What do Great Sounding Speakers have in Common?
Well they are rarely inexpensive thats true. To say that none on the list sound good or that they are all too expensive is to me absurd.
I am a DIY guy through and through but when I first heard the JBL 1400 Array I was bowled over. I followed up with another listen at the home of an acquaintance and without another thought bought a new pair. There are four or five others on the A list that I think sound great as well, some I could afford, a couple, not.
I thought I was getting pretty close to damn good in my DIY efforts and I still model, build, listen, measure and tweak, and it's very rewarding but, so far the Arrays and 4435's that I have for reference are better all around. That doesn't discourage me, just let's me know what I need to work on.
So, What do Great Sounding Speakers have in Common? I don't have the answer but I know what I like and I haven't found it on the cheap.
Well they are rarely inexpensive thats true. To say that none on the list sound good or that they are all too expensive is to me absurd.
I am a DIY guy through and through but when I first heard the JBL 1400 Array I was bowled over. I followed up with another listen at the home of an acquaintance and without another thought bought a new pair. There are four or five others on the A list that I think sound great as well, some I could afford, a couple, not.
I thought I was getting pretty close to damn good in my DIY efforts and I still model, build, listen, measure and tweak, and it's very rewarding but, so far the Arrays and 4435's that I have for reference are better all around. That doesn't discourage me, just let's me know what I need to work on.
So, What do Great Sounding Speakers have in Common? I don't have the answer but I know what I like and I haven't found it on the cheap.
err, what's "DDR"?
..how about hyper-fast depreciation. I'm pretty sure they ALL have that in common.
Err, everything ? What about dynamic range and freedom from thermal and frequency dependant dynamic range compression ?
Very important in describing how a speaker will sound, but not shown in JA's measurements, nor most other people's for that matter.
Small signal measurements like frequency response taken at one SPL are all well and good but don't fully describe the characteristics of a speaker when presented with music of varying amplitude.
A small limited SPL system and a larger more dynamic one do not sound the same no matter how much you EQ them to measure the same at one SPL...
Downward Dynamic Range (ref Allen Wright). The difference between a good piece of kit and a great one.
dave
i'd say good filtering.
i've heard very ridgid and very simple single-sheet plywood speakers, and both were simply enjoyable.
precision and control was better with the ridgid build, but the listening pleasure wasn't any less with the less well-built ones.
but, what totally ruins any speaker, how well built or how expensive the drivers may be: wrong filtering. phase-problems, combing, cone breakup, tweeter resonating.. all killing the music more then any other 'faults'. i'd say that even a mis-tuned basreflex is less destructive for the sound. speakers where you can hear a whoooosh (phasing) when walking past, sense a 'hole' in response when you switch between genres of music or clearly feel the frequency response devided and parts lagging are the worst.
i'd prefer a single driver with no bass or highs over all that crap. but thank god there are so many enjoyable loudspeakers and that i can hear the whole range at home.
once the filtering is correct and behaving well, only thén it's time to structurally improve cabinets, bring out the magnetostatic tweeters and diamond cones.
i've heard very ridgid and very simple single-sheet plywood speakers, and both were simply enjoyable.
precision and control was better with the ridgid build, but the listening pleasure wasn't any less with the less well-built ones.
but, what totally ruins any speaker, how well built or how expensive the drivers may be: wrong filtering. phase-problems, combing, cone breakup, tweeter resonating.. all killing the music more then any other 'faults'. i'd say that even a mis-tuned basreflex is less destructive for the sound. speakers where you can hear a whoooosh (phasing) when walking past, sense a 'hole' in response when you switch between genres of music or clearly feel the frequency response devided and parts lagging are the worst.
i'd prefer a single driver with no bass or highs over all that crap. but thank god there are so many enjoyable loudspeakers and that i can hear the whole range at home.
once the filtering is correct and behaving well, only thén it's time to structurally improve cabinets, bring out the magnetostatic tweeters and diamond cones.
Best to read what Allen had to say, but basically DDR is the ability of a device to reproduce very low level information even in the presense of the much louder main components of the music.
dave
Frequency response: Flat axial response down to lower midrange. Bass range response compatible with room acoustics and possible/desired speaker placement. No audible box/radiator resonances. Radiator size related to wave length.
Directivity: Smooth off-axis and power responses. Coherent/united/coaxial radiator (low lobing). Sufficient directivity index related to early decay time (EDT) and desired listening distance.
Timing: Excess group delay as low as possible.
(Stereophile should publish this curve, and reviewers should study to sense the difference in timing. Step response published but it's quite abstract).
Non-linear distortion: Harmonic, different kind of IMD distortions and thermal/power compression as low as possible.
Directivity: Smooth off-axis and power responses. Coherent/united/coaxial radiator (low lobing). Sufficient directivity index related to early decay time (EDT) and desired listening distance.
Timing: Excess group delay as low as possible.
(Stereophile should publish this curve, and reviewers should study to sense the difference in timing. Step response published but it's quite abstract).
Non-linear distortion: Harmonic, different kind of IMD distortions and thermal/power compression as low as possible.
Last edited:
Yes, the subtle pieces of information that make it easier to create verisimilitude & to reduce the amount that your brain fills in (so you can relax more). The relative phasing & timimg of these are important for imaging, and i imagine that the response down there should also mimic the (ideally flat) response of the big stuff.
dave
Well, showing those charts will give an impression that he who shows the charts (not talking about OP) are knowledgeable in this area. But the question is can he know the sound of the speakers (which one is better than which) from those charts? It is almost as useless as showing the brand of the capacitors used in the crossover.
Imo... arrrggghh
I'm surprised that Sonus Faber is not mentioned here. Because if I had to choose one speaker from those above and I had no information other than the charts, then I would choose Sonus Faber.
If the information I had were those charts and the price, I would choose the most expensive
If the information I had were the charts and the brand name only, may be I would choose Vandersteen (because I'm not familiar with other brands) but from the chart it seems the Vandersteen is subwoofered? Hmmm I'm not familiar with high end subwoofered system either... but knowing Vandersteen... its okay lah...
Will any of these improvements justify the cost? Sure as DIYers we can optimize anything and everything, but which improvements will be perceived as real sonic progress?
The goal is not to design to perfect a laundry list of technical parameters.
In a commercial world effort at optimizing every one of these will have an associated cost. For example, perfecting group delay will need either a more complicated stepped cabinet, an active design with DSP or complex passive delay. You might also go for a coaxial (UniQ) design and improve group delay while degrading axial and off axis frequency response.
The key is finding out what is important and what is not.
AES E-Library A Multiple Regression Model for Predicting Loudspeaker Preference Using Objective Measurements: Part II - Development of the Model
David S.
This is just a fancy term for distortion.... If you have low distortion drivers and they are used within their low distortion bandwidth, you can more clearly hear the details even through congested passages.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.