Is it possible to do 45Hz to 13Khz with one driver? Does the below graph answer your question ?
Raw uncorrected driver measurement without any EQ or BSC applied, of an almost 40 year old 8" dual cone driver in a 45 litre bass reflex box.
Below 230Hz was measured nearfield (cone and port summed) corrected for far-field BSC loss and then spliced above 230Hz with a 1 metre measurement with 4.5ms gate time. (1/24th oct smoothing) It's the best I can do in my restricted measurement environment... The ripple from 500-800Hz is spurious output from a front facing port which goes away if I block the port...(note to self, next time port on the back
By the time BSC is added and room gain is taken into account they work very well down to 40Hz. At the top end they manage 13Khz but there's a bit of on-axis cancellation from 13-16Khz that goes away a few degrees off axis...
So yes, it's possible.
Attachments
Coherence of a multiway design depends greatly on the skills of who designs the xover... and on his personnal interest/taste regarding this issue. Many multi way speakers obviously sound as if their designer have never heard a good FR, nor found anything of interest in their listening experiences with Frs...
On the other hand, Fr lovers tend to forgive and forget all the tonal flaws of their favorite drivers, because they have learned to love them...
So they just don't even need to Eq anything. Love is blind...and Deaf!
BTW, i personally eq as much as i can, FR o multiway, but Eq doesn't cure anything: Directivity is one of the issues uncovered by Eq...
Let me turn your argument on its head by suggesting that the speaker itself is one of the "parts" that you're putting into the circuit which is introducing colouration of the original signal.
Non-flat frequency response is one kind of colouration of the original signal which this component - the speaker, (driver+cabinet) is introducing.Unless you're using cored inductors or electrolytic capacitors, the colourations the driver is introducing are several orders of magnitude greater in severity than anything that a well designed BSC network would introduce.
(I'll skip over the fact that you're assuming I'm talking about a passive network, or indeed even an analogue one based on discrete components, as I'm not)
Of all the colourations that a speaker can introduce, non-flat frequency response is the only one that can be externally corrected for. If the driver adds non-linear distortion, tough, you can't do anything to improve that. If it adds high Q ringing/resonances or random cone breakup, tough, you can't do anything about that either, short of attempting cone damping modifications.
If the response has broadband non-flatness like requiring baffle step correction, this is one thing that you can correct for externally, which will reduce this form of colouration.
The audibility of a single passive component (cored inductors and electrolytic capacitors aside) is something that is highly debatable, while the audibility of a gross broadband error in frequency balance on the order of 6dB or more (the "or more" coming from additional non-flatness of the driver over and above baffle induced changes) is extremely obvious.
To me, getting the broad band frequency balance close to right is far more important than worrying about some tiny, perhaps undetectable (and highly theoretical) problem with a component, especially when a modern high quality digital EQ can do whatever correction is necessary in an almost zero distortion/noise way at line level.
Well again, I wouldn't call it a crossover. If you're doing line level EQ of the signal and the speaker is directly connected to the amplifier output, where is the crossover ?
As for frequency irregularities, that depends on the nature of them. Narrow band ripples in the response are generally inaudible as long as the trend line is flat, however even small broad-band shelf type errors such as uncompensated baffle step are plainly obvious. The wider the bandwidth of the error, the less the error has to be in amplitude to be obvious.
If clarity and timbre to you mean a grossly exaggerated midrange and treble, that's your personal choice but it's certainly not accurate. I don't find maintaining good clarity and timbre and getting good bass and a well balanced response a problem, especially when digital EQ makes it possible to adjust the response in any way necessary with as much accuracy as necessary.
The thing about BSC and frequency response in general is that the closer you get to a flat balanced response, the more critical it becomes to get things just right, or you risk "smooth and dull" as you call it. If you just apply a simple 6dB shelf there is a very good chance you'll overshoot the mark and it will sound dull, because 6dB is never exactly the right amount due to so many factors relating to the driver, baffle, room placement etc.
I'm sure more than a few people have tried a simple BSC network like this, didn't like the sound, and gave up. Use a fully flexible digital EQ that will do whatever you want with a high degree of precision and back it up with measurements and it's not hard to find the right balance.
You can have well balanced bass response and clarity and timbre at the same time
As I see it, the biggest problem with FR, and in my case the only one I care about, is the HF directivity. That to me is the only reason to still use tweeters from 10k and up.
But if you use two 3 - 5 inch drivers per channel in one of the various setups (angled horizontal, vertical, one pointing up, one backfiring on a hard surface etc.), this can be largely cured.
Other than that FR has far superior detail, transient response and cohesiveness, to any other speaker paradigm I know of (if you can call it that).
Only remaining problem being very wide dynamics, but that is mostly a psychological thing.
FR simply sounds better in every respect, once you've gotten over the shock of the radical difference in reproduction style.
Again it's psychology. A lot of people, even very knowledgeable people confuse "I'm used to this being the best" with actually listening. Really seriously testing a pair of good speakers takes weeks if not months, for these exact reasons.
But if you use two 3 - 5 inch drivers per channel in one of the various setups (angled horizontal, vertical, one pointing up, one backfiring on a hard surface etc.), this can be largely cured.
Other than that FR has far superior detail, transient response and cohesiveness, to any other speaker paradigm I know of (if you can call it that).
Only remaining problem being very wide dynamics, but that is mostly a psychological thing.
FR simply sounds better in every respect, once you've gotten over the shock of the radical difference in reproduction style.
Again it's psychology. A lot of people, even very knowledgeable people confuse "I'm used to this being the best" with actually listening. Really seriously testing a pair of good speakers takes weeks if not months, for these exact reasons.
Last edited:
Agreed. A multi-way design CAN sound as coherent as a full range driver if done properly, but only on axis, and only if you're very careful with the crossover design and tweaking. Rather small errors in the crossover (or poor choice of topology) can easily kill the sense of coherence, whereas the coherence is implicit with the full range driver.Coherence of a multiway design depends greatly on the skills of who designs the xover... and on his personnal interest/taste regarding this issue. Many multi way speakers obviously sound as if their designer have never heard a good FR, nor found anything of interest in their listening experiences with Frs...
I've done quite a lot of experimentation with the full range driver mentioned above using them both as full range (with EQ) and as a 2 way design crossed over with a ribbon tweeter, and one of my goals when crossing them with the ribbon tweeter was to try not to lose the coherence of the full range driver on it's own, and indeed see if it was even possible.
It put me in the interesting position that when trying different crossover types I was able to refer back to the exact same driver in the same cabinet running full range as the "reference" to what coherence should sound like, so as I tried different crossover slopes/frequencies, adjustment etc, I was always aiming for that coherence of the FR driver on it's own.
One thing that I found is that IMO for a crossover in the mid/treble region odd order filters (I used 3rd) sound much closer to the coherence of the full range driver by itself, possibly due to the flat power response.
2nd and 4th order crossovers didn't impress me at all, and didn't seem to have the same coherence, especially when listened to further away where more of the room influence exists. Since the majority of multi-way designs use 2nd or 4th order filters it comes as no surprise to me that many of them lack coherence.
I would call myself a FR lover, but not a FR purist, and I also like multi-way designs for other reasons. I am happy to go with whatever works/sounds best to me so if that means applying EQ to a full range driver I don't have any issue with that.On the other hand, Fr lovers tend to forgive and forget all the tonal flaws of their favorite drivers, because they have learned to love them...
I do like the fundamental concept of a full range driver though - I think trying to design a driver to be as wide band as possible (even if you don't use all of that range due to crossing it over at the extremes with other drivers) is a noble goal. I'm also of the opinion that crossover frequencies should never be placed anywhere between 300-3000Hz, so I always roll my eyes when I see designs where a tweeter is crossed over at 2Khz..etc... WORST possible place to put a crossover...
There is a lot of advantage to making a midrange driver as wide band as possible in a multiway design, so some of the lessons learned in making full range drivers can be applied to multi-way designs where the woofer and tweeter only handle the very bottom and top...
Yes that's true. On a large FR driver you really do need to tilt the treble up a bit (if it's not already in the raw response of the driver) to sound balanced at a normal listening distance - that's why I mentioned in when talking about EQ - some boost of treble is sometimes needed as well as bass to get the best result. Boosting/shelving the bass alone on a FR driver usually doesn't give a good result.
Agree. Too often search for best performance on ( design) axis leads to a very narrow sweet spot. But you can design indeed for a larger spot, especially if you are not the typical 3Dholographicpinpointimaging obsessed freak ...
A rising response to compensate for increasing directivity is indeed welcome to provide acceptably flat power response. But imho, even the best 1' tweeters are far from being perfect.
Issues....You mean like the gigantic 2.6Khz whizzer cone resonance on the FE207E ?
See left hand graph...I bought a pair years ago and never used them in a project because I couldn't stand that awful upper midrange resonance and it was too much trouble to ship them back overseas...and all the non-destructive mods I tried were ineffective. (As well as the whizzer cone there is also a bad uncontrolled breakup resonance of the main cone at 4Khz)
Compare that to the graph on the right of my Flat 8 II's whose whizzer cone resonance at 2.2Khz is almost undetectable, measured in the exact same measurement conditions...(no prizes for guessing which one sounds better! )
So what are the mods that will allow me to turn this sows ear into a usable silk purse ?
Attachments
Last edited:
If clarity and timbre to you mean a grossly exaggerated midrange and treble, that's your personal choice but it's certainly not accurate.
You can have well balanced bass response and clarity and timbre at the same time
Now where did I ever say "grossly exaggerated midrange and treble"? But many good points taken - cored inductors and electrolytic caps are more likely to degrade audibly. And no, I wasn't considering digital equalisation. Digital isn't my field so I can't comment. My all-DHT amplification is analogue of course, but I do use a CD player for input, so some equalisation in that part of the chain is no doubt possible. I wouldn't insert a digital equaliser between CDP and amp, though.
Andy
You didn't, my point though was that I consider a wide band error in frequency response that is up to 6dB over several octaves (low bass up to the baffle step frequency of the cabinet at a few hundred Hz) does constitute a gross exageration of the remaining midrange and treble frequencies if nothing is done about it.
The wider the bandwidth of a frequency response error the less it can deviate before it's noticeable. A narrow band ripple (say 1/3 oct) in the response on the order of 2-3dB might go unnoticed at most frequencies (maybe not in the presence region though) but a broadband error as small as half a dB over several octaves is noticeable, let alone 6dB.
My amplifier is all-analogue as well. I have many input sources though so I wouldn't put a digital EQ directly between CD player and amp either, as it needs to be in operation for all input sources, so I put it in the analogue signal processor loop of the amplifier - which comes after the pre-amp input source selector but before the volume control, so the equalizer is always being driven with constant line level signal.
(Connecting it between pre-amp and main amp is also a bad idea with a digital EQ because you lose a lot of SNR due to the low signal levels after the volume control)
It has a relay bypass so if I want to at the press of a button (or when turned off) the digital circuitry is entirely bypassed. With 24 bit 96Khz ADC/DAC the quality of the converters in the EQ are considerably higher than source material such as CD's though, so the only time I turn it off is if I'm taking raw speaker measurements.
Hi Simon -
I'm sure there's much to be achieved in the digital domain, and I look forward to that part now my amplification is pretty much decided. I've got some DAC boards for which I've built tube output circuits, but I'm sure it gets a lot more sophisticated than that.
So we come back to the basic premise of a full-range speaker unit. I know some of you guys add equalisation for that, though I guess that's a different forum than this?
andy
I'm sure there's much to be achieved in the digital domain, and I look forward to that part now my amplification is pretty much decided. I've got some DAC boards for which I've built tube output circuits, but I'm sure it gets a lot more sophisticated than that.
So we come back to the basic premise of a full-range speaker unit. I know some of you guys add equalisation for that, though I guess that's a different forum than this?
andy
Why add ? We do not add anything, just trying to remove spurious information that imperfect drivers add to the signal. Cleaning in a word...
I am afraid you constantly take "purist" and "minimalist" as synonymous and they surely are not. At least in audio...
Why add ? We do not add anything, just trying to remove spurious information that imperfect drivers add to the signal. Cleaning in a word...
I am afraid you constantly take "purist" and "minimalist" as synonymous and they surely are not. At least in audio...
No, I think I've made my point clear enough - introducing parts into a system, however important the goal, is not something that can be done without tradeoffs. Some trivial, some larger, and some well worth the addition. No system works without parts in it, and less isn't always more, but yes I do believe in minimum parts (and parts as good as you can get) as a design goal as long as the objectives are met. And as said, my objectives are clarity and timbre. Maybe I should add some kind of life, directness and vividness such as you can achieve with all-DHT amplification, as one example. An illustration of this is that I removed one stage in the amplification (3 stages instead of 4) and the increase in clarity was clearly audible.
I've been building hi-fi for most of my life, including several speaker cabinets, and the last years have been entirely devoted to all-DHT amplification. I've learned a lot about parts and how they contribute to the sound. In critical places I've auditioned 8 resistors before choosing the one that sounds best. And so on - I auditioned a good 20 directly heated triodes in numerous tests over 3 years before selecting the ones I found the most satisfying. Not everybody will do this kind of thing and many will say it's subjective, not verifiable by double blind testing etc etc. But with this iterative approach the benefits are cumulative.
So my question all along has been "how good do full-range speakers have to get to reach the point where we just don't need crossovers or two way systems".
But I don't want to put too much of my own situation into this thread, since there are some very expert speaker designers on this forum and thread, and they're discussing much more general theories around this subject. So please continue with that!
Andy
Last edited:
Buzzforb - have you tried Green Tea Frozen Yogurt? way much better than it sounds
Andy - to answer your original question reiterated in above post, since while I take a little pride in my construction abilities, I'm not an expert speaker designer, I'll just give a smarta$$ answer - good enough to satisfy your ears and bugger the rest of 'em
Andy - to answer your original question reiterated in above post, since while I take a little pride in my construction abilities, I'm not an expert speaker designer, I'll just give a smarta$$ answer - good enough to satisfy your ears and bugger the rest of 'em
Last edited:
No, i am just trying to understand what can be your expectations with loudspeakers considering what your taste with amplication is.
But it is not very easy to make an objetive assessment of subjective notions such as clarity, kind of life, directness, vividness, and even ...timbre.
Neither is it clear whether all these qualities are supposed to be present in the recordings or allowed to be added to it by the system, owing to positive euphonics effects for enhanced sense of clarity, sense of life, sense of..., sense of... etc...
My own (positive anyway) experience with DHT tells me that though your signal part is very short, there are very strong (sonically) ingredients in it, and i agree that a single resistor can audibly change the whole sonic signature of an amp...and my own (positive two) experience with FRs tells me too, that you can find among that family individuals with a strong personality which might marry well ( or not) sonically with your amps and according to your taste.
But i am afraid that this has little to do with the question of whether FRs are TECHNICALLY good enough vs a 2 way design, and i am afraid not being in a position to objetively assess the subjective issues in your decision making...
Last edited:
I agree with most of what Simon has said. The ANs also have some bad breakups in the 2k-5k region, but a notch filter done through line level equipment works wonders. Now, with a nearly flat response, they sound quite balanced and have been a joy to listen to.
From the popularity of full range drivers, one can infer that some people prefer coherence over tonality (or a flat response). How does that figure with Toole's research? How does that figure with the assertion that directivity control is most important parameter?
From the popularity of full range drivers, one can infer that some people prefer coherence over tonality (or a flat response). How does that figure with Toole's research? How does that figure with the assertion that directivity control is most important parameter?
Well, you're dead wrong, so there!
Actually, I've got better things to do, like studying the Russo-Japanese Battle of Tsushima. I believe nearly everyone reached the same conclusion on the outcome of that little disagreement.
Best Regards,
Terry
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.