Hi,

I am trying to find out how 'flat' is considered 'flat' (frequency
response) in a typical small listening room. I am searching for the
range SPL can vary over 20-20KHz to be considered flat in practice.

Any comments, useful urls will be very helpful.

Cheers,
Vimal

Vimal wrote:

I am trying to find out how 'flat' is considered 'flat' (frequency
response) in a typical small listening room. I am searching for the
range SPL can vary over 20-20KHz to be considered flat in practice.

"Flat" means just that - +/- nothing. That's why frequency response is
(properly) expressed with a tolerance attached.

For electrical devices, it's quite normal these days to have frequency
response flat to within a couple of tenths of a dB over the normal
audio range. Since you mentioned SPL, however, you may be talking about
the response of a room. There is no such thing as a "flat" room unles
it's an anechoic chamber. What we usually look for in a good room is a
frequency response that's flat within +/- 5 dB or so, but more
important, is even throughout the room. You don't necessarily want a
room that's flat within +/- 1 dB at one place, and a few inches away,
might have peaks or dips upwards of 15 dB at certain frequencies.

I always liked, and try to achieve, Julian Hirsch's ideal of a highly
diffuse sound-field exhibiting excellent octave to octave balance.

Smaller individual frequency variations are not as important, unless,
as Mike has suggested, those individual variations are really big.
Octave to octave balance is the key.

-glenn


Mike Rivers wrote:
Vimal wrote:

I am trying to find out how 'flat' is considered 'flat' (frequency
response) in a typical small listening room. I am searching for the
range SPL can vary over 20-20KHz to be considered flat in practice.

"Flat" means just that - +/- nothing. That's why frequency response is
(properly) expressed with a tolerance attached.

For electrical devices, it's quite normal these days to have frequency
response flat to within a couple of tenths of a dB over the normal
audio range. Since you mentioned SPL, however, you may be talking about
the response of a room. There is no such thing as a "flat" room unles
it's an anechoic chamber. What we usually look for in a good room is a
frequency response that's flat within +/- 5 dB or so, but more
important, is even throughout the room. You don't necessarily want a
room that's flat within +/- 1 dB at one place, and a few inches away,
might have peaks or dips upwards of 15 dB at certain frequencies.

Vimal, see if you can find any of Roy Allison's research papers from
the early 1970s--he was looking not only at the frequency response of
typical home listening environments, but also at the interaction
between loudspeakers and the room boundaries (walls, ceiling, floor)
when speakers were placed in typical positions within those
environments. The deviations in response that he found were more severe
than most people had previously assumed.

He documented and analyzed variations (peaks and valleys) as great as
12 dB in midbass response, due to reflections from room boundaries,
with cancellation and reinforcement occurring at various wavelengths.
He came to the conclusion that adequately flat response is possible
only if a loudspeaker is designed for a particular type of placement
relative to those boundaries, and then if the speaker is used only in
that specific type of placement.

For example, a loudspeaker could be designed to have the desired
response when placed in a corner of a room, but then it would have very
different response if placed on a bookshelf. Everybody knows that--but
what not everyone seems to realize is that the result is not merely
"less bass" as it is usually described, but also there will be highly
irregular low-frequency response overall.

--best regards

Vimal wrote:

I am trying to find out how 'flat' is considered 'flat' (frequency
response) in a typical small listening room. I am searching for the
range SPL can vary over 20-20KHz to be considered flat in practice.

If your narrowband response varies +/- 12 dB, you'll be doing better
than the vast majority of rooms out there. Third octave response will
always look better than narrowband, of course.
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Thanks everyone for useful suggestions and insight.

Would you have more details which journal was Roy Allison's paper
published in.

My room can be classified as a small room. I was reading 'Handbook of
Sound Engineers', they say that small room behave in two different
ways.

Below large room frequency ie f = K*sqrt(RT60/V) the room behaves like
a small room, and above f it may be considered as a large room.

My intial observation from measured impulse reponse is that for lower
to mid frequency I find harmonics developing, i.e. for frequencies
below 5KHz the harmonics of excitation signal are obvious in
spectrogram. However after 5KHz the harmonics reduce considerably.

Is this something obvious that I am witnessing, harmonics in lower and
not in higher freq.!


Vimal wrote:
Hi,

I am trying to find out how 'flat' is considered 'flat' (frequency
response) in a typical small listening room. I am searching for the
range SPL can vary over 20-20KHz to be considered flat in practice.

Any comments, useful urls will be very helpful.

Cheers,
Vimal

Along with the other stuff - remember that the louder it is, the more
compression happens in nature - so keep it quite
Vimal wrote:
Hi,

I am trying to find out how 'flat' is considered 'flat' (frequency
response) in a typical small listening room. I am searching for the
range SPL can vary over 20-20KHz to be considered flat in practice.

Any comments, useful urls will be very helpful.

Cheers,
Vimal

Vimal,

I am searching for the range SPL can vary over 20-20KHz to be considered
flat in practice.

A small to medium sized room without bass traps or other acoustic treatment
typically varies by 35 dB or so. With enough bass traps you can get the
window down to about 10 dB. Most people are doing well if they can get the
disparity between peaks and nulls down to 20 dB.

This article on my company's web site explains a lot about it:

www.realtraps.com/art_small_rooms.htm

Note that modal (low frequency) ringing is just as big a problem as a skewed
response. Most people don't consider this because it's not as obvious and
it's more difficult to measure. Here's another article that explains more
about the importance of ringing:

www.realtraps.com/art_etf.htm

--Ethan

Thanks you all again for valuable suggestion and urls.

Just wondered for a Listening room apart from flatness what are the
other important characterstics?

Ethan Winer wrote:
Vimal,

I am searching for the range SPL can vary over 20-20KHz to be considered
flat in practice.

A small to medium sized room without bass traps or other acoustic treatment
typically varies by 35 dB or so. With enough bass traps you can get the
window down to about 10 dB. Most people are doing well if they can get the
disparity between peaks and nulls down to 20 dB.

This article on my company's web site explains a lot about it:

www.realtraps.com/art_small_rooms.htm

Note that modal (low frequency) ringing is just as big a problem as a skewed
response. Most people don't consider this because it's not as obvious and
it's more difficult to measure. Here's another article that explains more
about the importance of ringing:

www.realtraps.com/art_etf.htm

--Ethan

Vimal,

what are the other important characterstics?

As I explained in my last post:

"Note that modal (low frequency) ringing is just as big a problem as a
skewed response."

It's also important to treat the first reflection points on the side walls
and ceilings (and even the floor if you don't have carpet), and to treat
large opposing parallel surfaces to avoid flutter echo. All of this is
explained in detail in various articles and videos on my company's site I
linked earlier.

--Ethan

Danny T wrote:
Vimal wrote:
Thanks you all again for valuable suggestion and urls.

Just wondered for a Listening room apart from flatness what are the
other important characterstics?

A picture of a hot babe always makes a room sound better

I assure you, it is a poor substitute for the real thing.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

A picture of a hot babe always makes a room sound better.

Chacun a son gout. A lifesize poster of Dan Haggerty would suit me right
fine. Heck, I'd settle for Gabby Hayes.

In all seriousness... In general, the room will "sound best" with the fewest
visual distractions of any sort.

Scott Dorsey wrote:
Danny T wrote:
Vimal wrote:
Thanks you all again for valuable suggestion and urls.

Just wondered for a Listening room apart from flatness what are the
other important characterstics?

A picture of a hot babe always makes a room sound better

I assure you, it is a poor substitute for the real thing.

A picture never says "No".

geoff