Can you update the first thread with locational information of the set up within the room, drivers and persons (speaker height, tow, etc)...and maybe room acoustic treatment status.
An EQ'ed full range ??
I've heard speakers like that - and they suffer hugely from extreme lack of detail and max SPL. One tiny driver, that barely hits the high notes or some middle sized driver that never gonna push any air - simply a dead end. Unless you are going to play very low - only - and accept all the breakups and added distortion.
Optimal hifi experience?? That requires both finesse and power. I have tried almost any type of DSP that you can find on the marked today - and even though it might look magical on the screen when you dial in the numbers and push the button - it cant defy the laws of physics. Unless you test it under some extreme condition, where people would never listen to music in the first place - which makes it close to pointless.
One thing I would agree with you on - is - that understanding a driver and using DSP-processing at a level where basic nature laws are not broken beyond saving. Then we can talk about using drivers, that without the help of DSP, would have been left unused, if only classical passive x-overs were available.
But simply because several people could not hear the difference in a - maybe - flawed test. Does not mean that suddenly we can just brutally force any driver to sound fantastic, because of the rise of DSP-possibillties 😉
Dont need such driver..... even before cheap DSP's - that driver was silly and annoying - like LP's and tubeamps - just something that was build to try and solve a problem that we dont need to fix or insist that everything was better in the old days 😀
Please define ''sound fantastic''...
what a DSP does is simply to apply filters, nothing more nothing less. You can unbalance the FR or make it flat (based on a specific weighting reference), as you wish or as you are able to.
You pay a lot for these drivers (like Lowther's, Supravox, etc..) simply because they are designed as the ultimate widebands and can reach decent SPL.
fullrange (wideband) drivers are probably the ones that benefits the most from an EQ.
Unfortunately, owners of these drivers are so used to their (very) raising FR, that they are confusing dynamics with faulty FR... Usually they looove jazz because when the trumpet hits, it swings right on the driver's FR peak. 😉
so once you EQ that little wild beast, they lose those peaks. Sounds better for all kinds of music, but it loses the ''artificial dynamic''..
But, really, an EQ'd fullrange driver can be very very surprising. You are, obviously, limited by the lower frequencies, and the realistic SPL you can reach is often below 100db, but it's perfect for smaller rooms.
I still have my pair of vented 22l enclosures with FR151 and once carefully EQd, they are flat from 28hz, which alone is amazing since the driver has a Fs of 50hz and a Xmax of only few millimeters...
Hotly debated topic I see here.
What I am taking from this thread is that this information is highlighting that frequency response over this band is the area that effects the human listener the most. Not that a cheap driver will match an expensive one. There are many areas like impulse and waterfall plots we can look at to distinguish differences.
I feel the whole thread is more showing how we all have our own self beliefs, but we can all better our craft if we can keep an open mind across the board.
Paul
What I am taking from this thread is that this information is highlighting that frequency response over this band is the area that effects the human listener the most. Not that a cheap driver will match an expensive one. There are many areas like impulse and waterfall plots we can look at to distinguish differences.
I feel the whole thread is more showing how we all have our own self beliefs, but we can all better our craft if we can keep an open mind across the board.
Paul
The idea is not to defy the laws of physics.
The idea is the get the most from those laws.
You just can't do much with passive crossovers. It's extremely limiting. It's limiting both for the final results, but it's also limiting for testing and studying drivers. With a DSP, I can test 10 or 20 different crossover points PER HOUR if I want. How would you do that with passive components?
Also, even with active crossovers, it's not enough if you're not using the EQ possibilities. Not only you can correct the room (to some extent) but also, and more importantly, you can correct the natural driver's FR... That means you can use a very efficient driver with a lot of SPL potential and wide bandwith, but with a bumpy FR, and make it.... ''perfect''.
A DSP is the audiophile's best friend.
JonBocani, is there any reasons that you do not want to publish each drivers' corrected FR response and distortion plot?
This bandwith was not selected because it ''effects the human listener the most'' but rather as a first step before testing both the woofers and the tweeters, later on.
Also, for practical reasons, it was the only bandwith that was listenable, taken isolated.
You cannot really test 20-640hz bandwith, no more that testing 1600-20000hz would be possible. Not on music anyways.
But 360-7200hz, yes. Very listenable. A lot of musical information still go through. It's it feels balanced.
And, yes, that's exactly the conclusions of the test: a cheap driver CAN match an expensive one, given said cheap driver works within his thermal & mechanical limits. That includes max SPL, bandwith and power response.
And of course, with a EQ/SPL matching.
Impulse, waterfall, THD... Yes. All differences on paper. But this test was about real human ears, not microphones collecting data.
The endgame is to pleases our ears/brains.
I know that a DSP can do almost magic. But I thought we were way past the question whether we used EQ with DSP. Of course I use EQ with my filters in my DSP - anything else would be beyond crazy 😀
You still cant iron out a driver with DSP/EQ in its breakup mode.... cause it will ring - just lower in level 😉
That is why you cant make a 10" play optimally at - lets say 5Khz 😎
I have an Accuton 5" that break up at 4800hz.... not gonna use that for tweeter duty - no matter the EQ 😀
Yes, it was not kept. I don't have any data. Can't do it now either because the audio lab is dismantled.
Like I said, we needed to find a threshold first (positive identification) before putting more energy and time into that, otherwise it was absolutely pointless.
Drivers could have been 20% THD mixed with 0.00001% THD, or even entirely un-EQ drivers, if there is no positive identification... There is no reason to look further.
See the logic?
Threshold.
BUT... still...
I EQd a driver in that very test right in the break-up zone (Dayton RS225)
😛
And no one could spot that ''corrected break-up mode'' from a... compression 950PB-Beryllium that is SURE not breaking up at 3khz....
OK, please keep the plot and publish it with the result next time you perform similar test. I wish to perform test like yours myself, and your plot would be the reference.
If you want to replicate the test, I would say your biggest challenge will be the ''rotative cube'' we built custom for this test, with the switch box, etc...
As I said you don't really need to go crazy about the EQ at first because the worst that can happen is: you will get a positive identification. Then, all you'll have to do is go back to the EQ and try to flat it even more.
But the good news is: you'll have your threshold at first. While I started ''too high'' and had to go down to find the threshold (which in my case was to intentionnally unmatch the SPL of drivers OR change the bandwith on one driver)
Most widebands are designed to be listened to off axis, cutting the high frequencies can screw up the power response.
I know. I'm still thinking about what to do with speaker location. It should be much easier to distinguish the speakers sitting next to each other, comparing to your method.
So you have convinced yourself, but it's not "balanced", obviously. You have drastically distorted the music and so shouldn't make any judgement about how the music "sounds the same" other than being equally distorted which I consider the most likely real result from the test and not at all surprising.
Just found a picture of those.
I wanted to use them in a blindtest against a pair of Bowers & Wilkins CM9, with an EQ that mimics the FR of the B&Ws...
Never did it though, but these FR151 are able to cover 28hz-20khz flat, so I guess that would have been interesting.
Agreed. What I was getting at was that we sometimes get caught up in areas that in reality don't matter as much or at all. In this case, we look at tiny details of a more expense driver when in reality, across this mid section, we don't notice this at all as humans. Something being statistically better and being humanly capable of recognizing it are very different things.
Paul
- Status
- Not open for further replies.