Hi srinath,
no need to read the whole thread, but take a look at this:
http://www.diyaudio.com/forums/mult...d-testing-need-your-help-104.html#post4803549
Basically it was clear from the start that we needed ABX identification in order to make any appreciation on any drivers.
Bottomline: at this moment people get confused much more than planned. So we need more golden ears. Only then we'll get to the yummy part everybody is waiting for...
no need to read the whole thread, but take a look at this:
http://www.diyaudio.com/forums/mult...d-testing-need-your-help-104.html#post4803549
Basically it was clear from the start that we needed ABX identification in order to make any appreciation on any drivers.
Bottomline: at this moment people get confused much more than planned. So we need more golden ears. Only then we'll get to the yummy part everybody is waiting for...
Please refer to post #1143
That would be the first official Matchup*
We are indeed starting with the bar set very very low. That's on purpose, for the reasons mentionned several times in the last 30 pages.
Once we get enough participants that actually have some minimum hearing capacities, we'll move on with a more interesting and challenging matchup.
* all other matchups were to test... the test (!) Now we'll go with that, as much participants as needed, and with 15 music excerpts each (or less if they fail early)
That would be the first official Matchup*
We are indeed starting with the bar set very very low. That's on purpose, for the reasons mentionned several times in the last 30 pages.
Once we get enough participants that actually have some minimum hearing capacities, we'll move on with a more interesting and challenging matchup.
* all other matchups were to test... the test (!) Now we'll go with that, as much participants as needed, and with 15 music excerpts each (or less if they fail early)
Last edited:
I expect this Matchup #1 to be an ''Easy'' one, with at least 70% of success or more.
According to wikipedia, we need 12 correct answers out of 15 to get a ''successful participation''
https://en.wikipedia.org/wiki/ABX_test
But we'll keep only the best ears for the next Matchup, as if half of participants are having 15/15 results, i don't care about participants who did any less than that. See the logic ?
We need the best ears available.
According to wikipedia, we need 12 correct answers out of 15 to get a ''successful participation''
https://en.wikipedia.org/wiki/ABX_test
But we'll keep only the best ears for the next Matchup, as if half of participants are having 15/15 results, i don't care about participants who did any less than that. See the logic ?
We need the best ears available.
I'd like to point out this:
https://en.wikipedia.org/wiki/ABX_test#Potential_flaws
Because of that (and my observations) the final 15 music excerpts are now from a mere 6 seconds lenght to 50 seconds.
I noticed that most of the time, when participants did choose correctly, they were able to do so very early in the playback. Sometimes almost within 1-2 seconds.
But on the other hand, i know that some people would want to have more time to listen... So i mixed short and lenghty in the final selection. I'll have a lot of data later about which is more easy to identify, etc...
https://en.wikipedia.org/wiki/ABX_test#Potential_flaws
Because of that (and my observations) the final 15 music excerpts are now from a mere 6 seconds lenght to 50 seconds.
I noticed that most of the time, when participants did choose correctly, they were able to do so very early in the playback. Sometimes almost within 1-2 seconds.
But on the other hand, i know that some people would want to have more time to listen... So i mixed short and lenghty in the final selection. I'll have a lot of data later about which is more easy to identify, etc...
Breaking news: NOT everyone spot a driver that miss ½ octave...
i don't even know what to say. Wow ?
Lol put Quad ESL 2604 on the mix, I think much easier to spot
Breaking news: NOT everyone spot a driver that miss ½ octave...
i don't even know what to say. Wow ?
Does your test include pink noise?
I suppose the one problem with pink noise is that it's a bit of a red herring, at least depending on what you're looking for in terms of results.
Take two drivers, say a SEAS Excel that is unequalised and a Scan speak driver of some sort with ruler flat frequency response. If you play pink noise through them you will easily be able to identify which one is 'better' or rather 'more accurate/neutral'. This is because the differences between the two are profound with one driver sticking out like a sore thumb and not in a good way.
If you however took those same two drivers and EQd them, as has been done here, it is more than likely that the pink noise would allow you to distinguish between the two but you probably would have no idea which one was better.
It'd be like well they do sound slightly different, that one goes wooosh a bit more than that one, that goes sshhhh a bit more, but who's to say which one is actually more neutral. And then these differences are going to be small and almost entirely related to small frequency response issues you couldn't EQ properly. Either that or the room acoustic and off axis differences. The latter aren't going to be a problem in a finished and competent design as the design will follow according to the drivers limitations and the former should also be compensated for but tends not to be, at least not down to a kind of +- 0.5dB scale.
Take two drivers, say a SEAS Excel that is unequalised and a Scan speak driver of some sort with ruler flat frequency response. If you play pink noise through them you will easily be able to identify which one is 'better' or rather 'more accurate/neutral'. This is because the differences between the two are profound with one driver sticking out like a sore thumb and not in a good way.
If you however took those same two drivers and EQd them, as has been done here, it is more than likely that the pink noise would allow you to distinguish between the two but you probably would have no idea which one was better.
It'd be like well they do sound slightly different, that one goes wooosh a bit more than that one, that goes sshhhh a bit more, but who's to say which one is actually more neutral. And then these differences are going to be small and almost entirely related to small frequency response issues you couldn't EQ properly. Either that or the room acoustic and off axis differences. The latter aren't going to be a problem in a finished and competent design as the design will follow according to the drivers limitations and the former should also be compensated for but tends not to be, at least not down to a kind of +- 0.5dB scale.
I wish somebody will record the pink noise from equalized SEAS Excel and a ScanSpeak and let me tell which one is the SEAS
No, I don't think that the pink noise, after equalized, would allow people to distinguish between the two. By equalizing the FRs, we neutralize the sounds. But the distortion is still there (we know how good the SS and how bad the Excel). And the sound of distortion is unique in a noise! A good noise is not fatiguing, while a bad one is fatiguing. A noise strengthens this behavior (it is easier to pickup distortion in a noise than in a normal music).
Hi Eldam,
I sold my pair of Radian 950PBbe with the intent to try the regular aluminum version. Meanwhile i'm using a pair of 3fe22 as midranges, therefore the RAAL 140-15D's is not really required and i'm using 64-10 instead. That is all temporary, but delivers very good results nonetheless.
To be honest, i miss more the 950PBbe than my 140-15D. I wouldnt say i stopped listening music, though