Boomy bass in Klipschorns

[DrWho]

"the difference between the high frequencies and 1kHz are the same through all volume ranges."

That

is just a cheap basic circuit. Like what you would find on a boom box.

Most people are more concerned with the compensation for loss of

bass or bass boost in and of itself and could care less about the

highs. Remember those of us who really care about the sound

quality of our music are but a small population in the universe of

listeners. Thus the Bloze customer.[]

Rick

you aren't reading the charts correctly.

Let's say the sound guy is mixing at 10dB on the chart. The way he

chooses mic placement and EQ will be such that 30Hz frequencies will be

50dB louder than frequencies at 1khz. And frequencies at 10kHz would

measure 10dB louder than frequencies at 1kHz. In other words, the

recording itself will naturally contain compensation for the curve at

that volume.

Now lets playback this recording at 90dB: 30Hz will still be 50dB

louder and 10kHz will be 10dB louder on the physical recording (in

other words, nothing has changed...just trying to stress what is recorded on the recording). Looking at the chart,

the 10kHz frequencies still need to be 10dB louder than 1kHz at 90dB,

so the relative balance of 10kHz will remain the same. However,

according to the chart 30Hz only needs to be 5dB louder when listening at

90dB. This means that the 30Hz frequencies from the recording will be

heard as 45dB too loud than 1kHz when listening at 90dB.

If we were to listen to this recording at 100dB: 30Hz according to the chart needs to be

1dB louder and 10kHz needs to be 5dB louder in order to be percieved as

the same volume. Taking our recording monitored at 10dB, the 30Hz tone

will now be 49dB too loud and the 10kHz tone will be 5dB too loud.

I don't know how to make it any clearer than this. You have to look at

the slope differences between different reference SPLs. You can't just

look at the chart and try to create an EQ that provides the opposite

effect to "make it flat." Doing so shows a complete lack of

understanding of what the curves are showing.

[3dzapper]

"Let's say the sound guy is mixing at 10dB on the chart. The way he chooses mic placement and EQ will be such that 30Hz frequencies will be 50dB louder than frequencies at 1khz. And frequencies at 10kHz would measure 10dB louder than frequencies at 1kHz. In other words, the recording itself will naturally contain compensation for the curve at that volume"

This would be one thing in a live venue. The sound tech may take F/M into consideration for the enjoyment of the audience. He would know the level that the PA system will fill the venue. This would be one of his considerations like assuring the primary sound arrives at all listers before the first reflection.

It would be quite another thing for a recording engineer to mix a recording using some arbitary line on the F/M chart. The recording engineer must assume that the reproducing equipment will comply with F/M as it varies over the range of volume control settings.

As a "sound man" you are looking at this from a different perspective than I, as an "audiopile".

Rick

[ricktate]

What if this problem is caused by what i found out re doing my 74 K-Horns...i found that the back air chamber was not airtight and it leaked pretty bad.Maybe some of us forum members who report bad bass have this problem and dont know it, its easy to fix and check,it does not take much of a crack or glue leak to cause a big leak.Rick

[3dzapper]

Good point Rick.

Rick

[dragonfyr]

After reading the tack the previous discussion has taken I have to grin!

Fletcher Munson is "out the window as far as the KHorn is concerned"? And a live sound engineer would refer to FM curves?

We have a major frame of reference confusion here (regarding which I am sure Heyser is rolling over in his grave doubled up with either laughter or a severe attack of gas!)
Look folks, first of all, no one mixing any sound has to refer to FM curves! These curves are 'abstract' regarding signals that you do not hear! Does this make sense?
If you are listening to them, you can't avoid them!!! As the psychoacoustic realm is integral to the listening process! So to say you should mix levels relative to the FM chart and not to what you hear is pretty ludicrous as that graph will not add any additional insight! I feel like MCEscher here as I draw my hand drawing my hand drawing my hand...well, it's self-referential![]

And referencing the possibility that the LF horn is not sealed!! I think that is, or should be, assumed! And if it is not, all bets are off![][]

So, back to the alleged 'boominess' of the KHorn woofer...
First, we don't have any information on which to act except lots of feelings. And running around with the usual cast of characters such as SPL and RTA meters will not help, as resonances occur in time, not frequency.
Now I am confident that 20+ folks are going to run out and say "Oh no! I can measure the gain at a particular frequency!" Great, but you can't tell WHY the gain is higher or lower except to make a random guess! You cannot identify the contributing component elements. Thus the value of the time domain measurements.

The simple measurement that would identify all of the resonances is an ETF (Energy Time Frequency) response. This is the classic waterfall plot. And it is a fundamental response that is NOT listed in the review! A significant omission.
ANY resonances are immediately apparent as the persistent ringing at a particular frequency is displayed in both amplitude and time!

Also, as there has been the USUAL references to EQing the mid bass peak...
No you can't! And this is also fundamental and would be apparent if folks would take the small amount of time to read Heyser or Davis!
If folks would refer to the Nyquist display in the KHorn review you will note the epicycles (whorls) from DC to ~175Hz (and again at +- 500Hz!!). The epicycles mean that this range is NOT minimum phase! You CANNOT successfully EQ this problem! You can only (successfully) EQ a direct minimum phase signal!

So, what do we have.
Well, if we had an ETF waterfall response we could identify any resonances in the LF realm and address them surgically! Anything less or more is pure folly, and you are on your own with your feelings!
As without them we haven't a clue as to what we are trying to address. Well, you do have your feelings, and for that you might call the psychic hotline or any of the pure subjectivists who don't find value in correlative measurements![][]

And as far as addressing the minimum phase issue... The only hope we have here is to use a minimum of a bi-amped system utilizing an active crossover with signal delay sufficient to bring the LF into a minimum phase condition.

I will try to post an ETF plot for those not familiar to that which I refer...

So, now I will let you folks go back to your debate...[]

Nyquist KHorn Nyquist Complete&DetailPDF.pdf

[dragonfyr]

For anyone not familiar with the ETF - Energy time frequency or cumulative spectral display, here is an example - pardon the 'fuzziness'.

Note, the 3 axes display: X-> gain, Y-> frequency, Z-> time

Thus the ridges that appear on the time axis indicate a persistence or resonance at a 'particular' frequency.

ETF.CSDExamplePDF.pdf

[DrWho]

"Let's say the sound guy is mixing at 10dB on

the chart. The way he chooses mic placement and EQ will be such that

30Hz frequencies will be 50dB louder than frequencies at 1khz. And

frequencies at 10kHz would measure 10dB louder than frequencies at

1kHz. In other words, the recording itself will naturally contain

compensation for the curve at that volume"

This would be

one thing in a live venue. The sound tech may take F/M into

consideration for the enjoyment of the audience. He would know the

level that the PA system will fill the venue. This would be one of his

considerations like assuring the primary sound arrives at all listers

before the first reflection.

It would be quite another thing for

a recording engineer to mix a recording using some arbitary line on the

F/M chart. The recording engineer must assume that the reproducing

equipment will comply with F/M as it varies over the range of volume

control settings.

As a "sound man" you are looking at this from a different perspective than I, as an "audiopile".

Rick

Oh my goodness! you are trying to argue that the world is flat!

IN CASE YOU ARE BLIND I AM TALKING ABOUT THE RECORDING STUDIO ...there

is that clear enough?!? The forum won't let me go any bigger. Perhaps

you have issues with the term "sound guy" - how bout you just

interchange it with the term "recording engineer". There, does that

make you feel better? (I have issues with the term recording engineer

because he's not exactly engineering the recording...he's just mixing).

When a sound guy ("recording engineer") mixes ("records") at whatever volume he chooses, the mix that ends up on the recording

is compensated by the loundess curve AT THAT PARTICULAR VOLUME! If we

were to playback at a DIFFERENT VOLUME, then the difference in RELATIVE

magnitudes between the curves is the amount of attenuation that a

loudness curve needs to implement. YOU MUST KNOW AT WHAT VOLUME THE

RECORDING WAS RECORDED AT! Or more simply, the sound guy needs to make

sure that he's monitoring at 90dB.

F-M CURVES HAVE NOTHING TO DO WITH LIVE SOUND!!!!!!!!!!!!!!!!!!!!

For the record, I have just as much experience with recording as I do

live sound as I do the "audiophile" world...don't give me this crap

that I have the wrong frame of reference. Learn how to read the graph.

F-M curves are the most elementary of concepts in the recording studio.

Btw, the recording engineer isn't using some arbitrary line on the

chart and then consciously compensating for it! He makes it sound good

to his ears at that level. But when the level of playback changes, the

percieved RELATIVE tonal balance changes. It is therefore important

that during tracking all the tracks are monitored at the same level and

that the final mixdown is monitored at the same level. Otherwise the

individual tracks won't sound cohesive because the tonal balances are

off. Eventually the recording is finished and makes its way to the

player in the "audiophidiots" home. He then choose to playback at

whatever volume he chooses. If he plays at a level lower than what was

monitored in the studio, then he will notice a relative lack in the

bass response (notice that low frequencies need to be RELATIVELY louder

than 1kHz as the SPL goes down).

The entire purpose behind the loudness button is to allow playback at different levels and maintain the RELATIVE tonal balance!

Here, try an experiment....gets 3 sine waves, one at 30Hz, one at 1kHz,

and one at 10kHz. Monitoring at 10dB make it so they all are percived

to be at the same volume (use your ears, not a meter). Now take this

recording and playback at 90dB and tell me what you notice. Guess what,

the 1kHz and 10kHz tones will still be percieved as the same volume

relative to each other, but the 30Hz tone will be percieved as 45dB too

loud! This is why the implications of the F-M curve are so important

and is the only application of the graphs! Until you conduct this test

for yourself you should make no further comments because you are

effectively misleading the uninformed.

(To take the experiment further grab the same recording and listen at

100dB. Guess what, now the 10kHz tone is percieved as 5dB louder than

the 1kHz tone).

The F-M curves are some of the most elementary concepts introduced when

recording. I am sorry that the "audiophile" mind can't comprehend them.

[dragonfyr]

"For the record, I have just as much experience with recording as I do live sound as I do the "audiophile" world...don't give me this crap that I have the wrong frame of reference. Learn how to read the graph. F-M curves are the most elementary of concepts in the recording studio."

Wow!

I'm not exactly sure what the tirade is aimed at, but for the record (is that a pun?), some of us have more experience in both realms.[]

OK, so the FM curves represent the weighting the human ear adds to sound. No one disputes that! And I am sure there is some relevance to the debate of boominess of the KHorn (althought the significance is lost on me!) Besides, I don't need a set of the curves, as my ears carry them around in case they forget how to hear! (mighty convenient![] )

Maybe we will have to get a laminated set of 'FM curves' to carry around, as they have never been used in a studio or a mastering suite that I have ever seen. They don't need them. That is why they use their ears and various playback sources! But neither are they designing electronics nor worried about RIAA curves, etc. which are incorporated into the electronics.

But they are a moot point regarding the KHorn response boominess. []

[DrWho]

[]

Btw, I know I'm not the most qualified or even have the most experience

in the field....but I do have some and have hung out with people that

do too.

I am also not advocating that we carry a set of curves around with us

either - the graph is just a representation of what our ears our doing

and we just need to be concsious of that fact.

In regards to khorn boominess,

I heard that the more recent incanations of the khorn are using MDF and

thicker panels in some sections on the khorn in an attempt to reduce

resonations and all that. The only time I've heard a newer pair of

khorns was at the pilgrimage last summer, but the klipsch dude in the

heritage room was more keen on demo'ing the lascalas than the khorns.

In fact, I think we got to hear the khorns for a grand total of 1

minute tops! I thought it was rather akward and I always look back on

it thinking he was trying to hide something...perhaps they were trying

something new and it wasn't working out to well, but they wanted to

show the speakers anyway (afterall, the majority of comments I heard

were about looks rather than sound!) []

[3dzapper]

I have been having difficulty pasting these images, is anyone able to open them full size? (800x600).

[tofu]

What if this problem is caused by what i found out re doing my 74 K-Horns...i found that the back air chamber was not airtight and it leaked pretty bad.Maybe some of us forum members who report bad bass have this problem and dont know it, its easy to fix and check,it does not take much of a crack or glue leak to cause a big leak.Rick

i understand you're talking about khorns, but could this apply to the la scala as well? when checking out the woofer, i noticed there was no weather stripping or any kind of sealant between the "cover" and the "opening." i'm pretty certain SOME air escapes from there, i'm just not sure how severe the issue is.

it's quite a pain to take these heavy suckers off of the risers, unscrew both panels, put them back on, and set them back on the risers again, this is why i'm asking before i start doing it.

[dragonfyr]

I would think a seal is important...[]

[3dzapper]

.

[DrWho]

.

The attenuation being described in that article is a REDUCTION of HF

information to make the LF sound relatively louder as the volume

control is reduced. Attenuation does not mean boosting the highs....in

this case it means a reduction of everything above 500Hz. It is also

describing automatic unadjustable attenuation as well and I still have

a hard time believing every amp does it this way.

If I were at home and had the ability to scan photos I would be busting

out the few textbooks I have that discuss exactly what I'm describing.

But alas I am not at home and I do not have a scanner.

Btw, PWK is great but no person building a loudness contour is going to

exagerate the EQ for speaker with less LF extension....so he's trying

to magnify the effects of the auto calibration. Also, most speakers are

in the 90dB range which puts the khorns only 10dB more sensitive. The

corresponds to a max difference of about 5dB at 30Hz (and much much

less as the frequency increases). I don't know anyone in the world that

would consider +6dB at 30Hz to be boomy. Heck, the cornwall is +6dB at

80Hz and I can't think of anyone that describes it as being boomy

either!

[tofu]

I would think a seal is important...

pff

[artto]

Hmmm, I'll have to go look at a few more circuit diagrams...I have always experienced that they weren't variable and we were always taught in school that they weren't variable (but wouldn't be the first time a professor was wrong) []

Your professors are definitely wrong (as they often are, if you could have only seen my site engineering professor when I slammed a book marked engineering manual on his desk and said READ IT A------ in front of a couple of fellow students after he had given me an F on one project. A week later it came back with an A). My Crown DL-2 preamp/control centers loudness control is variable and incrementally changes the F/M curve based on both gain and loudness compensation level. Its up to the listener to use it properly.

<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

As for mr-bs problem, its the room. It takes a very large room (much larger than mine) to overcome any room mode problems. For room modes to no longer become a major part of the equation the smallest room dimension should be at least several times greater than the wave length of the lowest frequency produced. The Klipschorns typical usable lower frequency response is 33Hz, which yields a wave length of 34 feet. This will obviously require a much larger minimum dimension than we can find in almost any domestic environment. Remember, Khorns are designed to take maximum advantage of room to increase its sensitivity, AND, extend its low frequency response, both of which require trihedral corner placement in order to keep the speaker to a practical size. In short, its capable of exciting any and all room modes. And lets not forget that this problem is also relative to the listeners position in the room, relative to the speakers position.

Other things, such as the recordings, are often EQd with a mid-bass boost in the 125Hz -180Hz range to make them sound better on the mediocre equipment they are most likely to be played on. Also, blending of the L & R channels to a partial or full mono condition will accentuate the bass and may produce tubbiness. Also, if you are using separate pre/power amps, even the way you set the gain between the two of them will affect tonal balance of the sound.

[Strabo]

When a sound guy ("recording engineer") mixes ("records") at whatever volume he chooses, the mix that ends up on the recording is compensated by the loundess curve AT THAT PARTICULAR VOLUME! If we were to playback at a DIFFERENT VOLUME, then the difference in RELATIVE magnitudes between the curves is the amount of attenuation that a loudness curve needs to implement. YOU MUST KNOW AT WHAT VOLUME THE RECORDING WAS RECORDED AT! Or more simply, the sound guy needs to make sure that he's monitoring at 90dB.

DrWho is correct and for an example of this phenomenon I point to Zep IV. It was mixed at a very loud volume. If you play it back at <85 dbs you will notice that there is no bass. The recording will always be that way unless they go back to the session tapes and remix it. But no one would dare mess with that mix or risk mayhem that everyone would hurl if they heard it mixed better. You can't mess with Stairway To Heaven, so it will stay this way forever. Even the Classic Records version is bass shy because of the volume during mixing.

Edit: what this has to do with boomy bass on a khorn I'll never know.

[DrWho]

Edit: what this has to do with boomy bass on a khorn I'll never know.

me niether []

Btw, perhaps the monitoring at 100dB+ was the band's way of saying "hey, turn it up if you want to enjoy our song!" []

(seriously, that music is meant to be CRANKED) []

[thebes]

Mr. B, first a caveat:

ANY TECHNCIAL ADVICE OFFERED BY THEBES IS HIGHLY SUSPECT AND ALMOST CERTAINLY BOUND TO BE WRONG.

However, first thing I would do is tuck those big boys back into the

corners and make sure they are sealed tight. If your pre/amp and

or reciever/ source etc. combo sounded good with the LaScalas then I

would expect that would eliminate your gear and probably your room as

the major culprits. However if you were running the LaScalas on

the eight ohm tap then maybe the Khorns would work better on the 4 ohm

tap. If not maybe the cans on the crossovers are just tired after

all these years. I've been led to believe that they do degrade over

time.

Since I see you are from the UK I would also check to see if you've

left some old fish and chips inside one of the Khorns, certainly

they've been know to cause "bloat". Just funing a little and

welcome to the forum and congratulations on acquiring one of the

world's finest speakers.

Hope this helps.

Oh yeah and ignore all the other stuff going on in this thread.

These people have so much technical knowledge they often overlook the

obvious. Like the time my amp was accidentally unplugged and they

told me to.......

[garymd]

Fish & chips............you're killin' me.[]