NHT C3 Speaker Review (Sample #2)

[amirm]

Right after I post the review of the NHT C3 speaker, the owner noticed that there was a dent in the tweeter [EDIT: mid-range]:

You can see the dent at 7:00 O'clock position. The owner requested an exchange from Amazon and another sample arrived. The new sample is perfect in that regard. So there went another 3 hours of measurements and here are the results:

I have taken the on-axis response of the last sample in bold red, and overlaid it on top of our new spin data. As you see, the response is essentially identical other than some flattening of the peak above 10 kHz. That could be the difference in samples. Or, it could be me having the speaker pointed slightly differently. I don't have a fixture that locks each speaker into position so how I have this speaker may be slightly different than the last one. Regardless, the difference is quite small and above many people's hearing bandwidth.

So it looks like the damage to the tweeter did not change its frequency response and tonality. In the interest of time and resources, I did not run a distortion test so maybe there was some impact there.

At high level, we see the high consistency of the Klippel NFS measurement system and by the same logic, the consistency of NHT speaker drivers, assuming that peaking is not related to them.

------------
As always, questions, comments, recommendations, etc. are welcome.

Appreciate any donations using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

 

[trl]

So it looks like the damage to the tweeter did not change its frequency response and tonality. In the interest of time and resources, I did not run a distortion test so maybe there was some impact there.

I expected zero difference, given that the bent was on the edge and the driver seems to be midrange related, but looks like the difference is quite big. Basically, now the trebles freq. response seems quite flat to me.

 

[amirm]

As I explained, I don't think the response difference is related to the ding.

 

[OdysseusG]

A while back I was chatting with some reps from a very audiophile approved electronics brand. I asked about what speakers they liked to test with and it led to a story about a dented tweeter on their listening rig. I think they popped it back out and didn't notice an obvious difference in sound, so they listened some more. Then, being a manufacturer and having access to the equipment they measured the two speakers to compare dented tweeter to normal and couldn't find a meaningful difference. I think the takeaway was kind of a "huh, go figure. Nobody's gonna believe this, especially with all the trouble the speaker guys go through" so they replaced the parts anyway even if only for appearance's sake, especially because of potential factory tours and whatnot.
I was skeptical, but why make that kind of story up? After seeing the results of ASR measurements of some of their gear (...that I own :/ ) I'm even more skeptical of their measurements, but that's another story.

 

[trl]

As I explained, I don't think the response difference is related to the ding.

Thanks, I thought the same myself too. However, it's great that this is a better sample.

 

[Beave]

Wasn't the ding in the midrange dome, not the tweeter?

 

[infinitesymphony]

Great evidence of both the Klippel's and @amirm's speaker measurement consistency and the case that a single small ding or crease in a dome-type driver is unlikely to change its response in a significant way. I also have a pair of speakers where one tweeter has a crease and I can't tell them apart.

 

[amirm]

Wasn't the ding in the midrange dome, not the tweeter?

Ah, good point!

 

[Robbo99999]

Right after I post the review of the NHT C3 speaker, the owner noticed that there was a dent in the tweeter [EDIT: mid-range]:

You can see the dent at 7:00 O'clock position. The owner requested an exchange from Amazon and another sample arrived. The new sample is perfect in that regard. So there went another 3 hours of measurements and here are the results:

View attachment 116271

I have taken the on-axis response of the last sample in bold red, and overlaid it on top of our new spin data. As you see, the response is essentially identical other than some flattening of the peak above 10 kHz. That could be the difference in samples. Or, it could be me having the speaker pointed slightly differently. I don't have a fixture that locks each speaker into position so how I have this speaker may be slightly different than the last one. Regardless, the difference is quite small and above many people's hearing bandwidth.

So it looks like the damage to the tweeter did not change its frequency response and tonality. In the interest of time and resources, I did not run a distortion test so maybe there was some impact there.

At high level, we see the high consistency of the Klippel NFS measurement system and by the same logic, the consistency of NHT speaker drivers, assuming that peaking is not related to them.

------------
As always, questions, comments, recommendations, etc. are welcome.

Appreciate any donations using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Ah, it's very cool just to see two separate units tested and overlaid on each other, and I'm pleased to see even a lot of the micro detail of both the tiny & larger peaks being almost perfectly mirrored.....which provides strength that intersample variation is not huge and that Anechoic EQ can be applied on the basis of your measurements to optimally flatten out frequency responses, even when it comes to the micro detail. I'm inclined to think the difference in the +10kHz was due to the positioning of the speaker on the Klippel as you mentioned, because the micro detail of the speaker is retained in the other parts of the frequency range when comparing the two. Hopefully this reliability between different units extends to other manufacturers and models because I see your Anechoic measurements as fundamentally useful in using EQ to create "perfectly" flat speakers......I did this with my JBL 308p based on your measurement and I do like the results.

 

[YSC]

Ah, it's very cool just to see two separate units tested and overlaid on each other, and I'm pleased to see even a lot of the micro detail of both the tiny & larger peaks being almost perfectly mirrored.....which provides strength that intersample variation is not huge and that Anechoic EQ can be applied on the basis of your measurements to optimally flatten out frequency responses, even when it comes to the micro detail. I'm inclined to think the difference in the +10kHz was due to the positioning of the speaker on the Klippel as you mentioned, because the micro detail of the speaker is retained in the other parts of the frequency range when comparing the two. Hopefully this reliability between different units extends to other manufacturers and models because I see your Anechoic measurements as fundamentally useful in using EQ to create "perfectly" flat speakers......I did this with my JBL 308p based on your measurement and I do like the results.

If possible it would be interesting to do the same for popular actives like Kali, jbl, Neumann and genelec. It’s interesting for how well manufacturers can get that right with amp component is integrated

 

[q3cpma]

If possible it would be interesting to do the same for popular actives like Kali, jbl, Neumann and genelec. It’s interesting for how well manufacturers can get that right with amp component is integrated

This post is worth mentioning: https://www.audiosciencereview.com/...on-axis-measurements.20050/page-3#post-662374

 

[flipflop]

This post is worth mentioning: https://www.audiosciencereview.com/...on-axis-measurements.20050/page-3#post-662374

This one, too: https://www.audiosciencereview.com/...udio-monitor-review.10897/page-19#post-306791

 

[Robbo99999]

This post is worth mentioning: https://www.audiosciencereview.com/...on-axis-measurements.20050/page-3#post-662374

Genelec are very serious & very active in making sure each speaker is the same then, they don't just build & QC test them, they actively EQ each individual unit to their reference unit - that's impressive!

 

[pozz]

If possible it would be interesting to do the same for popular actives like Kali, jbl, Neumann and genelec. It’s interesting for how well manufacturers can get that right with amp component is integrated

https://www.audiosciencereview.com/forum/index.php?threads/pair-deviation-list-s-r.18058/

Could answer some of your questions. This data is hard to come by so the list in the linked thread is pretty short.

 

[Rick Sykora]

As I explained, I don't think the response difference is related to the ding.

I concur and, although very valid to point out the difficulty in reproducing the test conditions exactly, the responses match well up until beyond 10 kHz. Since the midrange main contribution is under 5 kHz, the response above it is the tweeter. I have had (likely more expensive) dome tweeters that have sample variations in that range. So, IMO this seems more likely to be a sample variation.

Yes, it could be eq'd out, but you are right that it may not be all that audible anyway.

 

[Rick Sykora]

Genelec are very serious & very active in making sure each speaker is the same then, they don't just build & QC test them, they actively EQ each individual unit to their reference unit - that's impressive!

Vandersteen has been doing matching for decades (on passive speakers for that matter!). It is certainly commendable, but you usually pay extra for it too. In active speakers, it may be easier to automate, but doing every speaker is time. In manufacturing, time adds cost.

 

[YSC]

This post is worth mentioning: https://www.audiosciencereview.com/...on-axis-measurements.20050/page-3#post-662374

I am more wondering on the analog 80x0series though, the DSP series should be calibrated but with analog ones I really wanna know how well they did

 

[q3cpma]

I am more wondering on the analog 80x0series though, the DSP series should be calibrated but with analog ones I really wanna know how well they did

If they do it, I guess it's like Neumann: ASP via internal knobs.

 

[YSC]

If they do it, I guess it's like Neumann: ASP via internal knobs.

right, I am more of a mechanical geek, studying these are interesting to me as in me digging into the factory wheel alignment specs for different car models.

 

[q3cpma]

right, I am more of a mechanical geek, studying these are interesting to me as in me digging into the factory wheel alignment specs for different car models.

See a picture of the KH120A board, these are the blue knobs at the top: