List of Proposed Additions/Changes to Speaker Testing Methodology

[JBH129]

I am hoping to keep a running list of the additions/changes posted in the complaint "suggestion" thread. Once the list is reasonably complete, I will post the items in a poll so that we can see the relative importance of each in the eyes of the community.

Please respond with any proposals that i have missed. I would appreciate if you can limit the conversation here and just reply with additions/changes.

Draft List of Proposed Changes/Additions - Last updated 1/29/20 (10:44am EST)

1. Include distortion tests (on-axis sweep of selected tones at 90dB and 96bB SPL @1m)
2. Linearity and/or compression test
3. Maximum usable SPL (CTA-2034-A)
4. Informal listening tests at a standard level before measurements are taken
5. Reference axis determined by manufacturer

 

[hardisj]

Some caveats...

1. 90dB and 96dB SPL at 1 meter equivalent. A single point isn't as useful as understanding what happens as output is increased.
2. Ya'll better have a discussion on how you want this done. There are different ways Klippel can do it. Not sure if Amir has access to those modules.

 

[SimpleTheater]

I can't believe that its the year 2020 and everyone is still insisting on meaningless things like distortion test sweeps and SPL, when @amirm is missing the MOST important test of all:

• Every speaker must be listened to for hours equipped with Nordost Valhalla 2 speaker cables (and definitely not the crappy first version). Once this is done, I want to see detailed description of the sound, especially compared to speakers tested weeks and months ago, to compare the subtle nuanced sound differences using the human brains detailed ability to remember.

If anyone is wondering why the original Valhalla cables are useless, this is an excerpt from Stereophiles Review in which Brian Damkroger commented on the original version "I became more aware that what had seemed like precision was actually a slight overemphasis of the edges of images ."

This is PAINFULLY obvious to me when listening to the original Valhalla's. Sound is all about the edges.

 

[Hayabusa]

I can't believe that its the year 2020 and everyone is still insisting on meaningless things like distortion test sweeps and SPL, when @amirm is missing the MOST important test of all:

• Every speaker must be listened to for hours equipped with Nordost Valhalla 2 speaker cables (and definitely not the crappy first version). Once this is done, I want to see detailed description of the sound, especially compared to speakers tested weeks and months ago, to compare the subtle nuanced sound differences using the human brains detailed ability to remember.

Make sure you move to where Amir lives and be ready 24/7 to do these listening evaluations for us.

 

[Fluffy]

I can't believe that its the year 2020 and everyone is still insisting on meaningless things like distortion test sweeps and SPL, when @amirm is missing the MOST important test of all:

• Every speaker must be listened to for hours equipped with Nordost Valhalla 2 speaker cables (and definitely not the crappy first version). Once this is done, I want to see detailed description of the sound, especially compared to speakers tested weeks and months ago, to compare the subtle nuanced sound differences using the human brains detailed ability to remember.

The first version is crap!? You must be deaf. Please sell your gear to charity and delete your account.

 

[SimpleTheater]

The first version is crap!? You must be deaf. Please sell your gear to charity and delete your account.

I edited my original post to specify the issue. I sold my original Valhalla's for $10 at a garage sale to a guy who was going to use them for his Foxnovo soundbar he just picked up on Amazon. I told him, he's going to HATE the edges. He came back three days later demanding his $10 back, but I refused.

 

[Fluffy]

I edited my original post to specify the issue. I'm sold my original Valhalla's for $10 at a garage sale to a guy who was going to use them for his Foxnovo soundbar he just picked up on Amazon. I told him, he's going to HATE the edges. He came back three days later demanding his $10 back, but I refused.

You're so full of BS. I loaned the Nordost Valhalla ver.1 (the second pair I got, because they were so delightful) to my 12 YO niece, and she used them to replace the cable for her House of Marley headphones. When she first connected them to her iPhone6, she immediately told me about the dramatic difference in soundstage and clarity of midrange. She couldn't take off the headphones for a whole hour, relishing on the sweet sound the cables released from her otherwise sub-par equipment.

 

[frogmeat69]

 

[SimpleTheater]

 

[dshreter]

I became more aware that what had seemed like precision was actually a slight overemphasis of the edges of images

I'm tired of this objectivist bull crap. I like the sound of edges, and will make no apologies for that preference.

 

[Sancus]

Jokes aside, I think a list like this is a good idea, and I think Amir can just say 'yes/no' to each thing and leave it at that.

People tend to keep bringing up the same things over and over again, even when they've already been denied, and this must be exhausting and is also annoying for everyone else just reading those conversations.

 

[Severian]

This may not be necessary as a standard procedure, but as I mentioned in the LS50 thread, it would be very interesting to see an investigation of a speaker being run full range vs. high passed to simulate usage with a subwoofer.

 

[bobbooo]

Some of you might have already seen this, but I came across a whole list of clearly defined standards by the AES for audio electronics and acoustics measurements, experiments etc. on their site: http://www.aes.org/publications/standards/

As Sean Olive's AES profile says he's on the Standards Committee for Loudspeaker Modeling and Measurement, I think we can safely assume he used these standards when taking the measurements of speakers his preference formula is based on, and so following these same AES standards would produce measurements and so predicted preference scores from his formula that correlate most highly with actual preference. Here's that particular standards document: http://www.aes.org/tmpFiles/aessc/20200129/aes-05id-1997-s2019-i.pdf

I'm not an AES member, but if anyone who is wants to have a read of it, it may contain some useful information. These documents also might come in handy:
http://www.aes.org/tmpFiles/aessc/20200129/aes02-2012-i.pdf
http://www.aes.org/tmpFiles/aessc/20200129/aes56-2008-r2019-i.pdf
http://www.aes.org/tmpFiles/aessc/20200129/aes-01id-2012-r2017-i.pdf
(The last one probably isn't relevant as it seems to be about plane-wave tubes, but does mention mic placement in the contents so might have some useful nuggets of info.)

 

[tuga]

Would it be possible to re-upload all graphs of existing speakers tests with frequency response and/or directivity using a 50dB-wide* vertical scale?
As it stands, visual comparisons are misleading (i.e. Neumann vs. Harbeth)?
Other measurements should also have an identical vertical scale width across all tests.

Thanks!

* not sure if this is a standard but it is the width used by Stereophile and Soundstage

 

[Head_Unit]

I sold my original Valhalla's for $10 at a garage sale to a guy who was going to use them for his Foxnovo soundbar he just picked up on Amazon. I told him, he's going to HATE the edges. He came back three days later demanding his $10 back, but I refused.

That must have been the same guy that once bought an old dark green Ford Pinto from my dad. It had been sitting, I gave him the key and told him "take it to a mechanic or whatever" and hoped he just stole it because it was an eyesore. Bought it for something like $200, far less than the parts value as you could strip out just the windshield and sell it for $200. Had the gall to call a week later and piss off my dad asking for not just a refund but overpayment because it needed a valve job. Idiot!

 

[Head_Unit]

Back to the thread, I once upon a time wanted to run FFTs of multitone tests like the Audio Precision Fasttest or whatever they called it, which I think CLIO pioneered. Why? Because there are a lot of speakers that sound lovely with classical or jazz but if you try and crank heavy metal they turn into a hash of distortion. I felt a multitone test might replicate this much better than traditional swept distortion, especially as it would visually capture not just harmonic but intermodulation as well.

I also bemoan that cepstrum analysis never seemed to go anywhere as the one paper I remember seemed to show some interesting results regarding "horn sound." Really, the typical loudspeaker measurements totally ignore the time domain which we actually listen in.