That's strange, to my ears the most pleasing is the one "closest" to a reference. May be not close when comparing the graphs, but perception is different than graph. And to my ears, nothing is close enough to the reference quality, so I'm confused to hear these comparisons to the reference track...
I think it won't happen. The most expensive will win, the cheapest will loose...
I think this is the most accurate setup/recording XRK had produced. There is nothing confusing, like B80 with extra stereo effect. Also hard to believe a stiff cone like fiberglass could survive without crossover (ime, even simple crossovers do not work).
In the first listening (quick impression) our preferred file can change but usually the most disturbing file will stick forever...
At first quick listening, one file stood out being worse than the rest. TC9 just have a THD peak at 300Hz, not far from XO at 350Hz where TC9 rolls off at 12dB/octave.
Again, how do you know which is the TC9? Especially given that there are three drivers which have a similar response. Must be nice to have Goldenears. To be fair, the TC9 is a $12 driver being pitted against the best of the best in the $70 to $100 price range - and B80 costs closer to $170 in the U.S.
I do think that these are some of the best recordings so far. Using the 10F/RS225 Ref speaker as the test bed with a well worked out XO and source material in mono seems to have done the trick.
I do think that these are some of the best recordings so far. Using the 10F/RS225 Ref speaker as the test bed with a well worked out XO and source material in mono seems to have done the trick.
Last edited:
Fibreglass cones are not that stiff with very good inherent damping properties they are very well controlled by all usual standards.
This test has the very flat 10F and the very flat B80l. It will be interesting to see how peoples perceptions vary.
One driver definitely sounds like it has more bass than the others which I always find odd. I have narrowed it down to two. From there on it gets hard.
This test has the very flat 10F and the very flat B80l. It will be interesting to see how peoples perceptions vary.
One driver definitely sounds like it has more bass than the others which I always find odd. I have narrowed it down to two. From there on it gets hard.
Personally, I am rooting for the budget drivers to win. The budget TC9 beat out more expensive ones in the first round but the most expensive ones did seem to win rounds 2 and 3.
Cal, are you saying Bose is cheap and they have the biggest chunk of the market? I don't think they are cheap at all, rather mid to higher priced for what you get.
Cal, are you saying Bose is cheap and they have the biggest chunk of the market? I don't think they are cheap at all, rather mid to higher priced for what you get.
BMS,
You are voting for most accurate vs what you think sounds best? You are making quick work of this. Very tough for me. I am glad that clip 1 seems to do the weed out trick. It was mentioned in Round 3 that ACDC track was key for easy selection. It's interesting how rock over drive guitar which is full of distortion in the source is key for picking a driver that sounds accurate.
Yep, going for accuracy.
Got to say, it was the symbols that did it for me rather than the distortion/overdrive guitar. The guitar helped pick two drivers, the symbols gave me 1st & 2nd places.
Fiberglass sound is very unique (I think TG9 has less fiberglass or less stiffer than 10F. But they sound quite similar). TC9 paper sounds cheap and especially that distortion...
The mystery driver sounds "high end" but something is not right. Untrained ears will think that this is the best in term of performance to price ratio. I think its subjective tho.
Me too. Always a fan of the underdog.
I prefer to think of them as 'economically produced'
Huge
See quote #2.
I kinda like you are picking just the one spot on the tc9 where it actually performed better than its more expensive brothers/sisters. They are better at everything... exept that part of the spectrum. At least thats what I get from all the measurements I've seen so far.
Kudo's to you if you're right. I wouldn't be able to tell. I am listening trough 50 x tc9 to judge
Last edited:
I must admit that I'm having a hard time taking this subjective comparison seriously. The acoustic signal has been colored by a microphone and the acoustics of the room they are mic'd in, then it's converted to mp3, sent over the web, then converted back to analog, and then further colored by my computer speakers and room acoustics. Depending on the program material, and all its idiosyncracies, certain amplitude variations might sound better for that recording. I'm thinking my impressions could be way off with all these variables.
Perhaps if these speakers were in an anechoic chamber, with orchestra music choral music, piano music and small scale acoustic instrument music (guitars, mandolins, etc.) recorded by a very good mono measurement mic, then the recording could be played back through the same speaker driver and recorded again, and again, until the distortions have become so exagerated that a difference will be obvious to a listener. I hear it only takes about 3 times for the errors to be obvious.
You need to use music that is a sound that we know, like acoustic instruments, piano and voice. No synthetic stuff or distorted electric guitars. Choral music tests a system for I.M. distortion better than most things. Piano was used by Ray Dolby to show how bad 16 bit digital sounded back in 1986. An orchestra would highlight the ability of a driver to stay clean with lots of things going on at the same time.
Perhaps if these speakers were in an anechoic chamber, with orchestra music choral music, piano music and small scale acoustic instrument music (guitars, mandolins, etc.) recorded by a very good mono measurement mic, then the recording could be played back through the same speaker driver and recorded again, and again, until the distortions have become so exagerated that a difference will be obvious to a listener. I hear it only takes about 3 times for the errors to be obvious.
You need to use music that is a sound that we know, like acoustic instruments, piano and voice. No synthetic stuff or distorted electric guitars. Choral music tests a system for I.M. distortion better than most things. Piano was used by Ray Dolby to show how bad 16 bit digital sounded back in 1986. An orchestra would highlight the ability of a driver to stay clean with lots of things going on at the same time.
So you guys can spot a driver's signature when it has a fiberglass cone? I suppose aluminum and paper have their own distinct sound - why not fiberglass? The TG9FD has a very thin fiberglass cone - it is translucent. Whereas the 10F has what appears to be a thicker fiberglass cone with a substantially thicker damping coating on it. It also has a fiberglass dustcap whereas the TG9 has the same coated paper that the TC9 has.
Voted : )
Think its very hard this time so would like to spend more time but had to stop somewhere because of other stuff needing attention and get this rank done. Even without EQ/DSP/Electric/Acoustic mods think they all sound good and could be made to a good speaker system. X great thanks your big effort time after time put into this and look forward round 5 where names already put on four drivers. Attach my ranking and guess for which clips belong to drivers, but even i have TC9FD and 10F/8424 myself i couldn't point them out for shure as i could at round one and two, so have a laugh if guess is way off when revealed.
Jay,
Could i say it would suit you to hang on a txt file of your hard cut out prediction of what clips belong to drivers at a time before 15 of September 02:01GMT, so we know if you were right or not in predictions. Hat of of if you right, but its irritating if you were wrong and after drivers revealed continued at same track that blah blah that is because of paper and blah blah that is because of fiberglass, so please hang on a file : )
Think its very hard this time so would like to spend more time but had to stop somewhere because of other stuff needing attention and get this rank done. Even without EQ/DSP/Electric/Acoustic mods think they all sound good and could be made to a good speaker system. X great thanks your big effort time after time put into this and look forward round 5 where names already put on four drivers. Attach my ranking and guess for which clips belong to drivers, but even i have TC9FD and 10F/8424 myself i couldn't point them out for shure as i could at round one and two, so have a laugh if guess is way off when revealed.
Jay,
Could i say it would suit you to hang on a txt file of your hard cut out prediction of what clips belong to drivers at a time before 15 of September 02:01GMT, so we know if you were right or not in predictions. Hat of of if you right, but its irritating if you were wrong and after drivers revealed continued at same track that blah blah that is because of paper and blah blah that is because of fiberglass, so please hang on a file : )
Last edited:
Bob,
We have discussed this at length in the earlier 3 threads. I think you are coming to this later - a few key points: use good studio monitor grade headphones to listen to remove your room acoustics; it's immaterial that it is digitized and sent over web - digital data is unaltered in transmission whether by smoke signals, morse code or fiber optics; sure my room acoustics are there but the setup is exactly the same for all drivers so if you hear a difference it is due to the driver; choice of music is based on feedback from previous 3 rounds. Piano is there in tbe jazz clip 2, I avoided simple plain voices or instruments as complex richly layered music is what stresses a driver most for intermodulation distortion; we don't listen to music in an anechoic chamber and they are hard to come by, so I do the best I can.
Have you ever been in an anechoic chamber? Things don't sound good at all because we lose all the cues of reflected sound for ambiance. It literally sounds like you are in outer space.
I may not have the high end boutique audio chain but what I have is accessible by all DIY'ers: CD player, PC with Audacity, TPA3116 amps, miniDSP. On the recording side, I have shown elsewhere that the mic on the zoom H4 is quite flat - consistent with calibrations used on the measurement mic within 1 or 2 dB. I did this by recording pink noise spectrum on H4 and comparing that to calibrated mic.
One final thing - whether or not you take it seriously, these tests have repeatedly shown that listeners have picked the driver with the flattest frequency response each time. Despite all the pitfalls and problems that you mention, the character of a driver comes through and is discernible in a blind test.
Last edited:
Level-matched Ref Clips
I reduced the amplitude on the reference tracks to be as close as possible to the recorded sound clips of the drivers to make it easier to compare if you were to stick all tracks onto a playlist and listen via a player software. The files have a "B" in them. Change from .asc to .mp3 to listen.
I reduced the amplitude on the reference tracks to be as close as possible to the recorded sound clips of the drivers to make it easier to compare if you were to stick all tracks onto a playlist and listen via a player software. The files have a "B" in them. Change from .asc to .mp3 to listen.