An Attempt To Understand Spectral Content From An ETC Response. An Article from Nyan Mellor.

[ernestcarl]

Toole, E Winer had shown we heard up 1/24 for the FR. For other curves (waterfall, decay, spectrogram, group delay) what would be the optimal resolution ?

I meant adjustments primarily to increase legibiility of the graphs presented which obviously helps improve the viewer's ability to read and interpret them easily -- whatever true resolution the ear-brain system is operating at for any given frequency (brain states matter), it mostly heavily smooths out the perceived response either way. There is no direct ability to apply psychoacoustic smoothing or frequency dependent windowing to the decay graph itself in REW, but I don't think it matters there that much. 1/12 octave is a good compromise. The default 100 ms rise time setting is maybe okay, but 20 ms is better to limit focus within early reflections. There is no universal, optimal one-size-fits-all view settings for any graph -- but the defaults are probably very generally okay to start with. I cycle through different resolutions, window/scale ranges etc. all of the time.

 

[Trdat]

The decay plot is now much more revealing (compared to just the raw ETC) as we’re viewing in high resolution both frequency magnitude and time. That said, we do not really hear at such a high resolution as your graph settings seem to be set at for (by default?). Maybe some smoothing — 1/12 octave, at least to make things easier to read.

Getting the hang of this.

But this is the problem, I need to understand the basics of reading the decay graph. Any links to a explanation?

 

[Trdat]

Shouldn't the first 1-2 ms be "very clean" to have a good stereo image?

Are you talking about the problem of early early reflections?

I think that is something that hasn't yet been defined by science and is somwhere between 6ms and 15ms. Early reflections after this time gap are usually okay according to what principles you want to believe that is.

 

[Thomas_A]

Are you talking about the problem of early early reflections?

I think that is something that hasn't yet been defined by science and is somwhere between 6ms and 15ms. Early reflections after this time gap are usually okay according to what principles you want to believe that is.

One could stretch it to a bit longer, but yes, that is one of the questions related to stereo imaging. For longer reverberation times, this one relates to large hall acoustics and may not be applicable to small-room acoustics. But I still wonder if there are any conclusion that can be drawn with respect to ordinary rooms?

https://asa.scitation.org/doi/10.1121/1.3579145

 

[Trdat]

Yes, we use filtered ETC/IR - but what we then see makes no sense. When the bandwidth is limited, the graph expands in time (no surprise to those into mathematics), it looks like it will always be worse the lower you go in frequency, no matter how much we absorb and dampen everything, it looks horrible, and we conclude there is no way to fix it. But this is not real - looking at decay graphs, we see that it is indeed possible to achieve a consistent decay profile across the whole frequency range - same decay at 200Hz compared to 2KHz.

I am just reading through the thread
Effect of Loudspeaker Directivity Compared with In-room Measurements

and reading through all your articles on your website, there is a lot of good information but still very hard to understand the basics.

So when you say consistent decay across whole frequency range do you mean similar in the way the waves and the colours are across the graph?

Of course achieving that is the skill but what is the difference between achieving proper RT60 to a consistent decay profile? This will elp me understant firstly, what the difference between the two graphs is.

 

[Thomas_A]

I also wonder how to interpret them. From my room at LP, attached. It would be nice if they could be normalised.

 

[Trdat]

One could stretch it to a bit longer, but yes, that is one of the questions related to stereo imaging. For longer reverberation times, this one relates to large hall acoustics and may not be applicable to small-room acoustics. But I still wonder if there are any conclusion that can be drawn with respect to ordinary rooms?

https://asa.scitation.org/doi/10.1121/1.3579145

General conclusions or a definite on what consitutes as earl early reflections?

Not from what I understand, I have spoken to Mathew Pose who is a bit of an authority in the field as a researcher and he says nothing is set in stone as yet when it comes to defining early early reflections. I have read some of Greisengers basic stuff(most are too complicated) who at times seperates these issues between large concert halls and small rooms and he has it somewhere in this ballpark(from what I remember) as well stating that late reflections even in a small room are beneficial. There isn't much out there on early early reflections but I have come across some articles and early reflections not early early especially lateral ones are beneficial. But we have gone over this many times on ASR so nothng new.

Yeh, I would like to know as well if there is any conclusions drawn.

I'll check out your link. Thanks.

 

[Thomas_A]

This is the spectral decay when measured 50 cm from the speaker, using 100 ms, 20 ms and 1 ms rise time setting. But I do not know how to interpret them.

 

[Frgirard]

a bit of history to read

https://gearspace.com/board/showpost.php?p=15619598&postcount=13

 

[Kvalsvoll]

and reading through all your articles on your website, there is a lot of good information but still very hard to understand the basics.

So when you say consistent decay across whole frequency range do you mean similar in the way the waves and the colours are across the graph?

Of course achieving that is the skill but what is the difference between achieving proper RT60 to a consistent decay profile? This will elp me understant firstly, what the difference between the two graphs is.

Much information, but not necessarily always good, if it is difficult to understand. Also note that I now mostly use the forum-pages on my web site - less strict in presentation form, I do not need to do a complete article before starting to post, this unfortunately makes it even more difficult to read. But that is how it is.

I can try to give an example - one treated room with fairly good spectral decay, one room where both spectral distribution is compromised and overall decay is too long:

We see that the early reflection level is still quite good in the 2. untreated room - because it is larger, and the speakers are placed away from boundaries. But later energy increases at lower frequencies, and overall decay is much shorter in the treated room.

To compare graphs, scaling and signal processing must of course be equal. But when analyzing and trying to find out what is going on in the room, diffenerent scaling and processing is used to look into specific aspects - if very early reflections are of interest, we choose a shorter window and closer than this 20ms line spacing. And the ETC/IR can provide information on very early reflections at higher frequencies.

We also need to know what we are looking at - how the measurement was performed has huge impact on what we can see, such as frequency response of the 1. top line (direct sound) and very early reflections are affected by seating and other objects close to the mic.

 

[Kvalsvoll]

This is the spectral decay when measured 50 cm from the speaker, using 100 ms, 20 ms and 1 ms rise time setting. But I do not know how to interpret them.

The rise time is the window - a long window drags along everything that happened before, so that the second 20ms line when using a 100ms rise-time will not show what is going on after 20ms, because it contains sound from 100ms backwards in time.

 

[ernestcarl]

If I weren't so lazy I'd try looking at the decay of the exact same speaker in a (I would think) relatively "good sounding" room vs a really bad one (e.g. a particularly small/reflective bathroom and/or empty garage).

Nobody has really posted measurements made in subjectively (to themselves) bad or worse sounding rooms thus far. *edit: apart from Kvalsvoll, of course -- though didn't exactly elucidate how much the room sounded worse psychoacoustically -- only that measurement decay was objectively longer.

 

[JG2020]

Good info on Waterfall and Decay graphs - AustinJerry's guide to setting up REW with USB microphones and HDMI

Then followup with BASS INTEGRATION GUIDE - https://www.hifizine.com/2011/06/bass-integration-guide-part-1/

 

[Thomas_A]

The rise time is the window - a long window drags along everything that happened before, so that the second 20ms line when using a 100ms rise-time will not show what is going on after 20ms, because it contains sound from 100ms backwards in time.

So that would be - background noise?

 

[Kvalsvoll]

So that would be - background noise?

Not noise, just a different time window.

Each line - from the top down to the bottom - which eventually drowns into the noise floor - can be seen as a frequency response of what happens at certain time intervals. A longer rise-time means that each line contains more information from what happened earlier. If we choose a very wide window, all lines will be similar, because they are all made up from the same data. If we choose rise-time 20ms, the first 20ms line will no longer contain information from the direct sound, everything we see on this line now will be reflected energy.

Just to clarify - the rise-time setting decides the starting point of the window in use, then the window-setting defines the length of the window.

 

[Kvalsvoll]

Nobody has really posted measurements made in subjectively (to themselves) bad or worse sounding rooms thus far. *edit: apart from Kvalsvoll, of course -- though didn't exactly elucidate how much the room sounded worse psychoacoustically -- only that measurement decay was objectively longer

I have now had the privilege of listening to and measure the same speaker in many different rooms, and by looking at those measurements and combine this data with listening impressions, some general observations about room acoustics can be made. Basically, the sound will always improve with better acoustics, and a certain level of performance is required to get the full potential that the speakers can deliver. And a better speaker can not completely remove the room from this equation - a better speaker just has a potential to achieve better sound, in any room acoustic environment.

I would not rate rooms like bad or worse. It is possible to enjoy sound in a room with less than optimal performance. We just have to acknowledge that there is a potential for improvement.

The graph I posted in #30 (https://www.audiosciencereview.com/...an-article-from-nyan-mellor.26422/post-906298) is an example of measured differences from a good treated room and a quite typical room that actually sounded rather disappointing (look at the quite good early reflection attenuation - and later reflection level should give a nice sense of "space".. well, no..).

A major difference between quite-ordinary and good is this sense of presence and clarity down into the lower midrange and bass - intruments appear like one body, with defined boundaries.

 

[Thomas_A]

Not noise, just a different time window.

Each line - from the top down to the bottom - which eventually drowns into the noise floor - can be seen as a frequency response of what happens at certain time intervals. A longer rise-time means that each line contains more information from what happened earlier. If we choose a very wide window, all lines will be similar, because they are all made up from the same data. If we choose rise-time 20ms, the first 20ms line will no longer contain information from the direct sound, everything we see on this line now will be reflected energy.

Just to clarify - the rise-time setting decides the starting point of the window in use, then the window-setting defines the length of the window.

Ah, I get it now! Thanks!

 

[ernestcarl]

It is possible to enjoy sound in a room with less than optimal performance. We just have to acknowledge that there is a potential for improvement.

I would agree. Though, just for the heck of it, I finally dragged my S8 studio monitor off its butt into my tiny single car garage to get a feel of what is a (not so impossible) "worse" listening space scenario.

*excuse the mess!

Best listening position (where the speaker still sounded great) was a tad bit forward in the middle of the room. Walking nearer back the end of the closed garage door, the sound just got progressively worse.

In my dedicated small listening room in the basement of the house, decay of the bass is longer due to solid concrete walls existing behind the dry wall. Nevertheless, there is some substantial insulation behind these walls. Over the front of these walls, I also stuck some odd pieces of very ugly -- but cheap! from a bulk discount I got years ago -- 4-inch thick acoustic foam panels here and there (quite somewhat randomly at first, but I've begun moving them around). The effective absorption primarily is at 500 Hz and up.

Some of the results:

Decided to leave the sub running with my couch-wide FIR & IIR corrections in place.

Now, the sound wasn't exactly "intolerable" in the garage at all areas -- in fact, it sounded great the closer one got to the speaker. The flutter echo also wasn't too bad, except that at very loud volumes fatigue definitely creeps in faster esp. if one didn't shelve the high frequencies.


Overall, though, I think the biggest thing that affected my experience using the exact same (coaxial small horn waveguided) speaker in these two vastly different rooms was probably the alteration/mangling of the frequency response:

I can imagine the result being possibly even worse if directivity of the monitor was not as controlled.

 

[Trdat]

I have now had the privilege of listening to and measure the same speaker in many different rooms, and by looking at those measurements and combine this data with listening impressions, some general observations about room acoustics can be made. Basically, the sound will always improve with better acoustics, and a certain level of performance is required to get the full potential that the speakers can deliver. And a better speaker can not completely remove the room from this equation - a better speaker just has a potential to achieve better sound, in any room acoustic environment.

I would not rate rooms like bad or worse. It is possible to enjoy sound in a room with less than optimal performance. We just have to acknowledge that there is a potential for improvement.

The graph I posted in #30 (https://www.audiosciencereview.com/...an-article-from-nyan-mellor.26422/post-906298) is an example of measured differences from a good treated room and a quite typical room that actually sounded rather disappointing (look at the quite good early reflection attenuation - and later reflection level should give a nice sense of "space".. well, no..).

A major difference between quite-ordinary and good is this sense of presence and clarity down into the lower midrange and bass - intruments appear like one body, with defined boundaries.

A thought occured to me and it was likely born sunconsiously from a few things you mentioned in this great thread with superb responses. In the decay plot each line(frequency response) is the next response plot from reflected/reverberated time that reaches the microphone after a certain ms. So, am I right saying that one way to observe the spectral balance of the reflections is to match the subsequent frequency respones to the first one?

Essentially, 20 ms would be the reflection/reverberation after 20ms and its spectral content or basically frequency response can be just matched to the first one? Have I understood this correctly? My understandng of spectral content is the reflected frequency response.

 

[Kvalsvoll]

So, am I right saying that one way to observe the spectral balance of the reflections is to match the subsequent frequency respones to the first one?

Yes, if we normalize later sound to the first arrival, we see how the speaker-room changes the spectral distribution of the reflected energy.

If the speaker has smooth and reasonably flat off-axis response, and the room is not overly strange, the later decay lines will be more equal to the first. As we move further out in time, the lines will tend to look like the power response of the speaker.

Let us say we have a speaker with wider radiation around 5KHz, and collapsing highs above 8KHz. Then we will see that the later decay lines will have a boost around 5K, and then drop off above 8K.

If we mis-treated the room with too much high frequency absorption, we will see that the higher frequencies drops off in the later decay lines.

This will hold until we get down to quite low frequencies. In the bass-range, the room will be dominant, regardless of speaker radiation pattern.