i´m using minidsp now and no way i´ll ever go back to passive xovers . I like the flexibility active gives you and i never heard a good passive xover but that is my fault probably. the power stage in my 3 amps are mostfet and the preamp stage are tubes , the subs are class D . Took me a while to understand how the filter orders work but now my system sound better than ever , realy smooth 🙂
Last edited:
Heavy dynamic range compression and frequency response issues usually...
Loud recordings tend to be the worst of all though.
Loud recordings tend to be the worst of all though.
Yeah, I find it very difficult to listen to commercial FM radio at home. It's so compressed and EQed to death that it wears me out in a few minutes. In my noisy car, it's bearable - but just.
Wow, that's amazing! No wonder it wears us out.
Plus that fact that radio is mostly blah, blah, blah. I swear, people go into radio just to hear themselves talk.
Plus that fact that radio is mostly blah, blah, blah. I swear, people go into radio just to hear themselves talk.
LOL - Yeah, those presenters sure do love the proximity effect! Makes you wonder what their voices really sound like!
I find this a great example of my position on listening tests.
What changed in this "experiment"? The crossover? No that didn't change. Well then the only thing that could have changed is the listener's perception. But if that were actually stable then that shouldn't happen, should it? What's going on here?
This is why subjectivism doesn't work... When you try to hear something or expect to hear something you steer your hearing towards that aspect of the sound, you can hear things that were always there before but you just didn't hear before...
Try your experiment with an ABX test.
Try your experiment with an ABX test.
Perception is never truly stable.
There is probably a certain threshold for some-sort of awareness in this case.
Most of the time we think of awareness as being "instantaneous" (or something close to that), but the more subtle the effect - generally the less "instantaneous" that effect is perceived. Basically it's harder to process initially, so its effect is given less "weight".
Sort of a "look for the positive initially" and sort out the negative later - a defining element with life as we know it. *Hope*. (..where truly rational behavior is an impediment.) 😉
The one I am using is homebrew 'C' code, but I'm really just implementing general principles in a way that I find most convenient: lots of GUI stuff to help me make comparisons in real time. The main features relevant to the sound are:
1. Linear phase filters with adjustable crossover frequencies and slope
2. Phase and FR response correction based on measurements of drivers
3. Time alignment of drivers
4. Adjustable baffle step correction
Better than an off-the-shelf PA crossover, but I fully expect that the code is exactly equivalent to various other crossover filter calculator apps out there, coupled with a convolution engine. There is definitely no "black art" to it, I am finding. But attempting to do it by ear is a recipe for sore ears. Simply by following a fairly rigorous procedure, it all falls into place and produces the sweetest and most musical sound. I would have been sceptical before I talked myself into trying it: surely clinical procedures = clinical sound, DSP = harsh digital sound, but it's turning out to be the opposite. My early experiments were maybe similar to some people's experience here: wonderfully dynamic sound that fatigued after a while, but at first I was only implementing the basic filters - no FR & phase correction, no baffle step correction. Those refinements are what make it into really top notch sound, I think.
Another poster on here @Barleywater, got there first, as he has been a keen exponent of this stuff for quite a while.
Last edited:
I like to think of human perception as a combination of expectation reenforced by reality although ignorant of reality without expectation...
I think I'm getting a bit to philosophical here 😀
I think I'm getting a bit to philosophical here 😀
Subjectivism is no more than a conclusion through logic, not a process.
Ultimately everything is viewed through the "lens of perception" for that individual - even if we rely on information that might be considered more objective.
The irony here is that this conclusion was made from the perspective that we cannot trust our own beliefs. (.."sound" familiar Earl? 😀 ) We might believe we have perceived something, but have we rigorously applied enough logic to that perception (i.e. methodic doubt) to have basis in truth?
Descartes might well have used an ABX test in this context - his premise was most certainly "don't trust; doubt", but there would always be the understanding that interpretation of the results are ultimately a subjective conclusion. 😉
Yes ...
Personally, I'm after the "reality hit" - natural sound has a certain quality about it, for me, that makes it easy to listen to, it's stress-free - even at intense volume levels, with severe peaking of SPLs on transients. Most hifi comes across like an intellectual exercise, one has to digest it, sift it through the brain, make a rational decision about whether it's OK or not. As far as I'm concerned, sound working at that level has already lost, big time - it's not in the race by any measure, for me.
Sound reproduction should give one a huge "positive hit" straight up - any subtle "negatives" one can find by intellectualising the listening exercise will always vanish in the bigger, long term picture ...
Can this be true? Some composers want the experience to be stressful. If your speakers render every piece anodyne, then they're not doing their job.
Is a tension-building moment distressful, or eustressful?
I think more appropriately however is questioning the *cause* for that stress.
IF it's part of that piece of art, that's one thing. (Desired.)
IF it's simply a by-product of the reproduction process itself (a necessarily imposed *addition*), then that's quite another thing. (Undesired.)
(..and it's the later that I believe that fas42 was referring to.)
Last edited:
Uh, yeah... And what's wrong with that? We can't learn and judge over time? Never meet someone (like me) who you thought was intelligent and interesting at first - but with time just turns out to be an annoying loony? What has changed? The loony, or your perception of him? Which is impression was correct? The first impression or the one that came with more exposure?
Making up stuff again, I see.
Correct, of course ...
If the intention of the piece is to be, say, highly discordant, some bizarre experimentation on a piano perhaps, that might produce some type of 'stress', as in ... huuuuhh??!!. But that shouldn't get in the way of able to hear that it's a real piano being "mal-treated" ... 🙂
Sometimes it's more stressful hearing something that is almooost quite right, but still "wrong" in a disturbing way - then something that is obviously very wrong ...
Even totally synthesizer music, electronica from the 80's with huge, artificial reverb spaces, can be obviously "right" or "wrong" - the harmonics gell, or they don't, it's a thumbs up, or thumbs down process to evaluate the reproduction quality.
There are a million ways to justify what in the end is simply a humans inability to be objective in a situation like a listening test. You will never get me to accept a casual listening event as anything more than a totally biased subjective experience.
Sure, over a long enough period of time any blind fool has to sooner or later come to the truth (well maybe not for religion). But that is not what happens here, or almost anywhere else for that matter. That's why I say that measurements trump the subjective impression. I have never found the measurements to be subjectively wrong over the long run. No one else who studies this problem has either. But there remain those among us (not YOU of course) who believe that they can tell good from bad by "just listening". Yeah, OK. It's just that tomorrow they will have a different opinion, kind of like the wind, (which sounds like that is OK with you), but the measurements will still be the same.
Last edited: