X curve. Which varies from installation to installation in response to the size and reverberant properties of the theater/sound stage/studio/room. But in anything other than a near-dead room there will be from some to considerable high frequency rolloff.
IMO, for the average casual listener, it's most usually the speaker. But I don't think the users in this forum are the average listeners.... so if you have really good speakers then the room could be a relatively large component of "the problem". I think this is where lots of people on this forum are at; Is there still much room for improvement of the speaker or perhaps it there would be a better ROI to spend the $ on improving the room?
I for one would really like to build a dedicated listening room in my next house.... and build an even better pair of speakers for it. 😀
Yes . . . as far as that goes.The problems are (mostly) at the ends of the chain. Flat mics are (probably) a good idea, though some of the best sounding ribbons roll off on the high end. But *where* the mics are placed captures a sound field brighter than, and often significantly brighter than, what your ears would hear in the location(s) where you would likely listen. You can't buy a seat where the mics are for a "live" orchestral recording, and you wouldn't want your ears to be where the mics are in a studio. Seriously. They wouldn't last. In the studio there's a lot of "not flat" at the mix desk . . . way beaucoup a lot. I'm not going to say that you will *never* see everything flat on the board . . . maybe they get delivered that way. Maybe someone's playing a joke. Maybe the cleaning lady lines everything up to be "tidy". Stuff happens . . .
On the other end it's (probably) good to start with flat-on-axis loudspeakers, but there'a a general consensus that flat (in room) power response sounds . . . wrong. And flat-on-axis loudspeakers are rarely flat at 30,60 and 90 degrees (horizontal) and never flat at those angles vertical. Absorption in the room is not flat either, unless you've spent five or six figures to make it flat, and if you've spent that much it's a good chance that it was made deliberately *not* flat in order to get a decent sounding power response.
That's what this thread is all about . . . "flat", however good it is in theory or however good it looks on paper, sounds bad . . . the question is where to fix it, and how.
I suspect that the amount of roll off that may be desired is a function of the relative volume levels. At relatively low listening volumes, "flat" may be just about right for many people. However, at higher volumes like those normally found in a commercial theater, I think most would prefer at least some HF roll off in order for it to be perceived as providing a balanced tonal quality.
This concept really struck home for me when someone else earlier in this tread said, "I just turn it up until it sounds right". Perhaps this is also why there is alot of disagreement because the discussion about preferred frequency response curve most often neglects details regarding this listeners average volume levels.
Last edited:
In the "movie sound" world there is a "standard" (reference) level, and fr in various listening environments is tailored (and different) so that they all *sound the same at the reference level*. The notable point being that movie sound engineers understand that different listening environments have to be equalized (often very) differently to sound the same. They got over the "flat is best" nonsense half a century ago . . .
For "audio", at least when reproducing "live" performances, the most practical goal is that the reproduction have the same "sound" (tonal balance) as the original when played at the original level. "Loudness" compensation would then (attempt to) restore the correct(ed) tonal balance if the level is reduced. That can kinda-sorta work . . . sometimes . . .
Last edited:
so.... can you give us some more details regarding the "standard" reference level used for the movie theaters and the techniques used to optimize?
Hi came to this at a quite mature state....
Have to contribute a little... don't know if it has been covered before....
Loudspeaker systems are most often multi way..and especially the the crossover regions will stand out if totally flat...
If you look at humans we have two eyes and two ears...we can interpret dept in vision look at things in a 3D kind of manner...stereo vision so to speak...Hearing has the same ability to interpret the phase contents and thus create a 3D like hearing images. Phase content maybe even more important than frequency content..
The point is that frequency areas with phase shifts stands out...simply because of our sensitive awareness...hearing does not shut down while sleeping it's always on the watch..alert and ready to wake you up...
When designing loudspeakers this is is quite evident..whenever you have a crossover point and two different driver share the output the level must be a little down.. simply because if not your psychoacoustic hearing will have that area in focus..and have it stand out..
Point is that the target curve is not universal, but something that is very much speaker dependent...and highly related the the mixed phase response of the drivers in multi way systems.
Have to contribute a little... don't know if it has been covered before....
Loudspeaker systems are most often multi way..and especially the the crossover regions will stand out if totally flat...
If you look at humans we have two eyes and two ears...we can interpret dept in vision look at things in a 3D kind of manner...stereo vision so to speak...Hearing has the same ability to interpret the phase contents and thus create a 3D like hearing images. Phase content maybe even more important than frequency content..
The point is that frequency areas with phase shifts stands out...simply because of our sensitive awareness...hearing does not shut down while sleeping it's always on the watch..alert and ready to wake you up...
When designing loudspeakers this is is quite evident..whenever you have a crossover point and two different driver share the output the level must be a little down.. simply because if not your psychoacoustic hearing will have that area in focus..and have it stand out..
Point is that the target curve is not universal, but something that is very much speaker dependent...and highly related the the mixed phase response of the drivers in multi way systems.
Right. But let me restate it in a form which I hope will be helpful.
For a given room and set of conditions, there IS a concept of "sounds flat to everybody" or as dwardh says, "have the same "sound" (tonal balance) as the original when played at the original level". I think it is important in this discussion not to lose sight of that in a kind of relativism.
There is a kind of engineering fallacy which leads people to think that sounds produce perceptions in a kind of one-to-one way. Thinking that way makes it puzzling why ears and mics hear things differently. Better to think sounds feed perceptions.
The problem is, there doesn't seem to be a reliable method for setting up systems in rooms to hit that mark using mics. (Yes, we've all used some mic method in some room that did work... once.)
Footnote: sometimes people are talking about "flat speakers" when they mean "anechoic flat" response. And so they are flat only anechoically. The same can be said for polar charts. Helpful but not definitive when those speakers are plunked down in a room.
Last edited:
Indeed! And that seems to be the crux of the problem. Even if we have speakers that are anechoically flat on axis, how do we measure and translate that into a good tonal balance in our listening rooms? Getting them flat at the listening position rarely seems to sound right.
Allow me to ramble a bit.
It would be great if we could just set up a measurement mic at our listening position, run a sweep or two and then EQ it flat and be happy. But that rarely works. Can we EQ with a falling response and be happy? Yes, sometimes - but not always. There are automated EQ systems on the market, even Yamaha includes it in their HT stuff. But it's no miracle cure.
So how do they basically get it right in most cinemas? It's done with measurement mics. In fact I've done a few cinema THX and Dolby alignments myself, they work well. But the average cinema has several advantages that our music listening rooms do not. They have been mentioned in the posts above. Some of those advantages are:
The second two points hint at some of the problem. Our rooms suck. 😛 Basically because they are small and untreated, we hear far too much of them. And that is very difficult to EQ out.
Add to the room acoustic problems that almost all room measurements are done with an omni-directional microphone, and we have a mess on our hands. The omni mic works beautifully in large rooms or outside where all the energy of the small room is not hitting it from every angle. But in the typical listening room the omni is overwhelmed by the "room tone". In my experience I have found that omni mics work well in large rooms, but I get better results with a cardioid mic in a small room. There was a recent link posted to a very good paper on measurements of mastering suites. I was pleasantly surprised to see that they used a dummy head for the measurements. I.E. a polar response that actually mimics that of the human head. We don't hear as much behind us as the omni mic does. We don't hear the room the way the mic does. I've found this to be more important the smaller the room.
So if we had the magic EQ system that placed a mic at the listening position and did it's thing, how successful could it be? In a small room, I don't think it would work well if it does not take the room acoustics into account. Maybe it should run its seeps, analyze and tell us "Forget it, your room sucks." "Until you fix that awful T60 peak in the midrange and that huge room mode at 160Hz, there ain't much I can do for you." But it might also say; "OK, I've pulled a few dB out of the midrange to make up for the bad acoustics, it's not perfect, but the best I can do." "I've also tilted the response down a bit, because your room has too much high end energy."
From the above ramble, you can see that I believe that trying to measure in the typical small room is the problem. You can build speakers that address the problems of typical small rooms - give them a recessed midrange, a high end roll-off, for example, and that can work well. In most rooms. Or we can evolve our measurement technique beyond "Is it flat at my listening position?"
Allow me to ramble a bit.
It would be great if we could just set up a measurement mic at our listening position, run a sweep or two and then EQ it flat and be happy. But that rarely works. Can we EQ with a falling response and be happy? Yes, sometimes - but not always. There are automated EQ systems on the market, even Yamaha includes it in their HT stuff. But it's no miracle cure.
So how do they basically get it right in most cinemas? It's done with measurement mics. In fact I've done a few cinema THX and Dolby alignments myself, they work well. But the average cinema has several advantages that our music listening rooms do not. They have been mentioned in the posts above. Some of those advantages are:
- A standard to work with
- Films are mixed to that standard
- Cinemas are large (compared to most of our rooms)
- Cinemas have decent acoustic treatment
The second two points hint at some of the problem. Our rooms suck. 😛 Basically because they are small and untreated, we hear far too much of them. And that is very difficult to EQ out.
Add to the room acoustic problems that almost all room measurements are done with an omni-directional microphone, and we have a mess on our hands. The omni mic works beautifully in large rooms or outside where all the energy of the small room is not hitting it from every angle. But in the typical listening room the omni is overwhelmed by the "room tone". In my experience I have found that omni mics work well in large rooms, but I get better results with a cardioid mic in a small room. There was a recent link posted to a very good paper on measurements of mastering suites. I was pleasantly surprised to see that they used a dummy head for the measurements. I.E. a polar response that actually mimics that of the human head. We don't hear as much behind us as the omni mic does. We don't hear the room the way the mic does. I've found this to be more important the smaller the room.
So if we had the magic EQ system that placed a mic at the listening position and did it's thing, how successful could it be? In a small room, I don't think it would work well if it does not take the room acoustics into account. Maybe it should run its seeps, analyze and tell us "Forget it, your room sucks." "Until you fix that awful T60 peak in the midrange and that huge room mode at 160Hz, there ain't much I can do for you." But it might also say; "OK, I've pulled a few dB out of the midrange to make up for the bad acoustics, it's not perfect, but the best I can do." "I've also tilted the response down a bit, because your room has too much high end energy."
From the above ramble, you can see that I believe that trying to measure in the typical small room is the problem. You can build speakers that address the problems of typical small rooms - give them a recessed midrange, a high end roll-off, for example, and that can work well. In most rooms. Or we can evolve our measurement technique beyond "Is it flat at my listening position?"
Pano, it is hard to disagree with most of what you say. Let's extend the argument about the room problem and I will give just a couple of anecdotes and they should not be confused with honest to goodness data points.
Like others I have had difficult rooms (dens and living rooms) to deal with. As a weekend experiment I set up a Klipschorn bass bin in the corner so it gets its 1/8 space loading and it will do pretty ell up to 400 Hz or so. This was augmented by a midbass horn to bring me to about 750 Hz or so. I then added an Altec Mantaray horn (it was a long throw and series 2 that probably loaded down to 500 Hz and was constant directivity and narrow from about 1000Hz and up). IOW, this was a narrow dispersion CD horn.
These horns do best if you don't cross them too low. The crossover point was chosen so they crossed where the dispersion was comparable between drivers. So now the system was leaning toward minimizing room interactions by narrowing dispersion. But it was still somewhat "CD" so the on-axis and off-axis response was roughly comparable from about 700 Hz on up. Again these are approximations for my weekend expt. The crossover and eq were set to produce a roughly flat respeonse on-axis, perhaps dropping off a bit above 4kHz.
What are the anecdotes? Well, the location of the instruments was pin-point and precise (no surprise). Popular music sounded dry. Poorer than usual (agin, no surprise). This was the biggest drawback and the music did seem a bit "bright and forward". However ther was no sense of realism in terms of perceived spectral brightness. At higher volumes, this was not as bad (but you would also expect that from an understanding of Fletcher-Munson).
However, live music (Naxos recording that have little post-processing and are recorded in lively environments, eg, a church) sounded great. There was sense of geometry of the environment that it was recorded in. Whether this sense was accurate or not, I don't know. But along that dimension of "realism", they did well. Along the dimension of "spectral balance", they did fairly well, although just a bit bright.
Are long throw (narrow dispersion) horns fun to play with? Absolutely! (especially for recordings not made in a studio). Would I use them for daily use? No. Studio recordings, by and large, were just not much fun to listen to. Just as these recordings are not fun to listen to in an anechoic environment.
Again, this is anecdotal, so take it with an appropriate grain of salt.
Like others I have had difficult rooms (dens and living rooms) to deal with. As a weekend experiment I set up a Klipschorn bass bin in the corner so it gets its 1/8 space loading and it will do pretty ell up to 400 Hz or so. This was augmented by a midbass horn to bring me to about 750 Hz or so. I then added an Altec Mantaray horn (it was a long throw and series 2 that probably loaded down to 500 Hz and was constant directivity and narrow from about 1000Hz and up). IOW, this was a narrow dispersion CD horn.
These horns do best if you don't cross them too low. The crossover point was chosen so they crossed where the dispersion was comparable between drivers. So now the system was leaning toward minimizing room interactions by narrowing dispersion. But it was still somewhat "CD" so the on-axis and off-axis response was roughly comparable from about 700 Hz on up. Again these are approximations for my weekend expt. The crossover and eq were set to produce a roughly flat respeonse on-axis, perhaps dropping off a bit above 4kHz.
What are the anecdotes? Well, the location of the instruments was pin-point and precise (no surprise). Popular music sounded dry. Poorer than usual (agin, no surprise). This was the biggest drawback and the music did seem a bit "bright and forward". However ther was no sense of realism in terms of perceived spectral brightness. At higher volumes, this was not as bad (but you would also expect that from an understanding of Fletcher-Munson).
However, live music (Naxos recording that have little post-processing and are recorded in lively environments, eg, a church) sounded great. There was sense of geometry of the environment that it was recorded in. Whether this sense was accurate or not, I don't know. But along that dimension of "realism", they did well. Along the dimension of "spectral balance", they did fairly well, although just a bit bright.
Are long throw (narrow dispersion) horns fun to play with? Absolutely! (especially for recordings not made in a studio). Would I use them for daily use? No. Studio recordings, by and large, were just not much fun to listen to. Just as these recordings are not fun to listen to in an anechoic environment.
Again, this is anecdotal, so take it with an appropriate grain of salt.
This may be a good place to start for what ails us:
http://www.bobgolds.com/Mode/RoomModes.htm
To EQ the frequency domain to attempt to solve a time domain issue is still and always will be adding problem on problem. Of course if you've had a heart attack, a beta blocker is a good idea. The side effects are worth it. Never the less, it would be best to take care of your heart before hand. I haven't touched my EQ in many months and I listen to just about everything and movies too. All the oddiophile buzzwords are represented and I've only EQed the bass.
Dan
http://www.bobgolds.com/Mode/RoomModes.htm
To EQ the frequency domain to attempt to solve a time domain issue is still and always will be adding problem on problem. Of course if you've had a heart attack, a beta blocker is a good idea. The side effects are worth it. Never the less, it would be best to take care of your heart before hand. I haven't touched my EQ in many months and I listen to just about everything and movies too. All the oddiophile buzzwords are represented and I've only EQed the bass.
Dan
Last edited:
Of course. Not to put too fine a point to it . . . but . . . it can't be done. Part of why this thread is all over the place is that while most of us "get" that we still find ourselves having to explain/defend that to those who aren't there yet.
But this thread started with the observation that even when both "frequency" and "time" are (more or less) sorted out (if there even is such a thing, given the variation in recordings) things are still "not quite right". The whole "stereo" reproduction chain is still not well enough understood that "perfect on paper" (as far as we know) is enough to guarantee "sounds good" or "sounds real" or, or, or . . .
I still have to roll off the highs at home to get the sound I hear "in concert" . . .
You have a lively room with early pseudo-diffusion. A very different situation than the room your recordings were likely mixed/mastered in. I would hope you roll off the highs especially when your speaker's power response is also tilted in the opposite direction as mixing/mastering speakers. In this situation, tone controls are band aiding a gash as far as I'm concerned. The are doing this in the bass as well, but this area is hard to fix acoustically in a small room. You are right, this stuff is common sense and should be understood by audio enthusiasts. It shouldn't be thought of as HRTF compensation or that a flat speaker is somehow bright. By definition it is not. Is 'straight' 'crooked' as well? It would be great to see some objective data on your listening environment, but I won't hold my breath. We have it on your speakers and from what you've said about your listening environment, it should need toned down.
We agree on something,
Dan
We agree on something,
Dan
A lot of deep thoughts here; my +1 esp. to dewardh.
Footnote about cinemas:
I betcha if the movie sound guys took their best shot at a big symphony - dark screen - and we listened carefully in a cinema, it would sound awful. Just guessing. Anybody heard one of those Met (opera) movie-theater broadcasts to the hinterland?
Movie sound is a whole other world because having a picture on the screen lets your perceptual system "paper over" many flaws and suspend all kinds of judgments (earlier I referred to this as nonsensical sound... like hearing 20 foot tall faces kiss 60 feet away). I suspect a lot of movie sound magic (like radio sound effects) is really a kind of convention that we learn over the years of attending movies.
Similar issues with TV sound that is often "nonsensical" too. When PBS does a symphony and there's a close-up of the clarinet, the sound comes RIGHT OUT OF the bell, eh? (Even in mono.) And then a long-shot and whadda ya know, you can PINPOINT the clarinet sound again!
My point is that there are major differences between home and cinema sound and I can't say as any local theaters (some of which had some kind of input from Stan Lipshitz) impressed me too much.
Footnote about cinemas:
I betcha if the movie sound guys took their best shot at a big symphony - dark screen - and we listened carefully in a cinema, it would sound awful. Just guessing. Anybody heard one of those Met (opera) movie-theater broadcasts to the hinterland?
Movie sound is a whole other world because having a picture on the screen lets your perceptual system "paper over" many flaws and suspend all kinds of judgments (earlier I referred to this as nonsensical sound... like hearing 20 foot tall faces kiss 60 feet away). I suspect a lot of movie sound magic (like radio sound effects) is really a kind of convention that we learn over the years of attending movies.
Similar issues with TV sound that is often "nonsensical" too. When PBS does a symphony and there's a close-up of the clarinet, the sound comes RIGHT OUT OF the bell, eh? (Even in mono.) And then a long-shot and whadda ya know, you can PINPOINT the clarinet sound again!
My point is that there are major differences between home and cinema sound and I can't say as any local theaters (some of which had some kind of input from Stan Lipshitz) impressed me too much.
Last edited:
How would you know? Is it because that's what your flawed (and incorrect) theories (religion) tell you, so it "must" be so ? ? ? And even if it were true, so what? I have no intention of living in an anechoic chamber, and I listen to music where I live.
As I've said before (and you just don't seem to get it), I don't much care about the room they were "mixed/mastered" in. I care about the room they were performed in, and the room they are reproduced in, and in reproduction in the latter "sounding like" listening in the former (although if it's got to fail in one direction or another I'd prefer failure in the direction of "more clarity" . . . but that's just my personal preference, I like listening to the "on stage" sound).
If you're talking about how to get from Provo to Boulder yes, it is. Reality doesn't map well to your (over)simple formulations . . .
Cinema sound is how I put bread on the table these days.
In some ways cinema sound is more advanced and in other ways less. It is standardized via the cinema "X" curve but that hasn't always served it well. Since we are talking much larger spaces with more directional systems and deader acoustics, the balance difference between the direct sound and the steady state reverberent field is much greater. That is why the cinema X rolloff is so great (that plus the fact that it evolved from earlier EQ meant to curtail lows and highs to reduce optical track noise).
Still, the techs found that the X curve was never universal, that it had to be very different for smaller theaters and mastering suites, so now the specs call for a number of treble rolloffs based on theater size.
My belief: we aren't really measuring the right thing (in measuring steady state response) so there will always be uncertainty in large room EQ. The larger the room the greater the uncertainty.
David S.
In some ways cinema sound is more advanced and in other ways less. It is standardized via the cinema "X" curve but that hasn't always served it well. Since we are talking much larger spaces with more directional systems and deader acoustics, the balance difference between the direct sound and the steady state reverberent field is much greater. That is why the cinema X rolloff is so great (that plus the fact that it evolved from earlier EQ meant to curtail lows and highs to reduce optical track noise).
Still, the techs found that the X curve was never universal, that it had to be very different for smaller theaters and mastering suites, so now the specs call for a number of treble rolloffs based on theater size.
My belief: we aren't really measuring the right thing (in measuring steady state response) so there will always be uncertainty in large room EQ. The larger the room the greater the uncertainty.
David S.
B/c you've said so.
I've said a number of times that I get exactly what your point is. You don't get that it makes no sense. It's alright, I know it gets harder to learn as you get older. I still have faith in you. 😛 Anyone know the definition of faith? Since you are listening to a recording, you may want to understand the process.
As evidenced by what? Studio engineer hating the sound of their well designed studio? Or hating the sound of well designed speakers? Or audiophiles hating the sound of their poorly designed systems? Anyone hating a well designed room or speaker? I'd actually say the highway is fairly wide.
Here's a very old paper on the subject you might be able to read: http://www.bksv.com/doc/17-197.pdf
Should be an interesting. Funny the debate rages on for so many years.
Dan
Last edited:
- Status
- Not open for further replies.