David Polich wrote:Btw, in my experience, the instrument that
is the most challenging to get right in a mix is an acoustic
piano.
I completely agree with THAT! I've put a lot of time into studying what goes on with that. The problem is much more complex than it first appears.
To begin with, a piano is a percussion instrument. So is a celeste. The attack is ping-y, but that's not the meat of the sound. What we like to hear is on down the slope of the decay. When the competing sound of an orchestra, band, or chamber group rises above a certain level, it begins to mask the good stuff in the piano sound, but it has little effect on the "ping" of the hammer attack. This changes the perceived sound of the piano, making it more brittle and plunky. The nice, melodic decay is masked.
The old composers like Mozart and Beethoven knew this, and they managed it very carefully by clearing out the competing frequencies or controlling them for effect. They also doubled the octaves when the going got fast and loud. Rachmaninoff and other romantics began adding a 6th or 3rd, and sometimes even an 10th (3rd above?) in parallel to help distinguish the piano's sound from its competition.
Slow movements were all about hearing the tone. Fast movements were about technique. But what could you do when you wanted both? We have at our disposal another arsenal of tools that they did not have: compressors, expanders, and limiters.
This is where it gets good: By using a fantastic "compander" like Waves' C4, C5 or C6, you can actually alter the slope of the decay. This gives the ear a moment more to pick up the sound of the piano and distinguish it from the competing ocean of sound.
Imagine it this way: The sound of the orchestra is water. The sound of the piano is a fin sticking out of the water. Without the orchestra (no water), you see the whole porpoise (solo piano sound). But as the water rises, you see less and less of the porpoise, and when the water gets deeper and faster moving, you only see the tip of the fin as it swims excitedly in the waves.
In raw recording, the piano sound we hear above the loud orchestra sounds more like a xylophone than a piano. But adding a compander, with threshold set a little below the attack level, you actually can control the slope of that "fin" as if you're flattening the porpoise so that more of its fin and back rises above the water. (Pardon my jumping back and forth with the metaphor) The result is that more of the meat of the sound is perceived before it dashes below the waves. It's a beautiful thing!
However, you may not want to do this for Beethoven, Rachmaninoff, or others of the old schools. These guys knew what they were working with, and they arranged their music specifically for that sound. Some more than others. It may yet be effective on some composers who were less aware of the subtleties of this. You just have to use your ears and decide.
But for modern music, where the mix is done at the board more than in the score, this is a fantastic method of recording piano. When I first made this work, I was blown away by the sound. Maybe a better way of putting it was that I was less annoyed by the impossibility of mixing a piano. Suddenly it became possible.
The trick is finding the threshold. In the Waves C4 (and family), you move the Range (not the Threshold) ABOVE the mid-line by about 3 to 6 dB (give or take; depends on context), then set the Threshold a little OVER the average attack levels. Since you are working in frequency bands, you will emphasize the lower frequencies (the part of the sound that gets lost) and de-emphasize the upper frequencies. You may even want to flip-flop so that you get expansion below about 2500 hz and compression when you get up to 4000 hz. I don't remember whether I ended up using that setup much. Believe me, I experimented with all kinds of crazy curves. I do remember that simpler is better, but a single flip-flop doesn't necessarily mean it's not simple.
When you find the right threshold, you'll hear the richer sound of the piano hesitating an instant before decaying. It's not an odd sound; it just sounds like the piano your ear wants to hear, as if you've been given X-Ray glasses to see those porpoise fins just below the surface of the waves. You haven't altered the sound so much as "tilted" the decay. At lower volumes, the sound doesn't reach the threshold, so it has less effect, which is a good thing. At higher volumes, the effect increases, but since you set the Range, you control how much you can possibly get out of it. It's one of those things that just works.
This, to me, is the #1 reason for owning the Waves C4 family. (C4, Linear Phase Multiband — which I call the C5, and the C6 with side chaining) If I were buying only one, I'd get the C6. The new MOTU processor, whose name I forget, may enable this technique, but I haven't tried it, so I don't really know. Waves' C4, C5, and C6 are part of my standard toolbox, so I haven't really worked with the new MOTU plugin.
This would probably work on celeste as well as piano. I hesitated to mention it earlier, because I haven't actually used it on celeste. But I definitely have used it on piano (dozens of times) in loud mixes, and I recommend anyone to try it. So, you might as well give it a try on celeste, too, and see if it fixes some of the problems of mixing.
The complexity of recording and mixing a piano in an ensemble has many facets. This is just one of them. But this is probably the biggest one. Solve this, and I think you'll find most of the other problems easier to deal with.
Hey... you heard about it at MOTUNation!
Shooshie
A typical ensemble-solo piano compander setting. Note that the high band is compressing while the other bands are expanding. I'd fine tune this before actually using it. Also note that this is functioning as a 4-band processor, because I've got the two side-chain bands bypassed.
Edited by Shooshie, December 23, 2015, 7:00 p.m.
