Analog is Better than Digital!
My visit last Friday to the NAMM (National Association of Music Merchants) Convention in Anaheim included a stop at the Universal Audio Booth. They are one of the major developers of plugins for Pro Tools and are responsible for the Manley Labs Massive Passive EQ plugin that I talked about a few weeks ago (Check out the post here). There seemed to be a little irregularity in the specification department when Bob Ludwig (a very well-known mastering engineer) discovered the frequency capabilities of the UA plugin didn’t work in high-resolution.
The Manley Labs Massive Passive EQ processor is a very fine piece of analog gear…and the price is commensurate with the perceived value of the device. The retail cost is $6300 buys you a tube stereo mastering parametric equalizer with lots of knobs and capabilities. Is it worth $6300? Of course not. When you consider that the Universal Audio Plugin for your Pro Tools rig can be purchased for around $300, it would be foolish to blow that much cash on a one trick device.
The claim by UA (although I couldn’t find the information on their website) is that the device is high-resolution audio capable. In my world, this means that the plugin will handle 96 kHz 24-bits no problem. But as I pointed out in the previous post, the highest frequency that it can output is 26 kHz. This doesn’t square with the 96 kHz sample rate, which should get us to 48 kHz.
So I wanted to chat with someone at UA about this discrepancy. I found one of the sales types at the UA booth and explained my situation. He didn’t really understand what I was talking about so he introduced me to the product manager Lev. He didn’t have a card but did give me his email address.
After I explained what Bob Ludwig and Elliot Scheiner had discovered during the production of a high-resolution surround music production about the limitations of their plugin, he admitted that they invoke a roll off…a low pass filter…in the output stage of their plugin at around 26 kHz. So I asked him why? Lev said, “we have to roll off the higher frequencies because the digital filtering that we’re using results in some artifacts at frequencies above that. We needed to achieve the very best fidelity in the ‘audio band’ and thus the 26 kHz maximum.”
It is difficult to design and implement extremely high quality IR filters across an expanded frequency range. UA (and they’re not alone in believing that 20-20 kHz…the “audio band”… is sufficient to deliver all of the music that we hear) designed an emulation of an analog processor that doesn’t match the specifications of the physical device.
I told Lev that Rupert Neve designs his analog electronics to pass 100 kHz signals and that the Manley process certainly does better than 26 kHz (the specifications on the Manley website claims the frequency range extends up to 60 kHz). So shouldn’t the digital version strive to deliver the same kind of performance? He went back to the “audio band” deflection. Essentially, it doesn’t matter because we can’t hear up there anyway.
So I left him with this final suggestion. If your devices are striving to meet the specifications of the original machines or high-resolution audio digital standards, say something in the product description or in the printed manual when you come up short. Don’t let well-know recording and mastering engineers find out through experimentation. No one likes to have an unpleasant surprise. Better to inform people than let them hear through the rumor mill that somebody cut a corner.
