The LPS that I am using is my cheap lab supply and as Peter would say, it helps "get out the Ssss's" [...]
Yes I also find that sssss in the Silverstone and those hard highs are the bad part of the card, for the rest the sound is fantastic.
Anthony, Juan,
If this goes on a few posts more it will start to live its own life. I mean, the suggestion is put forward that I don't like the S'es (or something) about this card. Well, it is the opposite. I use the card *because* of it. Not S'es in particular, but the separation the card brings forward and what I describe as "more rough highs".
There is much more to this than what can be explained through words and possibly nobody will ever "see" what I'm after with this, nor might you be able to achieve it. But I'll try to give a few hints :
First of, take it that I am capable of more SPL in the highest regions than anyone else, unless you have the same mid-high horn as I use. For the tweeter this is 80W continuous power (over 300W peak). This is not to show off or anything, and it is also not about absolute qualities, but it IS about that I am really utilizing this power and that you can bet that if something is/goes wrong, it will express more easily than elsewhere and it will slamm you in the face.
What's already from quite many years ago (5 or so ?) is that a roughness in the higher regions (say 1000-2000Hz up to maybe 5000Hz) can be (or has to be) smoothened (filled up) with higher frequencies normally lacking. So, without this filling up you can perceived holes in the sound. This will be nothing strange or unusual, but notice that first all must be very speedy to perceive that in the first place; say that when the speed is not on par the sound will be more grey.
Side note : When I must listen to a paper cone driver for the higher regions this is totally unbearable. This is because paper is not fast enough and you will hear ... paper.
So when the speed is there, all is able to follow those 1K-2K frequencies, but because of the less energy in the way higher frequencies (5K and above) these holes appear. So, what I do is linearly increase the power to the tweeter so the higher frequencies present (in that same sound/music which implies the holes) now fills up those holes and sound gets smooth and how it would be in real life. In the mean time the highest regions of the tweeter are fed with close to 20dB more of energy; can be less and depends a but on other circumstances, like the operating system. Btw, I think it just have been the first W7 version where I explored this, because that was an OS with "holes" (SP1 solved that). And notice that W8 goes to a kind of extreme to the other side, where too much highs kills the cymbals; you know my stories about that.
You can already see that I'm operating in "illegal" areas.
But maybe not, because the energy of the high frequencies just *is* more low. Way more. This is because of microphones being less capable in that area and also because of how "digital" works. Well, never mind my (or the) reasons, but this is how I try to improve things, which normally are not even attacked by, say, loudspeaker manufacturers.
And let's not forget Fletcher-Munson curves, which is another reason. Ok ...
Without me changing those curves anymore, the sound of especially W8 is too silky. I can bring this down to the sort of fact that those holes are now overly filled, or better put in the realm of today : the output of the tweeter is too high.
Notice that it is not easy at all to "just" say that, because change a few dials in XXHighEnd and all goes to a very other direction. So, it is about what I perceive from it as a general characteristic and maybe it took those 5 years I apply the tweak to learn that the tweak itself is still OK. IOW, when I remove it, sound becomes quite "bad". Stuffed ears like.
What this Silverstone card does is bring back the holes but now to the extend it should. Maybe I shouldn't repeat myself, but it is about how "separation" is better, where "separation" actually is about resolution but at a somewhat lower frequency. So, on/off gets better of it (off is a hole). In the mean time, 1.186 is able to let the transients better "flow" and where previously a transient was a "one go bang", it now has a slope. This too is resolution, but of another kind and of the highest frequencies.
Still there ?
Although it wil be pure coincidence, what this card does is just right. I could add "to my system", but I don't want to say that because my system tries to do everything and all right and part of it is (thus) that a speaker should not show a flat frequency response curve.
Before we think that I now thus can not judge what will happen to your music reproduction - don't worry because so many other things are different, and this counts for everybody. For example, we could also say that what I am doing is not fair because of a 115dB sensitive speaker which you won't have. And if you only know how immensely much that does to speed ... If I turn this 115dB into 112dB it is already unlistenable for me because sound gets completely dull because of it.
Point could be that possibly I am better capable of judging this all (the higher frequencies) because if it goes wrong it goes wrong with me first. I mean, I just have way more output there and you can imagine that if e.g. an S is expressed wrongly in the base (say the software) over here this is amplified vastly and I will hear it before you ever can.
In the mean time, of course, you can wonder how my "S" won't get way more loud, but it does not work like that. So say the fundamental of an S is a fairly low frequency more square sound and that low frequency is not louder at all in my speaker. This is how the S won't get louder. However, because "square" means "high sine frequencies" it is exactly those higher frequencies needed to let develop the S and wonder oh wonder, how holes would exist with those lacking higher frequencies (of sufficient SPL). So because of the higher high frequency output my S is more smooth.
Got that a bit ?
This is all quite opposite from what intuitivity might tell you.
The Silverstone allows for better separation in the lower frequencies which sure makes the S more rough. Still this works for the better because in those holes now there, there's the opportunity to have higher frequencies sing. I know, this can't be explained by means of S-es but in the end you should be able to perceive them more real. Mind you, when these higher frequencies are there and a means of it is the Q3,4,5 = 1,1,1 setting. But what I told about a few weeks ago with cymbals, is the better example; while the more flattened grey sound is taken out and is turned into "holes", the singing of a ride cymbal comes forward, and right in the middle of smashing rock. This really is totally obvious and the card does it. And funnily enough, this does NOT need those higher frequencies to be louder because which cymbal could be represented well even on cassette tape ? the ride cymbal (hitting it in the middle as intended with a ride). So all that needs is this separation and this card does that.
Long story eh ? Well, it is meant to tell you that you should explicitly NOT try to smoothen that separation, but seek the solution elsewhere. Or just don't use the card of course.
But I think it is already clear from the reports so far that generally the card performs better, but you (or I) may wonder how, if in the mean time it is pereived harsh on the S-es. I don't mean "that is thus impossible" and I merely mean : so what is better then ?". So be careful because at least I never like to be subjective and better is only better when each of the technical elements is better. Each = all of them. And if one is not (could be an S) then this should express throughout though in other elements.
Ok, I'll stop here.
Peter
