I've said this many times: I think the SSL Fusion is one of the best pieces of hardware you can buy when you're starting a mastering chain. It gives you a little bit of everything in one reasonably affordable unit.

When I bought mine, they were around £1,800. I've since seen used units at considerably lower prices, sometimes under £1,000 in the UK. That's a lot of functionality for the money, especially with the mid/side insert.

My view of the individual sections is that they're generally good without necessarily being exceptional. The shelving EQ is clean and useful, the stereo width and depth processing are very nice, and the high-frequency compressor does its job, although it wasn't something I used very often.

The Vintage Drive was less convincing for me. On the units I owned, I found the noise level too high to comfortably use in mastering. I can't say whether newer units behave the same way, but that was my experience.

Where the Fusion really earns its place is its versatility. The mid/side-capable insert lets you bring another dual-mono processor into a mid/side chain, something that would otherwise require additional routing or hardware. Even if I eventually stopped using the Fusion's other sections, that insert alone would remain useful.

It's not the most sophisticated processor of the three, but at its price, I think it's brilliant.

I own two Rupert Neve Designs MBTs, and the reason I value them is very simple: Super Silk. That's what makes this unit special for me.

I love what the saturation does. The Red and Blue Silk controls let me shape upper- and lower-frequency harmonic character, and I think that sound is phenomenal.

There are other useful sections. The EQ and filter do their jobs, the stereo width is nice, and the opto-style Color Comp has applications, particularly on sources such as vocals. But it isn't the compressor I'd personally choose as the centre of my mastering bus.

That's where the cost becomes difficult for me to justify. When I'm mainly reaching for the Silk and occasionally the width, it's an expensive way to access those two features. I also think the lack of an external insert is a missed opportunity compared with the Fusion.

None of that takes anything away from the sound of the MBT. If the signature Rupert Neve saturation is your priority, this is the one I'd encourage you to explore.

I'll admit I'm still in the honeymoon period with the OmniBox. But what has impressed me is how much useful processing Wes Audio has put into one unit, and how well those sections work together.

It's a jack-of-all-trades style of processor, and I don't expect its EQ to replace every dedicated mastering EQ or its compression to replace every specialist compressor. However, compared directly with the Fusion and MBT as all-in-one processors, I think the OmniBox comes out ahead.

The shelving EQ sounds beautiful and remains relatively clean to my ears. I'd love to compare it with my Wes Audio Hyperion in more detail. The VCA bus compressor is the more useful option for my workflow than the MBT's opto compressor or the Fusion's high-frequency compressor; it reminds me of the kind of bus compression I enjoy from the Wes Audio Dione.

Then there's the analogue clipper. I recently used it on drums to gently catch snare peaks rather than aggressively flatten the signal, and it worked beautifully. The saturation is also good, especially with the drive and mix controls, although for subtle mastering saturation I still prefer the MBT's Silk.

One thing that initially caught me out on the OmniBox was its low-frequency treatment in the width section. In my testing, setting that frequency behaved more like a mono-maker or elliptical filter: stereo information below the chosen point was reduced, while I could enhance the width above it.

The MBT, as I use it, feels different. I can set a frequency region above which I want to widen the signal without deliberately collapsing the low end in the same way. Neither approach is automatically better, but they produce different results and you need to understand what the control is doing.

I'd love to see Wes Audio offer a switchable option in the OmniBox so I could choose between the low-frequency mono-making behaviour and a crossover-based widening approach. Whether that could be implemented in firmware is something I'd like to ask them.

The Fusion has a brilliant external insert with mid/side capability and a choice of insert positions. The MBT doesn't offer the same facility. The OmniBox surprised me by providing two insert points: a stereo insert at the input and a mid/side-capable insert at the output.

That means I can connect additional hardware both before and after the OmniBox's internal processing. Unlike the Fusion, I can't freely move these insert points into the middle of its processing path, but having access to both ends of the unit is extremely useful.

I've put my HUM Audio LAAL limiter on the output insert, alongside a Wes Audio NG Leveler for recallable gain control. By linking the gain changes, I can raise the level going in while compensating on the way out. Crucially, I can also recall the insert bypass from the OmniBox.

For my studio, that gives me a much more convenient way to integrate non-recallable analogue equipment into a recall-driven setup. The OmniBox also lets me change whether EQ comes before or after compression, which opens up further possibilities for processing order.

My main studio is built around digital recall. I want analogue sound without having to manually reset an entire rack every time I reopen a session. That's why I think hybrid hardware with software control is the best of both worlds.

Imagine using one OmniBox on a drum bus, printing the result, then moving it to a vocal bus and saving different settings. Later, you can reactivate the appropriate plugin instance, recall that processing and make a change without photographing every dial. It still requires sensible routing and printing, but the flexibility is enormous when you only own one unit.

This is also why I'd happily consider buying a digitally recallable SSL Fusion if SSL ever made one. I don't know whether they will. I simply think manufacturers who overlook digital recall are missing an important part of where professional studios are heading.