Do you really need all this science and research (bless their souls) to make informed decisions whether it be DIY or store purchases?
I think the opening question is designed to see how many heavy weights will enter the debate as opposed to a real fact finding mission.
I have a reasonably GOOD SOUNDING home system at a modest cost, I just think that we loose sight of what makes this hobby fun.
Larry
I think the opening question is designed to see how many heavy weights will enter the debate as opposed to a real fact finding mission.
I have a reasonably GOOD SOUNDING home system at a modest cost, I just think that we loose sight of what makes this hobby fun.
Larry
Flat and smooth FR is not all. Remember the Celestion SL-6 that has -1.5 dB / octave slanted FR, still among the best. I think minimal stored energy (shown by the waterfall diagram) is also important. So is the polar diagram, smooth transition from omni to beamed towards higher frequencies. Then we have thermal compression...
I'm afraid I must take issue with that. As the person who asked the opening question to which you refer, it was asked with the intention of getting feedback on what exactly counts as accuracy in a speaker. The question I asked is a further development of my curiosity regarding the claim that listeners tend to prefer more accurate speakers over less accurate ones. After reflecting upon the issue a little, I realized that I didn't have a clear idea of what indicators to look for in seeking an accurate speaker. Hence my question.
I don't think anyone is questioning your system or your likes and dislikes, and as far as I can tell there's nothing about my question or this thread that says you can't have fun pursuing your hobby. The question was intended to help me (and anyone else who happens to be interested in the issue) become better informed about what the primary determinants of accuracy in a speaker are, so that I can make a more informed decision when it comes to looking for reasonably accurate speakers to experiment with (as a long term listening test). Accusing me of asking a question specifically designed to attract your so-called 'heavyweights' is a little over the top, if you ask me, and so I just wanted to set the record straight.
Edit: p.s. I've really enjoyed following the discussion that has taken place in this thread and want to thank everyone so far for taking the time and effort to share their views in a thoughtful and respectful manner. It really is greatly appreciated.
Last edited:
Certainly true to its fullest extent but room exists between no attempt and live performers for all kinds of valuable experiments; common, familiar and repeatable household sound sources such as phones, blenders, drills, etc.. to mechanically actuated musical instruments. Grade schools participate in robotic completions, repeatable taps to a snare, kick or cowbell should be trivial to implement. Or a violin stroke.
While the tests would be boring and lack the advantage (?) of triggering emotional responses associated with music it would be no surprise to learn from the results. It does raise the question of the proper recording method to use for the comparisons but now we're really getting to the crux of 'accuracy'.
Member
Joined 2003
I think the reason for that can be traced back to the analog recording era, where bands would choose studios and venues for the character that the particular place imparted on their music. With digital recording, musicians still want a certain kind of characteristic sound that may not be in line technically with a clean & accurate recording.
With movies, intense realism (at least "real" according to hollywood) seems to be a selling point so a technically sound recording trumps an artistically influenced recording...and I bet there's a lot fewer sound studios doing work for major movies than there studios doing recording work for musicians.
Very nice, Mitch! So nice it hasn't gotten any comments.
Reading thru the past few pages I kept thinking "All this is great info, but how the heck do you measure it?" We all (should) know that a single point microphone measurement isn't likely to give a good representation of what we hear with a head and two ears. But we often forget to mention that fact when we discuss these things.
Back when I did Dolby and THX alignments in cinemas, we used multiple mics scattered around. Mainly that was just to check that everyone could hear the dialog correctly, but it also gave a decent picture of the overall response.
These days I'm happy with the moving microphone technique. It gives a decent representation of what I hear in the room. Doesn't do much for measuring phase, tho. There is a thread somewhere on the technique.
What you've shown on your plots is quite useful, it lets you see how the FR changes across your listening space, much like the cinema techniques. Good work and thanks for showing it.
Well, philosophically, first of all Hi-Fidelity is not about accurate because there is no measure of accurate available
The only such measure could be "how it sounded to the mixing engineer" but we cannot know how, no way.
Secondly, why should mixing engineer's subjective preferences be measure of anything allegedly objective? And accuracy is supposed to be an objective quality, right?
On the contrary, Hi-Fidelity is about realism, realistic sound reproduction - as a goal pursued of course, not anything already achieved.
It was the correct understanding from the very beginning. H.A. Hartley, who had invented the term, understood it in this way precisely.
The only such measure could be "how it sounded to the mixing engineer" but we cannot know how, no way.
Secondly, why should mixing engineer's subjective preferences be measure of anything allegedly objective? And accuracy is supposed to be an objective quality, right?
On the contrary, Hi-Fidelity is about realism, realistic sound reproduction - as a goal pursued of course, not anything already achieved.
It was the correct understanding from the very beginning. H.A. Hartley, who had invented the term, understood it in this way precisely.
Why do people like to complicate things when there is no need for that?
That was a rhetorical question. There is not much choice for a DIYer.
You can choose to trust sound engineers that made the recording and
design an accurate speaker (high fidelity in respect to input signal)
so you can listen to a recording the way it is or you can design it the way
you hope it will alter the input to whatever you feel is right. If I had a
means to actively equalize the sound, I would probably do so to improve
the bad recordings to my liking and use none for the good ones.
That was a rhetorical question. There is not much choice for a DIYer.
You can choose to trust sound engineers that made the recording and
design an accurate speaker (high fidelity in respect to input signal)
so you can listen to a recording the way it is or you can design it the way
you hope it will alter the input to whatever you feel is right. If I had a
means to actively equalize the sound, I would probably do so to improve
the bad recordings to my liking and use none for the good ones.
Well, there sort of is.
Most of us were not at any of the original recording sessions of music we seek to play thru our speakers and as such, we have no frame of reference regarding the original performance and thus cannot truly judge a speakers ability to accurately reproduce a performance.
However, in a truly technical sense, a measure of a speaker's accuracy is its ability to reproduce a signal fed to it. Below is described a way to do just that.
If you take good quality signal generator and a dual trace oscilloscope and a good quality microphone you can hook up the sscilloscope in such a way as to track the input signal to the loudspeaker (from AMP) and the output signal from the microphone pointed at the speaker. With the generator's capability to input a variety of signals ranging from basic sine wave to triangle and square waves, you can challenge a speaker's ability to accuratly reproduce those signals whilst watching the results on your oscilloscope. Some of the best speaker designs can do a pretty good job. Most can't.
But the signal is electrical and what the speaker does in a room is acoustical
How can one tell that the acoustical event accurately reproduces the electrical signal? Aren't they incomparable?
But Your proposition relies on an assumption that the accuracy so understood is relevant at all ie. that how human hearing works somewhat resembles what an oscilloscope with a mic does.
Science does not support such an assumption.
So - what's the point of accuracy so understood?
Dunlavy speakers were very good at such tests IIRC. What goes in comes out - on an oscilloscope, yes. But only on a specific measurement axis and at a certain minimum distance.
Well, where they "the best"? Are there any Dunlavy-like speakers today, any followers?
Why not?
Attachments
Last edited:
Such accuracy is quite relevant and science does support that notion: as O'toole noted, we prefer less distortion. This *must* be measured to prove which speaker has more or less distortion.
In that case they are comparable, because you are simply looking for consistency in how closely the original signal and recorded signal appear on the oscilloscope.
We know human ears have a different sensitivity curve to that of a mic, but we are not measuring how the speaker sounds, rather how accurately it reproduces a signal based on a known input. In other words, we know the signal should look like the tone that was fed to the speaker, and any deviation from that value indicates less accuracy.
I don't think any product, let alone speakers, can be quantified with a single test and declared to be "good" or "bad". I would apply this test as a measure of consistency, but the final "tuning" of sound would be done by ear using actual music and not test tones.
Speech of a known speaker can be a good source.
Movie mixing sets have sound quality standards and musical recordings do not. This is a huge difference that could easily account for what you say, which I completely agree with by the way.
Accuracy is definable and does matter. What often gets lost is that not all of what we can define as accurate is subjectively important in that it is not detectable by our hearing mechanism. One can define nonlinear accuracy, but we are not all that sensitive to it. One can define phase accuracy, but the jury is still out on to what extent this matters, most data says it doesn't, some newer data says that in certain circumstance it might. We can define frequency response accuracy and we know that this matters a lot. We can define polar response accuracy and we know that under certain conditions this too can matter
So the claim that we cannot define accuracy and so it doesn't matter is absurd. Some matter more than others and some not at all, but to throw it all out is ridiculous.
So the claim that we cannot define accuracy and so it doesn't matter is absurd. Some matter more than others and some not at all, but to throw it all out is ridiculous.
I am not sue what "a single test" means, but Drs. Toole and Olive would contend that sufficient tests of "good" versus "bad" do exist and I would tend to agree, although I might not agree with the exact tests and their interpretation.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.