Related question-- What tool do you use to measure max RMS?

--Jay

I had a recent piece which went from really thin ppp to really think fff and I tried to render it as natural as possible. I ended up with 40dB of program dynamics, and it sounded pretty real. But it was really hard to listen to on the stereo, so with additional mixing and some limiting I reduced it to 30 dB. (peak dynamics -30 to 0, RMS -40 to -10dB). Still it's a lot, some people say it's too much. I think it's good.

I work in Samplitude and it has very good metering.

It doesn't work that way, Jay. Perceived loudness is fluid, and the idea that 10dB sounds "twice as loud" makes no sense.

If you're asking about what the dynamic range of a recording should be, the answer depends on the music. In any case, you aren't going to capture the dynamic range of a live concert on a recording.

Okay what is the dynamic range for an extremely dynamic classical piece.

@Nick Batzdorf said:

...In any case, you aren't going to capture the dynamic range of a live concert on a recording.

Hello Nick (loving the magazine)

That can't be right - the VSL demo of Pictures at an Exhibition is bordering on uncomfortably loud.

@JBacal said:

Okay what is the dynamic range for an extremely dynamic classical piece.

As I said, I think 30dB is fine. And I observe this dynamic range in recordings, too.

Don't forget recordings that you buy are typically in 16 bit (around 90 dB max range). An orchestra can have a wider range (from single instrument pp to full tutti ff or fff. That means that they need either clever gain riding, upward compression (subtle ratio with low threshold) or more drastic (multi-band) compression. I can't believe they just check the loudest passage, adjust the board and hit the rec button...
24 bit would do a lot more justice to the orchestra.

Does your 30 dB range apply to a single instrument, a group, or the entire orchestra?

orchestra

mathis-- have you taken a commercial recording similar to your own track to take measurements of it as a benchmark?

yes, but not scientifically. I used to work in a mastering studio where I was cutting also classical recordings. I didn't make notes about the dynamic range back then but 40dB turned out to be hardly listenable and a limitation to 30dB sounded right. Still a lot though.

Sorry if this not totally on topic, but i thought it worth mentioning velocity which 'perceptively' can influence the impression of louder and softer. I don't know how you guys organise your working method, but i do velocity first, then adjust levels. It's been my experience that the range (ppp-fff) can often sit within acceptable parameters of tolerability, with velocity accounting for a proportion of 'dynamic' to a listener's ear.
I know this is basic stuff, but i thought it worth a mention, as i think it's related.

Two quick roubles worth.

Regards,

Alex.

all good points to which could be added that dynamics in the repetoire don't necessarily pertain to volume level. they can also speak to color and/or density. thus it is entirely consistent to have a densly notated score sounding forceful at a relatively low volume level. For example, the dynamic markings in Cesar Franck's organ chorals simply refer to the degree to which the swell box is closed (p) or open (f), where a lightly registered passage marked f is actually intended to sound at mp at most. this practice was common among french organists in the ninteenth century (franck, saint-seans, faure) who translated these colorist ideas to their orchestral writing, but more importantly to their academic jobs where their influence became pervasive as the new century dawned in paris. everyone was there and it seems unlikely that primarily orchestral composers (stravinsky, debussey, ravel) would have escaped the perception that orchestration might be less about volume than attitude.

the poor wind players with cymbals and trumpets behind them might not think so, but it's remarkable how quiet an orchestra really is from the conductor's (and hopefully the composer's) point of view...