SQ of 0.9z

[PeterSt]

Is it possible that arcPred has an invfluence to the sound even if no doubling(upsampling whatever...) is used?

If all is right, not. With Attended, very maybe ...

For a manual, at this moment I am afraid this comes down to reading the Release Notes from the newest till the oldest, until you are fed up with it. But all should be in there ...

[Flecko]

If all is right, not. With Attended, very maybe ...

I am not using unattanded yet, so it is possible. If it has no influence, it is a nice placebo

[GerardA]

Better late then never...
Thanks to you boys I tried the higher buffer settings (1024) with adaptive mode and normal Q1=1 and wow!
The first evening I tried to lower buffer and Q and everything went fine and low, thinking this will give the best sound.
A little disapointed I stopped to go sleep and never felt like trying anymore, maybe back to vinyl?
So now it is back to enjoying the music, and yes the best ever!
Maybe a little extra tweeking and maybe let John Kenny do some surgery on the hiFace and you'll never hear from me again! 

[Marcin_gps]

Maybe a little extra tweeking and maybe let John Kenny do some surgery on the hiFace and you'll never hear from me again! 

This trip never ends..., 

Remember me writing about buffer sizes and losing my rabbit? Guess what? After I deleted XXHE config files, made a clean install, I'm able to listen at 1024 samples again!!! Why? What happened? Peter would probably say - your PC is faster-  but no, it's not faster, I got 4 apps running, 40 tabs in Firefox open and my dpc latency varies from 200-350us. Maybe to low os latency doesn't allow for low buffer sizes with my drivers? That's magic for me. I won't reboot ever, afraid that my rabbit'd run to hole and I won't catch it again... The sound is 100x  better now. Again - why???

[arvind]

and of course it had to be you to come up with this strange change as the first person.

Peter

Hi Peter so am i made out to be the villian...he he...

Arvind

[PeterSt]

This trip never ends..., 

Remember me writing about buffer sizes and losing my rabbit? Guess what? After I deleted XXHE config files, made a clean install, I'm able to listen at 1024 samples again!!! Why? What happened? Peter would probably say - your PC is faster-  but no, it's not faster, I got 4 apps running, 40 tabs in Firefox open and my dpc latency varies from 200-350us. Maybe to low os latency doesn't allow for low buffer sizes with my drivers? That's magic for me. I won't reboot ever, afraid that my rabbit'd run to hole and I won't catch it again... The sound is 100x  better now. Again - why???

Hi Marcin,

Ok, to me it is clear that more might be going on than I can reason out. IOW I sure believe you.
Keep in mind though that I can not judge your "capabilities" of judging a system for its performance and everything. This can go down to stupid details even I don't want to know about, but can work (around) with anyway. Example : If I see my driver is behaving strange (and the RME Fireface can do so without (known) reason), I reboot or whatever it takes to let it behave normal again. This does NOT imply I solved the RME problem, but I solved the result of it. Why ? well, in fact only because I pay attention to nearly everything, so the leat I will know is what is "normal".

For everything is a reason, but it is the sport to find out what causes it. Again, this often means take away the cause, while the root problem never is solved.

So yes, something will be going on, but what it is I can't tell. But let me give you a hint, which is not such a common thing to come up with for a normal PC user, but which is a fact for me :
Your PC "lives" by interrupts. You know, the PCI stuff, but in fact everything. This all works by preset timings. Something like "if I want to read a block of data from disk, my request should be dealt with within 5ms". This 5ms is a setting and preset by the OS or drivers. The PCI latency program (which worked for XP) is an example of that those settings can be changed. But now think like this :
There is *another* setting, and it says that scanning a directory entry must be dealt with within 1ms. The pipe this goes through is the same as the other one, "reading from disk". Now, when the scanning is in order, this causes the reading also to receive 1ms, while the actual setting is 5ms. And thus, putting the scanning in a constant loop (consuming some cpu cycles) suddenly gives a 5 times faster disk access ...

This example is completely made up by me, so don't try any of it. But I always have the real life example of decoding of FLAC, which including the reading the FLAC file from disk, the decoding itself, and the writing of the decoded file to disk, goes way faster than I can achieve myself with the fastest means of a file copy in C++.
IMO there can be one explanation for this : the FLAC decoding keeps alive something (interrupt) towards the disk I/O, making that by itself way faster.

With this in mind, anything can happen up till "the more Firefox instances, the better it becomes" (up till a limit).

Also keep in mind the reporting of someone who found the latency (checker) dropping significantly while XXHighEnd is playing music. To me this makes perfect sense, although it may not happen to everyone like that, as the remaining latency can get worse of it (which would be the expected situation).

This is for your weekend !
Peter

[Marcin_gps]

Peter, I  get your point and totally agree with you, but assuming my PC "lives", I can't control it or predict its behaviour for sure. That's why PC audio is so complicated.

Another funny thing, now that's everything is back to normal (in my opinion), I like invert more. And all my findings are backed up by my girlfriend and she knows when sth is right 

Have a nice weekend!
Marcin

[Scroobius]

I delayed writing this -I wanted to properly listen to 09zb before commenting (in the past I have got carried away to soon). First thing is with 09zb I can run the buffer at 384 instead of 768 in adaptive. Why is that Peter? my sound card buffers have not changed of course maybe better interaction  with them.

I checked 09zb versus 09y with the following settings:  adaptive & 1024 buffer.

I really thought with 09y I was there, that I had arrived & there could be no more to come from my system as the sound was so good. So when 9z appeared I thought nice window dressing but underneath same engine same SQ.

How wrong - I started to make comparisons but 9zb is much better than 9y in my system so no need for further comparison. It is the whole sound that is better - clearer more dynamic there is a layer of hash removed. Hash that was not there to remove!! (and I don't mean hash in the Dutch way ha ha!!!). So I then just listened and listened for hours to all sorts of music - because it is just sooooo easy to listen to. I can't say much more the sound is so beguiling I just want to listen and listen.

I am right now listening to some 24/96 files and the sound is crystal, sparkling clear - voices are there in the room, voices oh my!! how good they sound - you can hear the real expression fantastic. A big improvement but how can it be? - I thought it superb before.

I look at my speakers and wonder how bits of paper, magnets and wire can produce such sound. And then I think what will it sound like with a NOS1 and maybe some nice horns? Scary.

So where do we go from here? how can it possibly get better? - but I know that if Peter has anything to do ** it will ** I don't know how you do it.

Best regards from a very happy bunny

Paul

[PeterSt]

A big smile Paul. Thank you very much !

[pedal]

I agree with Paul and the others. 09zb is the very best so far. I run Adaptive and 1024, like most of the others. Reducing latency to 512 make the sound (dynamics) somewhat "overdamped". 1024 is just perfect. Never before could I play so LOUD. Unlimited LOUD! At the same time the sound is extremely detailed and natural, without any boomy bass or hard treble.

Congratulations Peter!
 


EDIT: Also the soundstage is phenomenal. I have never had such a wide and deep soundstage. The sound has never before been so detached from the speakers.

[Marcin_gps]

Peter, you know, I've found the culprit of that mysterious buffer problems. Windows changed power settings without a notice although my power plan was still set to performance! I had to change to power saver and then back to performance again. Stupid, but that did the trick.
I've been experimenting more with number of activated cores and discovered another mystery. With 2 cores enabled OS' performance is somewhat better - apps load faster and dpc latency is about 20% lower in comparison to 4 cores. I don't know whether it's a system issue or my platform's rather. Anyway, one core sounds the best and I am able to achieve the same buffer sizes regardless of sample rate used. I also added a second PSU for 4P socket and all the dirty components (hdd's). My primary (better) PSU supplies 24P socket only - very good improvement, but that's just the beggining of my pc-audio trip. Dedicated super efficient audio server with linear power supplies to come...

[AUDIODIDAKT]

Marcin,

The "number of processors" should NOT be selected. (Thus Unticked)
If you selected 4 processors in the first place instead of unticked, yes, this slows things down.

Roy

[Marcin_gps]

Roy, I'm talking about "physical" number of cores running - in BIOS. This has nothing to do with XXHE or OS.

[boleary]

Never before could I play so LOUD. Unlimited LOUD! At the same time the sound is extremely detailed and natural, without any boomy bass or hard treble.

Yeah, I've now a new dilema: Whats the "correct" volume for the type of music playing? Rock, of course is easy, but classical is where I'm scratching my head. Hillary Hahn, Bach for Solo Violin, there is a point where I know that the volume is just so "unnaturally" loud, but it sounds so freakin' good at such high volume that I have to remind myself to turn it down to more natural levels, where it also sounds so freakin' good but different, not as revealing. Its like suddenly having an auditory microscope!

its all a bit dizzying over here.......

Thanks Peter.

[pedal]

Yeah, I've now a new dilema: Whats the "correct" volume for the type of music playing?

Yes, I have made som reflections about the same. My experience and taste, say that also classical and acoustic music can be played with higher volume "than normal", BECAUSE most instruments are recorded with mic's very close to the instrument.

Through our hi-fi systems we are listening to the recording from the perspective of the microphones (and how they are mixed). A few record companies use single stereo mics placed quite far away from the instruments. Then the "correct" playback volume is quite low.
But most of the time the mic's are very close to the instruments, and then I prefer to turn up the volume. I mean, if I was sitting on a chair 50cm away from Anne-Sophie Mutter, then the live volume would be high too.

Same goes for acoustic guitar. The pick-up/mic is very close. In my system I can play such recordings at +100dB without boom. It just sounds "big", like it would if I had the instrument just in front of my nose.