Freshly installed, only one participant did the test so far.
But i sure can tell you there is absolutely no audible distortion whatsoever at moderate volume 360hz crossover point (48db/oct)... But don't forget i have shims (spacers) on these units, that allows extra excursion. Wasnt even able to go down 450hz without.
couldnt care less about any long horn, the short horn + EQ just do the trick, without the long/exotic horn downside IMO. In fact, i'd personnally go with an even shorter horn or hornless.
Well... at low-moderate volume we will have, AGAIN, a case of where's Goliath ?
Our superhuman golden ears did a nice 7 out of 8, but wasnt impressed by the differences between the two. Both EQd and level-matched, of course.
what can i say... 10 bucks against 1100 bucks.
...but once you floor the gas pedal, Goliath crushes all and everything. I mean: we're talking heart-attack-loud.
So, yeah, if you need super high SPL, it's the thing to buy.
But kuddos to the tiny elevator driver ! 😱
THERE IS STILL HOPE!
yup.
Upcoming days, ladies and gentlemens, we'll throw on the battlefield the mighty Dayton RS225... we'll pimp his FR up to 7.2khz (if thats even possible??) and see how it handles the blind comparisons with the real mids...
stay tuned for more bizarre
yup.
Upcoming days, ladies and gentlemens, we'll throw on the battlefield the mighty Dayton RS225... we'll pimp his FR up to 7.2khz (if thats even possible??) and see how it handles the blind comparisons with the real mids...
stay tuned for more bizarre
Guess who's the BIG BIG winner so far ?
The DSP/EQ.
Looks like the Frequency Response is king of the mountain. And the EQ is the helicopter that can bring you pretty much anywhere on that mountain.
.... will it be the same pattern for the tweeters as well ? I have no clue as of yet.
But spending thousands on the mid drivers seems less and less attractive to me...
The DSP/EQ.
Looks like the Frequency Response is king of the mountain. And the EQ is the helicopter that can bring you pretty much anywhere on that mountain.
.... will it be the same pattern for the tweeters as well ? I have no clue as of yet.
But spending thousands on the mid drivers seems less and less attractive to me...
For your information:
So far we had 10 different participants and a total of 18 rounds, mostly with 3 different drivers (except last round thats used 2 new drivers).
Only ONE participant was able to positively identify, by an average 89%, of his rounds.
9 other participants combined makes a 52%, so it looks like a case of random 50/50 to me...
If that could interests anyone, that only participant who was able to identify something is liking the ATC SM75-150 the most, so far. But it wasnt matched up with the Radian or FR10 yet.
So far we had 10 different participants and a total of 18 rounds, mostly with 3 different drivers (except last round thats used 2 new drivers).
Only ONE participant was able to positively identify, by an average 89%, of his rounds.
9 other participants combined makes a 52%, so it looks like a case of random 50/50 to me...
If that could interests anyone, that only participant who was able to identify something is liking the ATC SM75-150 the most, so far. But it wasnt matched up with the Radian or FR10 yet.
Both groups may be affected in the same way. That was the hypothesis.
The 5% group can describe the differences. It's like a measurement tool describing a THD of an amp as 0.00001%, I believe even a deaf man will be interested 😀
In a short blind test, it is important that the listener can recognize the character of any of the DUT. It will help if the listener is familiar with any of the devices under test, long before the blind test. If he knows that the DUT consists of certain cone material, he can focus on the characteristics of this material's sound.
So there are many things a listener can focus on. He has to choose from these many focus area one by one until he finds one focus area that works. This can take time...
Why a listener can recognize a character of a cone material? Because he has lived with the sound for so many years! And it will be easier if he listen to music/song/singer that he has heard THOUSAND times.
So when I have to ABX two drivers with a music I'm not familiar with, I will put the music in repeat and listen 10 times or more with one driver, until the sound get registered into my 'sub-conscious' mind or memory...
Then when I suddenly switch to the other driver, I expect that my brain will EASILY recognize what is missing or what has changed. Knowing what to look for is important in a blind test.
BUT AGAIN: what is the importance of knowing which one is which, such as knowing which driver is using x cone material???
For me it is more important to recognize the enjoyment (tapping your foot, singing along, dancing along, etc) or boredom (wanting to turn off the music, getting annoyed, forgetting the music, etc) during listening. This is why we often heard that people come back again and again to a certain system. That's because they remember how much enjoyment they have had with that system before...
This, seen as uncontrolled listening experience, expectation bias etc. But there is something in such phenomenon, where you can relate all of these subjective experience with electron, if you have the skill and knowledge.
In an ABX, short term memory only work when there is BIG difference (sometimes not an essential difference, such as one is louder than the other).
Please note that most human's ears are sensitive to level difference or loudness (using short-term memory). Drivers frequency response is not only NON-FLAT, but they have variation from 20Hz to 20kHz. Somewhere along this frequency band, the difference in level can be fantastic. A listener who can find this critical frequency will easily detect a difference.
So by flattening the FR, those listeners have no chance... 😉
But what benefit you expect from knowing which one is louder than which?? You will benefit from knowing something else that has no relation with short-term memory.
From the personal tests I did, timing has quite the effect on that enjoyment factor. Especially the foot tapping part and still singing the songs long after the listening session. And while listening just drifting off in the music and not being able to judge a certain part (like you intended) because the enjoyment of the music takes over.
I always test long term, as it is easy to get wowed one time, it's much harder to consistently be in awe every time you listen. ABX is not going to tell me that. Tonality adjustments can be ABX-ed, but in my opinion you have be able to feel the timing part. Longer sessions help there to see if it "sticks".
(P)EQ is fun, but FIR is more powerful in my humble opinion.
Thanks for the comments, guys.
I'd like to point out something for you to read:
https://en.wikipedia.org/wiki/Psychoacoustics
I'd like to point out something for you to read:
https://en.wikipedia.org/wiki/Psychoacoustics
Wow, that was a very global write-up 🙂. I'm glad I've read a bit more on that subject. Quite a bit, actually.
So what's your opinion on mp3... is it good enough to fool us all? 😀
Have you seen the "virtual" test between different amps? Measurements between the different amps are very close. Fun test to determine if you can hear the difference between the different amps in the test: http://www.diyaudio.com/forums/solid-state/295286-virtual-audition-very-simple-quasi-mosfet-amp.html
ABX away to your heart's content 🙂....
Last edited:
Last big blind test i organized was MP3 v.s. AAC v.s. CD16/48 v.s. HD24/96
MP3 low bitrate was ''easy'' to spot, while AAC was very difficult to spot starting 128kbps.
Nobody were able to spot CD16/48 v.s. HD24/96, not even close.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
DAYTON RS225-4ohms WOOFER
So, i managed to make a flat EQ -same as the other drivers- even though i struggled a little bit more this time. Took all 6 bands and for the first time a double-digit dB correction. See below for the details:
342hz // +3.8db // Q1.8 // Peak type
588hz // -1.6 db // Q5.6 // Peak type
1600hz // +1.9db // Q6.0 // Peak type
2040hz // -2.9db // Q4.8 // Peak type
2925hz // +11.0db // Q2.2 // Peak type
6700hz // +1.3db // Q3.9 // Peak type
Also note that i'm always using 360hz/7.2khz crossover points from the beginning.
I guess i was lucky so far not to stumble upon a driver that couldnt make it that wide...
I guess i was lucky so far not to stumble upon a driver that couldnt make it that wide...
One can learn (maybe not all of us) what to listen for to spot the mp3 files, I have to admit AAC is way better as an algorithm. There is a difference between codecs though. I remember participating in several of Nero's AAC encoder tests a long time ago.
Point is, we can learn to hear certain things. That doesn't mean we can spot everything we need to know in an ABX test. So these tests are interesting by themselves, but do not supply all the answers needed. I remember on the Hydrogen forums you'd always needed to supply proof you could ABX any difference. I lost interest as some parts of audio were harder to find using ABX, but I noticed listening long(er) could make me like or dislike certain songs (whether it was from compression or otherwise). Ever since I'm not relying on ABX alone. Though it's still a useful tool in tests like these.
I bet everyone who has done some type of DIY audio, like building speakers, has noticed new things in songs they never heard before. Even in songs that are very familiar. The point is, we can focus on certain parts, and hear the material we listen to in another way. Basically training yourself. You can learn to listen for room effects, but it will be harder to "unlearn" after that. While our brain usually strips the room from what we hear without us consciously thinking about it. And each of us does it slightly different. We go to school to learn to speak, write etc, but not to really listen for specifics. Each of us can decide to train that part at any time. (Provided the hearing mechanism is up for the task)
Point is, we can learn to hear certain things. That doesn't mean we can spot everything we need to know in an ABX test. So these tests are interesting by themselves, but do not supply all the answers needed. I remember on the Hydrogen forums you'd always needed to supply proof you could ABX any difference. I lost interest as some parts of audio were harder to find using ABX, but I noticed listening long(er) could make me like or dislike certain songs (whether it was from compression or otherwise). Ever since I'm not relying on ABX alone. Though it's still a useful tool in tests like these.
I bet everyone who has done some type of DIY audio, like building speakers, has noticed new things in songs they never heard before. Even in songs that are very familiar. The point is, we can focus on certain parts, and hear the material we listen to in another way. Basically training yourself. You can learn to listen for room effects, but it will be harder to "unlearn" after that. While our brain usually strips the room from what we hear without us consciously thinking about it. And each of us does it slightly different. We go to school to learn to speak, write etc, but not to really listen for specifics. Each of us can decide to train that part at any time. (Provided the hearing mechanism is up for the task)
Last edited:
Isn't that the main point of expensive drivers though? Tons of speakers sound good at low levels but requesting clean realistic peaks always requires large or expensive devices.
Maybe.
Maybe not.
I'm not sure about anything at this moment 😕
Since it seems very obvious that once EQ'd and level-matched all drivers sounds very similar (or identical for most people) i'll have to question few things...
1. Is the ''learning'' thing could be the problem ?
2. Is the musical excerpts are not good enough ? Too short or too long ?
3. Should i move away the chair ?
4. Poor quality Source/DAC ?
5. SPL level too low ?
or else ??
... i'm just brainstorming again... Something in the set-up and/or methodology might be wrong. Or maybe it's just us, humans, that limits everything.
