Scott, ''balanced'' in a way that is listenable. Not balanced in a way that is enjoyable or even normal...
You can look at it this way: You have a 3-way speakers and you remove the midrange. It will sound bad, no? Now, you don't remove it but you replace with with another midrange driver. It will probably sound better than no midrange at all, but still will make the whole sound experience less interesting. NOW, you find that once you EQ and SPL balance the new midrange driver... it sounds the same as with the original midrange driver.
That's it.
Just forget about the whole thing of listening music on 360-7200hz, thats entirely irrelevant. And if you just can't get your mind around it, just make the same test but by using a common woofer and a common tweeter for all midrange drivers. Simple as that.
But, I think, not only you won't be able to spot the midrange drivers any more than we did, I think it will be even more confusing, and therefore the identification threshold would be even harder to reach.
Maybe yes, maybe no. You asked whether there was a flaw with your experiment, this could be one, really I think you should have addressed it. It seems fundamental to me that to make an accurate comparison using music one should use music that doesn't sound like it's being played down a telephone line. Now, ofcourse, if there were measurements we could see frequency responses of the drivers.
I don't know if I'd agree with that, Jon. I think bringing in a common woofer + tweeter on these mids might actually make the job easier. 🙂
The wider the bandwidth, the more things stand out. My son can pick out that I changed my DAC, SPL leveled to be the same. I did not tell him, yet he picked up on that. Yes I hear it too, but I did know something had changed.
The crippled bandwidth, the mono setup, that's making it harder to make any judgement. But don't underestimate our ears just jet. The circumstances do matter.
Same goes for tweeters. Listening to an EQ-ed tweeter comparing it to another with same EQ applied might prove very hard to do. Put it above a good mid and bottom end and it will get a lot easier to hear differences.
The wider the bandwidth, the more things stand out. My son can pick out that I changed my DAC, SPL leveled to be the same. I did not tell him, yet he picked up on that. Yes I hear it too, but I did know something had changed.
The crippled bandwidth, the mono setup, that's making it harder to make any judgement. But don't underestimate our ears just jet. The circumstances do matter.
Same goes for tweeters. Listening to an EQ-ed tweeter comparing it to another with same EQ applied might prove very hard to do. Put it above a good mid and bottom end and it will get a lot easier to hear differences.
Last edited:
Yes but anything that helps the listener identify A from B, other than what is originally planned, would make your test flawed.
In that case: spatial identification.
We humans (usually) have two ears, and spatial localisation is something we're pretty good at. I'm not sure exactly to which extent it's precise, but I would say that anything north of 3-5 degrees MIGHT be identifiable.
On-axis is one thing. Spatial position of the source is another one.
One way to go around that problem would be to ''scramble'' the listener spatial awareness by rotating his chair before each round. I thought of that at first, but I finally decided to build a more complex solution: ''the cube''. Which would put less stress on the listener and would involve less interaction with him, thus, letting him more focused on the exercise.
All in the details, my friend. 😉
I really like the sound of EQed fullrange. Mimicking B&W would be very easy, I guess. Much harder to mimic is something like large WE horn system, and it is probably impossible.
It doesn't stop me from trying. I definitely altered my "tone" after hearing an Altec setup. Granted I have a lot of full range drivers which gives me some headroom to play with. (in all honesty, I cheat with room sound too, having 'virtual ambience' drivers)
Maybe.
Maybe not.
The original plan was to start with the mids, then add tweeters using a common mid, then test the woofers but just with the common mid, then everything together against a commercial speaker (B&W or else)
I simply don't see the logic on which, adding common elements, would help. Quite the opposite.
But I'm not against the idea to test it.
That alone would be an earthquake of great magnitude, on Audiophilia planet.
😀
Do you have any idea of what that would mean?
But as I said, I'm not against the idea to test it.
Would probably not do it on purpose, but I might do it anyway if I decide to make an (informal) blindtest with tweeters... (because I would need a common woofer-mid)
I still have ''the cube'' and all my equipment, so I might do that tweeters blind test.
To be honest, it bugs me a little that a 3fe22 (or 3fe25) could be indistinguishable from a ribbon tweeter or an AMT... that costs 10 or 50x the price of the little 3'' cone... 🙄
To be honest, it bugs me a little that a 3fe22 (or 3fe25) could be indistinguishable from a ribbon tweeter or an AMT... that costs 10 or 50x the price of the little 3'' cone... 🙄
Bringing in a good bottom end on a good loudspeaker will change how you perceive that loudspeaker. It gets easier to hear (smaller) differences.
But then again, any common equipment (such a woofer) would bring critics and doubts.
''yeah but with woofer XYZ you would have get different results because blablabla''
Not to mention the tweeters have different crossover points that makes it difficult to compare in their ''optimum'' and/or ''normal'' range.
By example:
You compare a Beyma TPL150 and a RAAL 64-10..
The limiting one is the ribbon. So minimum 3khz crossing point, while the Beyma is usually used at a much lower point. That's why people are buying it, usually. Same for the 64-10's bigger brother, 140-15D, and so on..
So, if we cannot spot the difference at 3khz and up, it says nothing from, let's say 1.2khz to 3khz...
same applies for the woofers, which would be even worst because of the enclosure's and room's effects, and also because they are limited in both lower frequencies and higher frequencies in a more obvious way than mid and tweeters...
I don't think they would admit it even if it's very successful. In professional field, mimicking consumer speaker's FR as an option is becoming popular already.
https://barefootsound.com/wp-content/uploads/2014/06/Barefoot_MicroMain27_Gen2_Users_Manual.pdf
It's all about belief system. The true problem is many people do not admit "audiophile" is nothing but a belief system. Challenging against someone's belief system is probably somewhat unethical behavior, but I have to admit I enjoy it sometimes...
Jon, fwiw i think your test makes complete sense for what i understand its purpose was.....seeing if folks could simply identify under ABX, different mid range drivers covering the same bandwidth at the same SPL
The test was what it was. No more, no less, it seems.
Adding further drivers, a woofer or tweeter, would simply invalidate the direct comparisons, and any true identification test....imo.
I see no need for measurements for what you did other than verifying equal SPL.
I think realizing that the test necessarily underutilized the full SPL capability of particular drivers at particular ends of the tested bandwidth is implicit in the limitations of the test.
I think realizing that particular drivers were probably doing all they could to meet the full SPL being used across the tested bandwidth is implicit in the limitations of the test.
I think realizing that radiation patterns would be different with particular drivers in actual speaker designs is implicit in the limitations of the test..
I think how they actually sounded seems immaterial and is implicit in the limitations of the test...
My only wish would have been knowing more about how much, and how finely, each driver was EQ'ed.....for curiosities sake.
Now all that said, I wish I could have taken it too..
I'd choose pink noise, and be willing to bet on identifying....
And honestly, I'd be happy to loose.....🙂
The test was what it was. No more, no less, it seems.
Adding further drivers, a woofer or tweeter, would simply invalidate the direct comparisons, and any true identification test....imo.
I see no need for measurements for what you did other than verifying equal SPL.
I think realizing that the test necessarily underutilized the full SPL capability of particular drivers at particular ends of the tested bandwidth is implicit in the limitations of the test.
I think realizing that particular drivers were probably doing all they could to meet the full SPL being used across the tested bandwidth is implicit in the limitations of the test.
I think realizing that radiation patterns would be different with particular drivers in actual speaker designs is implicit in the limitations of the test..
I think how they actually sounded seems immaterial and is implicit in the limitations of the test...
My only wish would have been knowing more about how much, and how finely, each driver was EQ'ed.....for curiosities sake.
Now all that said, I wish I could have taken it too..
I'd choose pink noise, and be willing to bet on identifying....
And honestly, I'd be happy to loose.....🙂
- Status
- Not open for further replies.