Frequency response and human hearing

Yes.

no

The acoustician told me from the beginning on that considering the circumstances the measurements (first ones) were amazingly good. With these speakers I was in 5 different places and I never had big problems with the acoustics.

The 30Hz is the room resonance frequency. Actually even with the +7dB the sound was really fantastic.
Instead of the digital correction, I already put in 7 bass absorbers to correct it. Measurements will follow end of month.

Sorry but I don't understand what you mean. Having a flatter response under 50 Hz I heard the bass more intensively. In a certain way louder.
Though I must admit that in my case Loudspeakers make more difference than acoustics (flat frequency response).
I believe that for the human hearing a flat frequency response is to a certain degree not necessary.

 

Can you hear a difference with the 7 bass traps?

What I meant was that if you have a pair of small bookshelf speakers that only go to 50Hz, I would rather they had a raised low end rather than flat. If you are going to have a flat response it is better to have speakers that go very low. IMO of course.

If your room has a resonant frequency of 30Hz then it must be pretty big, that is probably the main factor in why it had pretty flat measurements anyway.

 

Yes I head the difference and I would say most people would have. When they were made at the carpenter next door his whole room got treated. We noticed that working there was another feeling as before. We wasn't used to that and he liked it.

About non linear frequency response on bookself speakers (>50Hz) I must say there are people who like speakers with different raises in the frequency response. There are developpers which make purposely non linear Speakers. For them there are other factors which are much more important.

When you hear an orchestra in a theoretical ideal room it has a certain (natural?) sound. When the same orchestra is playing in two different halls it will sound different. I hear this often because there are two main concert halls where I go to: Tonhalle in Zurich and KKL in Luzern. The first one was built in the 19th century, the second a few years ago by a star architect and with all the today's knowledge of acoustic. In Zurich I have the feeling that the whole hall vibrates and gives a coloration. In KKL there are far less vibrations and the sound in absorbed or diffused by a lot of acoustic elements. The ears adjusts them selves very quickly in both places so that the sound feels completely natural. I wonder if a linear frequency response over a certain level is for people who anyway care only for technique and not for the values in the music.

 

The acoustic requirements of a concert hall and hifi listening room are very different but never the less in both situation our ears certainly do adjust and compensate for room colourations (and hifi imperfection) very well.

Whether there could be too much of a good thing seeking the over accurate room-loudspeakers-interaction would be interesting as I have never hear a ruler flat hifi speakers yet. Bearing in mind that there are many more requirement for natural reproduction but never the less do you feel having a more flatter FR is an improvement to your ears?

 

Earl Geddes has some interesting ideas on room/speaker interaction, I'm working my way through his Premium Home Theater book ATM.
GedLee

 

A distinction should be made between equalising the frequency response (at the listening position) for bass or for mids and treble.

I cannot think of any reason that the bass response should not be equalised but equalising the upper frequencies will disturb the balance between direct and reflected sounds and result in a lack of realism. For example, if room reflections are causing a boost at, say, 1,000Hz, equalising will cause an increase in the ratio of reflected to direct sounds.

In any case, as has been said, our hearing will naturally compensate for imbalances.

Also, bringing the speakers away from the walls will cause the direct sounds to be heard significantly before the reflected sounds. The effect of this on our hearing is to diminish the importance of the reflections.

Although I wouldn't advocate the use of equalisation for the mids and highs, there is some value in the judicious use of carefully placed absorbent/reflective materials in the room. For the bass end, traps and the like are generally a good idea.

 

This depends IMO if the "problem" comes from the loudspeakers which don't have a flat FR or from the room acoustics.
Having a peak in a room that causes it and equalizing, will lower the energy coming out of the speakers for that frequency (no flat response from the speakers anymore). The room will then compensate this "fault" with a longer reverb time.
The energy at the listening place at time 0 (or slighty later) is higher when using aborbers (this in case of a peak in the FR) but for a shorter time. Equalizing the signal you'll have less energy at time 0 but a longer reverb time. Since the most important moment for perception is the first wave I believe a listener would hear a slight difference which depends how much was equalized. I believe you should correct the cause and not compensate it somewhere's else.

 

I should have added the proviso "unless the equalisation has the effect of driving the loudspeakers beyond their capability".

I think I follow most of what you're saying but don't see how parametric equalisation can affect reverb time? Are you referring to another type of equalisation?

 

You could of course not only equalise but try some phase adjustment at the specific frequency which has the longer reverb time. By letting that frequency play slightly before the rest of the audio you can get the highest amplitude peak (in this case reverberant sound built up after the direct sound) to be in closer time with the rest of direct sound.

This causes a 'pre echo' from the direct sound but in the case of bass with a lot of delayed room build up, is not very noticeable.

Anyone know if TACT offer phase correction? I know they do time alignment of the subs with the main speakers but do they do frequency specific delay to fix phase issues??

 

I think what Titian is trying to say is that the equalisation is compensating for the rooms longer reverb time by lowering the energy output. In comparison to a bass trap which will lower the reverb time so the energy output is the same but lasts less time.

 

I see.

In that case, make mine a bass trap please.

Just to clarify, at resonances, both the amplitude and duration of the sound is increased (think pre-fixed Millenium bridge).

 

Yes Tenson,
that is what I wanted to say. Nice how you said that in few words.
My major concern is the energy at the listening point at Time 0 which, depending on the solution you choose, will be different.
What will be interesting to find out is how much you need to equilize to be able to hear a difference.

 

By time 0 do you mean the time the direct sound reaches the listener?

In which case as I mentioned you can phase adjust (think minute time delay on the rest of the sound) the bass frequency so that the room resonance in fact gets to the listener at time 0 along with the rest of the direct sound. This is pretty complex though and frankly still doesn't work as well as real acoustic treatment, but its interesting to see just how far you can go with digital correction!

 

.. on the other hand with acoustic room treatment it is not always (hardly) possible to flaten completely the RF but with a equalization yes.
Now does that matter?

 

In lestening tests the smallest level change an average joe blogs on the street can hear, without being told it is there, is about 3dB. Whats more, thats the entire signal changing by 3dB not just a single freqencey band!!

Knowing this I think as long as the freqencey responce has a smooth curve and varies by no more than 3-5dB its as good to the human ear as flat.

 

That 3-5dB in the middle frequencies would lead to large differences between systems that would be easily detected by most Joe Blogs'.

 

Hmm thats true, I suppose our ears are mose sensitive when they have the reference level of the rest of the audio, where as when it all changes, the only reference they have is memory of what it was just like.

Then perhaps no more than 3-5dB high Q peaks/dips and an overall smooth response?

 

Where were the 3 dB measured from?

 

I really don't know! It wouldn't surprise me if it started at 80dB though as this is often the standard for engineers to listen at. The human ear has its flattest response around there.