Here is another status update ...
First of all, the parts I was waiting for, I am still waiting for. They are over due already, and day before yesterday I heard that they will be shipped July 16 now. Assumed that will be true, with another week of shipping time the first DACs should be ready for shipping near the end of July. If I'm only ready for the administrative side of things (never got around that anymore, so far).
So, I guess we are really almost there now !
In the mean time, again the additional throughput time has been the most fruitful;
A few weeks back the latest version of the DAC board was received, and only last week I got around to integrating the CPLD logic (made at a distance of which I told about I think) with the XXHighEnd logic, both providing the special 384/768 mode. And, after 2 days of work on the XXHighEnd side, I can say that it all works above expectations. Above ? how ?
On a side note : The whole technical and electrical idea behind this was created November-December last year (2009), and while this implied a complete new design of the DAC (board), while that was produced I got some other ideas that could nicely combine. And, now really having that DAC board, forced by some technical problems in the 768 area I again got some other idea. All the (three) ideas could combine into one solution, and that is what I have been listening to for the last two days ...
Let me start at the end, and an actual by now rather common story : yesterday I played one of the albums we run almost weekly for explicit testing, and I asked my wife "do you know what this is ?". Allright, you know the answer ... she did not.
What the heck happened this time ?
This time it is all about jitter. 3 folded.
1.
By means of a plain dirty trick, I was able to decrease the phase noise to -128dBc @ 100Hz. Very very good is -110dBc, and much more common is -60dBc. Mind you, this is for 100Hz, a common measurement point for phase noise. Also, jitter is the most audible in the lower frequencies, or better : it is the most applicable to the lower frequencies (something like : the longer the wave, the more time there is to notice fluctuations).
The trick comprises of a technical application, and the figure is derived from the base specs from the oscillators I use.
2.
By means of a "logical thinking" trick I reduced the jitter by another factor 2 (but think in dB, which would be 6). This makes use of available headroom when the 768 input is NOT used. Hahaha.
Like I said in the above, forced by technical problems in the 768KHz area (which is ~25MHz at the bit level and which operates outside of chips in my case) I thought to trade the (thus far) unuseable headroom for better jitter specs. And at this moment I think I don't even want the 768 input to operate anymore, because what is gained by (whatever it is) I applied is much much more powerful then shifting away high ferquencies further (and remember, 384 (352.8 ) was enough for that already). But, if 768 gets to work afterall, it is just a choice in XXHighEnd.
3.
Last but not least at all, something completely new, and let me be the inventor of it. I am not sure yet how it works or what "influence" is eliminated, but I tend to call it interchannel jitter;
We all know about the good merits of separating analogue devices like mono blocks instead of one stereo amplifier, separate PSUs for the separate channels or even separate DACs for the channels (inherently present at the chip level for the Phasure NOS1), but I found that there is also an "opportunity" at the plain digital level, where nothing in that path is routed to one channel anymore. The effect of this should be equal jitter for both channels. Or at least the way I set it up, should have that effect.
Huh ?
Sound wise ...Let us start with the latter ... interchannel jitter.
It is the most drastical sound change I ever met in any of the steps I ever applied, might it be in my younger days, or just the "XXHighEnd years".
It implies a channel "separation" in a dimension we didn't know it was there. So, there is channel separation, and when measured it can be 80 or 90dB or whatever figure we know it is inaudible. Or, when you listen to a really hard panned guitar from the left speaker and just can not hear it from the right speaker, well, that is good channel separation, and generally we are not worried we are bothered by it. It is too inaudible.
Now, in the above I put "separation" between quotes, because it is just about the opposite :
channel integration.
Let's keep in mind that I dedicated this phenomenon to what I perceive myself, and maybe there's a more official/known phenomenon for it. But now let's try to explain :
Let's first assume and aknowledge the audibility of jitter. I, or I think I can well say "we" don't care about low jitter specs of a device, and as long as XXHighEnd can influence *and* assumed that the carrier of this can be jitter only, we hear jitter. Oh yeah, we hear it very very well (the differences).
Now suppose that the differences we can create with XXHighEnd would be there between the left and right channel. That would be a quite messy sound, a bit depending on the extremes you found in XXHighEnd settings. Now go one step further, and let's learn that those differences just *are* there;
Also good to recognize is that we aren't used otherwise. It is all normal for us, because no tool was available to tweak that difference. The only thing fair to say (without real scientific knowledge) is that when the general jitter level is very low, the differences between left and right are also very low (and the other way around).
Now, what I did, was 100% elminate that difference between left and right, meaning : The jitter is still there (that assumed), but however it exhibits, it does this the same in the left and the right channel, and this is at the sample level. The audible effect ?
A so super full warm sound that you really can not imagine it is coming from the same speakers and amplifiers.
Yesterday I was explicitly listening for what to write in this description, but I can't describe it in a technical fashion. Oh, I could say that you will have the perception of a wider sound stage, but it is not that. Physically (for dimensions) is it not that. But there is a stage full of sound, instead 5 microphones at 5 places. There is totally no sterility left. It feels like how it was real, at the performance. It is also not like a mouth has become too big or something, but merely a far more exact projection of a small mouth which can't sound small at the distance you listen.
Let's take one of my albums I always try after a bigger change. Get Yer Ya-Ya's Out! from The Stones. Do you have it ? well, put it on now;
I think I talked before about the improvements achieved on this album, and that it is actually completely flat far away sounding. It will have vastly improved, but still, if you compare it with other good recordings, no, this is not it.
But now ? man, you would bet 10 million that I had a unique very well done remaster. There is completely *nothing* left of the flatness, and I even imagined Charly Watts to look happy, and applying variations I never saw him do in clips.
Btw, yes, I watched carefully whether the focus on voices or instruments was still there as should (instead of DSP-not).
Allright, on to the next two jitter subjects, far more normal.
Ehh ...
This is what makes your albums (again) unrecognizeable. In the end it is (at least to me) a very strange thing. Although I must admit that along the line the mids again came somewhat more forward - which has its own merits -, this time all kind of melody lines come forward, never heard before. It can be bass, it can be by background bells, it can be a far away guitar, but it is always about a melody line. Well, you can imagine that if you were used to listen to a track with a certain melody (whatever that was) and suddenly the melody changes because another line comes forward, well, your track sounds different up to unrecognizeable.
Part of it is created because of more energy which comes forward most in bass instruments. But it is also about "sub harmonics" (which don't exist) forming an attack layer for the particular instrument and even voice. To make this clear I better refer to drum veils and their attack. This was already very good, but got better again. There's more warmth under it now (the so called sub haronics).
Here's another example : Faith from George Michael. I mention this one, because it occurred to me that it wouldn't work as I wanted before. You could also say that I didn't get the feelings with it, I recalled from when being a tad younger, women involved. It was too sterile and had no slam. It didn't have the power the tough guy on the cover expresses, no matter how gay he is.
Well, you saw it coming ... now it works. But wait, what does that mean ?
Throughout the album a bass is used which really does the job on all of the tracks. It should be the most profound of all, and it should incorporate the synth created vibes in the mean time. So, very very powerful, but at keeping those vibes audible. That it goes along with the super duper warmth from the former subject is another matter, but together with those most hard drum rythm slams it draws tears (yes, it did). So, it is the combination at work here. It *is* a super tough setup album, but it *is* sung by a gay guy with gay warmth.
Without the slam the combination isn't there, and it's only one of those dozens of two gays behind a keyboard guys with women voices.
Wow, you really can't imagine what can happen when all has come together. I tried some of the Shulman albums, which is ambient and only 10 minute or longer tracks. I know so well how boring it actually is, because it is not only 10 minutes of the same, but a whole album of the same. Still I like it for a kind of background louder music (hehe ambient eh). But today ? today I can't stop from being attrackted to each of those seconds, the album throughout. It is one big pile of intelligent detail harmonically formed into a story you are sucked into.
Ok, unless 768 brings something additional, I'm done;
The Phasure NOS1 now works for all types of music, and I wouldn't know what to do further. It should be the best on the globe.
Peter
