Listening Volume - Emulated vs Acoustic Pianos?


#1

How much attention do you pay, when playing back samples, to the playback level (SPL).?

How does the volume of your playback through speakers compare with playing back a real acoustic piano?

For a long while I did not enjoy playing back sampled pianos, until I increased the volume, but I do not play acoustic pianos, so have no way to compare, if I’m playing at the right volume.

For those of you who play both dgital/sampled and acoustic pianos, do let me know if you attempt to match the volume at which you playback the emulation, with the real thing, or this volume matching is irrelevant to the realism of the emulated piano?

Would you know the dB SPL (and weighting if known - e,g. C weighting) at which you listen to your piano samples, and/or an acoustic piano. If yes please share in response.


#2

hi, I just saw this and have been thinking about loudness wrt the Bently piano samples.
dBSPL would be a great objective to use, but you kind of can’t. Every piano’s perceived loudness is going to be entirely dependent on the size of the piano, its age and resonance, the room in which it is being played and the comfort level of the player (for starters) and that’s going to be completely variable from day to day, song to song, note to note.
So, two thoughts:
if you are playing one of the sample pianos through speakers or headphones, for yourself, make yourself happy.
If recording your performance for mixing, you want to have a level with decent headroom in the recorded file, but you listening level is still subjective with the rest of the mix, room factors etc.
If it’s midi, who cares man? You can change it all in the mix at your lesiure.
As a point of reference, when mixing, a dbSPL of 80 is a common number. I find mixing in headphones as I do and in my small space, 80 dB is too loud. With sample instruments, I have discovered that mixing for long periods is worse on my ears than with recorded audio from ‘real’ instruments. Be careful.
: )


#3

Thanks Just Lisa,

So thrilling to get a response, finally.

Especially as most of us are working on audio/video/music, in the nearfield, by this I mean less than 2 meters from the speakers, every small change in our position, and every small change in volume of our audio, IMHO, has an impact. Which was why I asked the question about ideal volume levels, in the 1st place. I really wanted to hear what others thought about this, to also check on my own understanding of the importance of this issue.

This phenomenon of volume sensitivity is made the more poignant when we consider the Fletcher-Munsen variability in our perception of frequencies at different loudness levels (volumes) of sound.

The video above provides evidence that all vocations are perpetually a work in progress, and contrary to the thinking that there might be answers to these questions on loudness, it is also possible that there are no converged opinions, and everyone involved is simply “winging it”, like the Wild Wild West, no formulas, i.e just trial and error, and its more about JFDI than well thought out approaches to loudness… So maybe rather than try to seek esteemed opinions, sometimes we are the pioneers of our own understanding on subjects, cos there really is no answer to be gleaned. Or pianobook may not be the place to seek such answers.

So one thought is, is there are definitive consensus around loudness volumes… or none, or several different consensus/schools of thought?

https://www.soundonsound.com/techniques/establishing-project-studio-reference-monitoring-levels

The excellent article in Sound on Sound has been one of my references for gleaning an answer.

I do not yet have a sound level meter, and this is definitely one of my nice to have’s or birthday present to self, excuse to spend a bit of money for a good cause. Seriously, everyone in audio should have an SPL meter as their primary instrument, to judge things without bias. Part of my concern was also hearing protection, ensuring that audio was not unhealthily loud.

At the very least using an SPL meter on a smartphone device that is fairly consistent is a start, I understand that that the SPL meters on Apple devices have a more consistent result, as the hardware underpinning this measurement is well known and consistent, at least far more consistent than say a Windows Phone or Android based device.

My experience, if I may add this here, has been :

  1. If you have a good DSP process for tuning your speakers so that they sound more accurate than physically possible without any DSP, you can hear a lot better, and can lower the volume yet still hear very clearly. I tried Sonarworks but I think a good idea has been hijacked by capitalism, and while the results are pleasing, I found more accurate results IMHO using a combination of REW and generated correction impulses from REW.

All of the above applies to accuracy at the mix position, and caveat that it may not sound fantastic at any other position.

  1. There is a really nice location in front of the speakers where just a small lean in of no more than two or three inches allows you to pinpoint something a bit better - temporarily, without having to change volume levels. - like a mini magnifying glass.

  2. The ideal listening volume changes with ambient noise. I do not have a sound proof room, just closed windows and a closed door, in a bedroom studio. At night because the ambient noise from outside is a few decibels lower, I can listen, especially after 11 pm at lower volumes and still hear very well.

  3. Anything else is a bit of trial and error, swinging between higher and lower volumes and over time recalibrating to an ideal - and even the ideal changes a bit over time, as my ears have I think become better at listening and I do not need to listen as loud anymore… While the loudness settings on some of the audio flowpath are fixed, I can vary the playback levels of an audio piece or track using level plugins - which have one advantage over using faders - I can save each level as a preset…such as

a. Quiet listening - minus 8 dB
b. Tracking:- plus 3 dB
c. Post Recording - minus 2dB

Levels changes used above are purely for illustration

So depending on what phase I am at in production, I can listen to a track or mix at a relatively lower or higher volume, almost like having a monitoring dial but in software, rather than hardware. This point of using well defined levels which I can see and adjust in the DAW, has been a Godsend, to keep me honest on any new track, kinda like a preliminary gain staging,.for monitoring purposes.

In my DAW, I can save the chain of presets including the level plugin settings, so each time I need to use a specific sample library, I can call up the entire chain including the relevant gain settings, without having to start from scratch. Definitely a time saver, for libraries I use very often…The level presets are not an effective strategy though for libraries with thousands of sounds! (impossible to curate in this way), which typically have variations between audio levels from one patch/sound to another.

  1. Revisiting levels - After a few minutes of listening our ears acclimatise, almost like they have a compressor in them that gets turned on, so we become impervious to loudness. Taking any break for about 10 or more minutes, can help reset this sensitivity, and invariably I find that on coming back from a break I can turn down the level and still hear clearly.

I would estimate that I listen at somewhere between 55dB SPL and 70 dB SPL, depending on my mood, time of day, and whether I’m composing, patch editing, tracking or mixing,.or just enjoying some good old Spotify…

This is already really long, I’ll add the rest in another post.


#4

Hi. I haven’t read the entire post yet and I have to go to work! But funny you link that video – while watching it when the jump cut happened, at first I thought he was in a restroom!
One reason I work in headphones is realizing how subtle changes in body position can affect a mix (crazy right !?!?) and how bad my room is for mixing.
I use a Morphit plug in and my Beyerdynamic DT770 250 ohm and an interface to bypass the computer audio.
No, no discussion now. Work.
I promise to return.
: )


#5

Back now. Yes, an SPL meter is good to have, I got mine through Amazon.
A point of interest to me is that I took it to my work at a ballet studio and measured the ambient noise in each one during the day, no one around…72dB between outside noise and HVAC and the rooms’ cavernous reverb. The sprung floor is like a drum, the walls are concrete block and one side is all mirrors up to 7’
There is no drop ceiling installed which would be the norm for the type of construction and the sound baffling installed between the trusses is marginally adequate, but doesn’t keep out the outside noise one would usually never hear in a normal context for a building like that.
I take my mixes there and crank them when I want to do a type of ‘car test’ – if they sound decent there, then I’m doing well.
Even in my fantastic headphones (I can’t recommend them enough, amazing bass response and now that I’ve been using them with the tiny (Rode A1) interface I got for the purpose for all of my listening, I can hear things I never used to everywhere. And by everywhere I mean out in public – I had to step away from my mother who was talking too loudly in a crowded restaurant, she wasn’t really, but my perception is sensitive. Picked up some 30$ ear protection for live shows. Anything with any kind of mic’d PA, movies, concerts, shows: I must have them or be in pain.
In short – your ears learn to listen differently as you get used to whatever environment you are in, especially when it is intentional listening.
And I have given up trying to get straight or consistent answers about levels. I think, given how subjective it all is, that no one person or place can be like another.

Right now, editing samples, noise reduction process that I am unfamiliar with, it’s slow going and I’m turning it up a bit. I believe that six months from now with more practice, if the trend is true, I won’t need to turn anything up any more to get the job done.

Another interesting thing I’ve noticed; mixing recorded audio (songs) is less fatiguing than mixing music made with synths and samples. I think it has something to do with the ever changing harmonics in the ‘live’ audio. No matter what volume I’m using.
: )


#6

Just LIsa - you make an excellent note about the relationship of perceived volume to what I would consider as ambience(real or simulated).

Thinking through your observation, and attempting to derive both an explanation for this phenomena, as well as a best practice response to harnessing the observation with audio processing -

  1. As you have rightly stated ambience helps to increase the real or perceived volume of audio, and quiet sounds are brought up in volume, which can be a creatively positive enhancement.

There are obviously good acoustic reasons for the buildings in which we listen to live events,

I find that a bit of reverb using a combination of convolution via impulses as well as algorithmic plugins, provides the right backdrop to unsterilise the artificial(static) nature of sample libraries, which are for very good reasons captured with microphones positioned close to the sound source. It is something I do by default. has become a habit. The reverb is applied via an aux/send so I can adjust the amount of audio which is sent…to taste

I find it akin to the use of comfort reverb which is added to the monitoring chain of vocalists, to enable them hear their close miked voices, in a more enjoyable context. Singing in the bathroom is another relevant example…

I work in a bedroom studio, so with carpeting and clothes in wardrobes and the bedding/covers have fairly decent absorption so my challenge is the opposite, without appropriate room effects, sampled instruments can sound too dry and lifeless…At least while I’m composing/tracking the feeling of being in a room with ambience (albeit simulated via reverb) augments the virtual illusion. sufficiently.