• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Heavenly Soundworks

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,438
Location
NYC
Can you try it with their other graph showing the individual driver responses? That one is less smooth.
Ah yes, I see what you mean now. The individual driver responses show many of the same deviations as the NRC measurements:
Five17 nearfield vs NRC.png


But that just makes it weirder that they would get such a smooth on axis when the nearfield data is so off! There's no way I can think of to get the claimed flat frequency response from summing the above nearfield responses. The speaker would need extensive DSP to correct it, but there's no indication that's happening given the NRC results.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,244
Likes
11,477
Location
Land O’ Lakes, FL
Ah yes, I see what you mean now. The individual driver responses show many of the same deviations as the NRC measurements:
View attachment 122232

But that just makes it weirder that they would get such a smooth on axis when the nearfield data is so off! There's no way I can think of to get the claimed flat frequency response from summing the above nearfield responses. The speaker would need extensive DSP to correct it, but there's no indication that's happening given the NRC results.
What, you’ve never heard of 10/1 octave smoothing :p
 
OP
richard12511

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
Ah yes, I see what you mean now. The individual driver responses show many of the same deviations as the NRC measurements:
View attachment 122232

But that just makes it weirder that they would get such a smooth on axis when the nearfield data is so off! There's no way I can think of to get the claimed flat frequency response from summing the above nearfield responses. The speaker would need extensive DSP to correct it, but there's no indication that's happening given the NRC results.

Maybe the individual driver measurements are with internal DSP turned off(these are DSP speakers iirc)? But then that would mean soundstage somehow measured a version with no DSP.
 

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,438
Location
NYC
Maybe the individual driver measurements are with internal DSP turned off(these are DSP speakers iirc)? But then that would mean soundstage somehow measured a version with no DSP.

Yeah that was my thought. It really makes no sense. I mean, Soundstage network measured all three sound presets, and the only real change was in the bass amount. So unless the DSP is in some different system from the presets and somehow wasn't activated for during the NRC measurements...

Edit: Also, the deviation from linearity is pretty disappointing for the price @90dB, unless there's some DSP going on with that as well.

1617659696665.png

Credit: https://www.soundstagenetwork.com/i...&catid=77:loudspeaker-measurements&Itemid=153
 

nerdoldnerdith

Senior Member
Joined
Sep 10, 2020
Messages
495
Likes
690
Location
Chicago
I got to hear these things in real life. They didn't sound that bad. These measurements must be before DSP has been applied. C'mon man!
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,244
Likes
11,477
Location
Land O’ Lakes, FL
Yeah that was my thought. It really makes no sense. I mean, Soundstage network measured all three sound presets, and the only real change was in the bass amount. So unless the DSP is in some different system from the presets and somehow wasn't activated for during the NRC measurements...

Edit: Also, the deviation from linearity is pretty disappointing for the price @90dB, unless there's some DSP going on with that as well.

View attachment 122286
Credit: https://www.soundstagenetwork.com/i...&catid=77:loudspeaker-measurements&Itemid=153
The presets are just for loudness compensation based on listening level.
 
OP
richard12511

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
I got to hear these things in real life. They didn't sound that bad. These measurements must be before DSP has been applied. C'mon man!

But isn't it weird that soundstage would get a version for review that doesn't have DSP applied? I don't think there's a way for the end user to apply the internal DSP.
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,755
Likes
16,208
Imho nothing is weird, but the first impression of another overpriced audiophoolery company giving incomplete and incorrect measurements was confirmed. Those are active DSP loudspeakers, you can't just have the crossover without the other tonal correction filters unless you on purpose want to. Also if measurements were wrong the manufacturer would have reacted to that review and as written before the measurements they show on their website are not matching each other.
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,087
Location
.de, DE, DEU
Maybe the individual driver measurements are with internal DSP turned off(these are DSP speakers iirc)? But then that would mean soundstage somehow measured a version with no DSP.
Edit: Also, the deviation from linearity is pretty disappointing for the price @90dB, unless there's some DSP going on with that as well.
Imho nothing is weird, but the first impression of another overpriced audiophoolery company giving incomplete and incorrect measurements was confirmed. Those are active DSP loudspeakers, you can't just have the crossover without the other tonal correction filters unless you on purpose want to. Also if measurements were wrong the manufacturer would have reacted to that review and as written before the measurements they show on their website are not matching each other.

Wow! I am also very surprised about the results.

... and I love to speculate. So here is my speculation on "WTF happened there?"

The first thing I noticed is that a manufacturer whose cheapest product costs $10k uses a $300 measurement system for its PR measurements.
From the manual:
1617703993709.png

I have nothing at all against Dayton's measurement system, it's a great overall package. However, nowhere in the manual is a description of how to make reliable two-channel measurements with this system.
Something like this:
1617707792837.png

Without this, possible frequency response and phase errors of the measurement amplifier (and audio interface) are not compensated - which is a possible source of error in loudspeaker development. For professional work, therefore, other systems are usually used.


But there are bright spots, namely that SoundStage's measurements are very likely correct.
If you take the on and off-axis measurements given by HeavenlySoundworks in the manual and scale them to fit the SoundStage measurements, they are almost identical except for the smoothing.
1617708356000.png

It just seems that HeavenlySoundworks is not able to make gated measurements and therefore works with 1/1oct smoothing and truncates the measurements at 700Hz to hide possible room resonances.

The HeavenlySoundworks owners are certainly not going to heaven.
 
Last edited:
OP
richard12511

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
Wow! I am also very surprised about the results.

... and I love to speculate. So here is my speculation on "WTF happened there?"

The first thing I noticed is that a manufacturer whose cheapest product costs $10k uses a $300 measurement system for its PR measurements.
From the manual:
View attachment 122366
I have nothing at all against Dayton's measurement system, it's a great overall package. However, nowhere in the manual is a description of how to make reliable two-channel measurements with this system.
Something like this:
View attachment 122370
Without this, possible frequency response and phase errors of the measurement amplifier (and audio interface) are not compensated - which is a possible source of error in loudspeaker development. For professional work, therefore, other systems are usually used.


But there are bright spots, namely that SoundStage's measurements are very likely correct.
If you take the on and off-axis measurements given by HeavenlySoundworks in the manual and scale them to fit the SoundStage measurements, they are almost identical except for the smoothing.
View attachment 122373
It just seems that HeavenlySoundworks is not able to make gated measurements and therefore works with 1/1oct smoothing and truncates the measurements at 700Hz to hide possible room resonances.

The HeavenlySoundworks owners are certainly not going to heaven.

Even with 1/1 smoothing, I don't see how they can get that perfect of an on-axis measurement.
 

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Fresh NRC measurements show poor engineering as it was feared already from the cheap drivers for that price:
https://www.soundstagenetwork.com/i...&catid=77:loudspeaker-measurements&Itemid=153


From the measurements page:
Microphone measuring position: between tweeter and midrange


Is that the correct axis for this speaker? In general, I would assume "no". I can't access the company's product page at the moment to verify.
I don't expect it would make a huge difference at the crossover for this speaker but if that's the wrong measurement axis then that's a problem, too.
 

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,438
Location
NYC
Wow! I am also very surprised about the results.

... and I love to speculate. So here is my speculation on "WTF happened there?"

The first thing I noticed is that a manufacturer whose cheapest product costs $10k uses a $300 measurement system for its PR measurements.
From the manual:
View attachment 122366
I have nothing at all against Dayton's measurement system, it's a great overall package. However, nowhere in the manual is a description of how to make reliable two-channel measurements with this system.
Something like this:
View attachment 122370
Without this, possible frequency response and phase errors of the measurement amplifier (and audio interface) are not compensated - which is a possible source of error in loudspeaker development. For professional work, therefore, other systems are usually used.


But there are bright spots, namely that SoundStage's measurements are very likely correct.
If you take the on and off-axis measurements given by HeavenlySoundworks in the manual and scale them to fit the SoundStage measurements, they are almost identical except for the smoothing.
View attachment 122373
It just seems that HeavenlySoundworks is not able to make gated measurements and therefore works with 1/1oct smoothing and truncates the measurements at 700Hz to hide possible room resonances.

The HeavenlySoundworks owners are certainly not going to heaven.

As shown in an earlier post of mine, the on axis does not line up at all. The red line is what they show in the manual, scaled to match soundstage's scaling.

Five17 comp.png


The off axis and nearfield driver measurements show clear similarities, but I simply have no clue how they arrived at that on-axis. They say it is a quasi-anechoic measurement and I see no reason why it would vary so much. Summing the woofer, passive radiator, and mid drive would be fairly straightforward up to about 1khz and would definitely not give the above red result lol.
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,087
Location
.de, DE, DEU
Even with 1/1 smoothing, I don't see how they can get that perfect of an on-axis measurement.
As shown in an earlier post of mine, the on axis does not line up at all. The red line is what they show in the manual, scaled to match soundstage's scaling.

I didn't go into that again, since you had already pointed out the discrepancy - absolute agreement from me.
For me, the FR looks like a mix-up with another speaker ;)


I tried to normalize the SoundStage on- and off-axis measurements to be able to better judge the radiation behavior of the speaker - since all FR have the same color, this did not work without errors.
With 1/3 smoothing and if you ignore the "scan errors", the directivity of the speaker is not that bad - to say something positive (if you ignore the purchase price).
1617722128855.png
 

Heavenly Soundworks

New Member
Audio Company
Joined
Jun 7, 2021
Messages
1
Likes
3
Location
USA
Hello everyone, Heavenly Soundworks here.

Thought it would be good to answer some of the questions posed in this thread.

You are correct that there are some major discrepancies between the measurements of our system done at the NRC and our in-house measurements. For an explanation of those (and the improvements that we’ve made since) please see the article that we published on our website back in April, in response to the Soundstage review.
https://www.heavenlysoundworks.com/post/in-pursuit-of-perfection

Hopefully, this clears up some of the questions you have regarding our system’s performance and our integrity. By no means are we trying to be deceitful, we are sharing our measurements and these insights to back up what we claim.

If anyone still has further questions, please do not hesitate to reach out to us directly at [email protected]

Thanks!
 
OP
richard12511

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
Hello everyone, Heavenly Soundworks here.

Thought it would be good to answer some of the questions posed in this thread.

You are correct that there are some major discrepancies between the measurements of our system done at the NRC and our in-house measurements. For an explanation of those (and the improvements that we’ve made since) please see the article that we published on our website back in April, in response to the Soundstage review.
https://www.heavenlysoundworks.com/post/in-pursuit-of-perfection

Hopefully, this clears up some of the questions you have regarding our system’s performance and our integrity. By no means are we trying to be deceitful, we are sharing our measurements and these insights to back up what we claim.

If anyone still has further questions, please do not hesitate to reach out to us directly at [email protected]

Thanks!

Really cool! Thanks for the response. You at least have the right goal, even if you didn't get it first try. I really respect that you admit you initial measurements were incorrect and then went back and corrected based on the soundstage results.

I can understand where it's hard for a new small business. Anechoic chambers are expensive, though IMO I'd be buying a Klippel NFS before building an anechoic chamber. Better(imo) and cheaper :). NFS is still expensive for a startup, though. I would take a look at @napilopez 's guide for getting anechoic measurements by hand. It's a slow and cumbersome process, but his results match very well to anechoic results.
 
Top Bottom