Skip to main content

Hi. I'm an enthusiastic amateur, learning as i go. I have been recording school concerts and orchestras in large cathedrals and churches. Nearly all the advice and tutorials around the use of compression and limiters seem to be aimed at bands with drums, guitars, bass etc etc.

The dynamic range of the concerts i record is broad. Not only due to the nature of the music, but also because very young children feature in the performances, and the brass band for example, often gets stuck under my main mic array. In between the main choir (on x6/7 risers average 120 people) are various ensembles, middle school choirs, soloists etc etc. There is a lot of movement, fidgeting, and space is at a premium.

My question: can anyone suggest avenues of research that relate especially to the post production of classical music?

I track at levels of average -18 to -12 dBFS.

I have recently attempted this: Using a Sonoris Mastering Compressor at Threshold -40dBFS with Ratio of 1.15 - 1. This gives an average gain reduction always of between 2 & 5 dB. I get the added volume and a hint of more power in the CD but of course i get the unwanted sounds from the background, especially the turning of pages that sounds like frying chips.

I want to maintain a sense of 'difference' and 'scale' in the individual performances while allowing the quietest to be heard. So far i have resisted normalising all the individual tracks (the recording is split into items...i currently work with the whole mix as i have had no success in bouncing the mix to a stereo track and mastering that.
I did this because (Sonoris) my final CD's (for parental consumption, archive, and school funds) are always rather quiet, peaking at -12dBFS and averaging much lower.

I wanted to use a Limiter to catch the relatively few but strong peaks from timpani and brass that happen when i raise the overall levels.

Here is the equipment i use:

RME UFX interface
Mics include: AKG C414 xls's, MC930's, NT55's, NT4, with others.
Plug in's most used are Sonoris, Fabfilter, and NI (for reverb)

I use a large Cathedral stand and a Ambient Jumbo Boom to place my main stereo pair which is usually ORTF above/behind conductor. With ORTF over choir, but lately i have experimented with a AB spaced pair in front of choir and will trial a DPA stereo pair (4015C) next May for the 'Armed Man' in Cirencester Church/UK

Any suggestions to where i can find information pertinent to my work would be much appreciated
(p.s. my work is all voluntary)


pcrecord Mon, 03/17/2014 - 09:38

First, I would adjust the compressor so it lower some peaks, not to have a constant reduction. Go lite on it, slower attack and/or release.. But I guess it just may be that your threasold is to low (since you don't specify attack and release)
But you could try automation instead. I guess automation is a pain to do but may give you more control. I meen if you automate the volumes first (before you even compress) You may realise you don't need compression.
In the end, one the beauty of classical music is the dynamics. I would try to conserve most of it.
for the turning pages, you can address that a few ways, within the automation, with a compressor and sidechain or a multiband compressor. In the end automation is my first pick.

Boswell Mon, 03/17/2014 - 12:31

Welcome to RO!

Is this the event you asked for opinions on miking in the SoS forums last September? If so, you got some good replies there, so I assume your concern now is about the best way of processing the recorded tracks to even out the levels between the different "acts" whilst not squashing the dynamics of each act.

This is not an easy task, not least because a natural characteristic of young or amateur groups is a lack of sonic balance between numbers. It's something I battle with all the time when asked to record and produce CDs for events of this type.

One reasonably-straightforward method for attacking this sort of CD is a simple multi-stage process. The first of these usually involves measuring an average level of each track and setting a compressor on high ratio (or even as a limiter) to operate on the peaks solely on that track, irrespective of what the final track level comes out to be. After doing this for all the tracks, I would then listen carefully to the levels of each track relative to one another, noting what adjustment is needed on each track to get a reasonable relative loudness, bearing in mind the type of material in that track and the forces performing it. At this stage, the peak level could still average around -12dBFS. After all that is done, every track on the CD is brought up by the same amount so the maximum on the CD is around -0.3dBFS peak. Some tracks will peak at substantially less than this, but at least you know it's you as a producer that has made the artistic decision to make it that way.

Boswell Wed, 03/19/2014 - 17:14

I never use compression as such on classical works. You will note that I carefully talked about using a compressor as a limiter, which will leave all the real material untouched. It's the uncontrolled peaks that need to be tamed, just as they are in radio broadcasts of classical concerts by reputable organisations such as the BBC. You have only to experience the lack of dynamic range in most television audio and in FM broadcasts by a certain UK commercial classical station to know how unnatural true compression can sound.

With the proper settings, limiting can be undetectable to the end listener, and it enables higher average levels than would otherwise be possible.

tyrolean Fri, 03/21/2014 - 03:51

thanks to all for these thoughts. I found a whole new world opened up when i mixed down to a stereo track and worked on that. My mixing decisions were made and i could concentrate on an 'approach' to mastering (i hold no illusions of my abilities btw).

I found automation solved most of my issues, and the use of the limiter solved the rest.

However, I am still minded to produce a x2 disc set for the parents: a car & kitchen special with a more audiophile disc.

The responses here have confirmed my own thoughts and feelings on the subject.

to all


audiokid Sat, 03/22/2014 - 09:56

I don't know about all the math but from a working studio POV, if you are finding this thread in search for advice on what kinds of compressor some choose for classical music, I use a Crane Song STC-8 on every track that goes through my system. You don't hear it but you know its doing its job.

Its a wonderful touch for classical/ rich sounding acoustic music.

mberry593 Tue, 03/25/2014 - 10:46

Several points...

Yes, IMO you really do need to do some control especially with symphonic material. Getting it to fit (dynamics wise) on the media is a concern but an additional concern is practically making it work in noisy listening environments. Some of the previous comments are very negative about this. Please understand as you read the rest of this post that I am advocating only VERY GENTLE PROCESSING!!! It only takes a little to make a significant improvement and it is very easy to overdo it as evidenced by some RCA recordings of the mid 1960s.

Let me suggest from my experience that you consider parallel dynamics...aka New York compression. This is very frequently used in classical music. The idea is to split up the frequency spectrum into segments and process each independently.

This practice is now just about universal in radio broadcasting. It got it's radio start back in the late 1970s with Michael Dorrough and his DAP (Discriminate Audio Processor).

This was considered to be almost mandatory in 1976 when Disco kick drums started punching holes in audio.

Many believe that the practice was invented in the 1960s at Motown but Bob Olhsson sets the record straight on his site:

"It (parallel compression) was a fairly common technique especially in the classical genre.
We pioneered multitrack tape machine punch-ins and monitor switching but not parallel compression."

Indeed, it is my understanding that the technique was started in the late 1950s Deutsche Grammophon for use on Classical recordings.

This article states that it is "exclusive" to classical music, but we all know that is wrong!

On notes of personal experience...

I find that 2 bands isn't enough. IMO you need at least 3. More than 4 though seems to be unnecessarily complicated.

It is of some use to just do the processing with independent units...that does get rid of the 'hole punching' but the real power of this technique for me is to have different timing for the bands. The problem is that you need fairly fast timing to catch short-duration peaks but fast settings tend to mangle bass. The idea is to use longer attacks/releases on the low band.

You may find it helpful to pump pink noise through your setup and look at the results on a spectrum analyzer. BTW, an excellent & free spectrum analyzer is available here:

I have been using the Waves C4 for this for a few years with good results.

If that appeals to you, you might consider the C6 which adds side-chain access and is now on sale for $89 USD.

Recently however I have been playing with a different approach. I have the Blue Cat MB-7.

This allows you to split up the spectrum, insert plug-ins of your choice and put it all back together again. I highly recommend this plug-in!!! I have played a little with using 1176 limiters this way with promising results.

I have also in the past tried to do my own splitting with equalizers & filters. I had some hope that maybe I could get some interesting action by playing with linear phase filtering. That turned out to be very difficult and the results were not useful....maybe a larger investment in patience would have payed off.

Finally, here is a link to an opinion that is going to be controversial about loudness, compression and new listeners.

I don't agree with everything in that article, but it is interesting reading.