Realistic sound reproduction and percussion

Hi Edorr !

I love percussion, and what I see is not a deliberate mixing decision, What I think is a lot of times they don't care about the result, and when they care, the recording company don't. I have a friend owner of a recording Studio, who showed me some some examples about what was sent for pressing and the final result some times the difference is so big that it do not appear to be the same recording. Audiophile recording companies, use to care about that, but in this case when they put everything on protools usually good bye dynamics. The best percussion results use to be about direct recordings as mapleshade records, MA recordings and a lot of another small productions about who care about that.

If you have interest I can send for you some very good percussion recordings.

All the best
Luiz

I agree 100% edorr. If I had the time, I would list some great jazz recordings with realistic sounding drums, but here are two right off the bat. Both are XRCDs. Duke's Big Four which is simply superb. I use it when I test systems at shows. The other, Bill Evans at Shelly's Manne Hole, especially track 6 Swedish Pastry. Check them out if you can find them. They can be pretty pricey.

Good luck finding drums that haven't been compressed and limited in a recording. The exceptions are far and few in between especially with any rock or pop. Unless you can confirm how they were done you will rarely ever be wrong in assuming compression and almost as often limiting before you get to hear it.

edorr said:

But why? Music just does not sound like live music that way.

Click to expand...

Likely afraid to blow up stock car stereos and all the computer laptop, desktop and Bose Wave radios out there. Sad fact is that the great majority of consumer loudspeakers out there are lucky to get to 70Hz flat. Approved audiophile fare is only a small, small part of what I listen to. It's a big reason I went with loudspeakers with adjustable LFXO with active bass. It helps a lot in bringing the low end of percussions up to scale with the rest. I hate it when a kick or a Tom sounds smaller than a snare. This makes for a compelling argument for use of curves vs shooting for flat.

If you go and record with only minimal miking, the level of all percussion is way above everything else. If you mike it closely, then everything is so loud, though your other parts of the music don't have to be. It just comes down to matching levels. Percussion is a couple orders of magnitude louder. So one common method is limiting so you get very high average levels of the percussion, but the peak levels don't ruin the rest of the mix. Now not saying it is right or wrong. But it does happen the percussion is compressed and limited most all the time. Which really makes accurate percussion on recordings a bit impossible or highly improbable. If you have recordings that get close to a real live sound consider yourself lucky. Most of the time the real sound simply isn't in the recording.

esldude said:

If you go and record with only minimal miking, the level of all percussion is way above everything else. If you mike it closely, then everything is so loud, though your other parts of the music don't have to be. It just comes down to matching levels. Percussion is a couple orders of magnitude louder. So one common method is limiting so you get very high average levels of the percussion, but the peak levels don't ruin the rest of the mix. Now not saying it is right or wrong. But it does happen the percussion is compressed and limited most all the time. Which really makes accurate percussion on recordings a bit impossible or highly improbable. If you have recordings that get close to a real live sound consider yourself lucky. Most of the time the real sound simply isn't in the recording.

Click to expand...

Completely agree.

Minimally miking a drumkit, say with only two mics (i.e.; two LDC condensers in a widely-spaced A/B configuration or two SDC condensers in an X/Y or ORTF configuration) invoves a number of compromises the artist/producer/engineer may or may not be willing to make for several reasons. Although the above techniques will produce a more realistic and naturalistic representation of a drumkit compared with close-miking, there are pros and cons to all: A/B enables you to capture relative level differences and time-of-arrival differences depending on the polar pattern of the mics (enabling you to achieve a very wide stereo image but being more sensitive to phase), X/Y only gives you level differences (though can have very precise localization and precise imaging) while ORTF gives you both level and time-of-arrival differences in a phase coherent manner (with less focus, but better depth and ambience).

These techniques work best when bleed from other instruments isn't an issue (because the drummer is tracking in isolation or because bleed from other instruments is preferable), you have a good room, and most importantly, the drummer him/herself has control over their own dynamics. esldude has covered the inherent problems of these techniques nicely above; that is, drums drown out everything else when played at the upper-end of their dynamic capabilities (i.e.; "realistic levels").

Close-miking gives you much more control over the balance, EQ and compression levels of each individual drum, minimizes room sound and therefore allows the artist/producer/engineer to "fabricate" a particular sound, which in the case of the Hiromi Uehara album, has produced a very present and full-sounding kit, widely spaced. The problem is that in actuality, the resultant sound resembles nothing like a real kit being played in a real space, especially from the perspective of an audience member. Which is to say "Another Mind" sounds great for a kit that's been close-miked and nicely mixed, but still nothing like drums sound in real life, even when sitting behind them. (As a counterpoint, compare "Another Mind" with, say, Tim Berne's Bloodcount "Discretion" - the first track Is That a Gap? is a good example - which is minimally-miked, and presents a kit (and band) being played in a room much more realistically in regards to dynamics (greater dynamic range, less compressed), imaging (more amorphous, less spread out), and tone (leaner, more room tone).

If "realistic reproduction of percussion" is really the desired result, then minimally miking the kit is the best option but with the above caveats, and won't suit every style of music. Close-miking the kit achieves better separation, a fuller, more present kit, but needs greater fettling in terms of mixing, and balancing the kit relative to itself (due to the different output levels of each mic, their polar patterns and the way each drum "speaks") and the other instruments invariably invokes the use of some form of compression either in tracking, mixing or both.

I've mentioned the BIS CD of the Kroumata Percussion Ensemble here in a few threads. It's BIS CD-232, and has a huge warning banner on the cover concerning the dynamics.

Lee

edorr said:

Could have fooled me. To me, between may be 10 different MCH Jazz trio recordings this was easily the most "real" sounding, with what I am hearing in a live venue as my reference.

Click to expand...

I think that's the point - production techniques are meant to fool us, producing something that is an enhanced version of reality, which, once it becomes ubiquitous, leads us into thinking that that's what reality actually is.

edorr said:

In case you have it, what is your assessment of drums on Frank Zappa's Quaudiophiliac (4 channel DVD-A)? The track "Chunga Basement" is a reference for realistic drum recording to me (although can't hold a candle to to hearing Lake Palmer live in the variety playhouse).

Click to expand...

I've only heard a stereo version of it (does that count?) and I think it sounds great - lively, dynamic, relatively monophonic (drums panned hard left) but with depth and I'm guessing recorded live in the studio with the associated issues of snare buzz, bleed from the drums into other mics and vice versa. And yes, as a reference for a "realistic drum recording" it sounds great. But it sounds nothing like the drums on Another Mind. So which is more "realistic" to you?

Did you listen to the Tim Berne track? To me, that's what drums sound like from twelve feet in front of the kit, like would be the case if I was standing there as an audience member in real life. "Chunga Basement" sounds like a great pair of large diaphram condensors hung above the kit, perhaps with an additional mic for the bass drum, which, good as it sounds, is not ever where my ears can be, unless perhaps I'm the drummer. Another Mind sounds close- and multi-miked to me (with some EQ and compression), which again, is a significant departure from what my ears alone can ever hear, and therefore, is questionable as to its veracity apropos it being "realistic". That doesn't mean it can't sound great, or impressive, or immersive, or "hi-fi", but it's not realistic in the sense of what one hears when standing in front of the kit listening as an audience member.

So it seems to me there's a continuum along which "realistic" can be interpreted depending on whether one's ultimate perspective is A) as an audience member, B) what the drummer or engineer might hear, C) what the drummer and producer would like to hear. I'd say Tim Berne's Discretion is A, "Chunga Basement" is B) and Another Mind C). That folks, is what one might call an opinion, but there it is.

For what it may be worth, here's a list of albums I think represent the three points along the continuum of "realistic" sounding kits or drums:

Minimally-miked:

Kroumata Percussion Ensemble/Gert Mortensen - Xenakis - Pléïades
Cooder, Bhatt - A Meeting By the River
Tim Berne's Bloodcount - Discretion (In fact, most of Tim Berne's output on Screwgun)

Close-miked but minimally mixed:

Shellac - At Action Park (Steve Albini)
Nirvana - In Utero (As above)
Jon Spencer Blues Explosion - S/T (Ditto)
Nina Nastasia - Blackened Air (Ditto, again)
Ginger Baker Trio - Going Back Home
Billy Cobham - Spectrum
Tony Williams Lifetime - Emergency (In fact, most Tony Williams, esp. Miles era 1964-69)
McLaughlin, DeFrancesco, Jones - After the Rain (In fact, most Elvin)
Led Zeppelin - Presence (In fact, most Led Zep)

Close-miked and mixed well:

Joe Henry - Civilians
Gateway (Abercrombie, Holland, DeJohnette) - Homecoming
Dave Holland Quartet - Points of View
Torn, Karn, Bozzio - Polytown
Jeff Beck - Guitar Shop
Slint - Spiderland
Tool - Lateralus
Rush - Counterparts
Battles - Mirrored
Don Caballero - What Burns Never Returns
Soundgarden - Down on the Upside

Close-miked, room-miked, compressed, EQ'd, edited and over-produced to death but what the hell it still sounds awesome:

Metallica - Black Album

edorr said:

I spend some time zapping through my extensive collection of Piano Trio / Quartet recordings at high volume (wife was out) reproducing realistic (i.e. like live venue) sound levels. I did this with my multi channel recordings, but the conclusions hold for stereo as well. It occurred to me that of course all the usual suspects of good recording were important and differentiated between recordings, but the single most important variable differentiating realistic sound from non-realistic sound is pure and simple the dynamics and relative sound level of the percussion (i.e. drumkits). To sound like the real deal, you need very loud and highly dynamic drums - that is how a jazz ensemble sounds live, end of story. I found nine out of 10 recordings were falling short in this regards. Whenever I stumbled on a realistically recorded / mixed track, I was at the edge of my seat. This most notably happened with Hiromi Uehara's SACD "Another Mind". Other highly acclaimed SACD recordings simply did not make the cut - very nice sound, piano to die for, but not realistic because of lacking percussion dynamics. Do others share this experience / opinion? Why is it that most mixes simply do not reproduce percussion realistically? It is so blatant, that if it was not for the occasional recording that gets it right, I would think it a limitation of dynamics and low frequency response of my system. It is not. If one recording can sound right, they all can. It appears to be a deliberate mixing decision by the engineer.

Come to think of it - the same is true in concert (rock) blu ray's - most simply not have the drums recorded at realistic levels. No array of subwoofers can compensate for this mixing decision.

Click to expand...

Hi edorr,

In my experience, a lot of this comes down to judgements made by the engineer and/or producer that are the result of a combination of how they use the microphones and how they choose to monitor those signals.

My own take on it is to reject close mic'ing (once you've made hamburger out of it, it will never again be a steak ;-}) and to reject "stereo" (really dual mono) mic'ing of the drums. So rather than 7 mics on a drum set, or even 2, how many would I choose? My answer is zero. (Zen in audio?) Rather than mic the drums -- or any other instrument or voice -- I find the most realistic sounding recordings, the ones that I feel do the best job of "getting out of the way" are those where the engineer mics the *event* rather than the individual components therein.

If I may, two examples I can cite from my own work are Equinox (a quartet playing Haitian music with a jazz influence, selected as Stereophile's Recording of the Month in February 2011) and Americas (a full on jazz quartet). The latter even features a drum solo by percussionist Tony Jefferson.

Works the same with any type of music. There are some examples in our as yet small catalog with rock drums too. As with all of our recordings, there is no processing and no dynamic compression. I've long felt that dynamics are one of the last frontiers of making recordings that sound like real music, so I never use compression or limiting of any kind.

Best regards,
Barry
www.soundkeeperrecordings.com
www.soundkeeperrecordings.wordpress.com
www.barrydiamentaudio.com