9y--504

[boleary]

Hi Peter, don't know if I'm doing something wrong but 9y--504 just doesn't sound right. Both Engines 3 and 4 seem to smear the HF's. Also both seem to place vocals way down inside a hall where they sound very muted and incorrect on my system. I used the settings in my signature for both engines. Wish the news was better.

[PeterSt]

Thank you though !

[manisandher]

IT WORKS!!!!!

Thanks for getting this sorted out Peter - really appreciate that.

The lowest Q1 setting seems to be dependent on the buffer being used. With a buffer of 3072, I can take Q1 down to +3. Lower than this, and I get distortion... and the music plays at half-speed weirdly enough. With the buffer at 128 (as low as I can go at 176.4KHz), I can take Q1 down to -1 no problem. Below this, I get the same distortion as before.

So, with the buffer at 128 and Q1=0, I played a few minutes (only a few!) of music - just one track actually. I tend to start with an HRx 176.4KHz track to assure myself that everything is bit-perfect at 176.4... and indeed it is.

How to describe Engine#3 vs. Engine#4? Well, with Engine#3, you see the trees - hell, you see the branches and the leaves and the insects crawling around eating each other! With Engine#4, you see the forrest in it's entirity. It's a more 'organic' sound. It's less 'shrilly' and 'shouty'. It seems to have a nicer tonal balance - the harmonic texture of violins seem fuller and more pleasing. But whether this is more accurate or not, I don't know right now.

Just my very early thoughts.

Mani.

[PeterSt]

Anybody else having an opinion already (please) ?

I guess I gave to say this (rather pretentious it is though) :

From off Saturday I have been testing with sound (to listen for skips, repeats and everything), which merely goes by means of the same tracks, once one have been found which does something wrong. So, you can well say this was over 20 hours of listening to merely the same.
Saturday afternoon suddenly something seemed to have changed to the sound. It was without reason, but since I liked it and actually was "testing", I thought to address it later. But, as you can imagine, a bunch of hours later - well into Saturday everning, I forgot about it and I got used to the sound I guess.
Last evening (Sunday) I only played one album which was more or less familar to me, which was Sgt Pepper etc. - the remastered version which I never played before. I did play others from that collection before, which I liked better than the originals. This one (Sgt. Pepper) I did not like much. Too fresh or something.

First thing this morning I played a couple of well known albums, and the only thing I can say is that I seemed to like them better than ever before. But maybe it is a wrong thing for me to listen early in the morning.

But there is something else I noticed already, which a kind of was confirmed this morning : I seem to hear noise which I didn't hear before;
Yesterday and Saturday I already tried to justify it as normal background noise, and could come to no other conclusion than that it just is. Notice though, that one of the arguments for it being justified was ... Engine#3 does it too ...

Because this was again about not much known albums, in the end I didn't pay that much attention to it, although I related it to the suddenly changed sound for the better, at that time. Now :
To my surprise from two albums out of two I tried this morning, my speakers again pointed me to this "background noise", and this time I was sure this just wasn't there before. At the same time I heard details in the music not being there before ... (with Engine#4 even more than with #3).


I still think that what happened at Saturday afternoon was just my imagination, because I actually didn't change anything relevant. The change was too big to be "real". But on the other hand there's my experiences after that, which seem to confirm it just is so.

And so, boleary, you most probably are right just the same.

I know I can trust your findings, especially because you move in live performance circles. But, I hope you can twist your findings into a direction which is opposite from what you perceive from it now. Not because I want you to, but merely because it seems contradictionary to my own findings. Otoh I don't trust this background noise (IOW proove me wrong better than you already did, and I may start a search for this noise don't do that, and I think it's good. )

Thanks very much,
Peter

[PeterSt]

I tend to start with an HRx 176.4KHz track to assure myself that everything is bit-perfect at 176.4... and indeed it is.

Yeah Mani, it is and it was. It is with backround noise, and it was without background noise.

Now, who is telling the stories here, eh ? me at least !

[PeterSt]

The lowest Q1 setting seems to be dependent on the buffer being used. With a buffer of 3072, I can take Q1 down to +3. Lower than this, and I get distortion... and the music plays at half-speed weirdly enough. With the buffer at 128 (as low as I can go at 176.4KHz), I can take Q1 down to -1 no problem. Below this, I get the same distortion as before.

Interesting ...
I just looked it up, and with Q1 at -1 it means you're communicating with the driver at the level of 12 samples @ 32/176.4).
It also will mean that the bursts you'll imply @ 3072 consume so much of your system's resources for a so long time, that it will hold the (ultimately) 12 samples for too long and you will receive "skips". If one out of two doesn't get through, this indeed will express as half the sample rate (the driver just keeping up the sample rate).

For what it's all worth of course, but from a technical perspective it looks nice.
Btw, I didn't optimize the code for this yet, so maybe it can even go lower. BUT :
Once you (anyone) reached Q1 = -1, don't try lower with this version, because it will be the same as -4, implying 6 samples. So, -4, -3, and -2 all imply the same. I knew this already, but let it be because I thought 32 samples to be a physical limit. Something for a next version.

Peter

[Gerard]

Peter,

I did not did a-b but i haven the feeling Engine4 has some of Engine 3 in it. To me it sounds better than the old version.

 

(Q1 lower than 7 gives problems overhere) (But before this i never could get lower than -1) 

[Leon]

Hi Peter,

First of all I like to say that I'm happy we all now can start wondering what would be better engine 3or 4.
listening attended to the first album (flac decoded) I noticed that after a few tracks the music stops playing and a message pops up "No Track given" and than a little later the message Engine #4 did not start within the expected time! Pressing "play" again will start music again.

I have Q1 at 7 and the others at 0, processor setting at No Appointment and playerPrio and ThreadPrio at Normal, split filesize at 30. Setting the Q1 lower than 7 will start hickups in the audiostream.

Please let me know if one of these settings should be different, if you can say anything about that at this moment that is.

As far as the SQ, I love it but then again I love engine 3 as well and need to do a bit more listening before I can say anything further.

Well done so far .

Regards
Leon

[boleary]

I'll give it a longer listen tonight. Am off to work early today. I must admitt that what I posted was a first impression and I was in a "Super Bowl Sunday" frame of mind    (The Saints won!). But I'm fairly certain I heard "something" with both engines that was causing the sound to be "off".

[PeterSt]

Take your time, be honest, and in the end help yourself with it. I'll listen ...

Yea, I heard something last night from my wife about New Orleans guys being in the final. I asked here whether she was talking about "the" Super Bowl.
And this morning it appeared they even won.

[PeterSt]

Hi Leon,

No, I can't see anything specifically wrong, and that "No track given" will be something on my side; I don't know what it is yet, but I'll try to cover for it.

Important and useful or not : I can reach Q1 = 3 for my USB device (18/96, used as 16/44.1). I use Appointment Scheme 3 (if that matters at all).

Regards,
Peter

[manisandher]

I hate it when this happens. I should be working right now - finalising the content of my new website. BUT... I can't stop listening to music!

I've been using my 'reference' rig - PMII directly into Berning OTL 300B driving AKG K-1000 headphones. I know it sounds trite, but I've never heard a digital source sound so analogue. Listening to 'So What' from 'Kind of Blue', even the tape hiss at the beginning of the track sounds different. I know it's impossible, but I can almost hear the digital sampling with Engine#3 - the hiss has a granular texture. With Engine#4, this is gone. It's smooth tape hiss. I could swear that I'm listening to an open reel machine.

The music is just lovely. Not necessarily more 'life-like' (whatever that means)... but lovely. Nothing shouting out at you. Keith Jarrett's piano now sounds like a big and expensive piece of kit, rather than a small stand up made of MDF.

There is one caveat with all of this though. Although I haven't tested it, I suspect that Engine#4 needs QAP... on my system at least. I can imagine that without this, it will simply sound too smooth. But KS together with QAP is like peanut butter and chocolate - great on their own, but sublime together.

I'll put my neck out and say that for me, with my DAC (NOS at >=176.4K) Engine#4 may prove to be a greater step in SQ than even QAP. If indeed Engine#4 is adding some distortion and/or noise, then all I can say is that I like it!

Mani.

[pedal]

Manisandher's listening impressions are quite similar to mine:

Just installed 0.9y-5-02. The new "KS" alternatives poped up in the Output Device window, as expected. Engine #4 working just fine. When switching between Output Devices (with/without KS) XX automatically switches between Engine #3 and #4. Thats nice!

Listening: (Jazz at the Pawnshop, from original CD issue)
Engine #3: Slightly more "distinct" in the treble region. Sometimes I think the microdynamics definition is better, but sometimes it sound just "thinner".
Engine #4: Sounds "fuller". Vibes sounds more rich, more "kick" on the kickdrum. Drumsticks are more "woody".

[Calibrator]

Quick report here before I head off to 

Using Chesky Records "The Ultimate Demonstration Disc" as test album.

DAP in use.

Using K/S SPDIF output (Juli@ ver 1.21 drivers), I can operate with Q1 as low as -4 but this requires buffer latency size of 128 samples or less. As the buffer size is raised, so too must the Q1 setting.  If the buffer size is raised to 256 samples I get no sound until Q1 is raised to -1 or above. Buffer size of 512 requires Q1 at 0 or above. Nothing is apparently in error at these settings from a sound quality perspective, but it's late here and I can't have the volume up at normal daytime levels. I shall reserve judgement on overall SQ for a day or so.

The issue I do have though is that at the end of each track I get kicked out of UNattended mode back to the GUI interface. Simply clicking PLAY will resume playback at the next track. I've tried various changes to Split File Size & latency buffers etc. but haven't found any magical setting that keeps play going across tracks. In one instance the music stopped midtrack and kicked me back to the GUI, but that occurred when I had split file size set at 150MB. I usually have it at 100MB.

Will continue playing.

Thanks Peter

Russ

[GerardA]

The issue I do have though is that at the end of each track I get kicked out of UNattended mode back to the GUI interface.

Here the same.

BTW. KS works now (since prev.version) for Terratec phase 24FW with one entry for both SPDIF and loudspeakers! So no more apples and oranges.

@Boleary Did you already try the new drivers for Hiface: http://www.mediafire.com/?m2mj2twownq