I am thinking of the build-up of capacitance. It could be similar to "us" needing to reboot the PC once in a while in order to sustain a good SQ. I don't know ...
Does your DC-Offset remain the same throughout the days ? you could check each day before a first play session.

Regards,
Peter

Stanley, then I am afraid you have an other kind of issue. Some groundloop thing somehow. Try whether Switch#3 on your NOS1a/G3 matters. But generally, wheter a different Protective Earthing helps.
We could also send you an other Lush^2 to see whether that helps, but ...

Regards,
Peter

Stanley, does it help if you connect all the wire at both sides ?
(I expect not)

And otherwise it is time to send it over. I suppose it can happen something is wrong, never mind we test them all (happened one time before).

Best regards,
Peter

Hi Stanley,

The normal procedure with NOS1a/G3 is :

- Boot PC (if that was necessary for some reason);
- NOS1a/G3 does not connect;
- Switch Off NOS1a/G3 and switch On - connects now;
- DC Offset is now wrong, so witch Off/On with Switch#2.

This is 100% standard. Leave out the switch Off of the NOS1a/G3 and nothing works.

And then still something can be with your Lush^2 ...

Peter

Um Joseph, Maybe you noticed that you managed to put up the first "for sale" in the forum. We don't do that here ...
Sorry !

Kind regards,
Peter

So what's up ?
All seems fine to me. 

VJ, big grin here. Put very very nicely.

Proost !
Peter

Hi again Nick,


Yes, but it is to keep in mind that calibration takes place (has taken place) under the strict condition of 89dBSPL (no more but also no less in order not to fool ourselves). And *then* the excursion is within range in such fashion that no audible distortion occurs (and excursion is maybe 0.5cm (but I forgot)). Thus, 20Hz remains that (no 40Hz audible). 19Hz remains that. 18 ... all the way down. This is all arranged for in the DSP (per my preset of course). It could be a good idea not to mangle there and take very good care of understanding the principles first (which is exactly whay you asked of course !).

Kind regards,
Peter

Hi Nick - Question of the year !

The answer is, No, it is not allowed to use the NOS1 for this. This is because you wouldn't be using XXHighEnd to play the signal (well, you could, but using e.g. REW would be more logical and intuitive) ... and then the reconstruction filtering would lack, which the NOS1 requires (well, in my own thinking of course and I suppose every NOS1 owner happily agrees).
I fell in that pitfall myself after noticing the relatively heavy roll off of the highs. Never looked at the bass because I didn't get round to that. I just started using the output of my ADC (which is also a DAC).

What you could have observed is the relative more pronounced signal, the lower the frequency. Thus, the roll off with "genuine NOS" is there right from the start (lower frequency), increasing more and more towards the highs. Thus any swept signal would be loudest in the bass ...

Best regards,
Peter

PS: You might also want to consider the pandora's box when you realize that playing through XXHighEnd and its diverse settings and various Operating Systems for that matter, will let (physically !) feel the woofers different in each situation. Take distance of that or you will get mad. Thus, this signal generator you will be using (like from REW or hardware etc. etc.) WILL express only one of these 100s of situations only, and you can not change that. It will never represent your playback reality ...

Hi Richard !

Hmm ... The past week I thought of this two times and on the wrong moment (like driving the car). Point is, I guess, the order is by now from 5-6 weeks or so ago (so it went out of my sight) and for the same amount of weeks you also don't transfer the 45 you mention there yourself. So there too, no trigger. And UPS just left, thus too bad again.
Btw, I also asked you to send me something (5-6 weeks ago). Nothing to be seen yet.

So ? So by now we could make it formal ... your turn first ?
haha

Kind regards and thanks,
Peter

Well (if I understand you correctly), XXHE can be stored on the OS drive allright (no reason to avoid that) as long as we see that the location from where the music is played can be anything (but advice is to keep that local although that is not even a hard requirement).

Btw, this smells a little like what I put in my previous post but scratched again : We also have the RAM-OS Disk. That loads the OS into memory, boots from there, and after that you can remove the HDD/SSD. Now nothing can be in the way anywhere. Playback is now "from the XX Drive" which is pure RAM (16GB of RAM is sufficient but 32GB is more convenient). Load the music from the network somewhere ...
But costs 360 euros.

Peter

Laurence,


The writer of that software you use puts the wrong data in the header of the file. This data - via via - determines the length to play; Somewhere that is wrong. And apparently FLAC encoding can not even deal with it ... (and if other software can, you / that software is in luck).


It depends on the SSD, but generally I agree with that. If you search this forum for "SSD" I suppose you run into a bunch of posts dealing with this, over 10 years ago. But YMMV.

The approach is the exact other way than you envision : denote a "Playback Drive" of which you expect it sounds good and imply with it that also the WAV is always copied to that (XXHighEnd takes care of this). Can be a HDD (they all sound different), an other SSD (same), a USB stick, an SD card, ... anything.

Best regards,
Peter

Tim, 100% correct.

And an addition for further clarification :

A:[W]B-R, B:[W]B-Y-R

Now current is *not* flowing through Y because it is connected to one side only. Still this will bring a different sound as when Y is not connected to the B side; the shielding will work out differently.

Regards,
Peter

In the X3 log file, you see the number 4992 appear twice. That is the number of bytes found for "audio data". And this will be derived wrongly from for now too many possible "wrongness". N.b.: The various players (also XXHighEnd) use all kind of consistency tricks to find missing header data and/or correct wrong header data.

In this case, but only at brief glance, it could be the length of the audio data itself but also the "block align" (summarized : how many bytes are packed into a block of data to form one round of all channels for one logical sample).

Since I am not going to correct this, it could be an idea to convert this to FLAC (XXHighEnd may be able to do this just the same, no matter it won't play it really) to after that convert it back from FLAC (XXHighEnd could do that again for you). Now it may be OK (header data corrected, *if* FLAC was able to make it consistent / workable in the first place).


Kind regards,
Peter