When it comes to mixing, the debate gets more interesting. There are basically three different approaches:
- All-analog mixing
- In-the-box mixing
- Hybrid mixing
Before I tell you my approach, let me elaborate about why people would choose one over the others.
A good example of someone who used to do all-analog mixing is the very successful mix engineer Michael Brauer (he no longer uses that approach). Projects arrived to him in both analog and digital form. For analog tape, he had a tape machine and could mix directly from that (although it would be synced up to Protools so that he could transfer individual tracks there for editing when it was needed). For DAW project files, he (or more likely one of his assistants) would transfer it all to Protools first.
Michael used an 80 input SSL J Series Console together with racks, racks, and even more racks of great and very expensive analog gear. All the tracks were output separately from Protools and the tape machine to the SSL. On the SSL, everything would get sent through any desired on-board processing and, through the patch bay, to and from all of his outboard toys. The two-channel output of the SSL would then be recorded back to digital.
From a technical-purist point of view, that means everything in the mix ultimately has gone through an extra stage of conversion compared to if it had been mixed in-the-box. But out of all the things that affect the quality of the final product, I’d say that is one of the least significant ones. It does, however, have at least as much of an effect on the final product as the choices of what arithmetic to use (24, 32, 48 or 64 bit, fixed or floating point etc), which some people seem to care a lot about.
The main benefits of using this approach for Michael were that:
- It worked really well. Countless hit records have been mixed this way by the top mixing engineers in the industry.
- The gear he used is the best of the best, acquired over many years. He picked every piece because he tried it and liked it. In a way, it represented one aspect of his accumulated knowledge gained from years and years of work.
- Clients were always very impressed, and gear junkies had multiple orgasms when they attended one of his sessions.
- He was very accustomed to working this way, allowing him to focus fully on mixing, rather than on handing a new way of working.
The main drawbacks compared to in-the-box mixing were:
- Equipment Cost. All Michael’s gear took years to collect and I really have no idea how much his collection was worth. Needless to say, a young engineer starting out will not be able to have 24-hour access to such an array of equipment (unless he or she happens to land and assistant position for Michael or another mix engineer that has a similar setup).
- Maintenance. Keeping all that outboard gear, as well as the console, in good shape requires a skilled technician. That’s why engineers at this level usually have permanent residence in major commercial studios with in-house technicians and studio managers.
- Recallability. I know Michael had a nice system for recording and recalling every mix session, including the automation on the console and all the physical routing done in the patch bays, together with detailed notes of the settings for all of the outboard gear used in the session, but the time it took for the assistant to recall a session was still significantly more than the few seconds it takes to load an in-the-box session in Logic or Protools. Also consider that making many mix-revisions or alternate versions of mixes, often long after the original mixing session, became more common over time to the point when it became the norm .
- Cost. This is the biggest one. Housing all that gear, keeping it maintained, having an assistant, and paying for a studio with a maintenance staff all cost money, and naturally, that cost would have to be passed on to clients, resulting in only being able to work on very large budget projects.
In the box mixing means that the mixing, or summing, of all the individual audio tracks that make up a song is done through digital processing. I would include mixing on digital mixing consoles here too, since digital mixing consoles really only are glorified computers with very custom hardware user interfaces. Sure, the arithmetic used (24 or 48 bit fixed vs. 32 or 64 bit floating point) varies, but most of the various DAWs have options to adjust these settings.
Since this style of working became practically possible sometime around 1995, (depending on your definition of practical) many people have expressed concerns about it and made statements about the negative impact it supposedly has on the audio quality of the final product. Naturally, the validity of these concerns and the accuracy of the negative statements varies widely:
- “The bit rates used for audio transport and summing (arithmetic) are too low.”
- “Digital summing just doesn’t sound the way analog summing does.”
- “Plugin versions of classic outboard gear do not sound like the original counter parts.”
- “The delay added to tracks when inserting plugins destroy the phase coherency between multi-track recordings.”
- “Multi-bus mixing techniques cannot be used inside a DAW without loss of phase coherency due to plugin latency.”
- “The available audio headroom on the master track is not as great as it is on good analog consoles.
- Signals don’t distort in a pleasing way as the level approaches the ceiling.”
There are many many more of course, but let me discuss the ones I just mentioned:
The bit rates were a bit low in the beginning, and in many cases, lesser quality in one way for the final product while other quality aspects such as the noise floor were improved. Today, however, all DAW’s use at least 32 bit floating-point arithmetic which, as long as the software is used correctly by engineers who know how to gain stage properly, should not lead to noticable quantization errors or present a head-room problem. And today, most DAW’s can use 64 bit float or 48 bit fixed, which makes it hard for even incompetent operators to run into headroom problems and unwanted distortion.
The concern of the “sound” of digital vs. analog summing is very strange to me. First of all, different analog boards can sound very different from each other. In any analog design, the sound is highly dependent on the components used and their quality. None of them will sum signals perfectly. Whichever analog mixing board one prefers is totally a matter of taste, and for many people, their choices (if given choices) might vary depending on what they are about to mix. So, to lump all analog mixing together and contrast them as a group with digital mixing just doesn’t make much sense.
Inside a DAW, the summing really is “perfect” in the theoretical sense of the word. Not to say that every mix engineer will prefer the sound of it, but the summing in all the major DAW’s will sound very close to, if not completely, identical (if the same pan laws are used) to each other. Compare this one sound to the wildly varying sound of analog summing is kind of like picking one fruit, say a pineapple, and then compare that alone with all other fruits as a group.
Everyone is of course entitled to their own taste in what equipment they like to use when they mix, but to categorically say that analog summing sounds better than digital summing is just plain stupid. Here’s a good quote from the very successful mix engineer Dave Pensado: “Show me a guy who doesn’t like a particular format and I’ll show you a guy who doesn’t know how to use that format”. People who swear by the summing of one DAW over another and claim they can hear a big difference have most likely missed the importance of what pan-laws are being used in each DAW. This is a setting that can often be changed on a per-session basis, and if you were to input similar pan setting numbers for tracks in different DAWs with different pan-law settings, the resulting actual pan placements would of course be different too, thus most likely explaining what people mistakenly attribute to the summing algorithm used.
When it comes to plugin emulations of classic gear not sounding like the originals, I don’t doubt it. This is very different than comparing what is best though. It’s extremely hard to make a faithful emulation, and I don’t think they will ever be perfect. On the other hand, why does it matter whether the emulations are perfect or not? Maybe for someone who used a particular piece of equipment for every single song, then lost it, and is now trying to replace it with a plugin. He’s actually after a clone of what he lost. In that case, even finding the same make and model and paying a fortune for it on eBay might not solve the problem, since for old analog gear, sound can vary a lot from unit to unit. Most mix engineers however, are not after clones. They simply want to mix music, and are usually open to trying anything. If it sounds good, they’ll use it, and if doesn’t, they’ll move on to something else.
The delay, or latency, that happens when inserting plugins can definitely be an issue, and if not dealt with correctly, it often used to cause phase coherency problems. It was always possible (even though inconvenient) to deal with this through manually keeping track and compensating for latency though. All the major DAWs currently feature automatic latency compensation, so today it’s a non issue. If, however, you start working on a mix, and then proceed to do additional overdubs, this latency compensation might present a latency problem when monitoring your performance, in which case, workarounds will be necessary.
Multi-bus mixing techniques work great in any DAW that feature full automatic latency compensation.
The headroom in good analog consoles used to be better than what was available in DAWs and digital consoles, but today, this is not really a problem anymore. When 32 bit floating point is used, the engineer still needs to pay attention to gain staging and know what they are doing. When using 64 bit floating point, it can safely be ignored for any practical purposes and intents.
Now, what about the benefits of in the box mixing?
Complete and perfect recall-ability. No need for assistants spending time tweaking knobs while looking at photos of how a unit was set at a session that took place three weeks ago; no strange issues resulting in the patch bay not having been reset exactly as it was the last time the session was open; no need to use a different compressor on the lead vocal this time since the one originally used was sent out for repair, and so on…
Use as many copies of the plugins you like and your computer can handle (and with today’s computers, that number has gotten really really large). If you like the sound of a particular compressor on one of your backup singers, no reason not to try it on the other backup singers too.
Plugins don’t occupy physical space in your studio. Tons of outboard gear means having to house it all, preferably within reach of you while you work which might lead to acoustic issues in your mixing space (reflections from the flat surfaces of your equipment racks and so on). Outboard gear also means either less space in your room, or having to pay for a larger room. With plugins, all your virtual equipment is right at your fingertips – either through the use of a mouse or through control surfaces of various sizes. Whatever you prefer.
Plugins are generally much, much less expensive than equivalent outboard units, which of course means that whatever your budget is, you’ll be able to afford many more plugins and thus have a much larger arsenal of available tools at your disposal when you work.
With plugins (used in a DAW) you’ll be able to easily save complete channels strip settings in a library of your own, so that you can easily get that same great sound you had on the vocal in another project, for the current one. Just load it up with EQ’s, compressors, delays, reverbs and all. As you work, you can copy settings from one track to another within your current session.
I could keeping going…
The main point of hybrid mixing is usually to try and combine as many of the respective benefits of all-analog and in-the-box mixing as possible while leaving most of the disadvantages out. Another common reason to use hybrid mixing is to use analog equipment that you already own and like with an otherwise mostly in-the-box approach.
Some engineers that practice hybrid mixing already own a large analog console and have the space and financial means to support it, yet they want the ease of use that come with working in-the-box. So they might end up with a scenario where they bring all the tracks out of the DAW through individual outputs that end up on the inputs of their consoles, just like the all-analog approach. However, they’ll still have lots of plugins running on all the tracks inside the DAW, and they’ll use the automation of the DAW while leaving all the faders on their analog console at unity gain. That way, they never have to deal with recalling the analog settings since they always stay the same while all their plugin settings get recalled by the DAW. This approach, however, prevents them from using the EQ’s on their boards, and it might (so long as they don’t want to start documenting everything again) prevent them from using many great multi-bus mixing techniques. So in essence, they really only use their mixing consoles for the summing. If one happens to really like the way a particular console sums audio, then that’s a good reason to have a setup like this.
There are also some analog consoles that feature automation that can be controlled from a DAW. This can be done in a few different ways, but the goal is that all the settings for the processing in the analog console will actually be saved in the DAW project so it can be recalled together with everything that happens in-the-box. This improves the user experience of hybrid mixing to a large degree but does not include outboard gear.
Another approach that allows analog summing while making the whole setup smaller and more transparent, is the use of a so called “summing box”. Various companies basically make compact mixing consoles that don’t have any faders or knobs (and thus doesn’t need any hands on attention or have any settings that needs recalling) yet they still feature the summing and routing capabilities of an analog console. They also allow select pieces of outboard gear to be patched in before the summing (but then settings still have to be documented and recalled manually).
I would say this last approach is the only analog-summing approach that makes sense for a beginner that doesn’t already have a significant investment in analog gear or has a knowledge investment or an emotional attachment to analog gear. What it really comes down to is whether one prefers the sound of a particular summing box over the sound of digital summing, and how much money and complexity that preference is worth.
If it’s mostly about getting to use a few treasured analog pieces while mixing, there is a simpler way to do that while staying in the box. So long as your computer audio hardware has additional unused audio inputs and outputs, it’s easy to insert an analog piece of gear into a digital channel strip. Most DAW software even feature insert plugins that enables compensation for the latency that occurs during the roundtrip through the DA converter, to the analog piece of gear, and back to the DAW through the AD converter.
So, what do I do?
As you might have already guessed, I use in-the-box mixing. I don’t have any emotional attachment to a specific analog console or any particular outboard gear. I also don’t have anything against the “sound” of digital summing. There are so many ways to achieve “warmth” using various plugins, not to mention in the input stage during the recording, and so for me, the many many practical benefits of in-the-box mixing completely outweigh any benefits I might have gotten from an all-analog or hybrid approach to mixing.