As far as I know they measure a songs average level and then compare it to the songs peaks giving the songs DR.
They published a plug in to do that so anybody with a computer can add to the list.
In fairness it has gotten better since streaming services changed their practices and reduce average level meaning that over-compressed tracks sound feeble rather than louder compared to less compressed ones.
I find the DR database most interesting when comparing various releases of the same album.
The Clash London Calling for example started off with 13dB average on vinyl, same as he first cd pressing.
Fast forward a few years and the re-release had lost 3dB, a couple of years further down the line the HD Tracks 24/96 release only reached 9.
The Stooges Raw Power original release had a DR of 11 but when Iggy remastered it (the guy is practically deaf by now) he reduced that to 1(!).
The album is practically unlistenable.
Last edited:
Gosh, no - not on any commercial recordings I've seen. More like 18-24dB from average levels. But maybe that depends on how you define headroom. If you mean from the very softest music level on the recording to the peak, it could be high. But hopefully you are not playing those very soft passages at 80db.
I'll take look to see what I can find. I'm guessing 32dB max, but could be wrong.
Yes, my post could be misleading. But I was strictly responding to Overkill's post:
I believe it was meant as "dynamic range above ambient". I used it in this sense in response to Overkill's statements.
I have once measured it in my home using track 8 (Holst's "Mars") from Tacet's "My Audiophile Companion" - TACET-Website - english
Liner notes say: "For this recording no limiters or other devices were used (...). So it presents the full dynamic range acheved in the concert hall"
I came out to be around 60 dB from softest sounds just above ambient to the final fortissimo.
I believe that this is our "real life" reference, not a piano in a room, because audio is about reproducing recordings.
Last edited:
So it makes it a tool for relative comparison, fair enough. It would work less well in judging music with large dynamic contrasts over short and long periods. BTW CoolEdit/Audition has this built-in.
How could Raw Power ever be unlistenable.
I agree. We could just say that redbook audio has a dynamic range of 96dB and leave it at that.
What's useful to me, and I hope to other people, is the range between peak and the pianissimo sections of a recording. It's nice to know where the recording's noise floor sits, but it's not as important as the musical dynamic range.
What's useful to me, and I hope to other people, is the range between peak and the pianissimo sections of a recording. It's nice to know where the recording's noise floor sits, but it's not as important as the musical dynamic range.
I did some digging for you Graaf and sound a recording almost as musically dynamic as your estimates.
A Decca CD of Edward Elgar Enigma Variations. The softest passage I found was 39dB below peak. The noise floor of the concert hall seems to be at about -55dB. I also have some anechoic recordings were the floor is lower, of course.
A Decca CD of Edward Elgar Enigma Variations. The softest passage I found was 39dB below peak. The noise floor of the concert hall seems to be at about -55dB. I also have some anechoic recordings were the floor is lower, of course.
A simple test is to record a playback and do a difference comparison...
One major problem is shown in the CSD plots. These take a while to drop, at the same time you are playing along and the signals just get mixed. So in reality, you are getting probably less than 20db dynamic range.
One major problem is shown in the CSD plots. These take a while to drop, at the same time you are playing along and the signals just get mixed. So in reality, you are getting probably less than 20db dynamic range.
Last edited:
If you are talking about a small fullrange, the room has a very little inflience, if you are talking about a large woofer, the room is very influent.
The ability of a loudspeaker to achieve a low THD in a room should be considered as a part of the equation, especially if manufacturers gives anechoic measurement because they don't want to perform measurement in the buyers home.
Early on I tried a number of different drivers all in the same system configuration, but each set optimized with the crossover parameters. In blind subjective teats the differences were not significant even though the drivers cost was 10:1. Over the years I have come to conclude that the drivers are pretty much a commodity and that any driver set can be used with only small end result differences. There are differences, but small ones and they virtually always relate to linear response differences.
We may not have "perfect" drivers, but what we do have are drivers that are good enough so as not to make a significant difference in the end result of the total system.
I'm not disagreeing with any of that but that was hardly the original question as it is unanswerable since we all have different rooms.
My suspicion is that the OP was hoping for a number of straight up answers to compile a little list to use as info for driver selection, not a discourse about room acoustics.
The discussion has taken a number of interesting turns but it is probably not very helpful to the OP anymore.
Wait a minute....
The OP's question is a fair one (although really about systems, not just drivers and likewise, "distortion" needs to be broadly defined). And the influence of rooms is incontestable.
Therefore, the question becomes, what system plays best in the most rooms, is least influenced by the room*, and so on.
Which leads to (to me) one inescapable conclusion: you must have a DSP and a bi- or multi-amped system. If you the DSP-level control, you can address many issues of speakers and rooms and end up with a superior system.**
B.
*gotta admit, Earl has nicely thought this through apropos his horn tweeter
** likewise, unlikely to buy an off-the-shelf speaker that works in your room except by chance and passive xovers likewise
Last edited:
The key to learning is knowing the right question to ask. My point is that the question about what is the best "non"- linear distortion speaker is not the right one. But linear VS. nonlinear was not actually differentiated - it has to be, so someone should make that point. It was.
What is the best driver? Well, to me, that's not the right question, because the drivers are actually a small factor. What one should be asking is "What is the best loudspeaker architecture for the best sound." - now there is a question worthy of discussing.
If the OPs question leads to that question then it is a worthy thread.
But even that question begs another - "To do what kind of sound?" Accuracy? If not that then what? "What pleases me." - how should we know?
No thread that leads someone to a realization is unworthy.
Rather than getting into just the distortion numbers, one should really ask of a driver is what is the cause of such distortion trends. Since we can never eliminate the effects of diffraction and phase/timing related issues due to an imperfect wave front, the distortion figures can become polluted acoustically. So when you listening through speakers, much is a mixture of various sources which tends to fuzzy listening perception. This is why we need to cross compare different measurements and data processing methods to mine the real problem each at a time. Since the drivers’ own sound can mask the audibility of other imperfection, it seems that this aspect needs to be addressed first till the audibility of other aspects become more obvious. At least this is what I have discovered.
Room is very important, but it would be hard to required everyone trying to optimize their home for sound quality only, other things take priority. So the best we can do when designing speakers is to take care of the speakers themselves such that they can sound good in any room.
I recently visited a company which had an all glass listening room, but the systems there did not sound as fatiguing as I remember when I tried my design in an all glass room.
Room is very important, but it would be hard to required everyone trying to optimize their home for sound quality only, other things take priority. So the best we can do when designing speakers is to take care of the speakers themselves such that they can sound good in any room.
I recently visited a company which had an all glass listening room, but the systems there did not sound as fatiguing as I remember when I tried my design in an all glass room.
None of this answered the question which I think is what design specifications, within physical real world constraints, make for a good speaker driver? What specs make for a good speaker driver? Why and how did Gedlee pick the drivers for his speakers? The 12 is better than the 15 except for the bass. Why don’t they match except in the bass area?
Copper or aluminum caps to reduce eddy currents? Big magnets, magnet material, ferrofluid, cone shape, phase plugs, cone material, basket material, vc material?
We’ve all heard state of the art speakers that weren’t so good and inexpensive speakers that did quite well, and PA speakers that outperform hifi speakers. The question is why?
Is there any way to determine which speaker driver is the right one for your application?
Copper or aluminum caps to reduce eddy currents? Big magnets, magnet material, ferrofluid, cone shape, phase plugs, cone material, basket material, vc material?
We’ve all heard state of the art speakers that weren’t so good and inexpensive speakers that did quite well, and PA speakers that outperform hifi speakers. The question is why?
Is there any way to determine which speaker driver is the right one for your application?
The room is a loudspeaker enclosure... seems that the enclosure has already been chosen, the « best loudspeaker driver » is the one that distort the less in that enclosure (not in aechoic chamber).
I’m paricularly oriented on very large PA loudspeaker driver, because i’ve chosen the bruteforce attack (with a satin glove = low excursions), i’ve searched and find where my large multisubs are loading homogenously the room.
When your room is homogenously loaded all the waves are perfectly nulling at the center, the pressure variations of that center are extremely accurate and liberated of all standing waves.
I’m paricularly oriented on very large PA loudspeaker driver, because i’ve chosen the bruteforce attack (with a satin glove = low excursions), i’ve searched and find where my large multisubs are loading homogenously the room.
When your room is homogenously loaded all the waves are perfectly nulling at the center, the pressure variations of that center are extremely accurate and liberated of all standing waves.
Last edited:
Don't get me wrong I like this thread which is why I am following it.
It is a good thread with good and very worthwhile discussions.
I just think it may have evolved into something the OP did not anticipate as witnessed by his continued absence.