This link may be of interest to those discussing dipoles and listening distance. The theory is valid in the free field. In a room the correct way to analyze a dipole woofer system is as two out of phase monopoles separated by the dipole moment. You can get an idea of how a dipole behaves differently than a monopole in a room in this discussion.
Lynn Olson said:
Lynn, its fine to add terms, but you did not define it as I did mine. And I would further ask that the definition be such that one could see "how" it could be measured objectively. I have used a term which I believe might be the same as "resolution" and that is "transparency". I would define this term to be the inability to localize the loudspeaker itself when it is not visible (obviously if you can see it you know where it is.) This would be extremely difficult to measure objectively however. But it relates to to the criteria that the loudspeakers adds nothing of itself to the sound and as such cannot be identified as the actual source. In optical parlance, it would be called "resolution" and defined by the Modulation Transfer Function, which is a measure fo how much "resolution loss" a lens, for example, adds to the image.
Lynn Olson said:
Lynn I've mostly been lurking and reading this thread with interest. Do you have a link to any DBT with a reasonable protocol where there was proven an ability to reliably distinguish between 16, 20, and 24 bit recordings? I'm seriously interested since I have only come across the opposite. Looping a quiet passage at +60dB to compare noise floors is cheating.
Originally posted by xpert
...Don't trust phoney commercials on what a deviated pragma does better than common stuff. ...
"Commercials" meant advertising. There is no need for research on the topic of dipoles. Thankfully JohnK gave a link on his own work:
http://www.musicanddesign.com/Dipole-axis.html
Thanks John, I didn't new that before!
...Don't trust phoney commercials on what a deviated pragma does better than common stuff. ...
soongsc said:
"Commercials" meant advertising. There is no need for research on the topic of dipoles. Thankfully JohnK gave a link on his own work:
http://www.musicanddesign.com/Dipole-axis.html
Thanks John, I didn't new that before!
Hi,
This criterion is used often as a superlative in so called reviews. I would be very carefull. STEREO in itself is a not that well done concept. It should break down if the listener dares to move his head. Literally. I'm convinced that stereo listening takes some training like eating with cuttlery.
The intended illusion of STEREO may brake down by several reasons. Some after a while related to the speakers, others due to the listeners behavior, and may it be his attitude in liking the speakers for some reason or not.
But - dealing with illusions, an even wider field to discuss something beyond the ARIEL
so long
Audio versus Visual versus... NASCAR
Lynn, thanks for the comparison between the audiophile world, the photography world and the car world. It strikes me as noteworthy that, of the three pro/hobby fields, audio is the one where the hobbyists generally have the least amount of respect for the pros.
How many audiophiles have a handle on the pro audio world?
I suspect "we" have more in common with the "ricers". Sound familiar?
Good thing we have some pros contributing to this thread to keep the rest of us honest. Thanks guys!
Mike
"Do you want to fly so much that you will forgive the Flock, and learn, and go back to them one day and work to help them know?" - Richard Bach
Lynn, thanks for the comparison between the audiophile world, the photography world and the car world. It strikes me as noteworthy that, of the three pro/hobby fields, audio is the one where the hobbyists generally have the least amount of respect for the pros.
How many audiophiles have a handle on the pro audio world?
I suspect "we" have more in common with the "ricers". Sound familiar?
Good thing we have some pros contributing to this thread to keep the rest of us honest. Thanks guys!
Mike
"Do you want to fly so much that you will forgive the Flock, and learn, and go back to them one day and work to help them know?" - Richard Bach
gedlee said:
gedlee said:
I do agree to a large extent that its worth to look at a *lot* of deficiencies in speaker design from a mere scientific point of view – giving us hope for further improvement that can be generalised.
I also agree that there *is* respect for you.
But – I strictly disagree with the prayers of science , Earl
, putting science in heaven of religious believing.Let me explain before going mad – please.
First of all this whole "holly" combat between "subjective terms" versus "scientific truth" may - by some extent – be driven by the current situation.
When times get worse we look for anchor in leadership and also in believes (simply have a look into history on this subject)...
While scientifically comparing numbers is free of "believes" - to state that this does count (in one way or the other) definitely *is* believe - as it defines the lines on *subjective* improvement / preference.
Sure – for many aspects there is wide agreement about which way to go along – proven by scientific statistics – but not at each and any – and this ones are the ones in question – science doesn't help here – as the lines must be found and defined in layman's speech first anyway.
Second the truth about the terms "good / better / best" we are looking for here in the mini universe of audio are in fact subjective ones as any "truth" ever since has been – otherwise there wouldn't be different religions around the world.
Same for the "truth" as found by different courts over time and around the world. Simply put – "truth" is subjectively for anything else than comparing numbers like idiots (or "proving" axioms which isn't possible at all) – meaning it depends on individual / social / historical ... *experience* – and *judgement*.
The example given for the emotional impacts of cars comes in handy – any "truth" that really counts is not about comparing numbers (the scientific one) but about what includes *us* as humans (highly a matter of time and area we live in).
Last about the believe of anything having a measurable background in a scientific sense.
This *might* be correct if we assume to have all instruments available right now that ever will be invented in the future to come – but even then - I doubt.
Keep in mind that I refer *not* to the mere measurement (sampling) but also to the processing and visualisation that makes sense for explaining / displaying certain underlying mechanisms / assumptions / theories here.
To repeat – measuring in this context isn't meant as sampling only.
Measuring meant here is to take data – process the data – visualise the data - and *link* the data to a valid theory of explanation.
Merely from a philosophically point of view – its a problem to look at what drives what and what always comes *after* - and assuming that solutions rather be found *after* questioning the hunting after adequate measurement methods might be endless.
As for now – we are not even able to measure – and correlate – very simple effects to sonic perception. Example given by Lynn (bass performance of CD player).
For those that haven't done development of electronics it might not be as obvious – but once you did and have been confronted by – for example - the very simple and commonly shared fact that your audio does not play equally well over the week as on weekend – well – at least I haven't seen measurements (own or done by others) *in the audio signal* that give me a glimpse on whats going on (I have seen a lot – no doubt - but nothing that's meaningful enough in the frame I defined above).
There are a lot more such examples like layout, component quality and more "esoteric" ones like Bud likes to focus at. They all have in common that there *is* an effect that can be reproduced (and maybe even "proven" by differences in A/B sampling ) but can't be explained sufficiently "by measurement".
This is the big advantage of going with what we *hear* - its sampling / processing / visualisation and linking to known patterns (most of them "well known" for ages) – all in one !
(Even if we get fooled from time to time by what we think we hear)
Michael
Good point well made!
Hi Michael,
Here Here old Chap!!
I totally agree with you, there are are a huge number of unexplained audible differences, we dont know what causes them how to measure them or even what equipment to meausure them with.
Also your point about how to correlate and interpret data is a major one indeed.
At the end of the day I found having a small group of young musicians(they have the best ears and the lowest level of "sonic bias") to judge the real world in room sound quality was more informative (and the most fun!) than anything my Earthworks mic ever told me.
The down side is I sure used up a lot of hardboard and MDF with endless prototypes, but it was all FSC sustainable sourced!
Cheers
Derek.
Hi Michael,
Here Here old Chap!!
I totally agree with you, there are are a huge number of unexplained audible differences, we dont know what causes them how to measure them or even what equipment to meausure them with.
Also your point about how to correlate and interpret data is a major one indeed.
At the end of the day I found having a small group of young musicians(they have the best ears and the lowest level of "sonic bias") to judge the real world in room sound quality was more informative (and the most fun!) than anything my Earthworks mic ever told me.
The down side is I sure used up a lot of hardboard and MDF with endless prototypes, but it was all FSC sustainable sourced!
Cheers
Derek.
scott wurcer said:
This methodology is mentioned several times in the Newell and Holland book, and the authors imply that it is common practice at the higher levels of the professional world.
When I visited the SACD mastering studio here in Boulder, Colorado, I asked what method was used to create the reverb, since it had an obviously different character than the DSD sonics. Well, it turns out there is no DSD technology to create reverb, so it is done in 96/24 PCM, then folded back into the DSD mix. So even "pure" DSD recordings for audiophile labels, if they use digital reverb, actually use PCM reverb, not DSD. I found it interesting that this artifact - the sonic difference between straight DSD and an additional PCM step - was obvious enough in the mastering studio that I asked how the reverb was done after a few minutes of listening - and did so without any prompting. But that wasn't too hard to do; digital reverb, EMT plates, and real-room reverb sound pretty different, and the typical sound of digital reverb was pretty obvious in what was otherwise a very analog-sounding recording.
At the 2001 VSAC meeting, there was a mastering-disk comparison of the same Keith Johnson classical recording on 30 IPS Ampex analog tape, SACD/DSD, 192/24 PCM, 96/24 PCM, and 44.1/16 PCM. The 30 IPS analog tape and 192/24 sounded very similar and were hard to distinguish from each other, the 96/24 was a noticeable but not offensive degradation, and the 44.1/16 was a pretty obvious step down, with substantially degraded HF and altered tonalities on upper-mids and highs. I should add the ADC and DAC converters were studio-grade Meitner units (much better than consumer grade), and the down-resolution process was dithered in accordance with professional standards, so it wasn't a rigged demo.
The difference I heard there were similar to when the switch was flipped at different audiophiles' homes when playing hard-disk copies of 192/24 originals, and then stepping down in resolution. 192/24 and 96/24 are similar, but 44.1/16 is a big step down. It might not be audible on a mid-fi home theater from Circuit City, but it is certainly apparent and unmistakeable on professional-grade mastering loudspeakers, and reasonably competent audiophile speakers.
Where the difference might be masked is a mid-grade pro speaker with internal amplifiers and digital crossovers - in that case, the speaker can never sound better than the internal ADCs and DACs built into the speaker, and these are unlikely to be top professional-grade converters, like the Meitner. This warning is repeated several times in the Newell and Holland book.
I find it a little surprising that anyone in the pro world still believes that 44.1/16 PCM is audibly transparent. It isn't, and never was. The pro world stopped using it as a recording standard more than 20 years ago - and I don't think the entire industry, all across the world, spent hundreds of millions of dollars converting over to 96/24 as a fashion statement. The pros I've met regard the difference between 44.1/16 and 96/24 as similar to the difference between SDTV and HDTV - if both look pretty much the same on your display device, you don't have a high-definition television. The same applies to loudspeakers - if high-data-rate audio sounds just the same as low-data-rate audio, it's not a high-definition loudspeaker.
It's easy these days to compare 192/24, 96/24, and 44.1/16 recordings, if you have access to 192/24 originals, and have the software to do correctly dithered downconversions. This can be done at the flip of a switch. If no difference is audible between all three, that's a direct indication of the effective resolution of the playback monitors and associated amplifiers - no better than 16 bits, and maybe less.
If 20 and 24-bit depth sound the same (assuming correct dithering on the 20-bit downsample), that implies the effective resolution of the speaker/amplifier is somewhere between 16 to 20 bits, and if 20 and 24-bits are clearly distinguishable, then the speaker is at the highest level. That's what I heard at the 2001 VSAC; the same recording sounded different on 30 IPS Studer mastertape, 192/24 PCM, 96/24 PCM, and 44.1/16 PCM, with quality ranking in about that order. It was hard to assess where SACD/DSD fit in, since it had a sonic quality noticeably different than any of the PCM versions, and different from the analog master as well.
One of the nice things about digital PCM and software processing is that we really do have a "resolution" switch, provided that good dithering is carried out with downsampling - if it isn't, the buzzy artifacts give away the lower-resolution versions right away, and that's not a valid comparison. With lossy compression, of course, the moment-to-moment "resolution" is dynamic and program-controlled, not a fixed value. Continuing the speaker comparison, the lowest-resolution speakers would make MP3 and AAC lossy compression and uncompressed 44.1/16 PCM sound the same - in other words, iPod quality.
Lynn Olson said:
I was refering to E. Brad Meyer's study, "Audibility of a CD-Standard A/D/A Loop Inserted into High-Resolution Audio Playback" in the AES Journal. I would hardly say the difference is "obvious" anymore than wire directionality is. And yes it is entirely possible for the audio industry to overshoot necessary technology in the name of something "more" to sell. And BTW the people who design this stuff started laughing after 18 bits and 88.2K.
Lynn Olson said:
An ABX on the differentiation of PCM and DSD, performed with professionell musicians, sound engineers etc. gave 4 out of 145 listeners, who could differentiate with a probability of 75% or higher (with 75% being the threshold of significance in this case). You may look at Grafik Nr. 49 at page 73 of the PDF.
Obviously there is a rare breed of "golden ears", who can distinguish, but the overwhelming majority can´t.
Rudolf
Rudolf said:
Rudolf - this is much more in line with my experince. It's hard to say that the differences don't exist, but its very easy to say that they are not significant. I mean two sets of loudspeakers in a room could be differentiated by 100% of the people 100% of the time. That's significant. I am often accused of claiming that "there is no difference" in things, which I have never claimed. I claim that the differences are "not significant" when compared to those things that make a huge difference like the loudspeakers and the rooms. And, to me, it makes no sense to argue about the difference between 96/24 and 44.1/16 until you have a special built listening room with state-of-the-art speakers AND are one of those rare individuals who CAN hear the difference (and that is a very very rare individual).
Lynn - recording studios have to use a higher grade of sampling and storage for the very same reason that DSP chips have to use more bits in the multiply and add stages - roundoff errors when the signals are "accumulated" (mixed) down the road. But this is in no way proof that the final product in 44.1/16 is inadequite (when properly utilized, of course).
mige0 said:
Michael
You have some valid points, but the above is certainly false and clearly where I would disagree. There is NOTHING subjective about "truth". "Truth" stands alone completely independent of any human factors - "science is NOT a democracy".
Audio as a "religion" IS the problem.