Dynamic compression and classical music

[Keith_W]

I spent the evening doing a comparison of various versions of Beethoven's 3rd symphony and realized something I haven't noticed before. Modern recordings are dynamically compressed. Well I knew that it existed in pop recordings, but always thought that Classical music was free of such things.

Case in point: Furtwangler's 1950's recording compared to Claudio Abbado's recording in the 20th Century with the same orchestra. My memory of the live performance is the finale - when the full orchestra comes on song just before the close, the sheer loudness knocks your socks off. And so it should, when 100 musicians are playing full bore in a confined concert hall. When the same moment is reached in Furtwangler's recording, it compresses, clips, and distorts - which tells you that some recording engineer didn't have his finger on the volume to turn it down. But when the same moment is reached in the Abbado - the fortissimo is reached cleanly. But after careful listening (and the SPL meter on my phone) - the fortissimo is only 6dB louder than the preceding piano.

After listening over and over I am convinced that the Abbado performance isn't that much inferior to the Furtwangler, but the Furtwangler comes across as much more free and exciting - despite the sound quality issues. I think the failure to capture the dynamic swing is what makes it sound bland.

So who is to blame? Is it the media, which does not have enough dynamic range to capture an orchestra in full tilt? Or is it the recording engineer who is determined not to overload the disc?

 

[Bruce B]

This is typically what happens with Classical music. The label will say that we have to turn the volume up to hear the quiet sections and then during a crescendo, it's way too loud. This is where compression comes in and it's all dependent on the label. Some labels want it, some do not.

 

[microstrip]

Symphonic music is a predictable victim of compression. I found that Mahler and Beethoven, two composers with large audiences, suffer from it in some recent releases.

The Swedish classical label BIS recordings is known to use no compression and to maximize dynamic range in their recordings - if the recording starts with a pianissimo you should have your remote on hand when listening!

 

[rbbert]

I just checked out Abbado's Eroica from the box set (in Adobe Audition), and its compression is minimal and typical of any major label classical, even from the analog era. OTOH, Mackerrras' 2006 live set does suffer from some noticeable compression.

You may well be noticing a performance difference. However, one thing I've noticed consistently (and this is totally subjective) on all types of music; some older recorded music that has obvious harmonic and IM distortion seems to have more dynamic contrast, and I'm not sure why that is.

 

[Keith_W]

I recently attended a live concert of Bruckner's 4th symphony. I had premium seats. Out of curiosity, I decided to measure the SPL using my phone. Pianissimo = 80dB. Normal playing = 85dB. Fortissimo = 98dB. That is a 18dB swing between pianissimo and fortissimo.

I came home and pulled out my only copy of Bruckner's 4th symphony (Gunter Wand) and adjusted the volume so that pianissimo = 80dB. Fortissimo was only 91dB, i.e. a swing of 11dB. No wonder it sounds flat, boring, and lifeless compared to the live performance! It is not as if my system is not capable of loudness either! It makes me angry that I paid good money for this CD. You would not expect recording engineers to be working against you like this, yet this is precisely what they are doing.

 

[GaryProtein]

Whose phone SPL meter are you using?

 

[Keith_W]

Whose phone SPL meter are you using?

HTC One XL using Sound Meter. Yes I know the mic is not calibrated, but the point of the exercise was to compare the two scenarios using the same mic.

 

[GaryProtein]

I was curious because the JL Audio app in my iPhone STINKS. You can take it into the quietest closet in your house, hold it between your sport jackets with just enough space so you can see the phone, and it still reads about 53 dB. Their spectrum analyzer is equally bad, always registering at least 45 dB over most of the range above 200 Hz.

Moral: if you want to measure ABSOLUTE levels, you need real measuring equipment.

 

[Keith_W]

Moral: if you want to measure ABSOLUTE levels, you need real measuring equipment.

Of course I agree with you. I knew that I was not measuring it with any great precision ... the idea was to get an idea of the relative SPL. My ears tell me that the dynamics in the concert hall are much superior, but I know that people here on WBF will pooh-pooh my suggestion that my ears are in any way reliable. The next best thing was to use my phone to measure the SPL. I'm certainly not going to pack along my SPL meter to the concert hall.

 

[Bill Hart]

Aren't there multiple causes for this? On old recordings, was it the limit of the equipment at the time? Microphones? Tape machines? What can be cut onto vinyl for playback? When old records are converted to digital, old tapes?
On modern, digital recordings, aside from externally imposing compression, are there limits to intelligibility of information (again, microphone and the rest of the recording chain) when you have that many instruments playing at once, at full tilt?
Finally, aside from playback medium and the recording process, the constraints of the playback system?
Keith you use Acapellas, which can deliver the dynamics. My Duos can too, but I find that, for big orchestral stuff, or complex music that is dynamic (and I'm vinyl only, so that may be a limitation), I listen at a lower nominal volume to 'leave room' for big dynamic swings. The trick for me was not only to get the 'noise floor' (or whatever you want to call the ability of the system to interpret/convey nuances at low volume without sounding analytical) and make sure the listening environment is dead quiet. Part of it too may be the limits of the room, given the relatively smaller size of my listening room compared to a concert hall. This the area where the illusion can most easily fall down in my system.

 

[rbbert]

I would bet that because of the smaller space and more reflections in your listening room (compared to the concert hall), your phone is measuring a significantly higher level during pianissimo sections than it should (relative to the loudest sections), again compared to using it in the concert hall. It would take better measuring equipment and more refined software to make this comparison valid. That's not to say classical recordings aren't compressed, just that your anecdote really doesn't show that.

 

[GaryProtein]

Of course I agree with you. I knew that I was not measuring it with any great precision ... the idea was to get an idea of the relative SPL. My ears tell me that the dynamics in the concert hall are much superior, but I know that people here on WBF will pooh-pooh my suggestion that my ears are in any way reliable. The next best thing was to use my phone to measure the SPL. I'm certainly not going to pack along my SPL meter to the concert hall.

I wasn't disagreeing with you, just adding another point based on my iPhone SPL dB meter experience. From a relative dB standpoint, the iPhone meters do work.

 

[RBFC]

Conductors also have rather large dynamic variation between them! Having heard the Cleveland Orchestra under many conductors, I can unequivocally state that the individual style of the conductor can also impact the dynamic range of a performance. Lorin Maazel, for instance, got the orchestra to almost blow you out of your seats in tuttis. Christoph von Dohnanyi, on the other end of the spectrum, was quite reserved. I remember watching von Dohnanyi lead a performance of Beethoven's 4th. In the big moments, I'd see him clench up and drive with both arms, but the SPL didn't climb as expected.

Now, perhaps hall/atmospheric conditions and audience amount have a substantial effect, but my sample size is "1" since I only attended that one performance of the piece.

Lee

 

[24bit]

Fader ridin'

It is not all ways compression that is used on classical recordings but more often the old fashioned ''fader riding''or to put it simply, turning the overall volume knob up or down very gently in louder or softer passages.This is of course the simplest form of compression.Except it is done manually.
This has been done through out the whole recording history,normal consumer systems can't handle ppppp to fffff.
But also if you take a Mahler symphony with Tuba,Trombones and basses coming in with a fortissimo low C,some taming of the low frequency's has taken place,especially with L.P's or the needle would be jumping all over the place.That sensation of feeling acoustic instruments making your bones rattle,is almost only possible live.
So lets not forget to go out and support live music!