This is what Crescendo looks like on the Mac:
CrescendoLive is really just an AudioUnit host for several AU Plugins. The input source is indicated in the top left drop-down, while the output destination is the top right drop-down.
The bottom panel is really the Crescendo unit. The middle panel is a different plugin, called CLAS, which can spice up your listening if you like. But for professional use, as when mixing and editing audio, you would not want to have CLAS enabled because it would cause your mixed bass and treble to result too thin.
In the Crescendo (aka vTuning) panel, you can enable / disable the Crescendo DSP processing with the Process checkbox.
The HdphX is a special preprocessor for when you want to listen through headphones. You don’t have to use HdphX if you don’t want to, but it helps reduce ear fatigue when listening to strong stereo separation. It simulates the effect as though listening to loudspeakers at +/-30 deg azimuth, as well as providing some Haas effect in a Left/Right crossover that simulates hearing the opposite speaker in your ears, including some 5 kHz peaking to simulate the face wave propagation. I find that I use it almost all of the time, but when listening through loudspeakers, you should disable it.
There is a volume control slider and another called VTuning. That is just a nonsense name for the 1-knob adjustment for your best listening. Its readout shows what you estimate your threshold elevation is at 4 kHz. And you could start with it initially set to whatever your audiology report shows. But don’t be surprised if you end up at a slightly different level from there. Audiology is only accurate to +/-5 dB, and only then with good hearing. When your hearing is impaired, all bets are off. (more on this in a later post)
There is a little triangle symbol in Crescendo (aka vTuning) at the bottom, just to the left of Parameters. Clicking on that reveals several settings that need to be made the first time you use it.
Headroom is selected in 6 dB increments. For people with small amounts of impairment, they can just as well use 0 dB of headroom. But for persons like myself, where my 4 kHz threshold is 60 dB, I use 6 dB of headroom to give the DSP algorithm some processing room without clipping on the way out to the speakers. You can make up for that headroom attenuation by raising your amplifier volume control.
But that is the only time when you should touch your amplifier control. When you adjust volume anywhere after the Crescendo processing, the algorithm has no idea that you have done so, and that will end up making things sound incorrect. Whenever you need to adjust your volume levels, after first setup, you should use either the Volume slider or another source material volume control.
The Deemphasis drop box selects a corresponding outboard EQ that is used only in the most severe cases of impairment. In those cases we have to add so much gain at the high end that we would routinely clip the output. So instead of allowing that to happen, we internally apply deemphasis with a negative slope toward higher frequencies, and then make up the necessary restoration emphasis with a high quality outboard equalizer, like the Massenburg 8200 (very expensive, but equally fine!) For most people, just leave Deemphasis at None.
The Cal dBSPL and Cal dBFS are the “science” calibration sliders. If you have a sound level meter handy, you can perform an accurate calibration by suspending all processing in Crescendo, playing a 1 kHz sinewave or pink noise of known levels (Cal dBFS), and then measuring how much sound that actually produces in the room or headphones (Cal dBSPL). Knowing those two numbers allows us to know how loud your sound will be when we see the music samples streaming through the Crescendo algorithm.
As a default, we assume you will generally set your listening level at the output amplifier such that a -17 dBFS 1 kHz sinewave will show 77 dBSPL in a sound meter. You’d be surprised how humans just naturally gravitate to that level. This is also the level preferred by Mastering Engineer Bob Katz.
77 dBSPL corresponds to loud but comfortable listening, and is the targeted 0 dBVU level. By comparison, the DbX standard for movie theaters has 0 dBVU correspond to 83 dBSPL. And you know how loud that is…
Finally, for purists, you can select from among several high quality headphones, if you have them. Telling Crescendo what headphone is in use helps it estimate how much more, or how much less, correction is needed at each frequency, due to the non-flat nature of the headphone spectral response. In most cases you could leave this at None. But I’m a purist, and I use very fine quality headphones like the Sennheiser HD650.
Once these settings are properly established, you can close up the triangle again, and just operate on Volume and VTuning levels.
I’ll do another post on CLAS and what it is, and how it works.
Preamp, at the very top, is just another way to control the volume. I usually set all the volume controls to 0 dB everywhere, and actually control the volume either with that Preamp setting, or externally at the source.
When we deliver a CrescendoLive to you, you actually get this AU Host, as well as 3 different AU Plugins. Crescendo is one of those plugins.
That means, in addition to being able to listen through the CrescendoLive host, you could alternatively use the Crescendo plugin in the output channel of your DAW mixer. I’ll be showing how to do that in another post. That’s actually how I most often use Crescendo on my DAW computer. But on my laptop, for portable use, I’ll revert to using CrescendoLive to listen to recordings.
On a Mac, I use the freely available SoundFlower pseudo device to pipe audio from iTunes to CrescendoLive. The System Preferences are set up to make SoundFlower the default output device. Then the output side of CrescendoLive is sent to your actual playback rendering device – either the Built-in Output of the Mac, or an external Audio Interface, like my MOTU UltraLite.
So sorry, but if you insist on using Windows, at this time we can only offer a VST plugin for Crescendo, so you’ll need a VST host, like Sonar or Tracktion. However, that version of VST plugin has CLAS, HdphX, and Crescendo all built-in. It is in all respects the exact same DSP algorithm, just recoded to present a VST interface.
Whereas, back in 2006, the Crescendo algorithm soaked up 80-90% of available CPU cycles in a dedicated audio DSP computer, today’s multicore Pentium processors can execute this code in 10% or less of available compute cycles. So, loading should be very slight to imperceptible on most of todays computers and laptops.
And finally, of course, Crescendo is compatible with sample rates of 44.1 kHz, 48 kHz, 88.2 kHz, and 96 kHz. I haven’t ventured to higher rates yet, and really see no reason to ever do so. But that’s another story for another post someday…