To put it succinctly, stop worrying about non-linear distortion in speakers for the time being as there is a world of other issues that should be addressed.
Yeah, I was more talking of my experience (not with numpy/scipy, but with some more obscure shell script functions) where it took a bit of digging and installing the right 3rd-party modules into the python codebase.
My use of "public domain" and people sharing things was meant less for the algorithm itself than how that algorithm is best implemented, i.e. how to partition/pipeline the math to get through the processor/cache structures most efficiently. Heck, it's becoming more popular that an author will have to publish his/her code as supplemental material when submitting a paper (and to cross breed to another topic on null studies, have to share the code for doing one's statistical analysis).
All in all, I think we're hitting many of the same points with a bit different verbiage. Advanced math modeling is no longer held behind Mathematica/Matlab(Maple) or custom-built implementations.
I have Mathmatica and Scilab and I realize that they both have many of the more obscure functions, but neither of those are "code" languages like FORTRAN, VB and C++ since they are both written in core codes. Mathmatica has only recently added the more obscure spheroidal functions and Scilab itself is fairly recent. I coded the spheroidal functions in FORTRAN many decades ago and other had done it decades before me (but were not public domain when I needed them.) I am 66 years old and have no intention of learning a new language since none of them can do anything that I cannot do with my "obsolete" codes (and usually faster.) My son is learning Python and that's fine, but I have no interest in doing that.
I don't know who you are but you obviously have a uniquely fresh viewpoint on audio ... one that I agree with 100%. Preference is very problematic and most do not understand when audible, difference, etc. becomes simply preference. They are all quite different.
Yes, I would agree to your wording as preferential to mine, however, mine was much more concise and meant pretty much the same thing. At my age I don't have the time to waste words on forums.
I only have much experience with speakers having decided early on that speakers were the weak link in any system and that electronics could be made inaudible. That left the most interesting problems in speaker design and I still believe that to be the case.
The unsolved problems in speakers IMO are how to make good directivity control smaller. People want smaller and usually forfeit good directivity control to get it. But without good directivity control one cannot make a premier speaker, so "small" is contrary to "sound quality".
In speakers its linear distortion that is the main problem. Things like resonances, diffraction, polar response aberrations ... These are all linear, but quite audible.
The two speakers I have owned lately have addressed these linear distortions to a high degree and also have low distortion (non-linear). The newer Quad ESL's and the JBL M2. To mention just 2 but those are very few and not small nor cheap enough for the masses.
Many sit in the far or reverb field except studio and mastering people..... together with near field and DSP based EQ systems for ultra flat direct/near sound... can aid in many speaker and listening issues. Leaving resonances etal.
-RNMarsh
Many sit in the far or reverb field except studio and mastering people..... together with near field and DSP based EQ systems for ultra flat direct/near sound... can aid in many speaker and listening issues. Leaving resonances etal.
-RNMarsh
Last edited:
All the speakers that I have designed have these characteristics. The most common comparison that I get to my speakers are the M2's.
Certainly true for some, not true for others. For a good sampling, see here:
A Test. How much Voltage (power) do your speakers need?
It looks like some believe that self-proclaimed-expert listeners are the same thing as well-trained-skilled listeners. If there is a need or desire to use humans as a form of audio instrumentation, they should function in some ways more like drug sniffing dogs than human consultants. Trained, tested, and know the job they are supposed to be doing.
It seems to me that very recent advances (<10 yrs.) in cognitive psychology should make it practical to do.
It seems to me that very recent advances (<10 yrs.) in cognitive psychology should make it practical to do.
Last edited:
If preference matters they are their own experts. I recently talked to a chef who said that the average customer prefers the "Campbell's soup taste" of yeast extract and industrial flavor enhancers to real stocks made from scratch.
I know a few people who are much better at hearing things than I am. Pitch, tempo, tonal balance. It's not hard to test. The only thing I was ever precise at was feedback frequencies. Not that I could always name them, but I got rather good at reflexively reaching for the correct slider on a 31 band EQ.
This is one I hope will get solved. I think the technologies are all there, it's just whether the market will stand it. The fact that the Kii3 seems to be selling very well gives me hope.
12" wide at the baffle is the largest I can sensibly fit in my small room. Your Nathan would have been perfect and who knows one day a pair might come up. But I am lucky to have a music loving wife so can go larger than many
Regarding preferences and skilled listening, there could be preference experts, too.
For example, there are essentially such kinds of people who mix records. Basically they exist because their tastes line up well with the tastes of the masses. They "turn the knobs" until it sound good to them, and viola! Hits!
But, seems to me skilled listening would be a different job which could be trained and measured to suit its own requirements.
Measuring the performance of a preference expert would probably need to work differently, and would be more difficult.
For the purposes of designing audio equipment, my own sense is that we could use skilled listeners more than preference experts, for the most part anyway. There might niche areas for preference experts though. Maybe for some kinds of speakers.
For example, there are essentially such kinds of people who mix records. Basically they exist because their tastes line up well with the tastes of the masses. They "turn the knobs" until it sound good to them, and viola! Hits!
But, seems to me skilled listening would be a different job which could be trained and measured to suit its own requirements.
Measuring the performance of a preference expert would probably need to work differently, and would be more difficult.
For the purposes of designing audio equipment, my own sense is that we could use skilled listeners more than preference experts, for the most part anyway. There might niche areas for preference experts though. Maybe for some kinds of speakers.
Last edited:
Mark: couple of radical ideas for you
1. Please accept that mixing pop records and accurate reproduction are completely different and most of us honestly don't care what rubbish is used in a mix suite to check that a synthetic hit will sound 'banging' on a cheap car stereo.
2. Stand back and consider that a number of proper academics have done lifetimes of research, proper research not a collection of anedcotes and claims, and have come to broadly the same conclusions on what matters. Standing on the shoulders of giants would seem to be a very sensible thing to do rather than ignore that and chase some concept of 'expert listeners' based on some postings on a forum from people who hear things that aren't there.
1. Please accept that mixing pop records and accurate reproduction are completely different and most of us honestly don't care what rubbish is used in a mix suite to check that a synthetic hit will sound 'banging' on a cheap car stereo.
2. Stand back and consider that a number of proper academics have done lifetimes of research, proper research not a collection of anedcotes and claims, and have come to broadly the same conclusions on what matters. Standing on the shoulders of giants would seem to be a very sensible thing to do rather than ignore that and chase some concept of 'expert listeners' based on some postings on a forum from people who hear things that aren't there.
Sean Olive has shown that trained listeners, and they do rigorous training, have the same preferences as untrained ones. The trained group are more stable and arrive at the answer faster, but the average of a large number of either is the same.
Self proclaimed experts tend to believe that numbers of subjects are not needed. All one needs is their opinion.
Too bad these issues have not been appreciated by consumers. We need to educate them. I was using controlled directivity speakers from JBL back in the 1960's. using LE175DLH compression driver with 90 degree lens etal for my home system to reduce wall reflections near the speakers and I always sit earlier than critical distance. Just been what I do since forever. Your work has helped in this direction. How can we get more people to know the benefits??
-Richard Marsh
Last edited:
Sorry you're off base again, I have my personal preference and make no judgement about someone else's. What about that is ambiguous? Someone likes Seurat and someone else likes Munch so what? I like Munch of course.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.