Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
What's under the hood?
#1
Hey folks,

Well, despite the serious instability issues I've been experiencing, which I'm sure the people from Harrison are already taking care of, I'm starting to get Mixbus.

It really does something unusual, pleasant and very hard to describe to the sound, most specially to the spatial image of it. It's very easy to get a good mix out of it, you know, to make things "gel" together in a pleasant and coherent way.

Before trying Mixbus, I was wondering about how a mix made solely on Mixbus would actually differ from one made in any other DAW using quality analogue inspired plugins on every channel.

As an Industrial Designer, I'm well aware that the interface, the GUI, makes a huge difference in the perception of quality and usability of anything, including (and specially) software. Take the same plugin, let's say and EQ, with two different GUIs, one ugly and one beautiful, then ask people which one sounds better. I guarantee that people will tend to enjoy the sound of the beautiful one the most. That's just how we work. We're emotional beings. Our perception of things is rarely faithful to the real nature of things.

Anyway, I digress. To sum up, I used to think that Mixbus was a mix between placebo effect caused by the nice nostalgic GUI and cool but not extraordinary DSP coding. This is a very common phenomenon in the digital plugin world.

Now I realize that Mixbus does something apparently way more complex than what could actually be done on a plugin level. Maybe that's why they've decided to turn Mixbus into a standalone DAW, not a plugin suite like Slate's VCC or Waves NLS.

Certainly, the nostalgic analogue inspired interface plays its part on the way we perceive Mixbus, but mostly because form directly influences function. Mixbus' interface invites us to approach mixing peculiarly. It's very straightforward and doesn't provide any room for wondering around doing pointless tweaks. It helps to focus on the sound, not on tweaking stuff all day long.

However, I'm 100% sure that whatever is under its DSP hood, it's something extraordinary. It's definitively not the same thing as slapping analogue modeled channelstrips and tape emulation plugins on every channel of a regular DAW.

So, do you guys have any idea about what's really happening under the hood? Sure, the built-in compressors, eqs and tape saturation sound great, but there's something else going on. I'm not asking to reveal any industrial secretes here, but only a general explanation or insights of what Mixbus is actually doing to the sound.

Cheers,

Morgan.
Reply
#2
(07-21-2015, 08:47 AM)RMorgan Wrote: Before trying Mixbus, I was wondering about how a mix made solely on Mixbus would actually differ from one made in any other DAW using quality analogue inspired plugins on every channel.

An MB mix will not necessarily sound different from a mix done with another DAW, that's because it's very common to use tape saturation, distortion and all kinds of harmonic dist on mixes today. But making a transparent, warm Hi-fi-ish mix where every instrument can be heard (if the raw material is good enough) is IMO much easier and quicker in MB. The reason is of course the analog nature of mb's sound and also the nature of working. In MB, almost everything you want to do is under your fingertips.

But of course, you can easily make digital sounding mixes, but that is not the default setting.
Mixbus/Mixbus32C on Linux (Kubuntu)/KXStudio repositories.
GUI: KDE and Fluxbox
Reply
#3
(07-21-2015, 09:26 AM)Jostein Wrote:
(07-21-2015, 08:47 AM)RMorgan Wrote: Before trying Mixbus, I was wondering about how a mix made solely on Mixbus would actually differ from one made in any other DAW using quality analogue inspired plugins on every channel.

An MB mix will not necessarily sound different from a mix done with another DAW, that's because it's very common to use tape saturation, distortion and all kinds of harmonic dist on mixes today. But making a transparent, warm Hi-fi-ish mix where every instrument can be heard (if the raw material is good enough) is IMO much easier and quicker in MB. The reason is of course the analog nature of mb's sound and also the nature of working. In MB, almost everything you want to do is under your fingertips.

But of course, you can easily make digital sounding mixes, but that is not the default setting.

Yes. I can get the same mix sound regardless of which DAW I use, but it's much faster with Mixbus. I can get a mix sounding good in a matter of minutes with MB.
I really like their eq modules. They have done a good job picking default frequencies.
A lot less plugins are used in MB and the work flow is fairly streamlined.
Reply
#4
(07-21-2015, 09:26 AM)Jostein Wrote: ...The reason is of course the analog nature of mb's...

Yes, but what exactly is this analogue nature?

I did some tests here and Mixbus make things sound different (at least to my ears) even if you just import some stems, direct them to a few busses with minimum tape saturation and without any eq/compression.

I mean, something is done automatically, totally behind the scenes, unlike other DAWs, which try to be as transparent as possible until you start loading some plugins.

Mixbus' bus channel does not behave like a regular DAW bus track. It does something to the sound at the moment you route some tracks into it; What is it exactly? Some stereo image expansion? Maybe some sort of ultra discrete auto-eq or auto-multiband compression? I don't know.

All I know is that Mixbus seems to impart some character to whatever you import into it before you even start to tweak its buit-in effects.

Might be my imagination, though... Smile
Reply
#5
(07-21-2015, 09:34 AM)RMorgan Wrote:
(07-21-2015, 09:26 AM)Jostein Wrote: ...The reason is of course the analog nature of mb's...

Yes, but what exactly is this analogue nature?


Might be my imagination, though... Smile

To me Even if I don't work on analog console but comparing Harrison to others. The only way I can tell you how you can see the difference is by boosting any frequency that you like.

let say boost 5k on Harrison and it remains smooth. The other Daw will bring ooze out of your ear.

For people that I would like to say is don't get the Idea of digital when you see Harrison mastering eq still it sounds analog. It took me over 6 month to understand the difference.
Reply
#6
(07-21-2015, 11:49 AM)BHBstudio Wrote: ...let say boost 5k on Harrison and it remains smooth. The other Daw will bring ooze out of your ear.

Sure, I hear you my friend, I also think Mixbus has great EQs and Compressors. Very smooth.

However, if it was just for the EQs and Comps, Harrison could've just released Mixbus in VST format, as channelstrips, instead of a DAW.

There are a lot of great and very smooth sounding VST EQs out there as well. Check out the latest stuff from Boz Digital Labs, as an example.

I think there's something else. There might be some communication happening between Mixbus' channels and busses, something that happens on a deeper level than it's possible in plugin format.
Reply
#7
My investigations show this:

Everywhere there is a fader, there's a little low-level white noise. This can be seen with an analyer like Voxengo SPAN.

The "Tape Saturation" on the buses generates a third-order harmonic. Surprisngly, no other harmonics...Just the third. I would've expected a harmonic sequence myself.

While not explicitly measured, there appears to be some sort of HF softening/roll-off.
Reply
#8
Okay, for those who beleive that MB could have been a plugin:

Quote:Regarding the sound: The problem is that everyone expects a simple, clear answer why it sounds better. But this is impossible.

In digital processing, there are many inherent problems: aliasing, clipping, quantization noise and zipper noise, for example. These artifacts are arguably more bothersome, even at lower levels, than analog recording artifacts.

The goal is to pass the signal through the digital system in a way that avoids these artifacts. The techniques used are widely known to DSP engineers ... dither, staged summing, parameter ramping, etc.

However most DSP engineers only get to experiment with a few products, with a few collaborators of questionable abilities, in rooms of questionable acoustics. Our engineers have been able to develop multiple generations of analog and digital systems, on high-profile projects, with mixing engineers of the highest caliber in the most fantastic rooms available.

Furthermore, we have an opportunity to compare old designs and studio workflows with the more modern techniques. This is where we noticed the effect of tape saturation in the traditional recording workflow. It's not an effect ..... it serves an important stage in the recording, which is soft-clipping the loud instantaneous transients which are not important in the overall product, but have a significant effect on following processing such as compression/limiting. It has to be approached holistically, which is very hard to achieve with a purely plugin-based system.

Mixbus isn't the best mixer we can make. You have to spend over $100,000 for that. But the same people who designed the big mixers for many years have incorporated some of those seemingly insignificant, but collectively important, techniques into the design of Mixbus.

In other words, the techniques must be applied _appropriately for the product_, the techniques used in Mixbus might make no sense in another product. I hope this makes it clear why a simple answer is not possible.

Best,
-Ben
http://forums.macg.co/threads/mixbus-har...s.1181152/

and

Quote:According to Harrison, the sound of its consoles runs deeper than a simple EQ or compressor. The company found that running a plug-in version of its channel strip through the summing architecture exhibited in the aux buses and the main summing bus of most DAWs results in a sound that is unsatisfactory. Rather than building a DAW from the ground up, it seemed a better idea to take an existing infrastructure and swap out portions with newly redesigned components. Harrison chose the Ardour workstation (www.ardour.org), which was cross-platform (Mac/PC/Linux), had a large existing user base and offered the ability to modify the programming. Harrison had an existing relationship with Ardour, which helped design the software side of Harrison's Xdubber unit. This relationship was advanced through the development of Mixbus. Harrison's engineers took the foundation of this software, but made some serious changes.
http://www.mixonline.com/news/daw-record...iew/369475
Mixbus / Linux 64bit
Reply
#9
(07-21-2015, 12:43 PM)James Greenlee Wrote: My investigations show this:

Everywhere there is a fader, there's a little low-level white noise. This can be seen with an analyer like Voxengo SPAN.

The "Tape Saturation" on the buses generates a third-order harmonic. Surprisngly, no other harmonics...Just the third. I would've expected a harmonic sequence myself.

While not explicitly measured, there appears to be some sort of HF softening/roll-off.

Agreed. There's a perceptible amount of background noise. It's probably there as part of the analogue emulation scheme.

And yes, it feels like, besides adding harmonics, the tape saturation also adds some sort of soft clipping mojo.

I also suspect that there's a bit o crosstalk going on between channels, busses or both, which may be responsible for the apparent enhancements in the stereo image.

(07-21-2015, 12:48 PM)sonik Wrote: ...Okay, for those who beleive that MB could have been a plugin...

Thank you sonik,

I guess this partially answers my questions.

I still would like to know more about how the subtle nuances of these processes interact with each other to create this very interesting Mixbus "effect".
Reply
#10
(07-21-2015, 09:34 AM)RMorgan Wrote: Might be my imagination, though... Smile

Smile I would suggest not. All of us here apparently discern a certain 'something' out of MB - warmth/softness-or-something.
Unless it's mass audio-hypnosis. Smile
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)