@Kaffimann/CharlieLaub You are proposing something very different from what nas3000 wish to do. I believe most diy drivers are used without dsp, so their natural frequenzy-response is also an impotant subjective and objective quality. However, your idea will be great if there will be time and energy for such additional measurements and listening tests.
Cheers!
Cheers!
Point understood.
Just looked at aliexpress stuff - it is either not cheap at all, especially if I need to pay VAT + customs taxes or does not inspire confidence.
Last edited:
Yes, the idea is to find "the most nice sounding" speaker, not the most accurate. There may be option to make 2 tests (blind comparissons!)
a) Bare FR dirver, no auxilary subwoofers, purelly as is on the same baffle or cabinet. Well this one would be easy to guess exact drivers for seasoned builders.
b) As close as possible what xrk971 did - the copy/paste is doable by my limited skills
In both ways I do not have the desire to manage or flatten every different driver response with dsp. Drivers will be tested (compared!) as they are.
Thanks! The idea is very good - probably local guys can lend me some more interresting drivers or even dsp modules, which I am lacking. The bad thing is that the long time primary local diy forum site is gone couple of years and it will be hard to contact locals. Havent been in any of the meetings 🙁
I hope to do this for the fun of it, side benefits you mentioned are nice, but not primary ambition.
Yes, It does not. But this is the fun part of it.
Besides - I think it is better when participants have neutral unbiassed attitude and judge (compare) speakers by their sound only.
What I meant was that it would be good publicity for a HiFi-nerds club to partisitate by lending their ears. The value of recordings of speakers are so so.. Can give some idea of the sound but not much. Many several or many peoples opinion while blind listening is more interesting I feel
Cheers!
I have a pair of Sica 2" fullranges and pair of the TEBM35 BMR drivers (also less than 2"). All would fit in an envelope sized box. Would you be interested to test these as well? They are really on the small side, the use would be as inefficient mid-tweeter for small desktop monitors.
Yes, definitelly interrested.
I will PM you in couple of months to discuss the conditions, ok?
...It may be a separate project, as the major part of the speaker collection is my own or my close friends - they are not going anywhere. Besides, it would be very ez, as most of the speakers are so much different by the sound and some have many flaws.
I do not see the point of blind tests.
We are so gullible and easily disoriented that all of us should be deemed unreliable, and the measurements should reflect this:
"This driver produce sound equal to these highly detailed measurements, and here is the reason it has these charachteristics."
What is severely lacking, in my opinion, is something like for instance a thorough distortion analysis of a large selection of drivers, when driven lightly, and when driven close to their limits, in a repeatable and controlled environment.
We are so gullible and easily disoriented that all of us should be deemed unreliable, and the measurements should reflect this:
"This driver produce sound equal to these highly detailed measurements, and here is the reason it has these charachteristics."
What is severely lacking, in my opinion, is something like for instance a thorough distortion analysis of a large selection of drivers, when driven lightly, and when driven close to their limits, in a repeatable and controlled environment.
Last edited:
These blind tests are interesting, I just wish they came with more helpful data.
Lossy audio samples made with music playing through speakers in a room that you don't know what sounds like.
You do not really know what you like in the samples you prefer from a test like that. Like the thread title says, purely subjective.
I will follow the thread to see what happens, but silently, since there seems to be some objections to my interfering comments.
X and Nas, please consider making separate threads presenting distortion measurements on light volume and when pushing it a bit.
Best of luck to all involved.
Lossy audio samples made with music playing through speakers in a room that you don't know what sounds like.
You do not really know what you like in the samples you prefer from a test like that. Like the thread title says, purely subjective.
I will follow the thread to see what happens, but silently, since there seems to be some objections to my interfering comments.
X and Nas, please consider making separate threads presenting distortion measurements on light volume and when pushing it a bit.
Best of luck to all involved.
I see no problem, if someone else provides the files. Because of the size of the files I think we cannot afford to test some exotic lossless big bitrate audio formats, but there are very very tiny differences between mp3 320kbps and CD audio quality. Personally I can hear very easy differences between 128kbps and 192kbps, but need to focus to hear difference between 192kbps and 256kbps, and after that I can hear only on very good setup and in some specific songs and song parts. If someone thinks otherwise or can hear the difference after all the testing line (source->DAC->amp->speaker->mic->(ADC)->resulting file->DAC->headphone amp -> headphones...) - I have no problem with that and will not argue with.
The room will be the same for all drivers, I see no problems to make panoramic photo of it. Typical living room most of us have and spend time in. No room acustic treatment is made.
Point is taken about volume levels, thank you!
Mp3's have a specific sound signature because of the algorithms involved.
Easy to hear the difference between an mp3 and a flac file if you are familiar with the setup and the music, almost like the colour banding of overly compressed video on a half decent monitor.
I find it more difficult with ogg vorbis for some reason, but after some time you start noticing the small underlying details are just slightly more "muddy", most peaks are fine.
Same as with 16bit vs 24bit sound, when rendering some of the very lowest and softest sounds are just completely gone on 16bit with/without dither, on 16bit dither vs non-dither 24bit gets the minor details from low level sounds without sacrificing the details on peaks. You can also hear this on natural instruments, like a good cello recording.
I understand that people have some need to generalize and simplify, but I just think it's very hard to do a proper test where you want to chase after tiny details, but they are all missing, heavily banded or muddy.
And perhaps the most upvoted driver has a distortion signature that makes it sound extra sweet, as opposed to an unpopular driver that was actually much cleaner in real life.
But that's ok, this is about a subjective comparison after all.
Perhaps it would be possible to have drivers 1, 2, 3 etc. And also use SPL+distortion measurement data A, B, C etc. make people try to guess which is which.
Again, sorry for intruding, enjoy the test.
Easy to hear the difference between an mp3 and a flac file if you are familiar with the setup and the music, almost like the colour banding of overly compressed video on a half decent monitor.
I find it more difficult with ogg vorbis for some reason, but after some time you start noticing the small underlying details are just slightly more "muddy", most peaks are fine.
Same as with 16bit vs 24bit sound, when rendering some of the very lowest and softest sounds are just completely gone on 16bit with/without dither, on 16bit dither vs non-dither 24bit gets the minor details from low level sounds without sacrificing the details on peaks. You can also hear this on natural instruments, like a good cello recording.
I understand that people have some need to generalize and simplify, but I just think it's very hard to do a proper test where you want to chase after tiny details, but they are all missing, heavily banded or muddy.
And perhaps the most upvoted driver has a distortion signature that makes it sound extra sweet, as opposed to an unpopular driver that was actually much cleaner in real life.
But that's ok, this is about a subjective comparison after all.
Perhaps it would be possible to have drivers 1, 2, 3 etc. And also use SPL+distortion measurement data A, B, C etc. make people try to guess which is which.
Again, sorry for intruding, enjoy the test.
I agree with you. You used exact word - signature. But in most cases you have to pay atention to hear it. On the other hand the signature which have all of the bare drivers I hear is I think 10x more significant.
So, what is the go-to file format which all of us, using Windows 7-10, iOS, macOS can play with popular players? For practicality reasons we have to stick with the popular one AND probably with high data compression ratio (probably lossy). No worse than 320kbps mp3 and not significantly bigger file size
- Status
- Not open for further replies.