Of course you can´t,, you are "indigent" 🙄
Why would anybody?
Do your own homework.
In any case I suspect just one more case of "measurements don´t matter", so prevalent in DIY Audio Forum.
You are wrong.
NFS measures , period.
Results may be good, bad, average.
Speaker sound won´t change a bit before/after measurements, we are not talking Quantum Mechanics here.
If you expect that, you are in for a big disappointment.
straw man argument, nobody said that.
Please keep the discussion serious.
Not the only one who noticed that
Member
Joined 2003
To answer the original question, I think what most will want to pay is around $100 for a Klippel scan service, though the real price may have to be a fair bit higher to have any chance of paying off that $100k investment and the loss of a room in your home to house the contraption, and your time investment to learn how to use the thing and process people speakers and ship them back, etc.
The benefit of Klippel in my mind is rapid prototyping, and reliable repeatable measurement sets with automation. I think it would be neat to have that direct comparison of my DIY efforts to other well known designs, but I don't know that I'm really willing to pay all that much just for that. As a DIYer, I'm really only interested in my own perception of my designs, and shipping something to get measured and sent back provides no benefit in speed to a slow manual turntable process in my own home, it would be more of a final verification type of process, if anything. DIY measurement turntables can be built for a fraction of the cost of Klippel NFS, it's just a bit more work to set up and process far field, near field, port response, etc. I'm probably more likely to invest in a DIY turntable setup than I am to ship my speakers for Klippel analysis.
The benefit of Klippel in my mind is rapid prototyping, and reliable repeatable measurement sets with automation. I think it would be neat to have that direct comparison of my DIY efforts to other well known designs, but I don't know that I'm really willing to pay all that much just for that. As a DIYer, I'm really only interested in my own perception of my designs, and shipping something to get measured and sent back provides no benefit in speed to a slow manual turntable process in my own home, it would be more of a final verification type of process, if anything. DIY measurement turntables can be built for a fraction of the cost of Klippel NFS, it's just a bit more work to set up and process far field, near field, port response, etc. I'm probably more likely to invest in a DIY turntable setup than I am to ship my speakers for Klippel analysis.
I spent countless hours measuring with the usual DIY techniques, I also paid outside consultants for what turned out to be substandard measurements. Time and money wasted - lessons learned.
A couple of hours spent at a professional facility with a tech that supports his family doing this and I’m never looking back. We’re extremely lucky to have access to a very well equipped lab with nice knowledgeable people an hour away by car for our commercial efforts, and yes the sq is dramatically better.
If the DIY community, with its culture of sharing and free access to software like VCad, could validate their designs with in the depth analysis like Klippel offers we would all benefit.
A couple of hours spent at a professional facility with a tech that supports his family doing this and I’m never looking back. We’re extremely lucky to have access to a very well equipped lab with nice knowledgeable people an hour away by car for our commercial efforts, and yes the sq is dramatically better.
If the DIY community, with its culture of sharing and free access to software like VCad, could validate their designs with in the depth analysis like Klippel offers we would all benefit.
I'll chime in...
I'm dealing with some very heavy family stuff. I can't take on anything else that I don't already have on hand (which is enough to last me quite a while already). I'm doing the best I can to review and return the stuff I do have on hand and won't be accepting anything else until I have a better understanding of what the future holds for me. I'll never stop. But I'm going to be out of commission for a bit... maybe one week... maybe 10 months. No idea at this point.
I'll also add this:
I've tested a LOT of DIY designs and professional designs. The number of designers who are shocked at just how poor their accuracy of measurements below about 800Hz (or so) is high. Standard gated measurements aren't good enough. And many people don't have the means to perform alternative methods (like outdoor groundplane, NF summing, etc). Then there's the resolution of those measurements which is a separate convo. Many have said "my measurements don't show that" and after we dive into why, we understand their measurement system just isn't good enough and/or they don't have the knowledge of how to get useful data from the standard gated/quasi-anechoic methods. Not a knock at anyone. Just relaying actual experience and conversations I've had.
I've been a DIY'r for a long time. And I've been measuring and publishing data for over a decade. The benefit of the NFS is that you get true anechoic response with high resolution (1/20th octave is easy). I don't know a single DIY'r setup that can touch it. Including my own, which involved 140 measurements outdoors in the groundplane and on a 6 foot deck. You should see some of my measurement rigs I built trying to get that kind of resolution and response accuracy...
Last edited:
I am not sure if I am in the wrong boat, but my gut feeling is that a spinorama from ARTA with a motorized rotary table will provide the same data as the Klippel system. Can someone comment on this? Am I missing something?
The secret sauce of the Klippel is that it can remove the room from the measurements, so you get anechoic, high resolution results in a regular room.
You're missing an anechoic chamber or another way to get rid of reflections 😉
The Klippel "removes" the room by taking near-field measurements. I don't see why this can't be done with other systems. What am I missing?
It is nearish-field, I wouldn't call it nearfield in the sense we use where the mic is right at the DUT. The point being, the box diffraction is still captured just as it is in our far-field measurements.
The way it removes the room is not due to the mic distance, it is that it processes the data to remove the room. I imagine there must be an initial calibration where it measures some standard DUT, and then calculates the solution that would be true for all angles, any "leftover" response SPL is the room. Just my guess on how it does it.
I don't know exactly how the Klippel does what it does, but I have a thought experiment that can illustrate how it is perhaps working. Imagine you have a DUT that is omni, it has the same response in all directions. Now imagine you take many measurements at defined angles. They would all look different, but you could also calculate a solution that is correct for all of them after the "extra" energy is removed. That extra energy is room. It's processing the data with an algorithm to do that. Nothing to do with being near-field.
Yeah, also my best wishes for the family side of things of course! 👍
Those things are never nice 🙁
As for the second part, it really depends what you're looking for.
Developing a system is totally different animal than reviewing systems.
Klippel NFS shows it all nice in one nice ready package, which is neat and awesome.
There are however many other methods to get the same data.
Some of those methods won't show it as nicely as Klippel does, but for someone with enough knowledge and experience that's totally not relevant at all.
There is a difference in showing all the data (on a nice way) and developing a good speaker.
A good example of this, is showing the lower frequency part of a speaker all at once with the mid-high frequency part.
The exception here are like (big) line arrays, as well as some (active) cardiods, but for the most part, everything below 200-300Hz is extremely predictable. This has also been shown in a lot of literature, including your own measurements as well as what one would expect in the region were the woofer acts purely in the "pressure range".
(said on a simplified way)
A very similar case can be made for the directivity of speakers
After a while it's more like pixel peeping in combination of just how nice an complete a certain graph looks.
There is nothing wrong with that at all.
But those things can also be found in like near field burst decay/waterfall/impedance/distortion measurements.
It's just a very convoluted way of presenting data.
But to be very honest, Klippel is also not perfect.
Certain interference issues or resonances don't show up well at all, or it's really hard to figure out were they are coming from.
Because of that I have already seen (very) poor advice given by certain users when it comes down to EQ'ing and such.
Along those lines probably yes.
The thing is, there are many methods that work pretty well.
The win with a Klippel system, is that it does it all for you.
It's basically installing the speaker, hit start, and get a cup of coffee or lunch/dinner.
In some cases that's all nice and all, but it doesn't always give you much more information than you already knew when designing a speaker. The exception is when reviewing another speaker.
Since you don't know what parameters as well as issues, they started with.
It takes a ton of super precise measurements and serious math to separate room and speaker in the near-field.
'Near-field' doesn't have the common DIY meaning of "measure close and gate".
It's worth watching the entire first vid that tktran303 posted in #21.
The interview with Bellmann from Klippel.
It explains how vastly different the NFS is (imo), from anything we can do at home, even outdoors with spinorama ( lord knows I've made thousands of measurements that way.)
Here's an interesting post about what the Klippel Nearfield Scanner (NFS) does: https://www.audiosciencereview.com/...nderstanding-how-the-klippel-nfs-works.13139/
I don't know if this is the whole story. However, it seems to me the NFS does two things:
- Measure spinorama at high angular resolutions, in near(ish) field
- Fit acoustic model to measured spinorama, and use this model to predict the acoustic response at any position relative to the speaker
Doing the measurements in the near(ish) field will hide the room echoes relative to the direct sound of the speaker.
Well said. I truly respect DIY measurements and the extraordinary efforts the community puts in to them....but Klippel NFS is a whole 'nuther level imho.
It probably does require a re-calibration if you move it around or move furniture in the room.
Near-field also removes the box and diffraction. Klippel NFS is not doing that.
Replace the "serious math" with "serious and tedious amounts of programming"
The math behind those systems is pretty straight forward (at least for some here).
It's just an awful lot of work.
The use of "super" precise measurements here is kind of strange.
One can make "super" precise measurements on a standard wooden stick with a (well) calibrated microphone?
It's not automated, but that doesn't make it less precise, only less convenient.
It could be done as a DIY project but it would be a large one involving a substantial amount of programming and maths. The end result would be the ability to take a complete set of measurements faster and likely more repeatably. This is valuable to industry but less so to DIY hobbyists. The processing involved will lower the accuracy compared to a conventional measurement in a large open space. Possibly by not enough to bother about (I haven't looked) but accuracy is not why some in industry consider the high price worth paying. It is the ability to take complete measurements quickly and in convenient places rather than inconvenient ones like anechoic chambers.