XXHighEnd

Hi Peter,

I noticed that after changing from 09-z to 09-z1 I cannot play any track when setting the DAC to be 2x or 4x. When I select 2x or 4x I have message "Device cannot be allocated....Check your DAC settings....", and next message "Engine 4 did not start in expeted time...". My DAC is 24/192 and so was defined.

Cheers !
Jarek

Hi Jarek - But are you sure this is not about a 176.4/192KHz file (for 2x) or a 88.2/96KHz (for 4x) ? Because that (obviously) wouldn't work in your case ...

Peter,

I am trying to play 16/44.1kHz tracks, and there was not any problem to play them 4x or 2x on 24/192kHz dCS Scarlatti DAC using dual AES. This version simply does not allow any such settings. Only pure 16/44.1Khz is playing. Otherwise I have a message "Device allocated but Can't play. Check DAC settings..." . Obviously DAC settings are OK in XXHighEnd.

Log files will tell you more, I hope.

Jarek

My dac is 24/96. With a proprietary connection between the source and the dac it can play sample rates up to 192. I do not have this connection so I should only be able to play 24/96.

With z1, if I 2x the sample rate of a 24/96 file, XX shows 192 in the display and the file plays. ( how can that be!) If I then tick arc prediction the slider at the bottom of xx moves but I get no sound. This is with scheme 2. If I try scheme 3, I get many "Engine 3 stopped working" errors, but about 1 in 6 times I get the same results with scheme 2.

Because I have playlists that have both redbook and high rez files, z1, in its present configuration doesn't work for me, cause I use arc prediction for all the redbook stuff. Is there a plan to be able to choose what type of file gets upsampled?

Hi Brian,

Here is a quote from the release notes,

Roy

Thanks Roy. Guess I need to learn to read better!

Does this mean that even with a "band limited within itself" 24/96 file it will still sound better upsampled because of the increased resolution?  If so, wouldn't it make sense to rip a redbook cd as a 24/96 with a ripper like dbpoweramp?


Slow But Hopefully Not In The Way OLeary

Yes, with AP. It makes perfect sense if you read the threads about how AP works and the tests that Mani did with the samples.


I think that doesnt make any sense. On the contrary you would loose a few times of AP upsampling, i.e.:
44.1k - 16x to 752k (with the NOS1)
96k - 8x to 768k (")
scale down for 192k limited dacs (4x and 2x respectively)

Besides it could also be a non-integer resampling rather than a 0-stuffing what done by dbpoweramp (i dunno).
Short answer: no, rip the way the cd was recorded.

Thanks Testar.

If all is right this is because you told XX that your DAC is 24/192. If this is not the case (and you set it to 24/96 as should) please let me know.

Jarek, I am not sure, but wasn't it so the dCS is a "DAC Needs 24" ? I think you set that to 32 ...
Or ?

Yeah, dac is set in XX to 24/96 not 192.

Thanks, I see it now;
It goes wrong with Hires files. My bad ...

First off, very nice Peter!  The z version is maturing.  One issue though: With HDCD tracks, I’m getting a tick every minute-or-so through the left channel.  It doesn’t appear to be related to a cpu spike.  I’m using Quad Arc. 

Dan - Thanks.

Do you perhaps have this with more albums ?

It is my conclusion that this just can't be solved otherwise than specifying an output rate. So, this time not for conveniency (Release Notes), but because otherwise the Slider Setting will just be non-sense.

Hmm ...

Peter,

Version 8 and 9zb plays well with dCS on double AES. Version 09z1 is different. On my DAC it plays only 1x.

By the way - dCS accepts only 32bit transmission, swiching to 24bit I have message "Device allocated, but cannot play. Check...." not only on 9z1 but also on previous versions of XXHighEnd. So this is not the case...

At least latest version displays what intended to do correctly, that means, set to 4x, dispalys 32 bit 176400Hz which is correct. But DAC does not play displaying the message.

Thanks Peter, I wouldn't worry too much about it. At some point a new dac will just have to happen......

Peter,

my default configuration is PC with LYNX working in slave mode, accepting clock from external source, as well as DAC. To check things up, and identify the problem, I set PC to master and DAC to slave passing the external clock by. Now I understand why new version does not play 2x and 4x with external clock.

When I play 16/44.1 track and specify 2x, it plays 24 bit/44,1kHz, if I choose 4x, it plays exactly the same : 24bit/44.1kHz, instead of 24bit/88,2 and 24bit/176,4 kHz accordingly. Strangely, XXHighEnd shows 176,4kHz when set to 4x, but dCS dac identifies this as 44,1kHz. It must be true, because dCS does not accept anything more than 96kHz on single AES, so when XXHighEnd dpsplays 176400 - this must be the false value.

Jarek, this certainly will be true. The display is redundant to reality. The question is : why does this happen to you ?

OF COURSE (hehe) you are alone - well, so far.

Can't this be related to the double AES connection somehow ?

So :

1. The display *is* redundant;
2. It can happen that something the display allows, is not allowed at the "real playing side". But :

This would lead to your first post about this, not to "and I (XX) do what I want anyway".
So in the end I don't recognize this at all.

If others do, please let us know !!


Jarek, important : If you take 16/44.1 as the base, I can't think of anything that has changed *there*. So, are you very sure this didn't go wrong work in 0.9z(b) ? Not that I refuse to believe you, but I must have some clue to what is happening. If it changed from 0.9zb, I do NOT have any clues.

Thanks as always,
Peter

That is why I changed to single AES to be sure.
So, 16/44.1 tracks (I tried a loooot of them) plays :
1x -> XXHIghEnd shows 24bit/44.1kHz, LYNX MIXER shows clock 44.1kHz, dCS shows 24/44.1kHz
2x -> XXHIghEnd shows 24bit/88.2kHz, LYNX MIXER shows clock 44.1kHz, dCS shows 24/44.1kHz
4x -> XXHIghEnd shows 24bit/176.4kHz, LYNX MIXER shows clock 44.1kHz, dCS shows 24/44.1kHz
so, changing 1x, 2x, 4x does NOTHING except display in XXHighEnd.
Just checked.

Just to be sure. I returned back to b version, and it plays just right:
2x -> XXHighEnd shows 24bit/88.2kHz, LYNX MIXER shows clock 88.2kHz, dCS shows 24/88.2kHz
4x -> XXHighEnd shows 24bit/176.4kHz, LYNX MIXER shows clock 176.4kHz, dCS shows 24/174.4kHz
Cannot do anything wrong here.

Jarek

Ok ... It is my conclusion that you didn't denote any filtering means. So, just 2x / 4x without any of the filter buttons active. *Then* indeed the output sample rate stays the same ... (and this is NOT intended, hence a bug).
I hope you can confirm this.

About what the dCS accepts ... if that is 32, why does XX shows 24 ? I don't think it can. It can allright, if DAC Needs is set to 24.
Furthermore, what happened to your earlier posts about the DAC doesn't accept etc. ? ... I don't hear that back now ? Or is that related to the two AES lines ?
Besides, notice that it is the Lynx which determines what is accepted, not the dCS (unless you connect it USB).

?

Peter,

truly I do not try to use any filtering, since on previous versions I did not touch it. Any kind of filtereing in my setup destroyed the sound.

You are right about 24/32 bit. My mistake - I should have written 32 bit and not 24 bit in the post. Of course XXHighEnd displays 32bit instead of 24bit, and dCS does not accept 24 bit comunication.

LYNX is producing output on single AES when sample frequency is < 176 kHz. Only on 176/192 it uses double AES. So when I said that I'd switched to single AES = switched DAC to single AES, because second AES does not play at all when <176kHz. This is the main reason I used to play always on 4x and wanted 88.2 or 96 tracks to be upsampled 2x when XXHighEnd is set to 4x.

So, will you enable upsampling 2x and 4x without filtering ?

Hahaha, I am afraid you are getting crazy now ...

... There is no such thing as upsampling without filtering ...

If you apply the 2x or 4x and don't use any *upsampling* (which is the same as filtering !!) button, you're only doubling or quadrupuling the samples. It does nothing, but may have implications on jitter at the DAC side.
So try to understand : this is *not* upsampling, and there is no way this can incur for higher (fake) resolution, as there is no way it will filter harmonic distortion. It only sets the DAC to the higher sample rate and feeds more samples of the same.

But I will solve the bug that Double/Quad/etc. without upsampling works again.

Regards,
Peter

Very good, Peter.  I have latest version on both my computers, working well.  Only thing I have noticed (a VERY small thing) is the OSD time display doesn't reset the far right digit at each tick of a second.  The new number is superimposed over the last one, so by the time all 10 digits have ticked by, there is just a red square there.  This is on both computers.  I can live with this problem pretty much forever. :)  Thanks for all your hard work.

Hey Glynn, thanks. I guess not many people use the counter anymore (including me). Ok, will look into it !

Peter,

I have found some tracks (non-HDCD) that does not play properly when using 4x IntiImage upsampling. I do not know if it is already known issue.

On attended the problematic track plays till specific point but seams to omit some data then - skips some data in the track.

On unattended, this track plays to specific point and then track is skipped to the next track. Looks like some data recorded on the track causes problems. I can forward the sample track if you want.

Thank you
Jarek

Jarek, thank you.

From your description I knew this already, but I couldn't place anymore where it went wrong like this. But now I know again : HDCD.
I already knew it from before 0.9z, but I just couldn't find the problem anymore !

So as said, thanks !
Peter

Peter,

I have found a track that does not want to play at all on AntiImage 4x. When you press play, XXHighEnd does not react. This is 6-th track, Coming back to life, Pulse, Pink Floyd. Maybe this will help you to track the problem.

Ok, I have that album, assuming it's not an extract from the DVD. I'll look into it.

Peter

Using 09-z1 I tested to upsample a 24-96 file with 2x AP (24-192 DAC) and played OK- good sound.
I then accidentally played a 24-192 file with setting still at 2x AP and 24-192 file played, but at half the recorded speed.
I guess this will be prevented in 09-z2 as an unsupported output.

Thanks for all your good work

Bjorn 

I tested it with 09z not 09z-1. It looks like none track is played from this album. I tried a few.
Update: a lot of other albums is not played at all on AniImage 4x.

Which will be the situation you see below.
Answer : No, I don't think so. IOW, your DAC can just do it, so why not do it. One problem, the software must support it correctly, and not all was covered for in 0.9z-1. In fact, the problem in de software was kind of the same as our brains had with it : there was missing logic to do what you want with "one" dimension of settings. This now have become two dimensions : the slider for the factor, and the fx which determines which base to use (which in the picture below is the file itself <- fx not active).

Peter


PS: For those who don't understand what this is about, see here : 0.9z-2 Sample Rate Conversions  (http://www.phasure.com/index.php?topic=1286.0).

Peter

Sorry, but I did not express myself precisely. What I meant was I guess 09-z2 will prevent an attempt to 2x upsample a 24-192 file, if DAC max capability is 24-192 (as I tried by mistake).
I appreciate the 24 bit upsampling capability and am following with interest the evaluation of the sound quality btw 24-48 with 4x upsamping, 24-96 file with 2x upsampling and native 24-192 from forum members with more sophisticated systems than mine.
 
Bjorn

I don't know how I did it, but I completely misread your post (that one line in fact). I even read it several times to be sure. :fool:

Ok, so that now (indeed) gives the below situation. DAC is 192, file is 192, and the slider plus fx inactive says "double the input file".

I've just installed XP and I can't run XXHE. Instead the Phasure OCX installshield opens every time I click XXHighEnd.exe. Need you help...

Huh ? And of course you finished the Install 20 times by now ? ...

It won't help you, but here it works for XP.
:scratching:

Hey Glynn, I have a small problem ... it doesn't happen here ...

Can you perhaps tell me since what version this stopped working for you ?
I am merely thinking you installed some "monitor control" software, or perhaps general skinning software (on both PCs :)). Can that be the case ?

Let me know,
Peter


PS: Did anyone else see this ?

Jarek,

I just tried a .wav and a .flac version, but both work. Notice though that I just solved the general problems with AI, so it can be that I just solved this along the way.
Please try it with 0.9z-2.

Peter

Peter,
when working in the Library Area, I get te following message (in Dutch) on a regular basis. I guess this is the one you are referring to in the release notes of 09-z and it has to do with someting running in the background ?
Still, it is rather annoying to have to click the message away every time it appears. BTW: playback is not disturbed.
Cheers,
Eric.

"Er is een onverwachte uitzondering opgetreden in de toepassing. Als u op Afsluiten klikt, wordt de toepassing onmiddellijk afgesloten. De object verwijzing is niet op een exemplaar van een object ingesteld"

This is solved for 0.9z-2. Unless I said it was solved for 0.9z-1 already, then there is a problem somewhere. :)

Nope it works quite ok for me, at least.

Arvind

I am experiencing the same problem. Last digit of runningtime is superposed.
How did you solve it Arvind?
Cheers,
Eric

Peter,
after playing (part of) an album, XXHE adds a file named XXAnalysis.dat to the album.
How can I prevent that from happening?
I like to keep my backup as an exact copy of the original, that's why.
Cheers,
Eric

You can't. I am sorry Eric.

As a matter of fact I now saw it too. Only the last digit, and somehow it is related to the background. Also, it is (very very strange) not all digits which remain. So, the 4 (or whatever) resets it all.
If I have no background (just black) there is no problem, but if I have my drive screen (with "VU meters") as the beackground, the "whiping" of each digit somehow doesn't work.

No clues here ... :no:

Of course I can do it manually.
(using Windows search and select+delete them all in 1 go).
So I guess you can do it in an automatic way or at least provide a kind of "cleanup" button? Food for thought.

After making a (differential)backup, I always check the result (total number of bytes must be equal). The XXAnalysis.dat file makes it much more complicated to verify the correctness of the backup.

So here is the challenge.
Cheers,
Eric.

Hi Eric,

Frankly I have done nothing at all. It just works ok.

Arvind

Uhm - Eric - Why ?
Either a normal backup or a differential backup will also backup those XXAnalysis.dat files. I don't get this ...

Peter,
This is my usual scenario.
I have a base B (1000ths of albums) and an identical backup B'. These sets used to stay identical, so when I add some new albums A  to the base B and to the backup B', my new base B1 will be identical to my new backup B1'.
This has changed now because everytime I play an album, my base B will change. And everytime  I add new albums A, I will find that B1 is not equal to B1'. So I need to check the cause of the difference.
It is possible to do that but it is annoying. If I am the only one, then just forget it and I will live with the manual removal solution.
Cheers,
Eric

Ok, so it seems I understand it Eric. :)
But your "differential" backup is a manual one, right ? couldn't you change that into something which determines for you what is changed ? (ehh, like how it goes normally if we make backups) ?

You must understand that this is "my" means to store data with an album useful for a next playback. Whether you use that today or not is not all that important, because it could be tomorrow. So for example, it you would be using the Normalized Volume, you'd know you have to wait for the calculation of that for maybe 30 seconds. Not when you play the album the next time, because then this data is known. And you know where ...
So I think it would be a bad idea to let me make a parameter so you would be waiting 30 seconds each time, which would end up in you not using the functionality at all.

So just the question : wouldn't it be better for you to find this automated incremential backup procedure ?
Regards,
Peter

Peter,
now I see why you need to have that info in the music files.
I guess I want backup that info as well !!
So I will find a solution how to handle this.
Nothing for you to be done (in this area  :))
Thanks, Eric.