Why do you use a hybrid mastering chain?
The answer isn't simple, and there's a real variance on exactly what each hybrid chain mastering chain might include. I believe my chain makes a lot of sense and I'll explain why.
However, my chain is fairly simple. I start with a pitcher DAW, convert the audio from digital to analog, pass the audio through my analog chain, convert back to digital and on to my catcher DAW. Both DAW's employ digital plugins, while the analog chain is well...analog. Hence, the hybrid nature of the hybrid mastering chain.
Some mastering engineers work almost completely OTB (Out-of-The-Box), or analog, while other are completely ITB (In-The-Box). Still, others are using hybrid chains like I do. So what's the true advantage of each method? Why do mastering engineers use one or more of these methods?
Analog
Most every engineer who uses all analog gear will swear by it. It's clean, it's got character, it opens the music up, and only it can do what it does so well. For example, most old-school mastering engineers (we'll use the acronym, ME from here) will tell you a Sontec MEP-432 is the holy grail of mastering EQs. They're very expensive but still sought after. I've not heard one myself but I'll gander they're nothing short of sublime! I've never heard a Fairman TMEQ, either. But wait! Perhaps I have because ME's have used them and I've heard their masters. I won't argue over their venerable commanding value! These ME's produce some excellent masters, to say the least.
Analog has some faults, however. No, that's not true. I wouldn't say analog has faults (unless it's just bad hardware, or hardware gone bad). I'd say there are limitations. For example, you can't afford every piece of analog gear out there. If you could, you'd begin to understand why Steven Wright said in standup comedy, "You can't have everything, where would you put it?". Streaky, from Streaky Mastering once said (loosely quoted), "I've had so many pieces of gear at Metropolis, and it was just too much!" His current studio has only select pieces of equipment and that's what works for his hybrid chain.
Analog requires a significant attention to quality in cabling, power, and maintenance. For example, a tube compressor will change character over time, perhaps in the apparent level of increasing distortion in the audio. Other components may corrode or just simply age out like electrolytic capacitors which, hopefully no ME has those directly in the signal path in any piece of gear. Point being, analog gear can change over time. It can fail, requiring repair.
Analog is also limited in that you can only build what is possible with electronics. Barring outboard gear that has some digital functionality, such as mastering equipment from Weiss, for example, analog units must adhere to the physics of electricity and electronic components used to harness or alter the electrical signal transmitted from your music source. You can't simply (or, rather, compactly) build an analog compressor that can do what DMG's Multiplicity can do in a single box---possibly in a single rack! And you certainly can't do it without a significant expense for minuscule 'gain' (pardon the pun).
Digital
I slammed the door there on analog gear pretty hard. It was partially intentional because I also want to unleash the same fury with digital plugins. Read on.
Plugins have three key benefits that a large number analog units can't touch: instant recall-ability, quantity, and price. You can't easily compare an SSL modeled track plugin to an SSL analog console when the console is cost-prohibitive and you can instantiate unlimited numbers of the plugin on any DAW session, at any time--meanwhile not breaking the bank. You can't underestimate the power of having a thousand plugins in your arsenal taking up less space than a USB thumb drive. It's just not a fair comparison.
Open as many duplicate plugins as your CPU can handle for the price of one!
But this is one reason why the analog aficionados are such hardcore purists--and God bless them! They feel threatened by such compact replacements of what they love so well and invested thousands of dollars to acquire over time, whether or not they've got a real grievance.
You can also argue that never converting the audio to analog saves the music from issues with the conversion. I've read about an engineer converting audio in and out of the box maybe 100 times (?) and found no ill-effects. This certainly says a lot of the AD/DA converters for sure, but every single time the audio is converted, the result should be the same. Unless the converters are bad you should expect pretty much exactly the same result. The signal is regenerated fresh upon each conversion. It's not necessarily cumulative unless there's errors--again, indicative of a bad converter or even a bad cable or connection.
The other foot drops--and let's be honest for a minute. How many plugins today do you expect to remain in digital stores for the next 40 years? We'll see. There's a certain plugin company I won't name from whom I purchased a $1,500 linear phase mastering EQ from years ago. It was 32-bit only. It would only run on Windows. It was the most highly regarded linear phase mastering EQ plugin at the time. When everyone else was scrambling to offer their plugins as 64-bit, this company all but disappeared, focusing on other money-making ventures leaving their plugins in ruin. The software was fortunately revived by another company who now sells a redesigned and updated 64-bit version of the EQ for maybe half the price. However, this should illustrate the vulnerability of software.
At some point, your software will not work!
Analog gear might still be thriving in studios decades beyond the lifespan of a plugin.
Still, why the hybrid chain?
Now, I believe the answer is a little easier to state: I prefer some analog gear over plugins but I prefer the wealth of plugin choices over analog. Plugins have become alarmingly good, by the way!
My choices are fairly straight forward: I use plugins for corrective, surgical, subtractive EQ, most all compression, transient-shaping, addressing specific issues, and maybe for some character. I use analog gear for boost EQ, tube saturation, some spatial functions, and general character.
I also find that there's a few areas plugins can't faithfully emulate. Transformers and tubes. I feel transformers are close. Kazrog has a super amazing True Iron plugin that I think is well-reviewed, but still underrated. I often use it in mastering. Shadow Hills has a transformer character in their Class A Mastering Compressor plugin which I find useful, too. But I've not heard a tube plugin do what a real tube can really do. To me, if you've heard most amp sims, you can blast an amp sim plugin's signal into a solid state guitar amp, and it simply won't sound like a real guitar plugged directly into a tube amplifier cranked up. It will sound like a recorded guitar played back through a solid state guitar amp. The purity of beautiful energy and raw power just isn't there, in my opinion. Tube mics still aren't easy to model, either.
I believe that tubes are the hardest thing to model by any plugin developer and the reason is that the tube behaves a certain way that's sometimes unpredictable. It may behave one way one day and slightly different another day. A tube isn't likely going to be modeled again and again as it ages. How would you know it's the best day and aging to model it? Even more, most tube modelers aren't telling you the tube brand and models that were chosen. For example, I have a matched set of Amperex Holland 12AT7 tubes from the 1960's in my Black Box HG-2, having replaced the OEM 12AT7 tubes the unit shipped with. There's a difference. The plugin sounds more aggressive (perhaps forced) than what I'm using. It's quite difficult to model exactly what the Amperex tubes do. But it's truthfully blissful how the tubes sound, how they warm tones as they envelop your eardrums. This is why I do use some analog gear.
On the other hand, I don't have a single analog compressor in my analog chain. While I believe there are some undeniable gear out there that does something truly magical, this is an area I believe plugin manufacturers are more faithfully modeling, along with many equalizers. The problem with compressors and equalizers is that there's soo many of them that different MEs use and what works well on one song might not suit another as well. Having an instantaneous choice of plugins here is invaluable. No patching in a new compressor, no re-routing, and most importantly, no need to have an insanely huge mastering desk to accommodate every single one!
Try fitting this many venerable compressors into even a 4-row, 20-U set of racks.
I suppose if I were to purchase a hardware compressor it might be an SSL VCA compressor.
To add, many plugins can do things the analog gear they model can't. For example, most analog compressors are all-or-nothing when it comes to being added to the mastering chain. Conversely, most modern plugins employ a Mix knob that allows you to abuse the compressor with as much character and drive as you want, but just barely mix it in. You can't do that in analog so easily. Digital compressors also have the advantage of being ultra clean and transparent when you need that. Some compressors have absolutely zero coloration, which ME's actually need more than ever.
There's some things plugins can only do, such as dynamic EQ, multi-band limiting, and a myriad of other tools and useful processes that only digital can accomplish (at least, today). Take Soothe2, for example. I don't know a single hardware component that can do what it does to kill awful resonances. There's convolution reverbs that can emulate rooms you can never fit in a rack. There's other emulations, such as Waves Abbey Road Vinyl, which you can't duplicate very easily with hardware.
Recreating these tools with analog would be very difficult.
EQ is a different story. EQs are modeled or developed. But EQs can be linear-phase, which is unheard of in analog. You can't delay the audio as easily in analog to preview how to handle it. Everything in analog is reactive for the most part. Plugins can be made proactive, and can use oversampling to ensure the highest quality processing. I use EQ in plugins for transparency. I prefer the EQ to be out of the way in mastering. I almost never use plugins to boost EQ, either. I prefer hardware because if boosts EQ cause phase-shift I'd honestly prefer that to happen in the analog chain. Personal preference, even though there's probably some benefit to using a Chandler Curvebender to do that. The only other boost EQ I might use ITB is a Pultec style EQ, mainly because I do not wish to maintain the tubes of that EQ, and having two channels of a real Pultec requires a pair of them, consuming 6U of rack-space. The Puigtec EQs from Waves are fine.
I use a hybrid chain for only exactly what I want that's best suited in analog: tube saturation, boost EQ, and very slight coloration. I don't require it for every master but I've never had a bad result using analog. I've also never had a bad result staying ITB, either.
Finally, most MEs would agree, it's not about the gear. It's about the ear. Let's worry about that, because in my personal opinion, the ears and judgement of the ME is paramount.