that the old classic recordings are their “gold”, which they have known forever, just count all the remixes and remastered they have released over the years.

3. It doesn't have to be neither more time-consuming or harder to make a multi-channel mix than making a stereo mix, and when the record companies want their classic recordings released as multi-channel versions, pretty much any mixing engineers out there with a Dolby Atmos compatible studio will stand in line to do it.
And when it comes to how important or not it is to keep the mix close to the original stereo mix, it will probably be an ongoing dialogue between all involved about what will be done.

 

[krabapple]

Listening to an Atmos mix of Patricia Barber’s Modern Cool while responding: check out interviews with people like Steven Wilson, who has become well-known for doing immersive mixes of older well-known albums (not the Barber, though.)

Why? He was doing remixes before the Atmos flood, and they came appeared at a steady but not swift pace. I've already read about his process. It doesn't sound fast.

My question is about how the industry is getting this done in bulk.

 

[krabapple]

1. I would say most of that job has already been done, the recording companies have been releasing stereo remixes over the years so many of them have already been digitized because of that.

Stereo remixes have always been a very minor proportion of commercial releases, other than in dance music.
Mostly its been remasters, which don't require multitracks access

2. And many of them have been digitized for archiving reasons as the companies fully well know that the old classic recordings are their “gold”, which they have known forever, just count all the remixes and remastered they have released over the years.

This may well be. But again, remasters never required access to multitracks. I would bet the analog stereo masters of popular records have long since been archived to digital by now. But is that similarly true of the multis?

3. It doesn't have to be neither more time-consuming or harder to make a multi-channel mix than making a stereo mix, and when the record companies want their classic recordings released as multi-channel versions, pretty much any mixing engineers out there with a Dolby Atmos compatible studio will stand in line to do it.
And when it comes to how important or not it is to keep the mix close to the original stereo mix, it will probably be an ongoing dialogue between all involved about what will be done.

Which takes time, no?

 

[goat76]

@krabapple

As you seem to have a hard time believing that there are enough mixing engineers with Atmos-capable studios making all those Atmos mixes found on Apple Music and Tidal, in what alternative way do you believe those Atmos mixes were done, do you mean they were generated by some sort of AI process, or what?

 

[krabapple]

I'm not interested in engaging further with you on this.. You have nothing to offer me.


But people like me on ASR * who have actually been buying and listening to consumer digital multichannel consumer releases since the turn of the millennium -- that's DVDA, SACD, DualDisc, BluRay, not to mention the older DTS discs -- know that the pace of mch releases we are seeing now is highly unlike anything seen for most of those 20+ years.

I'm interesting in hearing from people actually know how that is being accomplished


*e.g. Sal1950 who like me has been on the QuadraphonicQuad forum for years.

 

[holdingpants01]

I'm not interested in engaging further with you on this.. You have nothing to offer me.


But people like me on ASR * who have actually been buying and listening to consumer digital multichannel consumer releases since the turn of the millennium -- that's DVDA, SACD, DualDisc, BluRay, not to mention the older DTS discs -- know that the pace of mch releases we are seeing now is highly unlike anything seen for most of those 20+ years.

I'm interesting in hearing from people actually know how that is being accomplished


*e.g. Sal1950 who like me has been on the QuadraphonicQuad forum for years.

it's done exactly like he said, it seems like you just don't want to believe it. Last week I did two stereo to atmos transfers of live concerts, both around 1hr, it took me maybe 4hrs in total for both, as I had stems from the stereo mix ready to go and there was no automation needed or any special effects

 

[goat76]

it's done exactly like he said, it seems like you just don't want to believe it. Last week I did two stereo to atmos transfers of live concerts, both around 1hr, it took me maybe 4hrs in total for both, as I had stems from the stereo mix ready to go and there was no automation needed or any special effects

According to Andrew Scheps, some aspects of mixing in Atmos is even easier to do as you don't have to struggle as much in finding space for the different sound objects in the mix, thanks to the larger pallet.
And not just that, thanks to the larger space for positioning objects, there's also less need for carving out space and less need for compression on individual sound objects as there are no longer as many problems with frequency masking, which otherwise is one of the main problems in stereo mixing. That is probably the main reason why a multi-channel mix doesn't have to take much or any longer time to do than a stereo mix, even if it may seem like a way more complex thing to do for someone who have no idea of how mixing is done.

 

[Chr1]

Gotta love technology.
Have to say that I am not particularly knowledgeable and mostly listen to upmixed two channel, but I love that the apparent simplicity of modern object oriented production has emerged.
I mostly listen to dub and electronic music upmixed to DTS:NeoX and reckon it works incredibly well. As good as most discreet multichannel mixes I've heard. Reckon this is genre dependant personally however.
Apologies if off topic.

 

[rirelien]

it's done exactly like he said, it seems like you just don't want to believe it. Last week I did two stereo to atmos transfers of live concerts, both around 1hr, it took me maybe 4hrs in total for both, as I had stems from the stereo mix ready to go and there was no automation needed or any special effects

Would you be in a position to share more details about your process and decisions? Which stems went to which channels, in what proportions? Same for the entire concert or different from track to track? Etc.

And… if you do this professionally, can you point us to publicly available examples to add to your explanation?

To be up front about my bias: I would think this would take 3-4 hours PER TRACK if you moved very fast indeed.

If you had tracks for venue and audience ambience already, and just had to recreate the front stage, I can see that this could be quicker, but not as fast as you say.

The mystery variable here is the quality of the end result, which is of course just opinion. Hence would love to have a reference recording or three you’ve produced to hear.

 

[krabapple]

it's done exactly like he said, it seems like you just don't want to believe it. Last week I did two stereo to atmos transfers of live concerts, both around 1hr, it took me maybe 4hrs in total for both, as I had stems from the stereo mix ready to go and there was no automation needed or any special effects

You had 'stems from the stereo mix ready to go' meaning you already had source in hand. Meaning, part of the 'work' was already done: someone supplied you with them and even had grouped multis to stems.

Was this a legacy remix, or something new, meaning, you were doing the first stereo and mch mixes of a recent recording?

Maybe you guys should reread what I wrote : I am talking about the flood of new multichannel mixes of old, previously mixed/released stereo material, going back five decades. Do you even know what I'm referring to? Have you ever even owned a DVDA , an SACD, a BluRay mch rerelease of some album first released in stereo in the 1970s? These MCH products were not typically whipped up on 2 hrs from conception to market.

 

[rirelien]

Maybe you guys should reread what I wrote : I am talking about the flood of new multichannel mixes of old, previously mixed/released stereo material, going back five decades. Do you even know what I'm referring to? Have you ever even owned a DVDA , an SACD, a BluRay mch rerelease of some album first released in stereo in the 1970s? These MCH products were not typically whipped up on 2 hrs from conception to market.

Also really interested!

I’ve been interested in MCH for (only) about 20 years, and have a DVDA of an old quad mix from the 70s (Randy Newman) as well as what for me really opened my ears: a recording of a large number of Bach pieces in the church at Kings College Cambridge (recreating somewhat the ambience of the venue), and Bjork’a Vespertine (the 5.1 mix created at the same time as the 2.0 mix I believe.)

But it doesn’t seem to me, an outsider, to be a small undertaking. Easiest if done at the same time as the 2.0 mix, for studio-driven works; for live recordings, you need the additional mic positions I believe.

(I tried the Miles Davis Kind of Blue 5.1 mix and don’t enjoy it, conversely. Ambience added after the fact. Not for me, for this kind of recording.)

 

[holdingpants01]

You had 'stems from the stereo mix ready to go' meaning you already had source in hand. Meaning, part of the 'work' was already done: someone supplied you with them and even had grouped multis to stems.

Was this a legacy remix, or something new, meaning, you were doing the first stereo and mch mixes of a recent recording?

Maybe you guys should reread what I wrote : I am talking about the flood of new multichannel mixes of old, previously mixed/released stereo material, going back five decades. Do you even know what I'm referring to? Have you ever even owned a DVDA , an SACD, a BluRay mch rerelease of some album first released in stereo in the 1970s? These MCH products were not typically whipped up on 2 hrs from conception to market.

I had stereo mix ready to go because I did it before. Basically all major label records have the mixed and digitized stems already in archives anyway. All the current work I do for bigger labels is finished with deliverables that include stems. All older rereleases like 5.1, 7.1, Guitar Hero, remixes, using songs for commercials or movies needed them anyway, so they should be prepared long time ago, even just for archiving purposes so it doesn't matter how old they are. There are some big projects like Michael Jackson's Thriller that was mixed from the start in stereo to make it sound like original and then transferred to atmos, those take time and money, but most really not and there are a lot of people doing it

 

[holdingpants01]

Would you be in a position to share more details about your process and decisions? Which stems went to which channels, in what proportions? Same for the entire concert or different from track to track? Etc.

And… if you do this professionally, can you point us to publicly available examples to add to your explanation?

To be up front about my bias: I would think this would take 3-4 hours PER TRACK if you moved very fast indeed.

If you had tracks for venue and audience ambience already, and just had to recreate the front stage, I can see that this could be quicker, but not as fast as you say.

The mystery variable here is the quality of the end result, which is of course just opinion. Hence would love to have a reference recording or three you’ve produced to hear.

In this case it was live recording with a video capture, so I just used the placement from the stage. Proportions were done in stereo and musicians accepted it so I didn't change them, only adjusted if something sounded very different from the stereo (I had it available in the same session so I could quickly compare). None of the tracks moved from track to track, just like musicians didn't, it was a live concert. 3-4 hours per track for a live material already mixed in stereo while using the session template for atmos (so absolutely no settings need to be adjusted to start the work)? That's just bonkers. Mixing a 100 tracks song from the scratch can take that long, not this.
Sorry, I won't show any of my work here, I prefer to stay anonymous

 

[Sancus]

None of this is answering the question, how are SO MANY Atmos mixes -- with actual content in surround channels, not newly recorded room ambience -- being generated so fast?

Is it really that fast? I think this is just an example of some investment, no magic at all. Apple Music mostly bans algorithmic content and while some has slipped through, they do remove it as soon as they find it, I've seen it happen. I can't find a recent source on exactly how many Atmos tracks are on Apple Music, but it's definitely in the low to mid single digit thousands of albums, or tens of thousands of songs. How many albums in total are released per year? Hard to be sure, but it's for sure more than 100,000.

So, being conservative, that's only 1% of all music being produced each year that's in Atmos. I would say 1% of very professionally produced music, because if you include literally anything that is uploaded to a streaming service it's more like *millions* of albums per year. I would guess Apple Music is adding maybe 1000 albums per year worth of Atmos music.

All you really need to mix an Atmos album is the correct gear and software from Dolby. It's hard to pin down exactly how much it costs to mix an Atmos album, but my guess is around $5-10K for 10 tracks. That isn't much in the context of popular artists. My favourite composer, Christopher Tin, did a Kickstarter 3 years ago and one of the stretch goals was to remix all of his albums in Atmos. The total additional amount was $5K per album($25K for 5).

If you do the math, that's about $5M per year invested in Atmos mixing. That is a pretty small number, so I wouldn't be surprised if they're doing more than 1k albums per year by now.

 

[goat76]

All older rereleases like 5.1, 7.1, Guitar Hero, remixes, using songs for commercials or movies needed them anyway, so they should be prepared long time ago, even just for archiving purposes so it doesn't matter how old they are.

Yes, most of the old music that people still recognize as old classics and hits is because you hear them everywhere, and in all types of situations like games, movies, commercials. It's not like all of them had to search the old archives for the original tapes and digitize them every single time, that has in most cases been done a long time ago and the digital tracks just have to be sent the Atmos mixing engineer who got the job.

Once in a while, it may be more digging to do which will be more time-consuming. But that is probably a rare occasion, and it always end up as a sunshine story on the interweb when the personnel involved are bragging about their achievement that took this forgotten gem and made it available for the generations to come.

 

[krabapple]

I had stereo mix ready to go because I did it before.

Which is rather unlike the cases I'm talking about. And you're apparently talking about a (recent?) live show recording, not, say, an old Donna Summer album. There is nothing to compare it to. Yours is not like the cases I am talking about.

Basically all major label records have the mixed and digitized stems already in archives anyway. All the current work I do for bigger labels is finished with deliverables that include stems. All older rereleases like 5.1, 7.1, Guitar Hero, remixes, using songs for commercials or movies needed them anyway, so they should be prepared long time ago, even just for archiving purposes so it doesn't matter how old they are. There are some big projects like Michael Jackson's Thriller that was mixed from the start in stereo to make it sound like original and then transferred to atmos, those take time and money, but most really not and there are a lot of people doing it

I'd love to compare your 'easy as pie' scenario with, say, Steven Wilson's workflow. I don't see him cranking out a classic album multichannel remix in two hours, though he can probably do an Atmos remix of *something he's already mixed in 5.1 before* pretty efficiently.

If you're really just saying, that amount of care just doesn't go into these Atmos remixes -- a lot of them of albums that have never had surround mixes at all before -- that's claiming something too.