.... and there sits the ME thinking "well I cant do like this, even if it sound good here, because it will sound sh*t for the consumer" ... it's the good ol' circle of confusion....
//
"Examining decay in terms of cycles reveals resonances and ringing issues clearly as additional oscillating periods" - Burst decay is a way to see Cycle time normalized to Periods. I can see the sense in it.... but in the case of keeping cycle time even across the whole Spectrum... Bass sets the bar, since it will always limit how low you can go.... and whos got a good even Spectral decay through out the whole spectrum? I don't even have that in headphones.
(normalized, mic laying on one of the headphone channels)
(normalized, mic laying on one of the headphone channels)
Last edited:
Actually, now I am thinking about it, weren't Linkwitz and co, as well as other people on the forum doing this like 10-15 years ago as well?
Jeez, it has been a while, but all of a sudden I am starting to recognize stuff again 😀 😀
If I remember correctly, the focus back than was mostly just room modes (< 250Hz), which never was very interesting to plot in such a way imo.
For reverberation in the room itself, you always have to do a whole bunch of averages on different positions to get a decent estimate. (used to do these a lot back than)
In that case such a graph is also not that interesting anymore I guess?
Jeez, it has been a while, but all of a sudden I am starting to recognize stuff again 😀 😀
If I remember correctly, the focus back than was mostly just room modes (< 250Hz), which never was very interesting to plot in such a way imo.
For reverberation in the room itself, you always have to do a whole bunch of averages on different positions to get a decent estimate. (used to do these a lot back than)
In that case such a graph is also not that interesting anymore I guess?
I was looking for this comment. This is clarity too, right? RT60 times are focused on, or only practical for certain parts of the spectrum>?
I mean, The wavelet graph seems to be the most useful for the task, not to say no other graph helps. Measurements around the room, would help explain what is happening and how to treat it, Measurements at the listening position will tell you how well you've done. Right?
I am not quite sure what you mean with clarity in this context?
Here is a graph of some measurements I did many many years ago (was working at an acoustic consultancy company back than). I can't remember how many positions this was. It must have been in an average slightly bigger office room with at least ten positions. (since the standard error bars are quite small)
Out of those positions I calculated the standard deviation (SD) as well as the standard error (SEM).
This is what I did typically back in the day.
You can very clearly see that the T60 is clearly spreading above and below certain frequencies.
(most acoustic companies call the graph RT60, but the reverb time itself T60 btw, I forgot why)
Correlating this back to the actual measurements, you will see this also in just the SPL graphs as well.
In this case this must have been an empty room, with enough furniture you will also see those spread lines going bigger.
Depending on the size and absorption coefficient of these things.
This is a typical graph, I have measured quite some rooms this way (also living rooms).
This is also the main reason why I still doubt that constant directivity is such a thing above 8kHz, since the local absorption can be really high but also the SEM will grow much bigger (= NOT constant in a room)
The reason I find a (R)T60 more useful is more because other companies use that method as well.
So you can compare your results immediately.
Something which can't be done with the other graph.
Reverberation needs to be good and constant through the whole room, not just on the listening position I think.
Attachments
lol 😀 😀
Well, it kinda depends on the genre as well I guess.
Although this is also a weird circle again.
Some very well done recordings are very artificial.
They contain far more details and subtleties that a person would never be able to hear when listening to certain instruments.
So yeah, what IS good to begin with? 😀 😀
Maybe we should just enjoy the music instead hey 😉
Periods, ms., doesn't matter, it's all the same thing. Yes, spectral decay would affect tone, but not to a significant degree, kind of a 2nd order effect. Except at LFs, most good loudspeaker systems would decay sufficiently fast so as to make this not a major concern.
Your points on reasonances being notable, yes, this is a bad thing. It would be and should not occur. But this is also obvious in the impulse response.
Last edited:
I am basically coming up with a list of things that affect tone perception which essentially is the judgment of specific frequency level in contrast to all other frequencies present. So for example, I went to my DAW and tried to create a situation where there was an exaggerated reverb tail, that was so, evenly, at all frequencies to see if it made tone, more apparent.... I think it works (wasn't able to ensure the Evenly across all frequencies part) but I really need to experiment with this in depth
What I am thinking is that I can exaggerate certain aspects to make certain qualities, readily apparent. While a very dry environment might lend its self to critique of dynamic processing, and imaging, a wet environment could be the environment that made Tone very apparent.
True Tone gets hidden by colored FR.... Tone does not get hidden by exaggerated decay if evenly spread through out the spectrum, is my thought. I think, emphasis on think, the Tone becomes more vulnerable when signal duration is increased. I think, resonant notes, become torturous in an acoustical situation where spectral decay is balanced but there is a long duration of decay.
Last edited:
Reproduction according to the source. Wrt a live recording: as if you were at the performance, or the performance takes place in your room. For studio recordings and especially electronic music, the 'black box' in the recording/production chain gets big quickly.
Every loudspeaker system suffers from coloration to a greater or lesser extent. Loudspeaker systems contain dozens of different materials. It's a myth to assume that those materials have no impact. The question is whether the - sometimes barely/not measurable - coloring is audible/disturbing.
....ignoring the influence of electronics, cables, etc. upstream.
Agree.... If most records sound nice and fine for you, your system has a filter in it 😉
//
That's a very silly and pedantic remark.
If something is (scientifically) not significant, it's simply not existing.
It absolutely doesn't matter if there is a microscopic difference. As soon as a person is unable to persive it. Or when other variables will just simply dwarf those differences.
There is a slight variation here depending per person, but certain types of "coloration", most definitely fall within being undetectable for human beings.
If there is any technical difference at all to begin with.
The question is never if there is a difference, the question is how significant the differences are in the entire context.
The two of you probably agree, it's not an easy point to make.
Sometimes speaker design compromises only allow you to just go under the threshold of audibility on the effects being compromised, lest one not be reduced enough while the other is reduced more than necessary. It doesn't have to be a bad thing when you get it in balance.
Sometimes speaker design compromises only allow you to just go under the threshold of audibility on the effects being compromised, lest one not be reduced enough while the other is reduced more than necessary. It doesn't have to be a bad thing when you get it in balance.
I never said it has to be a bad thing?
There is only a big difference in what someone finds subjectively pleasing because of believing in something and what's being said as an objective and audible difference.
If you feel better and think your audio sounds better by painting everything pink, please go ahead.
In fact, seen from a psychological and psycho-acoustics point of view, that argument is 100% valid.
Fyi, the opposite is also true; (strongly) beleving that nothing matters and therefor you won't notice a difference.
It's just about the nuance in these statements that's missing and often repeated the wrong way.
Saying that every little thing matters is a silly statement without knowing its context.
The only thing I do agree on, is that there is no perfect loudspeaker system by definition.
Since there are always a bunch of contradictory compromises that have to be made.
It's basically the law of preservation of garbage in physics.
Once you go into the direction of one variable or solution, the other ones will suffer from that choice.
So yeah, from that point of view there always will be some sauce by definition.
If you like it or not.
That's right, I wasn't talking to you specifically. However my point in saying so is that you can make successful compromises.
Logarithmic scales may not go to zero, but it's the fear of "how much is enough" that can cause people to go too far and ruin a perfectly good compromise.
That's basically what I was saying as well 😀 😀 🙂
Mostly trying to point out that acting from fear is never or very rarely a good thing.
Which definitely can ruin the whole experience, yes totally agree with that! 🙂
Understanding the nuances is one of the ways to prevent these kind of things.
Like said, move in one direction thats good, it will have consequences non the less. You have to figure out whats really important. At the beginning of my journey for this project I organized lists of points to satisfy as I learned them from the community. Sometimes its just a personal significance and others, felt on a community scale. Dispersion is likely the most important aspect to satisfy. Create a list of Dispersion traits to satisfy and from that point, many, desirable products can be formed, with various limitations of level and audience size. Thd and all the aspects that contribute to it, must be right behind Dispersion because both, directly limit aspects of audience size. Dispersion limits Height and Width, Thd limits Distance. Thats the 3d sound field in a nutshell.
Theres some abstractions in there😉
Dispersion and Thd, covers all aspects doesn't it? When these two are satisfied, the only thing left is, how it performs in the room.
Theres some abstractions in there😉
Dispersion and Thd, covers all aspects doesn't it? When these two are satisfied, the only thing left is, how it performs in the room.
Last edited: