Ehh ... huh ?

... would you be so kind to confirm (or not ) that -never mind with or without Digital Volume- you can't hear the difference between 96/24 files played with "DAC is 24 bits" and "DAC is 16 bits" (Settings Area) as long as you don't play at near full volume (but even then).

Of source I suggest that you won't hear the difference, which is my theory. But since I don't have a 24 bit audiophile DAC I like your opninions.

Thank you.
Peter

It is strongly advised to not use your system without a preamp (or the pre-amp at max volume for that matter) if you or your speakers won't be able to handle a situation that e.g. the file contains wrongly formatted data or otherwise - because of which cracks may emerge with an energy beyond imagination.

Edit : As it turns out now, 0.9u-2 is *not* good for Attended Playback ! (Error...).

Engine#3 only :

Added in this version :

• As the title says : Support for 96/24 official files on DACs that support 16 bits only but which support 96KHz samplerate.
  Maybe not many people can use this feature, but I hope those who do enjoy it since it took 3 full days to get this going;
  Here too Vista Exclusive Mode flaws because by normal means it doesn't allow for this.
  Anyway, do not hesitate to play 96/24 over 96/16 because I really doubt you'll hear it, playing at 16 bits only.
  
  
• The Analysis function (DAC Test for "DAC is" in the Settings Area) is extended with some more tests of which I hope they unveil the Supporting |24| settings as not supported at all. Please keep in mind, Support for *and* |32| *and* |24| cannot exist to my belief. Also, so far I've never heard of |24| really being supported.
  Please help and report back those who had the Analysis function report both |32| and |24| as Supported and if something has changed now for that matter. Thanks.
  
  
• Numerous messages have been changed / added to report better for rejected playback modes.
  
  
• At DoubleRate (or UpSample) (which works at 44.1 KHz and 16 bit files only for now) the time position cursor showed at the wrong position (and implied wrong play back commence when dragged). This is solved now.
  
  
• Support for native 96 KHz 16 bit files.
  
  
• Note that a 96KHz 16 bit files can be upgraded to 24 bits (or 32 bits) just the same as 44.1KHz files. Of course there's only logic to this when you have a native 96/16 file and a DAC that supports more than 16 bits.
  
  
• 88.2 KHz files are supported in theory just the same as 96KHz files (16 or 24 bit); also regarding the above topic. This all was *not* tested though.
  
  
• Sound Quality can have changed for any playback modes but for native (i.e. at 16 bit playback, though with - or without Digital Volume Control) 44.1/16 files.
  
• When the "DAC is" (Settings Area) is set to a 16 bit mode, Double (and UpSample) are disabled (unticked).
  
  

Warning-1 : By now it must be stated that all what has been incorporated for Engine#3 and playback modes, is *not* tested against Engine#1 and Engine#2. Quite some inconsistencies are to be expected (like being able to choose a 96/24 which just may error out on mentioned Engines).

Warning-2 : Since DTS pass-through never has been supported officially, but worked, *and* because many kind of official checks have been built in per this version while DTS pass-through still not is supported officially (or : explicitly), chances are high that currently things do not work anymore which worked before for this matter. Please report back on this, and I will try to restore things where needed.

Keep in mind : the bit rates shown at various places are *not* conform reality !

The answer is two folded :

1. I think I can manage to apply Volume Change on Engine#3 (Unattendedly running) with Remote Control;
The other "important" functions can -then- be controlled by itself similarly, but coincidentally what it takes to really apply it is more difficult than with the volume control.

2. When XXHighEnd is brought up during UnAttended Play, I sure will be able to change the Volume and in fact all of the other stuff.
The point here is that it takes the knowledge of where (time position) XXEngine3 is playing, so playing can commence where it left off.

That is what I expected. I can try to do something about that, although I wouldn't be able to test your situation. Let's see ...


I could but it's not for nothing that I do not apply that. Right now all keeps on being bit perfect (mind you, for the "audible" part) ... when I'd do that, it won't.


Yes, that will be applied in a next version.

Hahaha ... Johan ... NO.

I'm very sorry, but even with MP3 (where it is the most normal) this will not happen. It will be the complete opposite of "bit perfect".
Btw, never mind the support for MP3 which obviously is not bit perfect by any means, but I have reasons to believe that there too the sound wil degrade to a high degree because of it (normalization).

I understand that with the rough levels the Digital Volume operates, this can be a pain, and even more with UnAttended Playback; In a next version the Volume for UnAttended Playback will be able to change with less hussle though ("a" next, not "the" next).

Concluded, I fully understand your wish, but this time it "can't" be honoured.
Peter

Hmm ... that sure doesn't look normal to me and you are the first to notice this.
I don't want to blame this driver, but maybe it's an idea to try the original driver ? Just to know whether it makes a difference ?
The being "bit perfect" of the modded driver will be bit perfect for the unmodded version for XXHighEnd anyway (with Engine#3).

Btw, changing the volume in XX would be not different from pressing Pause and Play fast. Does it happen there too ? if yes, you could try to tell me at what "wait rate" things become normal (this could be something well over 1 second then).
Also, if fast Pause/Play vs. slow indeed makes a difference, chances ar high that the driver doesn't shut off the DAC properly.

Lastly, what happens when you press Next during playback ? same problem ?

Maybe in some future. The only close option you currently have is dragging a file from wherever (like Explorer) onto the Playlist Area of XX. The source must then allow dragging of course. Does that work with Total Commander ?

Peter

I 100% agree. Maybe we can both pay attention to this :

The low end of the bass (say under 40Hz) seems not louder, but I'm fairly sure it goes more deep or the deep end (~20Hz) is "louder". Can this actually be so ? it's hard to believe ... BUT :


Here you have a point I never thought about ...
Maybe you recall this post of mine were I talked about fumbling bass in the woofer area being able to shift downwards into the subwoofer area because of getting more tight ? This was a rather blahblah story trying to find explanations on a certain new XX version back then, and what I said sounded rather ok to myself, but the actual "why" kept spinning in my brain. Mind you, I measured (IIRC) something like 26dB more subwoofer output at that XX version. Later, at again another version the subs' output was back to normal again, and I judged the earler version wrong for it. But was it ? or is actually something else going on ...

If anything matters in the subwoofer area, it's the phase. For those who don't know : if you have more than one subwoofer, you're actually in deep trouble, because they will cancel out eachother or emphasize. This difference is a matter of "normal" versus house-brake-down activity.

Dave, your quoted text above brings me to this :

To me, by now, it is the most obvious that all what matters according to the changes of 0.9t vs. 0.9u is phase related. *That* this is so is fully logic to me, although I couln't predicty it really. Mind you, this is a Q2 potential.
Anyway, it will be clear to you what I'm heading for Dave : the phase has changed such that the subwoofers need re-adjustment.

For orientation, it is my own idea that (somehow) is is more about phase-alignment between the individual drivers of one speaker(-set, like Left) than that it is about left-right. Think again of my mentioned shifting fumbling woofer to tight subwoofer bass.
I know, everybody will say that it is impossible to mangle time alignment of drivers in a speaker (without XOver software like is *not* in XX), but I guarantee you it can. It is not explicit though.

So, in the above I say that the most attention should go to getting the subwoofer phase aligned with the woofer at the same side (e.g. Left). I think it is this what I hear going wrong.

Now listen to those VERY forward voices and guitars and all. IMO here's a left/right phase deal going on as well. If you indeed sense this as very much "more forward" you can imagine what two interacting subwoofers will do at the low end.

Mind you, adjusting the subwoofers via calibration software is sheer impossible now. It will produce the sound differently. Think of the difference between 0.9u and 0.9t, or Foobar for that matter. So, what springs to my mind is that I must create a set of test tones suitable for the job, so they can be played through XX. Similar to what Mani did in his post above. I must dive into that anyway.

Thanks.

Thank you LydMekk.

Coherent would sure be my description, as analog. The latter, tough, has been unintentional. Also, maybe I am not sure what gives us that "analog" feeling, because in fact it is not all *that* similar to, say, vinyl. But there's something like undisturbance which analog has somehow, and that's shown here too.


Sadly I can't judge on this otherwise than 18 bits (and I kind of refuse to play through the 24 bits Fireface as the DAC). Keep in mind though the presented harmonic distortion ( http://www.phasure.com/index.php?topic=372.msg2739#msg2739 ) which I still need to dive into myself and e.g. whether it can show under 22050 KHz and whether the bar at 20Hz is just hum or harmonic distortion as well. So, like I said in the other thread, it would not exactly be my way of improving sound quality.

Peter

Ok. It was late and I read Terje's post too fast. I read 48 kBps for 48 Hz, I read USB ship for USB, didn't understand the last half of hist post (soundcard output), and I guess I wanted to read what my expectations were. WERE.
Also, it is kind of confusing for me with a TwinDAC+ that does 96/18 over SPDIF and 44.1/16 over USB.

Obviously the speed of USB (even 1.1) is fast enough to do it, that's why I asked for "data", or maybe I should say asynchronous.
Also, I thought I read somewhere that indeed usbaudio.sys just doesn't allow for more. Well, this *is* all over the place, but for other reasons than natively not being able to.

I read a bit this morning, and if one thing is clear, it's that more people do not - than do understand these matters.
Also, by now it is fair to say that people must be very careful what they actually listen to, while Foobar just proves you get fooled without notice. This is not so difficult to see : when my DAC does 18bits at most, how can Foobar just play the 24 bits file ? How do people test it ? (and I used the 16 bit connection).
If your DAC actually allows for 24 bits, you *still* won't be knowing what e.g. Foobar makes of it ! Foobar, Vista, USB receiver chips ...

A response like that of Steve N. above says it all. He is asked a question by someone who knows what the culprits can be, is fed with the right stuff to test(ify), and the answer ... well ... shows that this DAC representative doesn't know anything of what is going on. Not to be harsh to someone like him, but the message is : you won't be knowing what you're listening to when you have to deal with these people who *do* make DACs but *do not* know about them. Well, to be honest, about the impact within the systems (OS) they are applied.

Mind you, it was less than one year ago that people like Gordon R. just couldn't tell himself what USB actually allowed and it may have been the Elias G. intervention that made others believe it was actually possible (96/24). Never mind if I twist some names or sequences, I only want to say that the TAS1020b wasn't doing the job either, without mutual exchange of knowledge on how to deal with the buffers properly. If I'd apply that chip but forget about the dedicated buffering I'd have 48/16 only.

Now what ?

I can tell you, I wouldn't care less. That is, about 96/24 files which -so far- hardly is provided in my music anyway. More importantly though, I wonder whether I'd really hear the difference. Pedal's post above ( http://www.phasure.com/index.php?topic=387.msg2746#msg2746 ) says some things maybe too, or anyway I recognize all of it. One thing though I wouldn't agree upon much :


... which is related to the technical impact of 24 bits vs. 16 and the derived functionalities from that. I refer to digital volume now which for me is just "under spec" if I had 16 bits only (and which is related to gain vs. sensitivity of speakers) and which is on par with 18 bits. But that's me and my system, and people would really need 24 bits to do it right when having more gain and less sensitivity.

On another note, aren't a few of those 24 bit supporting DACs non-oversampling ? If so, how would this happen while no ladder (or similar) DACs exist with 24 bits ?

You.

There are quite some contradictions in this thread now.
If USB specs tell that 48/16 is max, I don't see how a reprogrammed receiver chip can do anything about that. UNLESS you're not talking audio, but "data", which just is possibe too (at playing audio, might you know it or not ... my question is about that anyway).

Ok. Everything is again as unclear as can be. To me it is.

Possibly you are telling that you are not pushing "audio" as such through USB, but just data ? Or rephrased, do you play through an USB Audio DAC device, or just through an USB device ?

On behalf of everybody, thank you Terje.

All it needs is someone to make a higher grade USB driver. Maybe someone can call (IIRC) usb-audio.com ...

I'll try (a bit );

(don't try to understand ... I'm sure I can't myself when I read it back. )

Per bit you loose 6dB. You loose 8 bits when from 24 is cut to 16, and 8 x 6 = 48. Wow.

Loosing 8 bits is similar to attenuating 48dB when you started off with 24 bits. So 16 are left at this level of attenuating.
Now think the other way around :

With no lower 8 bits there from the beginning, all is equal to before situation when you start at -48dB going upwards towards -0dB. One difference : no bits are added. Now, the louder the music becomes, the more chance you have to hear the missing bits. At -0dB you will hear them at full power. Mind you, with soft music that is anyway, but with loud most possibly too, because it will cause plain (and loud) distortion.

AGAIN think the other way around : You start with -0dB with a 16bit file but 24 bit playback means. When you attenuate 48dB still not any bit will be lost. Attenuate more, and you start loosing bits. Please read the post of Calibrator and his receiver showing the bits in use, and you'll understand the principle better.

All depends on how loud -48dB already is. Is it way loud already, then no problem for the playback of 24 bit files cut to 16. The cut bits are down under in the noise. If it is not loud at all, and you need -12dB for an acceptable level, you will have shifted 6 MISSING bits into the audible level. But mind you, these 6 missing bits leaves still two more (18) than a 16 bit file contains. This however, is not compareable, because in the 16 bit file all necessary data is "compressed" in the 16 bits, whereas in the 24 bit file all is spread (AND WHICH IS ORIGINAL) over 24 bits.

Again, all is dependend on the SPL your system produces.
In my case, where -24dB is loud enough to never want to go louder (not so far anyway), I can attenuate a 24 bit file to -114dB and I still hear very faint music but *not* distorted. at -120dB I can't hear it anymore. So, no "room" for distortion. -114dB means a cut of 19 bits i.e. 5 bits left.
With the 16 bit file (played at the 16 bit DAC !) I hear destortion at -48dB i.e. 10 bits left. Why is this different ? One reason is that -48dB is way much louder than -114dB so I have the opportunity to hear the distortion better. Another reason would be that in the 5 bits left of the 24 bit data, still the resolution is in there for the part left (which is diffrent from micro data / changes which resides in bits which have been shifted out long gone). I'm not even 100% sure about the latter, and I just try to find an explanation for what I hear myself.