XXHighEnd

I've tried various settings and also read the manuals concerning these matters, but I keep getting a shut down of the amplifiers at ending a track.

Since it only occurs at some albums with high volume settings I can omit the problem by lowering the volume a bit before ending a track.

It would be nice if the NOS1 would have I slightly higher gain though. . . .


Not shure if I understand you. My amps are connected with XLR, since they are balanced monoblocks.

Stanley

Well, Stanly, since you know perfectly what this is about I don't see how I can further help you. What must I say ? dial in some harshening stuff so you won't like the higher volume ?


Replace my "while" with "if" or "when".

Peter

Maybe this helps you with understanding :

If you set the SFS to something like 400 and Q1 to 30 and xQ1 to 20 and DevBufSize to 4096. Do your amps still trip on that ?

Don't know whether this has anything to do with anything: My amp kept tripping its safety shutdown a day or two ago when the DragonFly was set to 50% volume.  40% was OK.  I don't recall whether I'd tried the DragonFly at that volume before.  It's the first time the amp has tripped that I can recall.

Settings as in my sig.  Unfortunately I won't be able to test any further for at least a week, maybe more.

No, with these settings the amps don't trip. Obviously sound quality looses its magic.

Dear Stanley & Jud,

Reading through this topic and you telling that your amp is going in to safety mode means that it thinks there is something wrong ...and the only thing that comes to mind is that your amp does not like DC at its input and that you should NOT use PA.

Or install an input capacitor in your amps and you're free to go..

Perhaps I am wrong but that is what my recovering brain is telling me...

Bert

Hey (recovering) Bert,

This is not necessarilly so. In this (Stanley's) case this is not about DC as such but about radiply changing DC. It is related, but not the same. We can see this by the sheer fact that Stanley's amps are quite OK when using Phase Alignment. So, it goes wrong at the end of a track; not when playing it.

Peter

Hi, Bert.  Yes, I've thought about this but did not have time to test.  All my PA tests with multimeter so far showed zilch, nothing, nada for DC at input.  And my PA settings are the same as they've always been since I began using it.  But when I get the chance to do so in a week or two, unless things have been resolved at that point, I can test with the multimeter again.

Jud

Hi Jud,

The same for you of course, if it's only happening when a track finishes. Otherwise it is still the same (story), but there's a parameter for starting the track (this is not the # of Rounds - I can't look into it right now - wanting to continue a beautiful - no - superb Who's Next  :) wow).

Regards,
Peter

It's the other way around : there's a sort of "Fade out"  parameter which may help Stanley. See the ToolTip on it.
But I think it is obsolete since one buffer cycle is used for it anyway ...

Peter

Hi, Peter.  Nope, mine happened near the beginning of a track, at a point about 10-20 seconds in where the music becomes louder.  (It was either The Boxer or Gone to Fortingall from Jerry Douglas' "Traveler" - can't remember for certain which one now, but I'm thinking it was the latter.)

Re Who's Next, one of my favorites of course.  Have you ever seen the 9 minute concert version of "Won't Get Fooled Again" from this film?  http://www.thewho.net/linernotes/TKAA.html

Not to be missed.

Of course.
But I have been thinking;
This is all (about) a technical thing, and actually it can't be so that a lower (volume) level album implies for this "crack" only because you pump up the volume. The technical thing here is that all is bound to the volume level (say, from the slider) and not to the actual level in the album. A bit complicated to explain (also to myself), but where the level in the album is lower the slider needs to be higher and when the slider is higher the headroom for the "fade out" is smaller. Now, this *has* to be rubbish because how can such a thing be related to a slider ? I mean, supposed the album is "gained" before it sees XXHighEnd, then the slider will be lower (as with the more normal (for level) albums) and then there is no problem ? that's strange. IOW, I feel this can be solved.

Ok, who could understand this ? Me not. :swoon:
Not yet.

Jud, looking at this again and after listening last night to a best rendered Who I ever heard before (I very regularly play Who's Next because it has always been very good for SQ) ... your problem is that of a PC/Driver combination which more people suffer from and I recall a topic from Arvind at least. But what I also recall is that he solved it, without real knowledge of how (ehm, at my side). I can't help to solve the problem, but at least I have some new information from lately ...

So, this is about "something" happening in the PC and the most logical would be something like needing to write to the page file. I saw this happening myself (your 20-30 seconds) and it looks to be related to needing more memory at a second load of a track part (per SFS boundary) and although this doesn't imply a "swap out" (it doesn't work like that from off Vista) it implies a "more memory needed" and let's say this always involves the page file. Why my focus to the page file ? well, because it receives the very highest priority of all, and it may stall everything else (when a bit in trouble so to speak).

Long story short : disable your page file just for trials.
Notice : even when disabled it will be there for a smaller portion but that is normal behavior and okay for this test.

Also, and similar to what I wrote two days or so ago, since this doesn't formally stop the stream hence I don't notice a thing of it, again this is about this real stall where all stops. It has to be, or otherwise I'd notice it. This again leads to this highest priority thing and the really only thing I know of which incurs for that *and* may take somewhat longer (and too long) is the page file I/O.

Don't try this with an SFS-Max (notice the Max) of around 400 and 8GB of internal memory only; this will choke your system sooner or not so much later.


And then about this PC/Driver combination ...
I have this PC which *needs* the "After SFS Rounds" to be set, or otherwise a crack will happen shortly after playback started. So, After SFS Rounds solves that, but it is a backdoor solution to something I don't know the cause of (better English perhaps : I have no clue what happens). Now :
This very same PC with the very same drivers and everything does NOT exhibit this with Windows 8. :huh: It hasn't a single time, just like my normal Audio PC never exhibited it under Windows 7.

This is quite important to know, already because of, say, half of all people suffer from this (necessity to use After SFS Rounds). But obviously it is OS related, or a combination with drivers which are different for W8 somewhere under the hood (yes, I said same drivers, but for that I merely refer to mobo / video card etc. drivers). Still, this too seems to be related to some high prio stuff, because I can't find the real relation to the happening in time (after the start) and it feels like a write cache which as some stage *has* to be flushed (to disk). This will not be at the precise same time, but at some stage the OS thinks "NOW there's some free time to do it !".

I know this is in the middle of a topic about something else, so people may not read it, but *if* someone reads this, for this too he might try to shut off the Paging and see whether it helps this issue (so, switch off Paging (Virtual Memory), set After SFS Rounds to 0 and see whether the crack at a few seconds into the track now stays away).

If someone needs help on how to switch off the Paging (Virtual Memory) just let it know.

Peter

Thanks Peter,

I think I understand what you say. I hope it can be solved.

For the record: I measured the dc of my amps and it was around 5 mV.

Stanley

100% OK.

More detail if it helps:

- No associated "crack" or any other noise except the amp power relay.  The power light on my amp goes off and the volume ramps down quite quickly with the amp power shutting down.  (The amp sits on the desktop in front of me, so I can't help but see the power light.)

- This happens when volume for the DragonFly, which is being used as a preamp/volume control in this setup (connected through an adapter to RCA cables that run to the amp) is set at 50 in Windows, but not when the volume is set at 40.

- With the volume set at 40, everything just plays fine, no problems at all.

- Edit: Also, attenuation in XXHE is set to -9.0db, as I undertand it the max volume (minimum attenuation) allowed for phase alignment to operate.

What I lacked to tell more explicitly is that when this happens actually the stream (towards the DAC) stops. But you won't be able to hear that crack of it anymore, because your amp is quicker than lightning to shut all off (well, the relais is fast and will cut the output).

When that volume is at 40% (XXHE still at -9 (which indeed is the max for Phase Alignment) you should be able to hear it I think. So, then some sort of "crack" should be there. When not it should be related to te music data itself, which you should be able to proove (but you seem to say that you actually did ?). And when it is the music data, it is the coincidence of the coming together of "my" DC and DC in the data (with the notice that DC already can be regarded that when only the frequency is low enough to let last the "offset" long enough to let trip something which protects from
But I can't imagine this much ...

Peter

I think I understand, Peter.  Happily, it appears I may have time after all to test tomorrow (unless I have a 7 hour conference call on my "day off" like last week, which may actually happen).

OK, a couple more data points today (everything's on Win 7, I haven't reinstalled XXHE on Win 8 yet - may not until there's a change in XXHE to test):

- I used a different track, "Rodeo Girl" from Rickie Lee Jones' "Flying Cowboys."  This is a natural track for such experiments: The CD is from way before the "loudness wars," so it's quite soft and I would actually like to hear it louder.

- DragonFly setting of 50 with XXHE attenuation of -9.0 resulted in amp protection kicking in after a few seconds.  I don't think there was any higher volume in the track itself associated with this, and I didn't hear any sort of crack or other noise.

- Tried a DragonFly setting of 45.  Attentuation of -9.0 caused the amp protection to kick in, as did -10.5.  Attentuation of -12.0 allowed the song to play.

So I can reach settings that allow playback.  But there are a few tracks I would actually like to hear louder.  Maybe not possible using my current hardware and XXHE settings.  Perhaps I'll try "After SFS Rounds," which I've never used yet.