Thanks for the explanation. Still, i believe there is information missing when i hear the samples... These days i'm doing similar driver's comparisons as you know and, in real, it seems the overall perception of the driver's differences is not quite the same as your recorded samples. Is it because i'm testing drivers with very different sonic signature ? Maybe.
anyways, i find your sound takes to be very good. Is there a page in this lenghty thread where you talk about your recording set-up ?
Thanks. It was probably discussed in the Round 1. But basically is this:
Source (CD player headphones out or laptop playing high res clip from Audacity) - TPA3116D2 amp - speaker.
On recording side: ZOOM H4 recorder set to .wav files 24-bit at 96kHz. In Audacity, trim length, set gain for peaks to be -0.5dB FS, convert .wav file to 48kHz mp3 at 320kbit.
That's it.
have you ever tested the accuracy of the Zoom H4 ?
at what distance was the H4 from the drivers ?
No, I never checked the recorder - it's built in mic's probably have a signature but I have personally found the sound to be quite neutral. It has the ability to choose a response curve of famous studio mics like a Neumann U87, etc. Distance was 1.0 meter.
I like very much the idea of doing such test at distance. However i'm not fully convinced we can draw any conclusions from such test, mainly because it's like a photocopy of a photocopy of a photocopy. And maybe even worst, such as a photocopy of a picture of a painting...
If my calculations are correct, you have 4x transducers conversions + a minimum of 4x analog-to-digital or digital-to-analog conversions.
We are far from sitting with Norah Jones... 😱
------
What would happen if you take one your sample recording and pass it through the process again ? and again, and again (...)
Maybe at some point we could have a more precise idea of what information is lost by this very process ?
(i.e. comparing original sample v.s. 5th level audiocopy)
If my calculations are correct, you have 4x transducers conversions + a minimum of 4x analog-to-digital or digital-to-analog conversions.
We are far from sitting with Norah Jones... 😱
------
What would happen if you take one your sample recording and pass it through the process again ? and again, and again (...)
Maybe at some point we could have a more precise idea of what information is lost by this very process ?
(i.e. comparing original sample v.s. 5th level audiocopy)
Last edited:
If you did the multigenerational copy - that is actually a good test of how good your copy process is. And it would show that a flat response driver like the 10F will pound a driver with huge coloration and peaks into the ground. Tom Danley I believe, suggested that as one means to check the neutrality and accuracy of a speaker.
That actually sounds interesting and I might do it between the 10F and A7.3. It will have to be done in a bigger room to avoid reflections from swamping the picture.
In any event, the argument that you can't get useful info doesn't hold. Just use headphones to listen to the sound clips. The only thing that changed is the driver. If you hear a difference it is due to the driver being different.
That actually sounds interesting and I might do it between the 10F and A7.3. It will have to be done in a bigger room to avoid reflections from swamping the picture.
In any event, the argument that you can't get useful info doesn't hold. Just use headphones to listen to the sound clips. The only thing that changed is the driver. If you hear a difference it is due to the driver being different.
That's one way to see it.
I organized similar blind test and, believe me, i heard every possible arguments against my tests (and test in general)... Justified and not-justified.
I'm just playing the devil's advocate here
From my ''scientist point of view'' there is few flaws in your test:
1. Not everyone is using the same final audio reproduction equipement/acoustics environment
2. There is a lot of conversions (as already discussed) most likely blurring audible information otherwise present after only half conversions.
3. Testees did not prove they're able to positively identify prior to chose their preferences
4. Poll should be blind until its closed, to avoid any kind of Group-bias.
5. Of course all testees must have listened equally the same number of files/drivers.
Yeah, i know it's very uptight requirements but that adds a lot of credibility since it's better controlled. (and i'm not saying your test has zero credibility!) like i said i'm playing devil's advocate.
I'm with you on that.
In that encapsulated reality, the only changing variable is most probably the driver.
Problem is: the ultimate goal of the test is (probably!) to help make a choice among said drivers. Therefore, that encapsulated reality CAN be a reliable guideline OR a total mislead.
To use it as a guideline, we must be sure it's very similar to a real-life use.
One thing that would help is to run the same test but directly at your place (reducing the conversions), same SPL-matching, and with at least half dozen testees so you can start to see if the results are the same as online.
That's one way to see it.
I organized similar blind test and, believe me, i heard every possible arguments against my tests (and test in general)... Justified and not-justified.
I'm just playing the devil's advocate here
From my ''scientist point of view'' there is few flaws in your test:
1. Not everyone is using the same final audio reproduction equipement/acoustics environment
2. There is a lot of conversions (as already discussed) most likely blurring audible information otherwise present after only half conversions.
3. Testees did not prove they're able to positively identify prior to chose their preferences
4. Poll should be blind until its closed, to avoid any kind of Group-bias.
5. Of course all testees must have listened equally the same number of files/drivers.
Yeah, i know it's very uptight requirements but that adds a lot of credibility since it's better controlled. (and i'm not saying your test has zero credibility!) like i said i'm playing devil's advocate.
Jon, these are all valid points you make. If you look at the results though, there is no clear winner. To my ears at least, a few loudspeakers sound grossly distorted and should never have any votes. And yet, people are picking them. I don't think the test setup is to blame here. People either do not hear the distortion or they prefer that sound. Either way, they find out what they like, which is I think the only thing this test is able to reveal.
.....like i said i'm playing devil's advocate.
Devil's advocate okay new hat other day judge for good wines 🙂....
As diy guy like the way xrk971 present these comparison it's not the journalist that do the review and point out what devices i shall buy and prefer, only cost is offer some time and energy. That the quality is as is for many reasons, hard as it can be judge between the sound clips one can also learn a lot in the listening process about ones own skills gear and what a job a live recording can be to get right.
No dout you have many good arguments plus open questions and sounds as more pro and more scientific way. That is good even it could be argued why those high demands in a diy thread, but as we like best and correct sound its of course the way to go raising the bar. Think xrk971 did raise the bar for round two in better quality sound clips verse round one, he is open minded and if proof turns up to raise the bar he do it if possible is my guess.
Over at comparison thread think OPC posted some very good points too and add post as picture copy below and link to same is here http://www.diyaudio.com/forums/full...in-5in-full-range-drivers-25.html#post4249709.
Attachments
Last edited:
Devil's advocate okay new hat other day judge for good wines 🙂....
As diy guy like the way xrk971 present these comparison it's not the journalist that do the review and point out what devices i shall buy and prefer, only cost is offer some time and energy. That the quality is as is for many reasons, hard as it can be judge between the sound clips one can also learn a lot in the listening process about ones own skills gear and what a job a live recording can be to get right.
No dout you have many good arguments plus open questions and sounds as more pro and more scientific way. That is good even it could be argued why those high demands in a diy thread, but as we like best and correct sound its of course the way to go raising the bar. Think xrk971 did raise the bar for round two in better quality sound clips verse round one, he is open minded and if proof turns up to raise the bar he do it if possible is my guess.
Over at comparison thread think OPC posted some very good points too and add post as picture copy below and link to same is here http://www.diyaudio.com/forums/full...in-5in-full-range-drivers-25.html#post4249709.
i'll simply put a quote from my favorite movie of the decade:
“The two most harmful words in the English language are ‘good job.’ ”
?Good Job? ? Are These the 2 Most Harmful Words in the English Language? | TLNT
Neither Byrtt nor OPC mention the phrase "good job" so not sure what you are opining about.
I am pretty sure we are all aware that audio is a series of choices and compromises. The trick is to manage our choices so that in the end we are left with something that is still good to listen to. After all, it is about the sound in the end.
I am pretty sure we are all aware that audio is a series of choices and compromises. The trick is to manage our choices so that in the end we are left with something that is still good to listen to. After all, it is about the sound in the end.
And just to show that taste is personal...
There are those who absolutely HATED Whiplash. Saying the drumming and the overall quality of the music played was sub-standard, plus the overly exaggerated portrayal of the "mean" teacher/director was cartoonesque.
For me, I enjoyed it as what it was... a movie.
But I understand that everyone has their own taste in movies, music, etc.. preferences. I will not bash another because they didn't a movie I liked. Don't force me to listen to hip hop, and I won't force you to listen to progressive rock!
There will always be naysayers and people who will denigrate the efforts of another.
To them, I should say: "Try to do a better job and come back here to tell us about it."
Assembling all these drivers and testing them in our environment and listening with our own ears would definitely the best. Unfortunately, for most of us, that is not a possibility.
Having someone taking the time to do it, and us being able to listen to that effort, is what the DIY community is all about.
If someone complains about the lack of transparency, the music being tampered with all the steps taken before it reaches one's ears... well... go out and buy the drivers to do your test at home.
Otherwise, enjoy these "blind listening tests" for what they are.
I have since found out that most drivers (unless seriously flawed) will be able to give a very good level of listening experience, if we put effort into it. It's not enough to shove a driver into a box and say "that one is good", or "that one is crap".
With all the tools we have at our disposal these days, we are able to extract some kind of magic from any of the drivers mentioned in these threads of subjective listening, and from so many more still...
New carbon fiber cone mid (almost full range) 4.5in driver from Dayton. Has anyone tried this driver or know where to find its frequency response plot? Looks interesting.
Dayton Audio CF120-4 4-1/2" Woven Carbon Fiber Midwoofer 4 Ohms
Dayton Audio CF120-4 4-1/2" Woven Carbon Fiber Midwoofer 4 Ohms
>>> With all the tools we have at our disposal these days, we are able to extract some kind of magic from any of the drivers mentioned in these threads of subjective listening, and from so many more still...
I like that, perceval. Thank you.
Additionally, I did purchase the cheaper (Vifa) driver and gave it a listen. My teenage boys and I really like it. Ultimately, after about a month of listening we've decided we prefer the Tangband 1320 if only because it sounds more mellow with the music we enjoy... in my apartment... on an open baffle. The measurements done by xrk confirm exactly what we hear. We simply prefer the less accurate driver under the current conditions we have. We also enjoyed the Vifa. It's crisp and detailed. Vocal textures are excellent and I'm really happy to have the driver. I will buy another pair since, for the money (these are inexpensive drivers), they are superior values. The Vifa sounds better than my much more expensive pair of Fostex 168z on open baffle. That was certainly annoying to discover.
The arguing during these comparison tests is totally unfounded. I applaud xrk for his dedication to this experience and hope he continues testing more drivers. If we help by supplying drivers, comments, encouragement, participation and suggestions on improving the entire process we all win as a diy community. No one should be afraid of what the measurements show. We all have flawed systems and rooms. As perceval said "we are able to extract some kind of magic from any of the drivers"... I agree with that. The drivers provided are all 'top notch' products with their own personalities built in. I love that they all exist and we as a community have so much to choose from. Even the drivers no longer in production provide good enough sound for many. I'm glad they were tested, thanks! A third test (if conducted) will further reinforce the many options we are fortunate enough to enjoy.
Some like Klipsch and some like Bose. Polk may measure flatter than JBL but each has delighted owners. Many love Lowther! Everyone has the right to select and enjoy anything available for purchase. It's nice to have measurements from a third party whose intent is to help everyone.
I like that, perceval. Thank you.
Additionally, I did purchase the cheaper (Vifa) driver and gave it a listen. My teenage boys and I really like it. Ultimately, after about a month of listening we've decided we prefer the Tangband 1320 if only because it sounds more mellow with the music we enjoy... in my apartment... on an open baffle. The measurements done by xrk confirm exactly what we hear. We simply prefer the less accurate driver under the current conditions we have. We also enjoyed the Vifa. It's crisp and detailed. Vocal textures are excellent and I'm really happy to have the driver. I will buy another pair since, for the money (these are inexpensive drivers), they are superior values. The Vifa sounds better than my much more expensive pair of Fostex 168z on open baffle. That was certainly annoying to discover.
The arguing during these comparison tests is totally unfounded. I applaud xrk for his dedication to this experience and hope he continues testing more drivers. If we help by supplying drivers, comments, encouragement, participation and suggestions on improving the entire process we all win as a diy community. No one should be afraid of what the measurements show. We all have flawed systems and rooms. As perceval said "we are able to extract some kind of magic from any of the drivers"... I agree with that. The drivers provided are all 'top notch' products with their own personalities built in. I love that they all exist and we as a community have so much to choose from. Even the drivers no longer in production provide good enough sound for many. I'm glad they were tested, thanks! A third test (if conducted) will further reinforce the many options we are fortunate enough to enjoy.
Some like Klipsch and some like Bose. Polk may measure flatter than JBL but each has delighted owners. Many love Lowther! Everyone has the right to select and enjoy anything available for purchase. It's nice to have measurements from a third party whose intent is to help everyone.
Perceval and Godzilla: great insightful comments - thanks.
@Godzilla, so sorry to hear the 168z got trounced by the Vifa. 🙂 I am sure the 168z will be quite formidable in a BIB or other BLH though.
@Godzilla, so sorry to hear the 168z got trounced by the Vifa. 🙂 I am sure the 168z will be quite formidable in a BIB or other BLH though.
The 168z has been well used and will continue to be enjoyed.
https://speakerprojects.wordpress.com/2015/04/27/fostex-168z-projects/
https://speakerprojects.wordpress.com/2015/04/27/fostex-168z-projects/
Well said, perceval and godzilla! X has put in a lot of work so that we may enjoy from the comfort of our sofas the sound of many drivers.
dayton cf120-4, rsponse graph
Hi there X: PE has the FR response plot available, looks like good low mid, not much hi-mid , where response is peaky, probably due to break-up from cone matrix material. ...regards, Michael
Hi there X: PE has the FR response plot available, looks like good low mid, not much hi-mid , where response is peaky, probably due to break-up from cone matrix material. ...regards, Michael