I've asked a question. That is not the same thing as 'putting words in people's mouths'.
You don't explain what that 'DDR' really is but instead go on with speculation.
You confuse a speaker driver with the human brain. A driver can not interpret a signal and 'know' its meaning. It also doesn't need to. A driver is a electro-mechanical device that follows the input signal. Everything about frequency response and dynamics can be measured. You are obfuscating the topic with a meaningless term.
Hmmm... I don't want to add more dross to this thread, but I have to say that there are folks who enjoy color/flavor in their system (by sheer preference or might be cases of ignorance). And no sin in that...
I understand that most likely you and Xrk971 are advocating to start with neutral and then add color if prefrerred (through EQ or other means). There are folks who would start from the other end - start with a certain color/flavor and change from there...it might sound counter intuitive though... but each to his own. 🙂
Guess nothing new that I have added and getting off the
:Back to measurements...
Last edited:
For CHN-70. Factory data (Earthworks M30 in anechoic chamber on IEC baffle using LMS) scaled and laid over top XRK's
dave
What excitation signal was used? Is there any information about the anechoic chamber? How large is it? What's the lower cutoff?
There's something wrong with the Markaudio data. No driver should exhibit 15dB swings under anechoic conditions unless the driver is really badly broken. This has already been discussed. Please see http://www.diyaudio.com/forums/full...surements-different-measuring-techniques.html
Starting with colored sounding drivers is fine. But how do those people know the driver has the color they are looking for if there's no data available that would objectively describe driver performance?
You have selective memory?
For the upteenth time, the lower limit of the Dynamic Range and what happens to that when you add a strong signal .
dave
Adding those two signals simply creates another single electrical signal the driver has to reproduce. If the driver's frequency response is flat then it does it more accurately than a driver that adds ringing. You might like the latter better than the former but sound reproduction is about reproducing a signal and not about creating new signals.
Last edited:
Especially poor when using fullrange drivers.
Measure at different input levels. You'll notice how distortion goes up until the driver runs into thermal compression. It happens very early with wide-band drivers and is the number one reason why multi-way speakers even exist.
All drivers. A bass driver does not do HF at all well, and subject a tweeter to too much bass and it tries to launch itself across the room.
dave
That tells you the top limit, we are interested in what is happening at the bottom.
I already said that too.
dave
For CHN-70. Factory data (Earthworks M30 in anechoic chamber on IEC baffle using LMS) scaled and laid over top XRK's
dave
Thanks for taking the time to compare (and the Vifa and X's measurements as well)... That's not a remarkable difference to me. The Vifa measurements showed a difference in the highs too as well as higher output around 2.5-3 KHz in X's measurements.
It is obvious we need the graphs on a similar scale to be able to compare them. Things get overlooked fast if we don't.
Still wondering how this driver would do if equalised. I like the off axis response compared to all of the other drivers posted in this thread so far.
I would love to see the higher range MA's to see if they do something similar.
There is no way in Hell that I will EVER post another measurement on this forum, nor will I allow further posting of measurements from my web site. That is copyrighted material.
I set a microphone at my listening chair and compared responses with REW. I EQ's the CHN70 until it approximated what I already had for the A7.3. I then went through my test track alternating speaker. There was a significant difference in the sound of the two drivers.
Since it is now well established that the CHN70 was designed for a specific niche in the Japanese market, and Mark Audio has canceled their diyAudio account, isn't it about time to drop the issue?
You seem to be confused why multi-way speakers exist and how tweeters and woofers are used.
Tweeters are optimized for reproducing higher frequencies and woofers are optimized for lower frequencies. Using them for tasks they aren't designed for would be plain stupid. Like designing the best hammer but using it to drive a screw.
Wide-band drivers aren't capable of reproducing all audible frequencies within the dynamic range of human hearing without major distortion and compression. That's why tweeters and (mid) woofers exist. They are optimized for a limited bandwidth. In return they can go much louder with way less distortion and compression than any full range driver.
*sigh* At this point I could only repeat what I've already said a couple of times and it wouldn't help anybody. So I stop.
That's not how it works, unfortunately. Please see http://www.dirac.se/media/12044/on_room_correction.pdf
No, I'm well aware that a lot of people are obsessed with marginalia while successfully ignoring the elephant in the room. That's still no excuse for such behavior.
https://yourlogicalfallacyis.com/bandwagon
Last edited:
For CHN-70. Factory data (Earthworks M30 in anechoic chamber on IEC baffle using LMS) scaled and laid over top XRK's
dave
When reconstructing data pulled off a log graph with a 120dB vertical scale, and photoshopping it to overlay with data on a graph with a 45dB vertical scale, one has to wonder what the errors might be in such an operation. The signal to noise ratio in the data that I present, actually allows one to see variations of 5 or 10 dB on a quantitative scale. Whereas the scaled MA data superimposed, has a much lower SNR and high level of vertical scatter. There are some rather large deviations at 7kHz and 15kHz. Is that a result from errors/noise in the photoshopping or is that real? It would be useful to have the scale and tickmarks of the MA data superimposed as well so we can compare.
Here is the MA data on the 120dB scale with 10dB divisions - the scaled data from above doesn't resemble this factory data at all:
Last edited:
>>> You have mistaken the majority of wideband driver users for people who give a toss.
LOL, Scott. I will never forget (many years ago) comparing the Fostex 168z with a two-way I built using Dynaudio dome tweeters (that cost more than the Fostex 168z). I expected the treble to be much better with the domes but was surprised how the overall effect was much more pleasing with the full ranger. I was (and remain) shocked how treble is reproduced by such a large driver. After sevearl months of listening and comparing I eventually sold the Dyns and have been enjoying the full range route ever since. I'll add a tweeter when I feel it's necesarry - I'm not a glutten for punnishment when it comes to good sound.
Interestingly, tho not directly related, using Xbaffle to simulate open baffle results really produces some horrid looking charts and a few inches variation can make a chart look terrible. But in reality, once the speaker is built - if it's a successful project and sounds good - charts be damned, the music fills the room and all is right with the world.
I guess the lesson is (and will remain) that you can't judge the sound by the chart and must listen. Subjective opinions are just as important as objective data. I'm a big fan of having both prior to making a purchase.
Thanks to all for your input and effort on this thread.
LOL, Scott. I will never forget (many years ago) comparing the Fostex 168z with a two-way I built using Dynaudio dome tweeters (that cost more than the Fostex 168z). I expected the treble to be much better with the domes but was surprised how the overall effect was much more pleasing with the full ranger. I was (and remain) shocked how treble is reproduced by such a large driver. After sevearl months of listening and comparing I eventually sold the Dyns and have been enjoying the full range route ever since. I'll add a tweeter when I feel it's necesarry - I'm not a glutten for punnishment when it comes to good sound.
Interestingly, tho not directly related, using Xbaffle to simulate open baffle results really produces some horrid looking charts and a few inches variation can make a chart look terrible. But in reality, once the speaker is built - if it's a successful project and sounds good - charts be damned, the music fills the room and all is right with the world.
I guess the lesson is (and will remain) that you can't judge the sound by the chart and must listen. Subjective opinions are just as important as objective data. I'm a big fan of having both prior to making a purchase.
Thanks to all for your input and effort on this thread.
- Status
- Not open for further replies.