How do great sounding domestic listening rooms measure?

[Senior NEET Engineer]

Speaker spinorama preference seems to be well understood in general, but what about the rooms? Is RT60 suitable for assessing liveliness of small room? What is the preference for RT60 curve? Should the curve be flat from 20hz to 20khz, or downward slope, etc? Do most prefer similar value for same room size? Does stereo vs. multichannel matter? Does music genre matter?

 

[mikewxyz]

I think the cognoscenti would say that RT60 metric does not apply to small rooms because the diffuse requirement is not met. RT60 is also a function of room volume so one value doesn’t fit all. My point is that there is a certain amount of preference to all this.

There is at least one website where you can hear different concert halls and reverb times. It’s interesting to listen to. I didn’t bookmark it so you’ll have to search.

I’ve been trying to reduce/ level out my Topt curve to unmask some lower frequency sounds but have not been successful.

 

[amirm]

General target for RT60 is about 0.25 to 0.50. The lower value is good for home theater and non-acoustic music. The latter is better for orchestral. The value is also room size dependent. In every large, open spaces you will have a hard time getting down to 0.5.

As for shape of it, that is a more involved topic and not one that I would focus on.

If you are using REW, RT60 has different names depending on metric it uses for the final attenuation. Topt is a good one to use.

 

[andreasmaaan]

I don't know of any clear, simple answer to this question, but I can tell you what I do know.

Firstly, as mentioned above, the idea that RT60 should be applied to domestic listening rooms has been questioned, since its definition is based on a uniform distribution of energy and random directions of propagation, which are far from the case in small domestic rooms.

Normally in recording studio environments, the goal is constant reverberation times at all frequencies, with a tolerance of (but not a preference for) some rise in reverberation time in the bass. For concert halls and other large performance spaces, a rise in reverberation time in the bass is considered preferable.

As the volume of a room increases, permissable (in the case of studio environments) or desirable (in the case of halls/auditoriums) reverberation times also increase.

Perhaps from this it can be inferred that some rise in RT60 in the bass in listening rooms may be desirable, and that larger listening rooms should have longer reverberation times. But I'm not sure either of these inferences is entirely valid.

IEC 60268-13, which relates to rooms for conducting listening tests on loudspeakers, specifies a maximum reverb time of around 0.6ms above 200Hz, rising slightly in the bass. ITU-R BS.1116–3, which relates to rooms for subjectively testing degradations introduced by perceptual encoders, specifies a maximum of just over 0.3ms above 200Hz, with a target of 0.25ms.

Taken together, these arguably set up a tolerable range for "critical" domestic listening of between 0.3 and 0.6ms (above 200Hz). Interestingly, this range coincides with the average range of most domestic listening rooms (around 0.4ms, with some country to country variation).

 

[Wes]

 

[amirm]

One other thing to consider is that in western languages, the duration of combos of consonants and vowels lasts about 0.5 seconds. If your reverberations don't decay fast enough in time for the next combo of consonant and vowel, intelligibility can suffer.

Conversely, if you absorb reflections enough to get the reverberation time too low, then you lose energy that is needed for someone to properly hear the consonant and vowel.

Take these two factors and you wind up with the target range of 0.25 to 0.5 seconds I mentioned.

 

[MZKM]

As mentioned, RT60 is not valid for domestic rooms.

EDT, C50, and C80 are more appropriate curves in REW to use.

Using a Spectrogram/Wavelet with normalized peaks is also good.

 

[QMuse]

As mentioned, RT60 is not valid for domestic rooms.

EDT, C50, and C80 are more appropriate curves in REW to use.

Using a Spectrogram/Wavelet with normalized peaks is also good.

REW actually does quite a good job with RT60 and the reason is because it doesn't calculate it the same way for small and for larger spaces.

From REW help:

"Where the EDT is much shorter than the T30 RT60 figure REW's Topt RT60 calculation uses a start point based on the intersection of the EDT and T30 regression lines, to determine a point that lies within the diffuse field region. It then tests each possible end point in 1 dB steps and picks the one that gives a regression line with the best linear fit. That produces a more reliable RT60 figure."

 

[tuga]

Conversely, if you absorb reflections enough to get the reverberation time too low, then you lose energy that is needed for someone to properly hear the consonant and vowel.

Are you referring to large rooms without amplification like classrooms, theatres or auditoriums and the ratio between direct and reflected sound?

Because we can properly hear the consonants and vowels in the desert if the source is close enough.

 

[Kvalsvoll]

RT60 can be misleading in small rooms, because the decay is too short and consist of discrete reflections rather than a diffuse sound field. But I admit I look at RT60, to get some rough estimate of how live a room is - small rooms often end up around 0.15 - 0.2s, larger room will have longer RT60.

Decay should be similar across the frequency range, but often it is difficult to achieve low decay times in the bass range. A room with too much thin absorption will have very short decay at high frequencies, a speaker with collapsing polar response will also tend to show similar curves.

A better way to visualize decay is to look at the decay graph, or the spectrogram.

The IR plot shows how sound energy dies. The IR is weighted towards higher frequencies, so it does not give a good picture of overall sound decay.

The ideal room has a very fast drop in sound energy down to completely dead, then the sound energy rises again after a short time interval, and then continues to decay rather slowly. This is difficult to achieve in a small room, where early reflections tend to be too loud and early decay is too slow, while late decay is too fast.

The radiation pattern of the speakers have significant impact on the shape of the decay profile - both in time and frequency. Some speakers are born with a better early-to-late reflection level, and will perform much better especially in small rooms.

To analyze acoustic performance, it is therefore necessary to look at several different graphs and representations, and RT60 is not the most useful information.

This post needs some pictures; some IR graphs seems appropriate.

First is a speaker in a small room, very little or no acoustic treatment. just like most of us have:

Then a slightly different speaker, in the same room:

We see that this speaker gives a huge improvement towards the ideal - there is a fast drop in level early in time, and then the sound rises slighly before it continues to slowly disappear.

The same speaker in Room2:

Now we see some improvement of the early gap, and a huge reduction in decay after the initial rise around 10ms.

The RT60 graph for this last situation:

The RT60 can be misleading here, because the shape of the decay profile affects the RT60 values, giving different RT60 depending on how we choose to calculate. Though in this example here, all values shown are quite similar.

 

[andymok]

Shorter for a monitoring environment / multi-ch, gives better localisation

Longer for a more lively sound

~ .5s is the figure for the large rehearsal room we built that can house an orchestra as well, considering the music genre and modern taste.

 

[Eddy]

I listen to Geithain RL 906 speakers in a small corner of a L-shaped room. The size of the corner is only 1.8 x 1.2 x 2.2 m (LxWxH). Distance speakers - hearing position ~ 1.1 m.
I have some panels type 244 and 242 from GIK Acoustics (broad band absorbing > ~80 or 200 Hz).
Using REW, I have measured RT60 of ~130 ms @ >200 Hz and increasing values up to 500 ms from 200 to 50 Hz at listening position.
Removing the panels resulted in an increase of RT60 from ~130 ms to ~150 ms. Still a very short RT60. My conclusion is that the room size has a major impact on RT60, even more than the acoustic treatment.
My listening expericence was that I heard indeed more details with acoustic treatment. (But this effect was not very clear and can be some kind of placebo.)

What I do not understand: Many people (like @amirm) say that 250 to 500 ms is a favorable RT60 value for hifi. Would that mean that listening via headphones is unfavorable because it lacks reverbaration by nature? ("RT60 of headphones" should be more or less zero, or not?)

Shouldn't be the "favorable amount of reverberation" part of the recording/mastering? Why do I need extra reverberation via my listening room?

 

[MakeMineVinyl]

Beyond numbers and charts and graphs (oh my!), its very important that you don't have 'slap' or 'flutter' echoes anywhere in the room. You can simply walk around the space while clapping your hands. The decay should be completely smooth, even, and brief (the previously mentioned 200 mS - 500mS).

When I built my dedicated room, I aimed for it being a very good 'echo chamber' when completely empty before acoustic treatment. By that I mean a decay which is totally smooth without slap echoes or flutter echoes with a decay time around 1 second. Then I applied acoustic treatments to bring the RT60 down to 450mS in the mid-range. The room's size is 20' wide by 30' long with cathedral ceiling/open beams with some large irregular surfaces on the side wall/back wall.

 

[dasdoing]

View attachment 51634

where is this from? it doesn't take into acount the direction of the reflection? A -10dB 10ms reflection of the right speaker coming from the left wall wont create space, but confusion. I guess this was made using a single wall and speaker?

 

[youngho]

Speaker spinorama preference seems to be well understood in general, but what about the rooms? Is RT60 suitable for assessing liveliness of small room? What is the preference for RT60 curve? Should the curve be flat from 20hz to 20khz, or downward slope, etc? Do most prefer similar value for same room size? Does stereo vs. multichannel matter? Does music genre matter?

May I suggest reading the following as a starting point, otherwise it's going to be "No, no, it depends on multiple factors, probably not unless one likes brightness and dislikes bass, no, yes, yes"?

http://www.aes.org/e-lib/download.cfm/17839.pdf?ID=17839

http://blog.acousticfrontiers.com/storage/AMS%20for%20Stereo%20List.%20Rms.pdf

https://users.aalto.fi/~ktlokki/Publs/mst_laukkanen.pdf

 

[seedragon]

One other thing to consider is that in western languages, the duration of combos of consonants and vowels lasts about 0.5 seconds. If your reverberations don't decay fast enough in time for the next combo of consonant and vowel, intelligibility can suffer.

Conversely, if you absorb reflections enough to get the reverberation time too low, then you lose energy that is needed for someone to properly hear the consonant and vowel.

Take these two factors and you wind up with the target range of 0.25 to 0.5 seconds I mentioned.

This is fascinating. Does it imply that languages with shorter phonemes (Mandarin?) have tighter decay requirements?

 

[risandipra]

I know this is an old discussion, but I bumped into this whilst trying to find "appropriate RT60 value for home listening". Then I put the number Amir mentioned into Amroc Calculator. Tweak the Roon's EQ according to the room modes in relation with my listening and speaker's position.

IIRC about a year ago I measure using REW and the results is pretty close with Amroc Calculator.

Been tweaking with speakers position since, and got satisfying result after trying to replicate Floyd Toole's Sound Reproduction suggestion per Chapter 15, then finished the EQ according to the RT60 value suggested + Amroc Calculator.

Many thanks for the RT60 suggestion.