NOS1a (v1.0.4b) : Fifo errors

[JohanZ]

Hi Peter,
Yesterday evening I replaced the standard 1 meter USB cabel for the new short Clairixa USB to connect to the Intona Isolator. During the listening sessions i clearly hear some sound disturbances in the music. When checking NOS1a interface (driver v1.0.4b) i see that there were a lot of Fifo and USB errors. Restarts of the PC and NOS1a and reconnecting the USB to the silverstone did not helped. I will replace this evening the short Clairixa for the standard USB cable to see if this helps. Any suggestions for me what caused this errors en solutions? Thanks.

Best regards

Johan

[PeterSt]

Hi Johan,

Any suggestions for me what caused this errors en solutions?

Not yet, but a very strict advice :
Write down the current value of each (!!) setting you are going to change. If you don't do that I will predict the next issue : with the old cable it also does not want to work any more.

Now first a question or hint - if not a suggestion :
Is this Windows 8 ?
Coincidentally, when I was testing a few of the last fabriqued cables I was running Windows 8 and it was may more difficult to get it right. I needed to set the SFS to 4.00 because under that it showed what you describe ...
Do notice that I tested all the short cables as intended : with Intona in the chain and a longer Clairixa at the other end.

If you have laying around a W10 install (for the info to others - you don't have the RAM-OS), you can test that and I guess you will see the difference.
Anyway, and of course still assumed this is W8, you can test with the higher SFS (like 4.00).

If from this comes that the Intona somehow can't bear the Clairixa under Windows 8, then we must see how to proceed. But I am fairly sure that Daniel from Intona will be willing to help out.

Btw, are you talking about the Industrial version of the Intona ? because remember, that is the one I advised (read : tested and technically approved).

I hope you can answer upon all the questions and that I was not toooo specalutive ...

Regards,
Peter

[JohanZ]

Hi Peter,

Is this Windows 8 ?

I'm using Build 10074 of Windows 10.

Btw, are you talking about the Industrial version of the Intona ?

Yes the industrial version.

Johan

[PeterSt]

Hmm ...
Clearly I did not expect this.

Another attempt :
Already for some longer, it has become quite difficult to work with 16ms for NOS1 Driver (Control Panel) buffer. So I have that at 4ms. That this should not be 16ms (I recall that this is since Windows 10) was discussed but is very easy to miss and forget. With all the testing I mostly forgot since the default at installing is 16ms.

Can that be the culprit ?

Peter

[JohanZ]

I'm using the 4 ms for NOS1a control panel.

[Arjan]

Hi Peter,
Same issue here, I highly suspect the short Clairixa. Without it no problems!
But SQ is super.....
Regards, Arjan

[JohanZ]

Hi Arjan,
For me the Fifi errors are audible. This evening I have changed the Clairixa connection from the Silverstone USB port to the the computer USB port. So far no errors and strange disturbances in the sound anymore. I never installed the drivers for the Silverstone card in W10. I don't know if thats the solution for this problem. I have the impression that the sound via the Silverstone is better.

Someone a download link to the Silverstone driver for a W10?

Regards

Johan

[Arjan]

Hi Johan,
Audible for me as well, but with a cheap cable no problems.
I will try the other usb port tomorrow.
Regards, Arjan

[Robert]

Hi I woke up this morning and no sound whatsoever. I have just received the Intona and blamed that initially.

The software for the Silverstone was not working.

I downloaded the latest W8 drivers from the Silverstone Web site and all working again. I installed both the firmware and driver and restarted, it all worked. Don't use the driver from the install disk its a problem I believe for W10 although it did work but stopped so not sure what happened.

Here is the Silverstone ECO4-P driver link under downloads for anyone: http://www.silverstonetek.com/product.php?pid=364

Robert

[PeterSt]

Guys,

Something is going on here and Robert, thank you for your post because it only adds more of it.

With the explicit notice that we should not see ghosts (as we say it over here), I am now at the count of 5 with sudden strange problems. The common denominator is the Intona.

The common denominator is also that it seems that things "break" and can't be reversed. Well, technically and very far sought that may (electrically) be possible, but so far I can't believe that. This is how Robert's post is crucial, assumed he did not see ghosts (from another angle coming in). Point seems to be :

It is the most easy to draw wrong conclusions as the problems are not logical (or don't seem to be as long as we don't understand). But example : At testing the cables I myself had once all out of order and could only get it right after a reboot of the PC (after switching off/on the remainder). To me this tells that things also can work halfly, like in too many errors (FYI this is not far-sought at all).

This is only a lead-in, as I had some more prepared already, but Robert's post must let me re-think a bit first. Anyway I will be having some questions, and I hope you can deal with them precisely.  No matter what, write down your previous settings and I say this because of seemingly irreversable problems I run into just the same (remember, count is 5). Again see the importance of Robert's post ...

Thanks for now,
Peter

[PeterSt]

Both feet back on the ground, first another observation :

I know from 10074 that there will be many USB or FiFo errors for sure. Notice though that I never checked that so I can only know for real if I start looking at it, which I sure could.

In between the lines I reported about this like "it needs 30 seconds at least to stabilize and then still audible anomalies might occur once in a while".
This has been there right from the (10074) start and all it takes is a lower SFS (I forgot how low but say under 4.00).
The sound of this is a sort of scratching DC-ppplop and people not familiar with the music or even the system (but know the music) may not notice.
Mind you, already this I read about exactly in Johan's post. Also, the fact that you now hear these things can theoretically imply that things are better audible. This really is so.

Then, at writing the very last sentence above I realize that Johan *is* using 10074 (I really did not think about it). So for this (his) situation I can already turn it upside down : Johan, if you can proove to me any setup WITHOUT errors and an SFS of say 0.2 or lower ... be my guest.
Keep in mind : I listened to that for months myself, as it slowly goes away after pressing Play.

That is 10074;
With the others I don't think such a thing is in order, but this very well can be about the sound it exhibits and which is a kind of special with 10074. The anomaly smears there and therefore makes it audible easily. But look :

I already told about my W8 experience and how we had to set the SFS to 4.00 in order not to see errors. Yeah ?
Well, not quite, because with 10586.0 I had to set the SFS to 1.00 in order to not see errors. Mind you, see; I don't hear them anyway (oh, maybe if I want to hear them, but that is not me).

So the danger is - and about seeing ghosts - that if we now all start to measure the USB/FiFo errors, we indeed see them and think things are flawed, while they have been there all the time. But I must be careful with such expressions just the same, because "all the time" for myself would be since 10074 which has been with Clairixa right from the start, and with W8 not because I used an SFS of 4.00 because lower did not sound right.
Do we get the message ?

Sadly you both (Johan and Arjan) see that with a normal USB cable the errors are not there.

Ready for the first question ?
What happens with either of your situations when you set the SFS to 4.00 ?

And Arjan, which OS and how (RAM, Unattended etc.) are you using ?

Peter

[PeterSt]

Question #2 :

What happens to either of you when you swap your both Clairixa's ? Thus, the long one coming from the PC, the short one going into the DAC ?

[Gerard]

I see this. But changed bugger from 4 ms to 2 ms and they where gone.

Just now going back to 4 ms and no errors apear

SFS 0.10

 

[PeterSt]

Thank you Gerard.

Another question I have (and it relates to Gerard's post) is :

How long does it take before you see things stabilize ?
So remember, I talked about that for "audibly and 10074" but I am also confident that this happens to you (Johan, Arjan) and everywhere (OSes). Example :

You press Play and right away a pile of Streaming errors pop up. A few 100 more after a second, more afer 5 seconds, maybe a handful after 10 seconds and then it is quiet.

Or :
You press Play and right away a pile of Streaming errors pop up, and that's it.
In this case, think of an audible tick when you press Play. Regard this a Driver error (and virtually my bad but no issue).

Or:
Right away a few dozen USB Stream and FiFO errors show, and that's it.

These situations per definition happen. Here they do and I have to regard them normal.
Still they can stay away just the same, and the higher SFS will "cause" that, usually.

I hope it doesn't need explicit telling that the errors don't automatically reset when you press Play. So keep track of the last score at pressing Play for a next attempt.
Otherwise changing the sampling rate (in the Control Panel)  will reset the count, but always change to a multiple of the sampling rate you play with (like to 88200 (and not 96000) if you use 176400). Otherwise for 100% sure errors will emerge right after pressing Play.

[Gerard]

Well,

I got back to buffer 4 ms. Other buffer size give usb and fifo errors too.

But i played a song right now.

1 track no errors

going to the second track first second 742 error counts and the next second 1198 errors.

From second track to the third in the first second 2400 errors.

So in my situation it only happens in the first seconds of the track. max 2 seconds.