The Mojo of Universal Audio Plugins: An Interview with UA’s Will Shanks

By • December 8, 2010

I have made no secret on many forums and among colleagues of my unabashed love for the plug-ins Universal Audio has created for the UAD cards. You would have to pry them from my cold dead hands. There are tutorials in both of my “Going Pro With Logic Pro” books. I honestly feel they have totally elevated the quality of my mixes.

So I decided to have a little discussion with UA’s Software Product Manager and Marketing guru Will Shanks. Will also owns and operates Ear to the Ground Studio in Santa Cruz, California. His bio says that he “specializes in working with independent artists and independent labels who are budget-conscious, but want the best in experience, quality, and expediency”.

I would say that fits UA’s mission statement pretty well also.

Thanks for doing this, Will. As a longtime UAD user, I have made no secret of my love for your plugins in my books and on various forums and how much they have improved my mixes.

First of all, can you give us some insight as to how UA decides which hardware to model? Sometimes it is for pieces that others have not modeled but other times, it is for ones that already have been done by other companies.

The decision to make a plugin model can come from various avenues—customer demand being the first and foremost. For example, tape and tape machine modeling was the number one request with our customer base for several years. That’s a big request that cannot be tackled in a short time, but of course we’ve finally made it available. Even with existing models in the marketplace in some cases, our customers want to UA to tackle their requests with our own approach to design. Other times a decision might be based on filling a hole in the product line with something that is special, but might not have been a huge seller in the hardware realm. The Cooper Time Cube is a relatively obscure device, but has a fanatical cult following for a good reason. Being able to resurrect a device like that for plugin use made for a compelling story, and became popular with our customer base selling many more than the hardware ever did as the Cooper Time Cube MkII!

Indeed, the Cooper Time Cube is one of my personal favorites. Years ago, I was a staff writer for Casablanca and when I would go to do demos, there was frequently no time to spend doubling vocals and since this was before there were digital delays, we would use this piece. I fell in love with the sound of it and still do to this day. But it certainly is an odd piece of gear.

UA has had a lot of success partnering with some pretty famous companies like Neve, SSL, Roland, and now Studer. One the one hand it is surprising because it seems like it might undercut their sales. Do you feel that perhaps it is because they figure somebody is going to do it anyway and they trust you to get it right?

In our experience and according to our partners, plugin sales do not noticeably affect hardware sales—keep in mind we are all in that boat together with emulations of our own hardware too. I’d further it to say in some cases the plugin can be seen as a promo tool for the hardware. It appears each market tends to feed the other—you have folks with no experience with the hardware who try the plugin and decide they also want the hardware, and you have experienced hardware users who hear the plugin and want both for whatever their reasons. But I do think the notion that UA will do the best job possible in the modeling accuracy is also a big factor in our partnerships, as is our custom copy protection.

Let’s talk a little about EQs. Back in the days when I was regularly in recording studios they would of course have a board, perhaps a Neve, Harrison, SSL, etc. and maybe 1 outboard EQ like a Massenburg. So the engineer mostly used the boards EQ and got great results. Experienced engineers who worked at multiple studios no doubt got to know different boards and their different characteristics. As a composer doing his own engineering, however, I do not have that wealth of knowledge to draw on. So my dilemma is, is there anyway other than hours and hours of trail and error to know when I should reach for the Neve 1073, the Trident A-Range, the Manley Massive Passive, etc? It almost seems that there are too many choices.

I got a tip from Michael Brauer recently, who has been using the UAD-2 extensively. Even he had not been afforded the experience of using every console EQ he now has available to him with the UAD platform. He has been doing mixes where he limits himself to just one console EQ to really understand and learn what it is about—once he gets a general idea what the EQ does and how it might apply to the song or genre he’s currently in, he then restrains himself to just that EQ, just as if he were in front of that console with only that available, thereby giving that record a particular console’s EQ sound. You can even set your DAW to adopt the pan law of that given console. Do that a few times, and I think mixing and matching EQ’s within a session for a particular application will become second nature, just like reaching for the right mic for a certain sound or application.

That is a terrific approach! I swear as soon as I complete my current projects I am going to do just that.

Some engineers have said that essentially all digital EQs are capable of arriving at the same sound if one knows exactly how the GUI reacts and that in a controls view, eventually given time any EQ can be made to sound like any other, except for Linear EQs. Is this so?

This may be somewhat true if you’re comparing digital EQ’s of a simplistic design that have the typical textbook-perfect response. But if a plugin’s aim is circuit emulation of something unusual like a Manley Massive Passive, a Pultec, a Trident A-Range, then stringent modeling of the circuit and measurement data must be employed for the unusual filter shapes, parasitic component behaviors, inter-band dependencies and other unusual behaviors to be recreated from the analog world to a very exacting degree. This absolutely cannot be recreated with a basic digital EQ. Just like any device–software or hardware, design decisions dictate the device behavior.

Thanks for clearing that up because it is widely repeated as fact, obviously erroneously.

Some EQ plugins seem to impart a sound the minute you open it up even without adjusting or loading a preset. This seems to me to be particularly so with the Pultec. Is this a modeled feature from the hardware?

I would consider this simply a byproduct of accurate circuit emulation rather than a feature. While in the 21st century this is considered a cool thing that the Pultec does, ironically, I would imagine that less-than-flat frequency response in the minimal state would not be considered a desirable trait by the original designer’s standards–but it is an attribute of the device nonetheless that is represented in our plugin.

Are there others that do this by design?

Almost all of our EQ plugins have what we call a “soft bypass” which provides the frequency response of the unit when using the unit’s circuit bypass. Most are very subtle, the Pultec is a bit more obvious.

Which plugins have been the most popular to date?

The 1176, LA-2A and EMT 140 has a huge lifetime achievement as does the Precision Limiter and Fairchild 670. The 1073, the 88R as well, and of course now the latest have been hugely popular: the EMT 250, FATSO, Massive Passive and Studer A800.

Do you have some personal favorites?

I can’t mix without the EMT 140, EP-34 Tape Echo, Roland Dimension D, SPL Transient Designer, LA-3A, dbx 160, Fairchild 670, Cambridge EQ, 1073 and 1081. In some cases we modeled my own personal units like in the case of the dbx 160 and the EP-34.

Interesting. Among my go-tos are the LA-2A, EMT 140, and 1073 also. The Fairchild does not work for me on a ton of stuff, but on the stuff it does work on, it sounds amazing.

For some reason I only ever use it on drum busses, and I rarely use anything other than time constant “1”.

Has UA ever tackled an emulation and then scrapped it because there were just too many anomalies to reproduce to your standards? If not, is there hardware that you would not tackle because you can anticipate that?

Well, you could say we’ve “deferred” plugins—specifically, we knew we’d be better off waiting until the UAD-2 to tackle something like Manley Massive Passive or the FATSO and even the Studer. At this point, I can’t think of anything that we couldn’t broach having tackled intentionally interactive and “circuitous” circuits like Massive Passive and FATSO.

The Studer A800 and EP-34 Tape Echo emulations are killers IMHO and the most “analog” of any plugins I have heard. I always assumed that partly what made analog sound like analog are the differences that result from the fact that no two tubes, capacitors, resistors, pots, and other parts can be manufactured to be exactly the same and they adjust to periods of use differently, etc. where digital is because it is just code. And yet, somehow you guys have captured this. Is that programmed in or are my assumptions just incorrect?

There are more interesting behaviors in the circuit design itself and the way the parts interact without necessarily needing to go into parts tolerance variation. In general, our approach is measuring a new unit, or a freshly serviced and calibrated unit that behaves as closely to the original manufacturer’s intentions as possible. This ensures the best measurement data which then is compared to plug-in’s design based on the circuit analysis. If I may speak for our engineers, if there’s a discrepancy between the plug-in and target hardware device, either there’s an error in the diagram, or a behavior is not necessarily represented in the diagram, in which case we would make a model based on the measured behavior. Each is a safeguard to the other, as is the “user” listening and measurement evaluation (my job, among many others in the process). However, it is interesting to note that a few circuits we’ve modeled were more susceptible to parts variation and aging than others—another reason to thoroughly service the units before beginning work.

Thanks Will for taking the time out of your busy schedule to do this.

A final note:
I just want to add that I am not saying that software emulations bring all that expensive hardware brings. Tubes age, hand crafted parts will vary somewhat, a compressor sounds different on the way in than on the way out, and so on and so on. But having the ability to totally recall mixes, use many more instances than even successful studios could afford with the hardware are enormous plusses for those of us working in a DAW. There are a lot of fine plugin developers and choices these days but in my opinion, the UAD platform belongs on anyone’s short list for serious consideration.

Comments

By George Leger III on December 8th, 2010 at 12:55 pm

Regarding your comment on how using the UAD plug-ins made your mixes better: I had just completed mixing a CD when I purchased my first UAD-1 card, and as a test I decided to remix one song with the card, using the basic plug-ins: 1176, la-2a, Pultec, CS-1, and a few others. My client was so blown away by the difference that we went back and remixed a whole 11 song CD, and in the end it became so much better than it was.

I now own 4 UAD-1 and a UAD-2 Duo, and I refuse to work without them. I won’t mix at another studio, regardless of what they have (since most studio’s have about 5% of the hardware the UAD’s emulate). The UAD family of plug-ins, and the additional power I get in my systems to mix, are one of the best things out today in my opinion.

George Leger III

By Hequin O'Mas on December 10th, 2010 at 11:03 pm

The Studer A800 is awesome! What it does to the low-end is so sexy.

Trackbacks

Leave a Comment