Could it be possible to derive a meaningful figure of merit for complex graphical data such as FR plots and spinoramas? Perhaps, but ultimately any single descriptor of an information-rich graph like those things will have an averaging effect. There are probably some mathematicians among us who could give that some thought.
That is the number I hope to compute. I need to setup an analysis script but there is not enough info in Sean's AES paper. I plan to contact him soon once I have some measurements to analyze.Toole describes Sean Olive‘s predictive preference rating here (~50min mark):
Just thought I'd throw this in though many of you know it already.
Probably helps if you have read Floyd Toole's book to have the basic vocabulary and ideas in regards to spin-o-rama testing the Klippel system can do. If you don't have the book here are a few short articles on the web explaining various ideas about spin-o-rama testing and what a desired result is.
I see already people asking, "hey Amir, can you show this or test this". That is good, but some of what the Klippel will show makes sense if you know about the Harman research and how it became a speaker testing standard in the ANSI official audio standards.
One thing to look for is Directivity Index or DI from the spin-o-rama data.
Here is a good slideshow presentation with some explanation.
https://www.sausalitoaudio.com/wp-content/uploads/2018/07/Interpreting-Spinorama-Charts.pdf
Here is a 7 page article by Floyd Toole explaining DI and spinorama results.
https://www.edn.com/design/audio-de...ring-the-essential-properties-of-loudspeakers
It is explained here also.
https://www.audioholics.com/loudspeaker-design/understanding-loudspeaker-measurements
Sorry for the scattershot approach some explanations are easier for a given person than others.
Here is another explanation in terms of the ANSI standard.
https://speakerdata2034.blogspot.com/2019/02/spinorama-cea-2034-2015-ansi-data-format.html
This patent filing from Sean may have more info.That is the number I hope to compute. I need to setup an analysis script but there is not enough info in Sean's AES paper. I plan to contact him soon once I have some measurements to analyze.
How much time I have to do this, I don't know. But my goal was to test blind and use my youngest son who has excellent high frequency hearing (and overall sensitivity). I pursued help to get an automated turntable to switch speakers quickly but got too much for me to handle. If someone can help build this, it would be great. It is just a simple table that you put 3 speakers on, and it rotates them into position using external control. My thought is to limit it to smaller speakers right now. Once there, I can put an acoustically transparent curtain in front of it and perform ABC tests while someone else listens. I can use my dedicated theater which is extremely quiet and has the space to hold such contraption.
Failing this, I just plan to turn on the speaker and give it a quick listen in my main system area/lab. This would be a casual test but sufficient for people who say, "did you listen?"
Problem is that Rting has not correlated their ratings to listening tests as Sean did. So who knows what they show is representative of listener preference.Even though how they rate and how they weight is debatable, I feel an approach of scoring similar to Rtings would be worthwhile in considering.
The controls are easy and I can do that. I also wrote up the requirements. It needs to handle 3 speakers, each with say 20 to 30 pounds, and spin them around in 2 to 4 seconds without them being thrown to the walls. If someone can build the turntable with servo motor/controller, I can do the rest.Designing one from scratch is not difficult at all. As long as the functional requirements and space envelope is clear.
Would need someone to sort out the electronic controls though.
In fact I started donating today through the upgrade account option. Hope it'll help.
That is the number I hope to compute. I need to setup an analysis script but there is not enough info in Sean's AES paper. I plan to contact him soon once I have some measurements to analyze.
For what they can afford to pay, it won't make sense to use my time for that instead of testing commercial products with broad appeal. Exceptions can be made of course. FYI service using the same instrumentation as this one cost about $1,000 for one set of measurements. You think any DIY would want to pay anything close to that?Perhaps it is mentioned elsewhere, but there may be a possibility to charge som money for measuring speakers for the DIY community. It could help financing the whole thing as well.
Above a few hundred hertz, a speaker sets the overall tonality of what you hear, not the room. Below that, it is another ball game and requires optimization that is completely room dependent, not speaker.This is all good, but I have one question. Is there a way to adjust scoring according to the intended placement of the speaker and its use (near-wall, on-wall, "free field", toe-in degree, room size/dimensions)? A speaker can sound quite different depending on placement of speaker, listener position and room dimensions.
For what they can afford to pay, it won't make sense to use my time for that instead of testing commercial products with broad appeal. Exceptions can be made of course. FYI service using the same instrumentation as this one cost about $1,000 for one set of measurements. You think any DIY would want to pay anything close to that?
Above a few hundred hertz, a speaker sets the overall tonality of what you hear, not the room. Below that, it is another ball game and requires optimization that is completely room dependent, not speaker.
A speaker that measures well will behave well in a room. Those with odd off-axis response are the ones that requires a lot of messing around to get them to sound good.
Well, if it’s made available, you could see what the correlation would be to the subjective ratings in Harman’s listening tests and see if it does better than the 86% that they achieve.Problem is that Rting has not correlated their ratings to listening tests as Sean did. So who knows what they show is representative of listener preference.