SVS Ultra Bookshelf Speaker Review

[amirm]

usually obtaining high resolution distortion measurements of speakers requires very low noise in the mic, system, and room. I assume this is still true.

It is fortunately near-field measurements produce the same results as far-field (assuming the mic doesn't saturate). That is how I measure it. Indeed Dr. Klippel recommends putting the mic right in front of the speaker.

In addition, log CHIRP signal significantly increases signal to noise ratio and isolates harmonics without noise (within reason).

 

[amirm]

so in the CSD those ridges look to me like noise. Is that your interpretation. I am admittedly looking at this on a phone so I might be missing something, but I didn’t see an associated peak in the response. It would seem highly unlike to have such energy tails as that would make the speakers non-minimum phase

The tails at 1 and 5 kHz did look odd to me. As I noted in the review, I didn't have time to investigate it further. Somehow I assumed this speaker review wouldn't get much attention (!) and require a lot of investigation.

 

[thewas]

It is fortunately near-field measurements produce the same results as far-field (assuming the mic doesn't saturate). That is how I measure it. Indeed Dr. Klippel recommends putting the mic right in front of the speaker.

Only there is a problem in that method when you measure loudspeakers with several drivers so placing the mic somewhere close to the baffle can be at a lobe/ partial cancellation which will reduce the level accuracy and thus the distortion measurement precision. Also you won't fully measure the level influence of baffle (unless you use the NFS multi point measurements and mathematics) that's why many still do their distortion measurements at 1 or 2 meters (usually at least three times the baffle width).

 

[briskly]

anechoic simply means no reflections. No no room noise. This approach also has little to do with CSD which behaves the same on the Klippel as any other. It’s still dependent on a quiet room.

NSF is only relevant to the obtaining of the frequency response minus room reflections and extrapolating the full 360 degree sphere of sound. Temporal and non-linear issues don’t get special treatment. I’m unaware of any way to do that.

The field separation module and a second scanning layer are only required to isolate reflections at low frequencies. Klippel's NFS necessarily determines phase information by fitting a multipole expansion to the measurements, an analytic reconstruction of the sound field. The phase accuracy of near-field holographic measurement is a key selling point of the Klippel NFS over conventional distant anechoic methods that contend with phase over long distances.

I assume this is still true.

The near-field measurement helps. We would still see the negative effect of noise on fitting near the woofer cutoff, which is why Amir still performs several averages with the NFS. At least, that's what he said a while ago.

 

[MerlinGS]

text deleted to comply with @Thomas savage's request supra.

 

[Matthew J Poes]

The tails at 1 and 5 kHz did look odd to me. As I noted in the review, I didn't have time to investigate it further. Somehow I assumed this speaker review wouldn't get much attention (!) and require a lot of investigation.

It’s a review of SVS. Any review of Klipsch or SVS gets a ton of attention. You can’t speak Ill of these brands either or else the army of supports will sacrifice you to their gods.

by far the most popular videos I’ve ever done involved SVS speakers.

 

[Matthew J Poes]

The field separation module and a second scanning layer are only required to isolate reflections at low frequencies. Klippel's NFS necessarily determines phase information by fitting a multipole expansion to the measurements, an analytic reconstruction of the sound field. The phase accuracy of near-field holographic measurement is a key selling point of the Klippel NFS over conventional distant anechoic methods that contend with phase over long distances.


The near-field measurement helps. We would still see the negative effect of noise on fitting near the woofer cutoff, which is why Amir still performs several averages with the NFS. At least, that's what he said a while ago.

Im not sure my point is being understood. I understand what it does. But that has little to do with distortion or CSD. Let’s say we are trying to measure a speakers distortion and the 3rd harmonic is at -65dB. The speaker output is 75dB at 1 meter. That is 85dB at 1/3 meter and just the harmonic lies at 20dB. That is at or above the noise floor of most rooms and many Measurement mics. What if we want to look at the 5th harmonic. At some point the sweep method and distance isn’t sufficient help and the method is no longer valid. The highest S/N for distortion is usually the stepped sine method and that is still only good to about 10dB below the noise floor of the room. I had been previously told that anything below the mic noise floor is not valid.

 

[briskly]

It is perfectly relevant to the CSD in that the CSD is only a visualization of the time-frequency plane. The NFS w/ SFS is perfectly capable of determining the impulse response independent of external reflections. On that note, what parameters were used for the CSD?
We can thoroughly correlate periodic components through random noise, though the returns are logarithmic wrt time. It would be faster to test at a higher level that would produce more significant nonlinear distortions anyway.

 

[richard12511]

Somehow I assumed this speaker review wouldn't get much attention (!) and require a lot of investigation.

I expected it. SVS is one of the most popular ID companies. I'd expect a similar(though smaller) response if you reviewed another Ascend or Chane.

 

[amirm]

I expected it. SVS is one of the most popular ID companies. I'd expect a similar(though smaller) response if you reviewed another Ascend or Chane.

For subs sure. But I didn't think there was a lot of following for their regular speakers.

 

[Matthew J Poes]

It is perfectly relevant to the CSD in that the CSD is only a visualization of the time-frequency plane. The NFS w/ SFS is perfectly capable of determining the impulse response independent of external reflections. On that note, what parameters were used for the CSD?
We can thoroughly correlate periodic components through random noise, though the returns are logarithmic wrt time. It would be faster to test at a higher level that would produce more significant nonlinear distortions anyway.

But that simply allows a reduction of room reflections not system noise. What about the process lowers the noise floor of room or mic (or random system noise)?

 

[whazzup]

It's not the credentials that are under attack, it's the fact that he doesn't seem to understand that they are irrelevant, as they do not include the (superhuman) ability to nullify (mostly psychological) biases that are inherent in every sighted and subjective test. His skill for hearing low level distortion and other artifacts doesn't relate in any way to the required skill of controlling for and diminishing those heaps of intervening factors unrelated to actual sound.

The objective comparison between SVS Ultra and M106 doesn't suggest anything that would knock the former's rating two panthers down, so you can fill the rest of the thread with countless hypotheses about why he heard what he heard, but considering that blind testing speakers is not on the agenda, those hypotheses will remain untested. So what's the point in going on with that?

To reply your response on my comment, but not aimed at you personally, here are a set of 4 simple questions with yes/no answers:

1. Do you accept that a professional odor tester can detect odors in perfumes / fragrances that regular, untrained humans cannot?
2. Do you accept that a professional pantone color expert can detect minor color variations that regular, untrained humans cannot?
3. Do you accept that a professionally trained marathon runner can run for longer distances that regular, untrained humans cannot?

4. And finally, do you accept that a professionally trained listener can detect acoustic variations, AND have reduced biases during sighted listening, that regular, untrained humans cannot?


For anyone with a bone to pick with Amir's subjective opinions, in answering the above 4 questions, I believe we will know on a bias spectrum, where the person stands in his trust level of anything produced by Amir.

But note that yes, I do agree that audio science is imperfect and currently cannot reconcile 100% the results of subjective tastes to graphs and charts.

Sorry @Thomas savage if this was not appropiate. Just think that there's a better way of holding (and reading) discussions without having to wade through the 100th comment attacking what Amir heard/preferred. Unless maybe that's a snakey way of increasing forum readership....

 

[jazzendapus]

4. And finally, do you accept that a professionally trained listener can detect acoustic variations, AND have reduced biases during sighted listening, that regular, untrained humans cannot?

Pretty sure I'm addressing this very point in the post you're replying to - I don't see how the skills of trained listener are related to the skill of reducing biases, what is the mechanism for their relation? So a trained listener might pick on a real issue if such exists, but the final subjective impression/assessment is a sum of a real issue with addition of 0% to 99.99% unrelated "noise" (not in the literal, acoustic sense, of course) that caused him to dislike the speaker. The fact that we can't properly determine the amount of that noise and can't separate it from the detection of a real acoustic issue is the essential problem.
It's weird that this needs to be repeated on this forum of all places, this realization is exactly why I believe most people are here to begin with and not on whathifi and stereophile forums and the like.

 

[amirm]

But following industry standards is not science.

Don't get confused. We are not inventing science here. We are just evaluating an audio product. It is testing. Both objective and subjective. Your doctor is not inventing science either when he asks you what is wrong with you, tries medication, etc. Learn how the real world works and don't throw big words around that show lack of appreciate for the task at hand.

For now I think we won't agree, because again you are not doing science, and I thought wrongly that toi were. I think a lot of people are ok with that, it annoyed me, but it's ok, that's my problem.

Definitely your problem because you are not familiar with how the world turns in real life.

There are no infinite resources. Not here. Not in the industry. We take calculated risks to get quick answers. Wanting zero possibility of error would invalidate all the research we are relying on. After all, how would tens of people be good enough to estimate what billions of people would prefer in a speaker?

Know the limits of your knowledge. Don't keep protesting.

 

[amirm]

Pretty sure I'm addressing this very point in the post you're replying to - I don't see how the skills of trained listener are related to the skill of reducing biases, what is the mechanism for their relation? So a trained listener might pick on a real issue if such exists, but the final subjective impression/assessment is a sum of a real issue with addition of 0% to 99.99% unrelated "noise" (not in the literal, acoustic sense, of course) that caused him to dislike the speaker. The fact that we can't properly determine the amount of that noise and can't separate it from the detection of a real acoustic issue is the essential problem.

By your logic, the odds of your doctor being right about what illness you have is also 0 to 99.99%. Why would you go to him then?

Answer is that your doctor is likely correct. And trained listener is likely correct. Both can make mistakes. If they make enough of them, then they are not qualified to do their jobs. But if they don't, then they earn their unique status of being above the general population. What gets people to pay for their expertise is precisely because they give far more correct answers than wrong ones.

I asked my doctor once why I would not suffer from the side effects of a medication as listed by the drug company. He said he has prescribed that medication to hundreds of patients and based on that experience, that side effect is very unlikely to occur. Can he guarantee it? No. But is his guess more correct than any lay person's? Of course. This is the power of training and experience. It makes you far more right than wrong.

Did my doctor ever run a placebo controlled test of that drug? Of course not. He did not therefore follow the "scientific method" to determine the odds of side effects for that medication. Yet, I bet on his experience and he absolutely was right.

I give you these examples because I can't make you understand what a trained audio listener is until you become one. After tons of training, all of a sudden a curtain is removed and reality becomes so clear to you. You will make mistakes some of the times but vast majority of time you amaze the people around you by identifying what is wrong.

It's weird that this needs to be repeated on this forum of all places, this realization is exactly why I believe most people are here to begin with and not on whathifi and stereophile forums and the like.

This is why the discussion is continuing We have people like you who assume that what I say is all "noise" to use your vocabulary. You are putting yourself with no experience or training in my shoes, wanting us to be equivalent. If you were doing my job, I wouldn't trust what you said subjectively whatsoever.

Now if you showed me your ability to be a critical listener, have industry credentials and demonstrated ability to find impairments in audio products, then I would absolutely put weight on your opinion. This is what I have done yet to keep acting flippant about it. You have no evidence or argument in your favor, just a plea for agreement. Talk about non-scientific method...

 

[whazzup]

Pretty sure I'm addressing this very point in the post you're replying to - I don't see how the skills of trained listener are related to the skill of reducing biases, what is the mechanism for their relation?

At the risk of Tho(r)mas' hammer....

In the context of a marathon runner, so what you're asking is, what is the specific exercise/training the runner did to for example, increase his lung capacity? I wouldn't know if Amir actually hung weights on his ear lobes with The Final Countdown playing in the background while training his hearing, but sure, I'll admit I do not know and did not care (note that I've only started playing audio measurements recently, way more concepts to learn). Unless I misunderstood your comment.

So a trained listener might pick on a real issue if such exists, but the final subjective impression/assessment is a sum of a real issue with addition of 0% to 99.99% unrelated "noise" (not in the literal, acoustic sense, of course) that caused him to dislike the speaker. The fact that we can't properly determine the amount of that noise and can't separate it from the detection of a real acoustic issue is the essential problem.
It's weird that this needs to be repeated on this forum of all places, this realization is exactly why I believe most people are here to begin with and not on whathifi and stereophile forums and the like.

So you're saying because you don't understand how the listening training was carried out, so you have to object to Amir providing his subjective opinions? Like if you didn't understand how a marathon runner trained himself, so you have to object to his winning a race?

Apologies in advance, I honestly re-read this part a few times and still can't say I clearly understood what you mean to say.

 

[jazzendapus]

By your logic, the odds of your doctor being right about what illness you have is also 0 to 99.99%. Why would you go to him then?

If he would say something like "you look kinda pale, you probably have an X disease, take some of those pills" instead of sending me to do proper tests to determine what I actually have, then no, I definitely wouldn't go to that doctor again.
And this scenario is pretty much exactly what you pulled here.

 

[TimVG]

Instead of focussing on a single score metric it would be wise to look at the data as summarized by @hardisj in post #242

The only conflict is between the listening impression and the score - not the objective data in itself. Would some of you prefer Amir lies about his impression to suit te score? How would that help anyone along?

 

[justcheeze]

Prime Pinnacle Tower measurements, the retained energy off-axis around 4kHz I’ll for sure make it sound bright.

Yeah I've seen that. I personally haven't heard the Prime Pinnacles yet, but I've been reading that SVS has managed to tone down their tweeter some with the Pinnacles.

In comparison to the Prime Towers here, as you can see - this would sear your ear drums out. This is old Klipsch RF series harsh.

SVS needs to go back to the drawing board instead of rehashing their tweeter and crossover network for their speakers. It was a surprise to me that they even released the Pinnacle Towers in the first place without addressing the elephant in the room first.

 

[justcheeze]

I expected it. SVS is one of the most popular ID companies. I'd expect a similar(though smaller) response if you reviewed another Ascend or Chane.

Frankly speaking, their subs are no longer the king of value and performance as they once were either. They've long been beaten by Rythmik, Hsu and even their former owner who has now formed PSA.

I was critical of them on AVS and I was called a hater, I told SVS that they're going down the route of Apple if they're following their trend of clinging on to blind non-brand agnostic customers and not chasing for excellence like they used to do. Not saying they no longer make good subs, they still do but they're not as great as the price they're asking for.