Is a house curve accurate ?

[Vaughan Odendaa]

Hi there,

I've been thinking about this and I would like know your input. When calibrating speakers to reference level, the sub level (in order to be accurate in line with the mains) should be kept about 2 dB's below that of the mains according to the Radio Shack SPL meter.

This is an accurate calibrating of levels. But in terms of frequency response, our ears gradually taper off in the low end. I've heard that in order for our ears to better perceive low frequencies due to the equal loudness curves, the low end should be boosted gradually and then the response should drop higher off flat.

What I would like to know is, is this accurate ? Is this what one should do not because of personal taste but because of the way our ears function and so in order for the film soundtracks to be faithfully reproduced as intended, a house curve is necessary ?

What are your thoughts ?

--Regards,

 

[avaserfi]

What I would like to know is, is this accurate ? Is this what one should do not because of personal taste but because of the way our ears function and so in order for the film soundtracks to be faithfully reproduced as intended, a house curve is necessary ?

What are your thoughts ?

--Regards,

House curves are personal, just like the rest of this hobby. Many people do not enjoy a perfectly flat bass response because as you said our thresholds for differing frequencies taper with lower the frequency. On the lower frequencies, below audible frequencies, house curves can be used to increase the chest pounding feeling.

There is no perfect house curve it is just how one likes the bass response that is the 'right' way to do it. Some people like perfectly flat responses while others enjoy a very extreme house curve.

You must remember, there is no perfect specification for sound don't get caught up in this idea of "perfect" try to find what you like.

 

[Vaughan Odendaa]

So there is no standard on how the response should look like ? I assumed that the SPMTE had specific standards for home theater. Do they not specify a target frequency response for home theater ?

Otherwise, what is the point of calibrating for "accuracy" then ? THX products claim to faithfully reproduce the content as the director envisioned. That is accurate and I assume that it is accurate according to the SPMTE (which governs pretty much everything in theater except for music recording).

At first I thought that a flat low end response would be the target frequency response but if that is just personal taste then does the word "accuracy" lose all meaning ?

--Regards,

 

[avaserfi]

So there is no standard on how the response should look like ? I assumed that the SPMTE had specific standards for home theater. Do they not specify a target frequency response for home theater ?

Otherwise, what is the point of calibrating for "accuracy" then ? THX products claim to faithfully reproduce the content as the director envisioned. That is accurate and I assume that it is accurate according to the SPMTE (which governs pretty much everything in theater except for music recording).

At first I thought that a flat low end response would be the target frequency response but if that is just personal taste then does the word "accuracy" lose all meaning ?

--Regards,

I don't know too much about the standard but this link should help you out with your questions. It goes indepth on exactly the types of house curves as well as a standard they call the X Curve.

Hope this helps.

 

[mtrycrafts]

So there is no standard on how the response should look like ? I assumed that the SPMTE had specific standards for home theater. Do they not specify a target frequency response for home theater ?

Otherwise, what is the point of calibrating for "accuracy" then ? THX products claim to faithfully reproduce the content as the director envisioned. That is accurate and I assume that it is accurate according to the SPMTE (which governs pretty much everything in theater except for music recording).

At first I thought that a flat low end response would be the target frequency response but if that is just personal taste then does the word "accuracy" lose all meaning ?

--Regards,

I am not sure if SPMT get into home theater standards unlike THX.
Yes, you should boost the subs a few dBs and cut the highs. I think even Harman may have something on this through their research efforts.

 

[Vaughan Odendaa]

Thanks for posting. I am getting my SMS-1 today and I have been thinking about a house curve for some time. The thing is, I would like to be as close or as faithful as possible to the film recording.

If that means that a house curve needs to be implemented then that is what I will do. I brought THX into the discussion because they strive to give you what the director envisioned.

The thing is, don't THX subscribe to a flat response ? I guess the purpose of this thread is ask yourself whether or not a house curve is accurate and is required for the faithful reproduction of movie soundtracks.

You mentioned Harman. I haven't read any articles (from Harman international),on house curves or anything as of yet but perhaps you've read an article that I missed ?

Thanks.

--Regards,

 

[jcPanny]

LFE level

If you want to preserve what the director intended, then a flat frequency response is appropriate. If the director/editor/sound mixer wanted to increase the LFE a few dB then they do that on the sound track.

It is likely that your SMS-1 will reveal how far your in-room response is from the "ideal" flat frequency response.

 

[Vaughan Odendaa]

Are you suggesting that a house curve would go against the directors original vision ? Then it would seem that it isn't accurate.

All that I have read on movie soundtracks suggest that a flat frequency response is what is accurate. But then there is the equal loudness curve.

It seems that I am getting somewhat confused here. Going with standards, is a flat frequency response what is required to reproduce a film soundtrack at home faithfully, or is a house curve needed to "correct" our ears insensitivity to low frequencies ?

Thanks !

--Regards,

 

[avaserfi]

jcPanny is right, if you want to have full faithful reproduction of what the director envisioned you should try and get the frequency response as flat as possible end of story.

Now many people find that they are not happy with a perfectly flat frequency response and decide that the point of watching a movie isn't to make the director happy, but to enjoy it themselves. This is how the housecurve was invented. Since humans are less susceptible to low frequencies than high people decided to bump up the lower frequencies to get that extra effect they like. Just because it doesn't faithfully reproduce exactly what the director envisioned doesn't mean its bad, especially if you like it more.

 

[Vaughan Odendaa]

Thanks for your reply. I understand that having a house curve isn't bad but I guess I wanted reassurance that it either was supporting (or not) a more accurate representation of the film soundtrack.

It seems that it does not, or does it ? I understand that it's just a subjective modification of your frequency response. Apparently Harman has done research on this but I'm not sure how their research fits into all of this. It does make sense, at least to me, that boosting the very low frequencies and then flattening the upper bass frequencies should make for a more perceived flat response.

Thanks again for the article. In that article, it was said that a measured flat response is not the same as the perceived flat response and I understand that now. But if the measured frequency response is not perceived by us to be flat, then why is it that engineers aim for measured flat response ?

Thoughts would be much appreciated on this ! Thanks.

--Regards,

 

[avaserfi]

Thanks again for the article. In that article, it was said that a measured flat response is not the same as the perceived flat response and I understand that now. But if the measured frequency response is not perceived by us to be flat, then why is it that engineers aim for measured flat response ?

Simply put it is because there are too many variable involved with a perceived flat response. Similarly to everyone liking different speakers for different reasons one person might perceive a slight slope as a flat response while another might perceive a much larger one. Human perception is extremely fickle and often times incorrect thats why we have so many gadgets to adjust our HT. If we could trust our perception who would need an SPL meter after all?

Some people could go through all the trouble of testing what the average perceived flat response is, but that still wouldn't satisfy many and would likely end up being a waste of time. As I have stated before audio is an extremely personal hobby and everyone has the way they like things and need not attempt to change that because of a arbitrary standard which this situation would be.

 

[Vaughan Odendaa]

Fair enough. What you say does make sense. But if each person adjusts his or her frequency response to achieve the flat perceived response then, to me, it would seem that that would represent the ideal flat response. Which makes everything subjective. I know that.

The only question remaining is whether this is how movies are mean't to be watched because of our ears roll off at low frequencies. I mean, if you are not going to percieve the very lowest frequencies properly due to our ears inherent roll off, then it would seem more accurate to use and modify the Fletcher/Munson curves so that you can faithfully reproduce whatever low frequency content that is recorded.

Has anyone here achieved this flat perceived response and what are their thoughts ?

Thanks.

--Regards,

 

[tbewick]

'So there is no standard on how the response should look like ?'
Vaughan Odendaa

There are standards which make recommendations for the in-room response, see -

http://www.aes.org/technical/documents/AESTD1001.pdf

Hi

EDIT: I thought that it was also worth mentioning this -

'7. Steady-state measurements of a loudspeaker in a room generate data that is not open to interpretation without detailed knowledge of the loudspeaker and the room of a kind that is not commonly available. At frequencies above about 300 Hz, the room curve is an uncertain combination of direct, early reflected and reverberant sounds; at lower frequencies it is dominated by position-dependent interferences and standing waves. Such measurements appear to be of some value for 'room equalization' at low frequencies but, at middle and high frequencies, the quantity being measured does not necessarily represent the quantity that is of greatest importance to listeners. This makes target functions for room curves unreliable except in cases of well-defined rooms and loudspeakers'.

'The Loudspeaker and Headphone Handbook', 2nd edition, edited by John Borwick. Page 489, Chapter 11: Subjective Evaluation, by Floyd E. Toole. Focal Press 1994.

 

[Vaughan Odendaa]

So they are saying that there really is no ideal curve ? Anything goes ?

Am I interpreting this correctly ?

--Regards,

 

[avaserfi]

So they are saying that there really is no ideal curve ? Anything goes ?

Am I interpreting this correctly ?

--Regards,

As I have told you repeatedly a house curve is just like the rest of this audio hobby a preference! There is no ideal because some people like a steeper curve than others. Just like the many other things in this world there isn't a set and forget way to make a house curve you will like. Experiment and see if you like the house more than a linear curve.

Edit: If you want further elaboration on why there is no real ideal curve read everyone of my previous posts again and the link I sent you one more time it is mentioned in all of them.

 

[Vaughan Odendaa]

Wow. Thanks, I guess.

--Regards,

 

[mtrycrafts]

jcPanny is right, if you want to have full faithful reproduction of what the director envisioned you should try and get the frequency response as flat as possible end of story.

Now many people find that they are not happy with a perfectly flat frequency response and decide that the point of watching a movie isn't to make the director happy, but to enjoy it themselves. This is how the housecurve was invented. Since humans are less susceptible to low frequencies than high people decided to bump up the lower frequencies to get that extra effect they like. Just because it doesn't faithfully reproduce exactly what the director envisioned doesn't mean its bad, especially if you like it more.

The question then becomes: how did the studio/director/thx/etc EQ their speakers before they mastered the sound. That is the key to all this.
Was their speakers and room EQed to flat? If so, then their preferred sound will show up if yours is also EQed flat. Were their speakers EQed first to a house curve? Then their mastering would be based on that curve which would be tough to duplicate or achieve no matter what.

 

[avaserfi]

The question then becomes: how did the studio/director/thx/etc EQ their speakers before they mastered the sound. That is the key to all this.
Was their speakers and room EQed to flat? If so, then their preferred sound will show up if yours is also EQed flat. Were their speakers EQed first to a house curve? Then their mastering would be based on that curve which would be tough to duplicate or achieve no matter what.

I completely understand, but you are still running into one problem: For the most part these are questions you will never get an answer to. Many movies will have different house curves depending on the variety of variables you listed and more.

Seems a bit silly that you would need to EQ your low end every time you watch a different movie, doesn't it? Then you get into the world of what if the speakers they had in the studio are colored on the high end or what if one of the people involved with the audio mastering for the movie has a slight hearing loss at 1200Hz? There is an infinite amount of variables that affect these situations and that is why I have been stressing do what sounds best to you not copy some house curve that calls itself the best.

 

[mtrycrafts]

I completely understand, but you are still running into one problem: For the most part these are questions you will never get an answer to. Many movies will have different house curves depending on the variety of variables you listed and more.

Seems a bit silly that you would need to EQ your low end every time you watch a different movie, doesn't it? Then you get into the world of what if the speakers they had in the studio are colored on the high end or what if one of the people involved with the audio mastering for the movie has a slight hearing loss at 1200Hz? There is an infinite amount of variables that affect these situations and that is why I have been stressing do what sounds best to you not copy some house curve that calls itself the best.

THX all the way!!! That would solve it

 

[MDS]

There's been a lot of good information and advice in this thread so far. I just want to add one thing.

I get the impression that the desire to use a specific 'house curve' is to compensate for our ears insensitivity to lower frequencies. But the Fletcher-Munson equal loudness contours do not address that specifically and should not be used as a basis for justifying, one way or the other, whether you should or should not boost the low frequencies.

The Fletcher-Munson curves show the SPL required in order for the human ear to perceive two different frequencies to be of equal loudness. I don't see any correlation between that and the desire to achieve a flat frequency response or one with boosted bass or rolled off treble. Flat is desirable in the sense that the playback system isn't adding anything to what is there - but one's preference to boost the bass or not is perfectly acceptable.

Measuring SPL while balancing the channels will cause the subwoofer channel to read 2-3 dB lower (if RS SPL meter) than it really is and that leads to all kinds of 'rules' that you should set that channel X dB higher. I say do whatever you want. I personally don't bother. I don't care if the level is actually a few dB higher than the others because it sounds fine to me.