Joe Grasso is at the forefront of the immersive music scene. Over the past three years, he’s already created Dolby Atmos mixes of iconic songs from The Killers, Marvin Gaye, The Weeknd, Norah Jones, and other well-known artists.
I recently had the chance to chat with Joe about his immersive audio journey, the complexities involved with remixing classic songs in Atmos, and where he thinks the industry is headed.
How did you first get into audio engineering and mixing? Who introduced you to the idea of mixing music in Dolby Atmos?
I was a high school kid who loved hip-hop music and wanted to be part of that scene in some way. Production and recording ended up being my lane, and I went on to college to do a four-year program for music business and audio recording technology.
I started off interning for Daddy’s House recording studios in Manhattan, and then took a job at Columbia Records. At the time, Columbia didn’t have any in-house engineers. It took a while to build trust there, but eventually they started leaning on me for some recording and mixing opportunities. The turning point was in 2019, when I got to work a little bit on Lil Nas X’s “Old Town Road.”
In 2021, I first got connected with my now-mentor Ken “Duro” Ifill–who’s a legendary recording and mix engineer. He was the first person to tell me that there were going to be a lot of work opportunities involving Atmos, so I started to learn how to mix in the new format. The following year, I left Columbia and became a fully freelance mixing engineer.
When it comes to mixing music in Atmos, what’s your philosophy? Some mixers seem to prefer a more conservative approach, while others are more experimental and place isolated instruments behind or above the listener.
My number-one rule with Atmos is to honor the integrity of the stereo mix. What I’m primarily listening for are things that I would have done in stereo, but couldn’t because there are only two speakers.
For example–with background vocals, we’ll often use a lot of processing tricks involving reverb and phase to try and move them behind the lead vocal in the stereo picture. In a 14-speaker Atmos environment, you can actually position those vocals behind or above the listener.
I know some other mixers take a gentler approach to the format, placing just minor accents in the rear or ceiling speakers, but I’m a little more heavy-handed with it. I look at Atmos kind of like pizza dough. It's a question of how far you can roll the dough before it falls apart–and that varies depending on the DNA of the song–but overall I like the idea of putting the listener in the middle of the action.
So I’ll usually keep the ‘meat and potatoes’ of a song upfront–like the drums, bass, and lead vocal–but other elements like the background vocals and sound effects can go all around you. I try to make it more of a 4D experience in relation to the traditional 2D stereo mix.
I’m always frustrated by mixes that keep all the instrumentation upfront and then use the rest of the space just for reverb returns or the odd sound effect. What’s the point?
You’re right. To really show off what this new technology can do, you have to be a little bold with it.
That said–I've worked on projects where the artist preferred a more conservative approach, or there was a really limited stem count. If it’s a simple acoustic piece with live drums, a single guitar, and vocals, there’s only so much you can do. I’m not going to push that guitar all the way to the rear.
So you really have to take the song’s DNA into consideration. But when there’s ample stems, that's the perfect opportunity to really immerse yourself.
In the past several years, you’ve worked on so many Atmos projects across a wide variety of genres. The bulk of it appears to be more recent music, but some of it also is vintage catalog material from past decades. What kind of multitrack assets do you typically receive from the label? Is it a DAW session with all the original mixer’s effects committed or tracks frozen? Is it pre-processed stem groups that add up to the original stereo mix at unity gain? Or in the case of older music from decades past, are you getting completely dry analog tape transfers?
That’s a great question, and the answer is it’s a bit of everything. For songs already mixed in stereo, I’m usually getting a set of pre-mixed stems and the stereo master to match them to. So the first step in the process there is to balance out all those stems, then figure out how to add back in that extra layer of processing that happened during the stereo mastering. Once I have the ‘sound’ of the stereo master, then I’ll start moving things out into the immersive space.
There are some catalog titles I’ve worked on going back as much as 30 or 40 years, and for those I’m getting a completely-dry multitrack from analog tape transfers. So for those projects, I have to actually remix them in stereo first. In order to match the original stereo mix as closely as possible, I’ll spend a lot of time researching what kind of outboard gear they used in the original sessions–tape machines, consoles, reverbs, etc–in order to recreate those sounds.
For the newer projects where you’re being sent pre-combined stems, is it possible to ask the label or stereo mix engineer for more of the individual elements broken out? For example, if you were sent a single stereo stem of the entire drum kit with all the processing burned in, could you ask for separate prints of the kick, snare, overheads, and reverb returns, or do you just have to make it work with whatever you have?
It depends. Some of these projects have super-tight turnarounds, and sometimes the stems have been printed by the stereo mixer in a very specific way so they could be run through the mix bus appropriately. So there are circumstances where I probably couldn't get them broken out further, but all the same there are plenty of times where I’ve told them that I need more to complete the Atmos mix.
I definitely like them to be as individually broken-out as possible. I don't want a drum group stem or an instrumental group stem. I want everything separated–even the reverbs and delays, if possible–so I can make the mix really immersive. The best scenario is when I’m mixing the same song in both stereo and Atmos, so I can make sure I have every individual element broken out.
I suspect that the underwhelming quality of a lot of Atmos mixes I’ve heard on the streaming services has to do with limited stem counts, but it’s also difficult to translate certain genres into the format effectively. I think of a lot of early-2000s rock music–bands like The Killers or Green Day, where that big wall of distorted guitars is a key part of their sound. You can’t just put the drums in the front and then stick all those guitars in the back, it’ll sound weak and diluted.
You’re right, certain genres are actually tougher for me to do in Atmos. I find rock to be especially hard, because each element is supposed to bleed a little bit into the next. As you say, pulling the guitars away from the drums can take away a lot of the ‘punch’ in the stereo master. It can start to feel hollow and empty.
When working on that kind of music in Atmos, I’m very careful in making sure I haven’t ‘unglued’ a record. If I was in a circumstance where all I had to work with was the drum kit as a single stem and all the guitars as another stereo pair, I might opt to put the guitars more in the wide speakers than the sides so those elements stay better connected.
So your studio is 9.1.4, with front wide channels?
Yeah, at Republic we’re fortunate to be working in a 9.1 room. I also use a lot of phantom points when mixing in Atmos–like rear wides, top center, and rear center–to make sure that I’m really covering each quadrant.
That brings me to another question. When working in Atmos, do you primarily only use objects or a combination of beds and objects?
Yeah, I’m working mostly using objects or object beds.
I’ll usually first try to get coverage in the fronts, sides, and rears. I think those three spaces form the basis of an Atmos mix. Once they’re covered, I’ll start to get creative and fill in the blanks. For the wides, I’m looking for elements that connect my fronts to my sides or my sides to my rear.
But again, it's super-important to me that no sound is placed randomly. It all has to relate back to the original stereo experience in some way. If there’s anything super reverb-y or space-y in the stereo mix, to me that implies there was an intentional choice by the mixer to set that element further back in the picture. So in Atmos, it’s logical for me to move that same element further away from the listener. All these decisions have to connect back to the stereo.
A big point of controversy with immersive music, going back even to the introduction of 5.1 in the early-2000s, is how to best utilize the center speaker. I’ve heard some Atmos mixes with a totally isolated dry vocal in the center channel, but those seem to be more the exception than the norm. How do you typically involve the center?
I’m very light on both center and LFE. With the center, I achieve most of what I want using the phantom imaging in the front channels. I’ll use the actual center speaker for some minor flourishes, but not every mix calls for it to be used. Same for the LFE–I’ll only send a lot of information to the sub if the song calls for it. Overall, it’s not my go-to.
Going back to what we were talking about before with older catalog remixes, what’s an example of an album you had to rebuild from vintage analog tape transfers?
There are quite a few classic Marvin Gaye releases I’ve mixed in Atmos, and those came in as completely-dry multitracks in need of restoration. It was really an honor to get to work on those albums. Fortunately, I worked with a great team that helped prep those sessions–they’d start off with some basic leveling and panning, so I’m not just looking at a completely raw tape transfer.
The interesting thing about these projects is that I have to go pretty heavy-handed on the processing, because back in the day they weren’t so gentle with these things. It was all outboard gear, and they’d turn those knobs pretty aggressively in order to get the tones you hear on the final mix. So once everything’s balanced out, I’ll lean pretty heavily into the compression and EQ.
After that, the reverbs are really the key to those old records. They’d often send multiple elements through the reverb chamber at the same time, but today we can set up different aux returns for each instrument and position them all throughout the room.
Being from the early-70s, I assume those Marvin Gaye records were recorded on 16-track or possibly even 8-track?
Some of the earlier stuff was 8-track, but a fair share of it was 16-track as well.
The thing that’s fascinating to me about these older records from the ‘70s and ‘70s is how they’d have to commit certain effects during the tracking process. I understand that it wasn’t uncommon for drums to be recorded directly to a single channel on the master reel back then, often with outboard EQ and compression. So I bet that a faders-only rough mix gets you most of the way there on those projects, because a lot of the ‘sound’ of the finished stereo mix is baked into the recorded tracks.
Yeah, they’re not that far off. What tends to be missing a lot of the time are these subtle bits of dynamic mixing, like volume rides or swells, that add excitement.
The tones they captured on tape at Motown were amazing, but that isn’t always the case with older recordings. I imagine it must be tempting to try and enhance some of these vintage albums using digital tools–like layering in samples to reinforce the drums, or using de-mixing software to further separate the instruments.
This is something that came up when we were working on Marvin Gaye. The goal was still to make it sound like the original, but add in more of that low-end and body we’ve become accustomed to in 2024. If it ends up on a spatial playlist on Apple Music, I don’t want it to sound thinner than everything else.
A big part of the “Motown sound” was heavy use of mix bus compression, in order to give the finished product that trademark ‘punch’ and grit. When working in Atmos, it’s not really possible to do this because objects can’t be sent through a bus. So how do you approximate that effect?
In rare circumstances, if I’m really stuck, I’ll actually take all the stems in my session and re-print them through a stereo bus compressor in the same way I would for a traditional stereo mix. Then, I’ll do the spatial positioning using these new stems.
That said, sometimes it’s nice to have the option of letting a song be a little more dynamic in Atmos. As much as I like to get the mix sounding as identical to the stereo master as possible, I also do keep in mind that this is a whole new format and a whole new mix.
If there's a song where the low-end is really blown out and distorted, maybe the Atmos version doesn't need to be as blown out and distorted. Maybe this is the dynamic version of the mix. So again, it's whatever the song calls for and whatever the situation calls for.
Another big project you worked on recently was The Killers’ Rebel Diamonds compilation, which collects many of their greatest hits from across the past two decades. Some of their earliest and most popular records, like 2004’s Hot Fuss and 2005’s Sam’s Town, were released at the height of the ‘loudness wars.’ Since fans are so used to hearing the super-compressed stereo versions of those songs, did you find it difficult to make them work in Atmos without diluting the ‘punch’ and overall power?
Absolutely. Sometimes you have to sacrifice a little bit of immersion to make sure that the integrity of the song is still intact. With those songs, the Atmos soundstage would end up being more like a half-circle from the fronts stretching out towards the sides. From there, I’d pick a few elements and start pulling them further back all while trying to keep everything glued together.
With the introduction of Apple’s “Spatial Audio” format, it seems most listeners are experiencing a binaural approximation of Atmos over headphones rather than the full immersive experience on a home theater setup. Is it difficult to simultaneously achieve effective results on both speakers and headphones? Do you check how your mixes sound with both the Dolby binaural rendering and Apple’s Spatial Audio?
I'll try to keep this as brief as possible, but there's a lot to digest. I do think it’s important to make sure the mix works in headphones, because that’s how most listeners will first experience Atmos music.
So I’ll make sure the headphone rendering is as identical as possible to the stereo master, though it should be a new version of the song with its own spatial details. In general, I’m using about half of my time mixing in the room and then the other half is spent checking on headphones. Sometimes I do have to go back and make changes to the room mix so it all translates.
If Atmos was only a speaker-based format, I would probably do a lot of things differently. I might turn certain sounds up louder, and play around a bit more with movement or panning. Once I think I’ve achieved a good balance between the binaural and room mix in the Dolby Renderer, I’ll bounce out an MP4 and check how it sounds in Spatial Audio on my AirPod Pros.
After much trial-and-error, I think I've got a pretty good formula now where the mixes translate really well in all DSPs.
I’ve done a few mixes myself and I find that it really is hard to achieve an effective compromise between the speakers and headphones. For example, it might sound really good in the room to have a harmony vocal line pushed entirely into the height speakers. But then when you put on headphones, that part sounds distant and buried in the mix.
Yeah, it’s a really tricky balance. But even with stereo, it’s paramount to ensure your mixes sound good everywhere. I can’t say “well, this mix sounds great in the studio but it doesn’t work on bluetooth speakers.” That’s just not an option in today’s market.
The ultimate win for Atmos mixes is when you find that balance–a room mix that sounds really exciting, and then is translating accurately on headphones. So like I said before, I always start in the room. I am not a headphone-only mixer. The room mix is really important, and will be even more so once Atmos becomes more prevalent in cars.
If you think a room is immersive, just wait until you're in a car. It's a smaller space with the same speaker layout, so you're getting quite a bit of immersion.
Similar to what you’re saying about cars, what has me excited is the growing prevalence of simple wireless speaker-based solutions like the Sonos ERA-300 or ARC soundbar. These systems can’t exactly replicate the full 7.1.4 or 9.1.6 experience with in-ceiling monitors, but the results are still impressive–especially for listeners who’ve never experienced any form of multichannel audio.
I think anything that enhances the sonic experience is a win. Even if it’s not the same as 7.1.4, it’s still cooler than playing music off a flat stereo device. It’s amazing technology, and we’re still very much in the early days of all this. It’s been less than five years since Atmos was introduced in the music world.
I think consumers are gonna get really excited when they have their first-in room Atmos experience, and that's why I always try to get as many people in the room as possible. Approvals on headphones are fine, but the room really sells the experience. I feel like once you hit play, you don't have to explain Atmos anymore.
I was going to ask–do the artists or their managers/producers often come to your studio to hear the Atmos mixes? If so, do they ever request revisions?
It’s a mixed bag, depending on the project deadlines. If the artists themselves can't come in to listen, I at least try to get someone close to the project that can speak intelligently to it–like an A&R rep or producer.
In the case of Norah Jones’ new album, her producer was in the room for the mix and actually directed a lot of the spatial placements. He had a very clear and specific vision of how he wanted that record to sound in Atmos.
I read recently that something like 30% of the entire Apple Music catalog is already available in Atmos, which seems hard to believe.
Yeah, that’s crazy. When all of this started, there was a huge rush from all the labels to get as much product out there as possible.
There was almost a full year where it felt like I was in the studio 24/7 mixing Atmos. I'm actually glad that the boom has calmed a bit, because we've caught up. Now we get to be a little more personal with each track.
Yeah, a lot of those early singles that came out in 2020 or 2021 have since been pulled or replaced with revised versions. For instance, I remember listening to an Atmos mix of Styx’s “Renegade” where the harmony vocal and synthesizer parts–which are so critical to that song–were almost absent in the mix.
Yep, that's what happens with anything that you do too fast.
Sadly, there are still artists who haven’t experienced Atmos yet. When working on these songs, I think it’s really important for the artist to have a voice and say in how the mixes turn out. Some of them may not have paid enough attention early on, but they're starting to see the importance of it now.
So it's one of those things where the early adopters really won out, especially the artists that were involved hands-on at the very beginning. I think those are the mixes that stick around, and those where the artist wasn’t passionate about the process or wasn’t involved at all are the ones that tend to get pulled or redone.
It’s my understanding that the labels often greenlight these mixes without notifying the artists, especially when it comes to older catalog titles. So it’s perfectly understandable why some are a bit hesitant about the format at first.
Yes, and that's why we spend so much time focusing on maintaining the integrity of the original stereo mix. So even if it’s the most immersive mix in the world, it still sounds familiar and cohesive.
So when an artist toggles between their stereo and our Atmos on headphones, they'll at least be able to say that it sounds the way they intended. To mw, that’s absolutely paramount.
Do you have a particular favorite or specific title from your discography that you’d recommend listeners check out first?
It’s an early one, but Shawn Mendes’ “When You’re Gone” came out really great. I still use it as a demo in the room a lot.
When Atmos streaming first launched in 2020, I remember one of the first tracks I’d heard was Shawn Mendes and Camilla Cabello’s “Senorita” - it had one voice in the front and the other in the back, which was pretty cool.
Oh, that’s interesting. I actually haven’t heard that one in the room.
If you could remix any older record in Atmos, which would you choose?
In either stereo or Atmos, my dream would be to work with Eminem, 50 Cent or Jay-Z. I would be absolutely honored to touch any song they worked on, but I really think Jay-Z’s “Empire State of Mind” and 50 Cent’s “Many Men” could both be amazing in Atmos.
I’m actually surprised a lot of that stuff hasn’t been done yet. There are still some really big new records, like Beyonce’s Cowboy Carter, that aren’t out in Atmos.
Yeah, I saw that Eminem’s latest track “Houdini” came out without Atmos too. I wonder how long that’ll last. Maybe they just had a bad early Atmos experience, and it's only a matter of time until they have a good one that brings them back on board.
Just like anything else, there's a lot of great examples out there and also some poor ones. We’re still kind of in the wild west, and everyone is figuring out their own unique way to work with Atmos.
There are so many classic records I’d love to hear mixed in Atmos, but none more so than the Led Zeppelin catalog. The movement in songs like “What Is And What Should Never Be” or “Whole Lotta Love” lends itself perfectly to the format, but I wonder how much work is required to restore those multitracks. Like we were talking about before–if you were to just put up the dry tape transfers, how much would be missing relative to the finished stereo mix?
It’s an interesting question. I’d be willing to bet that a lot of those psychedelic pans and effects were done live at the mix, so someone would need to go in and recreate all that stuff.
Can you tease any upcoming immersive projects you’re involved with?
Sure. I’ve done some really cool stuff with LL Cool J that should be released soon. There’s also a new Yung Gravy song with Shania Twain, and I'm really excited about that one too.
