Hi,
measured at a distance of 3ft only will for sure show more lower bass volume, that won't be there any more at greater distance due to acoustic phase cancellation.
Then, beeing dipole radiators there'd be no output at all below the lowest room mode ... hence a higher lower bandwidth limit, the smaller the listening room dimensions.
So if You equalize the FR it's either the base resonance of the speaker that counts or the lowest room mode .... whichever is higher in frequency.
Pushing beyond these limits You'd just wasting tremendous amounts of power for the worse.
But then ... every Fullranger ESL wastes power for the worse.
Deeper, more precise bass can be had with more compact dimensions when using dynamic bass drivers ... and if You aim for true sub-bass going dynamic is rather unavoidable.
jauu
Calvin
measured at a distance of 3ft only will for sure show more lower bass volume, that won't be there any more at greater distance due to acoustic phase cancellation.
Then, beeing dipole radiators there'd be no output at all below the lowest room mode ... hence a higher lower bandwidth limit, the smaller the listening room dimensions.
So if You equalize the FR it's either the base resonance of the speaker that counts or the lowest room mode .... whichever is higher in frequency.
Pushing beyond these limits You'd just wasting tremendous amounts of power for the worse.
But then ... every Fullranger ESL wastes power for the worse.
Deeper, more precise bass can be had with more compact dimensions when using dynamic bass drivers ... and if You aim for true sub-bass going dynamic is rather unavoidable.
jauu
Calvin
mark100 -
Any chance you could take Calvin's good advice and put your mic where you sit? And use 1/12 smoothing? And show THD?
I mount my calibrated measurement mic on a camera tripod (legs kept locked in height) that I place where I sit, at ear height. I'm within an inch or two each time I do it. While it is best to average a few FR runs, I rather like the stability and ability to make comparisons over time by using one fixed location.
Ben
This would be a very common experience with DSP correction. It's not easy to improve reasonably good speakers with EQ unless the user is experienced with EQing. EQ is an art, not something that automatically can be applied based on the frequency response measured in the room unless the speaker or the room is horrible.
Most of the anti-DSP people in this forum seem like they just have tried one of those cheap automatic Room correction DSP, or manual EQ to make the response flat as possible without enogh knowledge or experience with EQ. I don't say they are wrong, but it's not DSP's fault.
Nicely put.
I thought I had a million good reasons to dismiss DSP and EQ.... until I started doing careful REW freq responses and adjusting slightly accordingly. I don't say everyone should have a DSPeaker (not sure it works all that well) but your system is just charmingly rudimentary without some DSP power. And with a cone sub, it is all but essential (for reasons detailed earlier).
In audio, the enthusiasm for self-delusion and thinking (falsely) that you are some great judge of sound because you washed your ears that morning, is always around. As well as coming to take present sound as the new-normal (known in psychology as "adaptation level").
Ben
Last edited:
I have to admit I give very little attention to in-room measurements of speakers.
I've spent too much time making too many measurements not to realize that the room removes any chance at good measurements at anything other than gated HF, or at a single mic position.
So when folks start talking about higher rez (1/n octave) in-room measurements, I kinda roll my eyes in my head.
THIS is the foremost reason I gave up on home HiFi....the room.
Now I build all my high-end aspirations (deep end?) on measurements made outdoors. Or if I win the lottery, maybe i'll build an anechoic chamber.
But really, in-door measurements and room correction ??? Not for me...I'll just tune by ear...
Anyway, enough of that rant.
Calvin, I don't get where you're coming from saying acoustic cancellation will dominate as mic moves further back. I mean, who knows where reflections / modes will combine constructively or destructively.
Otherwise, we just follow an inverse square law...Yes?
Ben, like I say, I shun indoor measurements, but here's one more at listening position, both speakers running..just for you
To me, everything about this curve says good bass, other than rather high 2nd order distortion below 100hz. IMO,it has good bass because it keeps up with the rest of the spectral output...that simple really...
And guys, pls don't get me wrong...I totally agree if you want either bass volume or extension, you have to go to cones.
That's the second reason I gave up on home HiFi.....killer bass...
and now run 4-6 labhorns...
I've spent too much time making too many measurements not to realize that the room removes any chance at good measurements at anything other than gated HF, or at a single mic position.
So when folks start talking about higher rez (1/n octave) in-room measurements, I kinda roll my eyes in my head.
THIS is the foremost reason I gave up on home HiFi....the room.
Now I build all my high-end aspirations (deep end?) on measurements made outdoors. Or if I win the lottery, maybe i'll build an anechoic chamber.
But really, in-door measurements and room correction ??? Not for me...I'll just tune by ear...
Anyway, enough of that rant.
Calvin, I don't get where you're coming from saying acoustic cancellation will dominate as mic moves further back. I mean, who knows where reflections / modes will combine constructively or destructively.
Otherwise, we just follow an inverse square law...Yes?
Ben, like I say, I shun indoor measurements, but here's one more at listening position, both speakers running..just for you
To me, everything about this curve says good bass, other than rather high 2nd order distortion below 100hz. IMO,it has good bass because it keeps up with the rest of the spectral output...that simple really...
And guys, pls don't get me wrong...I totally agree if you want either bass volume or extension, you have to go to cones.
That's the second reason I gave up on home HiFi.....killer bass...
and now run 4-6 labhorns...
Attachments
A few things that learned,
1. Most importantly, the correction should be started with low Q, and stick to lowest Q possible. Surgical High Q EQ should be avoided. In my experience, notch out room mode in bass frequency digitally has some advantage as well as some audible negative effects. Ideally, room mode should be corrected acoustically.
2. Less EQ point would sound better, even in digital domain. 1 dB boost / cut difference should be heard clearly. If it can't be, just stop EQing at that frequency point.
3. This is a common sense, but for bell curve, cut is much more inaudible and much less harm in general. For shelf EQ, boost is usable as long as the EQ algo is high quality. With cheap digital EQ, boost always sounds horrible.
4. Manufacture published curve is most likely more trustful than your own measurement in your room, but room measurement also can be useful as a reference, NOT THE GOAL.
5. In my experience, FIR minimal phase (or analog emu) type sound better than IIR, ONLY if FIR is extremely long (>60000). Also don't ignore the window function parameters. They affect the sound a lot.
1. Most importantly, the correction should be started with low Q, and stick to lowest Q possible. Surgical High Q EQ should be avoided. In my experience, notch out room mode in bass frequency digitally has some advantage as well as some audible negative effects. Ideally, room mode should be corrected acoustically.
2. Less EQ point would sound better, even in digital domain. 1 dB boost / cut difference should be heard clearly. If it can't be, just stop EQing at that frequency point.
3. This is a common sense, but for bell curve, cut is much more inaudible and much less harm in general. For shelf EQ, boost is usable as long as the EQ algo is high quality. With cheap digital EQ, boost always sounds horrible.
4. Manufacture published curve is most likely more trustful than your own measurement in your room, but room measurement also can be useful as a reference, NOT THE GOAL.
5. In my experience, FIR minimal phase (or analog emu) type sound better than IIR, ONLY if FIR is extremely long (>60000). Also don't ignore the window function parameters. They affect the sound a lot.
Last edited:
The Acoustats' low end is a function of the area of the speaker surface, the Q of the diaphragm, the resonance of the diaphragm, and the placement/size/type of room.
Typically they are good to 30-40Hz, usually a solid 35Hz. Which is lower than a majority of "cone" speakers. Not lower than subwoofers purpose built to go low.
The larger Acoustats, up to Model 6's give a higher output and more apparent LF, since they have 6 cells and are approximately 8 feet tall.
Of the "standard" models my ranking for bass is: 4, 3, 2 and 1+1. The 2+2 (rare) is on par with the 4... the best compromise all round for imaging and bass was the Model 3.
Acoustat published both the freq response and the impedance curves...
Since the Acoustat uses Strickland's matching system, the main requirement is voltage swing, and the Z is fairly level across the range. You want a "200 watt" class amp.
So there is "real bass" from the Acoustat, unlike the ML and most other ESLs... that was their "big selling point", imo. I've never heard bass from a Quad 63, or a ML that was even good, and certainly not close to what Acoustat did.
Otoh, the Acoustat is an imperfect speaker in all regards, but still better than maybe 98% of all speakers available today anyhow...
Oh, measuring speakers at your "listening position" is problematic because you are combining direct and reflected energy. I'd agree that if you can compare the nearfield (windowed/gated) response with the listening position response and identify and room problem, that might be valuable. I've never been a fan of EQ to flat (or a curve) at the listening position, certainly not without taking into account the room in some significant detail.
_-_-
PS. you're not going to EQ out "floor bounce", for example.
Typically they are good to 30-40Hz, usually a solid 35Hz. Which is lower than a majority of "cone" speakers. Not lower than subwoofers purpose built to go low.
The larger Acoustats, up to Model 6's give a higher output and more apparent LF, since they have 6 cells and are approximately 8 feet tall.
Of the "standard" models my ranking for bass is: 4, 3, 2 and 1+1. The 2+2 (rare) is on par with the 4... the best compromise all round for imaging and bass was the Model 3.
Acoustat published both the freq response and the impedance curves...
Since the Acoustat uses Strickland's matching system, the main requirement is voltage swing, and the Z is fairly level across the range. You want a "200 watt" class amp.
So there is "real bass" from the Acoustat, unlike the ML and most other ESLs... that was their "big selling point", imo. I've never heard bass from a Quad 63, or a ML that was even good, and certainly not close to what Acoustat did.
Otoh, the Acoustat is an imperfect speaker in all regards, but still better than maybe 98% of all speakers available today anyhow...
Oh, measuring speakers at your "listening position" is problematic because you are combining direct and reflected energy. I'd agree that if you can compare the nearfield (windowed/gated) response with the listening position response and identify and room problem, that might be valuable. I've never been a fan of EQ to flat (or a curve) at the listening position, certainly not without taking into account the room in some significant detail.
_-_-
PS. you're not going to EQ out "floor bounce", for example.
No mic measurement of any sort is definitive. Nor is any criterion like "flat" measured in any way. While Toole and chums were heading towards objective measures, a long way to go to define good personal sound in your chair.
But anybody who has lab experience knows, is that Step One is settle on some reliable, stable, meaningful, and otherwise helpful methodology, record some benchmarks and modify from that point of reference.
And all those who arrogantly (and mistakenly) think they know what is coming out of their speakers and reaching their ears will be very very surprised the first time they run a REW FR curve.
B.
All generalizations are false, Ben.
Just because we might not know what we are hearing, does not mean that applies to everyone. I have measurements of speakers and speaker systems that were tuned entirely by ear (not mine) and they hit their targets better than most I've seen. Most people can't do that, but with training and experience some can and do.
I like to measure, it's important and helps me see what I'm hearing, or often not hearing. But measurement isn't easy, we all know that. Sometimes you have to let your ears guide you to better measurements.
Recently I had a listening room that was near anechoic. Call it hypo-echoic. Placing the mic in different locations didn't make much difference other the the tweeter off-axis changes you would expect. I have not experienced that in any normal domestic setting.
The Moving Mic Technique has given me good, consistent results in the low end that match well with what I hear. Not useful for crossover work, of course, but very handy for general tonal balance and room EQ.
There's a whole thread about it here.
http://www.diyaudio.com/forums/multi-way/269864-caveman-speakers-troglodytes.html
I can post some more ETC and decay charts if you like.
http://www.diyaudio.com/forums/multi-way/269864-caveman-speakers-troglodytes.html
I can post some more ETC and decay charts if you like.
Bear, I only have experience with the acoustat x and model 3, both of which I can easily agree with you that they produce good bass as you describe. Jim recently rebuilt the X's servodrive amps...I'm hoping to keep enjoying them for some time still.
Also agree about ML CLS (little bass) and your other good points about measurements and in-room response.
Pano, I sure agree that good ears and good measurements can get us to the same place, even indoors.
I'v just given up trying to start with measurements indoors anymore.
I believe in divide and conquer...fix speaker outside ... then bring known speaker inside, and see what can be done with placement and room acoustics...
Below is an example of that process on latest project...
..first plot is 3-way active box tuned outside (for x-over at 100hz)
..2nd curve is with the box taken inside and crossed with sub.
2nd curve is close miced and not fully representative of listening position,but it still shows that good tuning outside has real value inside..
Also agree about ML CLS (little bass) and your other good points about measurements and in-room response.
Pano, I sure agree that good ears and good measurements can get us to the same place, even indoors.
I'v just given up trying to start with measurements indoors anymore.
I believe in divide and conquer...fix speaker outside ... then bring known speaker inside, and see what can be done with placement and room acoustics...
Below is an example of that process on latest project...
..first plot is 3-way active box tuned outside (for x-over at 100hz)
..2nd curve is with the box taken inside and crossed with sub.
2nd curve is close miced and not fully representative of listening position,but it still shows that good tuning outside has real value inside..
Attachments
If you have a deviation of less than 3db, don't worry about it; you won't hear it.
Also, you can have dsp chart saying all is adjusted, things have been tweeked, all is flat and perfect, but when you listen, the speaker will lack that 3d sound.
Don't think I'm anti DSP as I have audessey dsx in my theater room with 11 Realistic LX5II speakers, a speaker whose freq plot is as jagged as a hillbillie's smile and I love the job it does with them and my two subwoofers.
But,that's a different purpose. With two full range speakers that have an inherently flat response, what's the point of using DSP? What are you trying to achieve?
Also, you can have dsp chart saying all is adjusted, things have been tweeked, all is flat and perfect, but when you listen, the speaker will lack that 3d sound.
Don't think I'm anti DSP as I have audessey dsx in my theater room with 11 Realistic LX5II speakers, a speaker whose freq plot is as jagged as a hillbillie's smile and I love the job it does with them and my two subwoofers.
But,that's a different purpose. With two full range speakers that have an inherently flat response, what's the point of using DSP? What are you trying to achieve?
Olton, as you noted, pre-DSP era ESLs were perfectly able to project a very good sound field. I can imagine that with DSP you can fine tune and possibly improve the crossover between the ESL panel and the sub, but having no DSP need not be a show stopper at all. So you really have a lot of options.
One comment: you may want to lower your DAC budget; in this time and age, with, as an example, the fantastic Benchmark DAC3 just out, 4.5k seems a lot. I think you get more bang for your buck if you shift half of your DAC budget to your speaker budget. My 2cts.
Jan
Sent from my SM-T800 using Tapatalk
One comment: you may want to lower your DAC budget; in this time and age, with, as an example, the fantastic Benchmark DAC3 just out, 4.5k seems a lot. I think you get more bang for your buck if you shift half of your DAC budget to your speaker budget. My 2cts.
Jan
Sent from my SM-T800 using Tapatalk
This notion is really an important one; even the best DSP has correction limits. Furthermore, if the device"redigitizes" analog outputs, the results are often audible.
The best implementation I have heard did all of the adjustments in the digital domain, with low distortion drivers on the receiving end.
I use the DSpeaker 2.0 up to 160hz, and run Quad ESL63 full range, without EQ or DSP. As recommended above, a properly positioned, low distortion driver with sufficient amplification doesn't benefit much from corrections.
Sent from my Nexus 6 using Tapatalk
The best implementation I have heard did all of the adjustments in the digital domain, with low distortion drivers on the receiving end.
I use the DSpeaker 2.0 up to 160hz, and run Quad ESL63 full range, without EQ or DSP. As recommended above, a properly positioned, low distortion driver with sufficient amplification doesn't benefit much from corrections.
Sent from my Nexus 6 using Tapatalk
Yes, very good indeed, no DSP needed. The better your speakers and room, the less useful is DSP.
Sent from my mind, using Mindspeak.
I've done both. The current iterations of DSP, those that are affordable, trade one set of problems for others.
If you have a well designed ESL, many of the problems DSP was built to solve don't exist.
In my opinion DSP is best with complex crossovers and waveform management while the signal is still in the digital domain.
There maybe versions that are more effective, but they cost multiples of all the gear i use.
Until it's simpler, it's not better.
Sent from my Nexus 6 using Tapatalk
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.