OK, any issues I had with a thin sound have totally disappeared with W7 SP1. The sound is beautifully full-bodied now and the NOS1 sounds simply superb.

Mani.

PS. Cats 'purr' when they're happy.

I have my new copy of W7 x64 and will 'upgrade' from W2008SP1 later today. I will use a 4GB RamDisk for sure, and assume it's still important to apply all the 'Tweaked to death' tweaks.

Any other recommendations for OS optimisation (uh, uh, Marcin, Nick, Josef and others!) would be appreciated.

Mani.

The PC, really? If it's an Atom or a laptop, this would be fine, but for a mega-watt monster, this seems OTT. What is the advantage of this? Does the PC 'settle down' with time? Or does it help the NOS1 interface burn in?

Mani.

I think this will always be the case. But maybe the NOS1 will continue to 'bloom out' and at some point perhaps 4xAP will start sounding too full...?

Mani.

Absolutey no chance of this... it will be exactly 14 days tomorrow, not today

Mani.

Yes, this is certainly one reason.

But another is because of my findings over the weekend. HOWEVER... I may have begun the testing a week or so too early - I have the NOS1 playing in the background right now and I think its metamorphosis into the 'World's Best DAC' really is starting to happen... and this is with 16/44.1 and 8xAP.

I think it's ready for 'Round 2'...

Mani.

Hi Peter, I think we're pretty much in agreement. But let me just clarify a few points.


No, absolutely not. Vanilla 16/44.1 is headache-inducing and simply needs upsampling.


Your graphs are interesting here. Your 8x graphs are actually identical to the 4x graphs (due to the limitations of the ADC). No doubt the 'real' 8x graphs would be better, but I think 4x is already good. My point is that additional upsampling of my native 24/192 seems to do more harm than good. I'm not sure, but I think that an analogue filter is applied by the Model Two during the ADC. For ADCs that don't do this, I can see how upsampling even 24/192 would be beneficial.


Yes, I agree 100%. For 16/44.1 material, I use 8x AP.


Yes, and why I called this post 'Round 1'. I want 'Round 2 to begin as of tomorrow and 'Round 3' in a few weeks time. But I have to say that the NOS1 already sounds great with my 24/192 recordings (with no upsampling)... although 16/44.1 with 8xAP still sounds a little edgy.


You've hit the nail right on the head here. I'm definitely not using my vinyl rig as an absolute quality measure. Like you, my feeling is that a DAC should be able to replay my recordings and exactly match my vinyl rig with total ease - adding nothing and taking nothing away. Hell, the dynamics aren't great and the freq response is limited... And yet, most of the DACs I've tried simply can't do this. The NOS1 can, and the Model Two can. So, they've passed the 'entrance exam' in my eyes, that's all.

EDIT: You know, it's so easy to get fooled into thinking that something sounds great, when it's actually just adding more HF or LF, or whatever. Using my vinyl rig as a reference against which to compare is just like using a 'control' in a scientific experiment.

As always, thanks for taking the time to share your thoughts Peter.

Mani.

At the moment, I have the ability to apply 2xAP to upsample 24/192 up to 24/384. If/when Peter comes up with 768, I'll have the ability to tp apply 4xAP to upsample 24/192 to 24/768. I won't need to start all over!

However, I don't like what 2xAP is doing to my 24/192 files - it's changing the sound. And they sound absolutely fine through the NOS1 with no upsampling whatsoever.

Mani.

PS. I've found a way to stream these 24/192 files over my wireless network.

Hi Joachim. Here's my total chain:

Technics SP-10MKII (mounted on a solid metal plinth, which itself is mounted securely to a supporting brick wall)
SME 3009II (pre-improved)
Denon 103 (on high mass headshell) with Kimber armtube cabling (no connectors from cartridge to phono stage)
AQVOX phono stage (using moving coil balanced input)
Analysis Plus Solo Crystal XLR cables
Pacific Microsonics Model Two (no digital or analogue attenuation or gain)
Pacific Microsonics AES cables (with ground-loop isolators built in)
Weiss AFI1
Belkin firewire cable (with power wires disconnected)

At the moment I'm using Sony's 'Audio Studio' software. This costs €30 from Amazon! I have no idea how well it compares to professional software, but I can say that using the Weiss's ASIO drivers, Audio Studio is good enough to capture 24/192 files that I cannot distinguish from the original vinyl source.

If you're thinking of doing something similar, I'm happy to help in any way I can.

Mani.

Thanks boleary. I think this could definitely help when using AP with the NOS1.


Well, I have a tentative theory that's actually been brewing for quite a while now (>1 year) and that has only been supported by having the NOS1 to play with. My theory is that all oversampling/upsampling (i.e. filtering) performed in the digital domain is bad. And this includes Arc Prediction. BUT... AP does the least damage of all the filtering methods I've tried (in HW and SW). But it still seems to change the sound. Leading edges are over-emphasized and I'm sure extra HF content is 'created' that may not be there in the original analogue. But AP is so much better than any of the old FIR-type filters or the newer 'sinc' filters found in all oversampling DACs (and in all upsampling SW)... these simply kill the sound IME.

You know, I would love to be able to report that the NOS1 with 8xAP applied to 16/44.1 is just perfect. BUT that's simply not how I'm hearing it.

However, in the last couple of days, I've proved to myself that the NOS1 with 'properly' recorded 24/192 (which actually includes analogue filtering by the ADC) and no additional upsampling/filtering is as close to perfect as I need.

Mani.

Yep, I tried Special a week or so ago. I can get down to 2 samples totally glitchless! Amazing. I'm not sure about 'darker'.  Maybe that's correct. But I'd describe it as sounding more 'rounded' to me.

In any event, I prefer Adaptive at 1024... like I always have.

Mani.

Yesterday I started my quest to digitise all my vinyl records. The idea is to firstly thoroughly clean (using 'Disc Doctor' liquid and a Loricraft cleaning machine) and then record 7 records per week. I have ~5000 records, so it's gonna take me a while!

I've spent the weekend comparing various formats because I want to choose a single consistent format for all the files. The size of the files isn't really important as I have a large NAS, and can easily increase the capacity when needed. One consideration though is the ability to stream these files wirelessly around the house - my current wireless router just about handles 16/44.1 OK.

But my main criterion is of course SQ with the NOS1. To this end, I've been listening and comparing various recorded formats through the NOS1 and comparing to the original vinyl. And I've decided to go for 24/192 files. Through the NOS1, these sound indistinguishable from the original vinyl. With these files, the NOS1 and the Model Two sound pretty much identical.

But there is one interesting thing that I discovered. It is 'better' to play these 24/192 files natively through the NOS1, i.e. with no DAP to take them to 24/384. If DAP is applied, the sound gets brighter and looses some of it's bottom end weight.

So, just to conclude: vinyl = NOS1 (at 24/192 with no upsampling) = Model Two (at 24/192 with no upsampling)... with these 24/192 files.

I'm absolutley over the moon with the sound I'm getting from the NOS1 playing these files back.

Mani.

For the last 10 days or so, I’ve been playing the NOS1 through my pre-amp – not because of any glitches or whatever, but simply because it has sounded too ‘sharp’ going directly into my power amp. Also there was just not enough low-end to provide a decent overall balance. But a couple of days ago, I bypassed the pre-amp and am now getting a very nice sound. Still not as full-bodied as I would wish for, but very nice nevertheless.

With no pre-amp, using Engine#3 (WASAPI) with OAP (8x Arc Prediction) is pretty much unlistenable. Engine#4 is an absolute must.

On a final note, I’ve tried the NOS1 with the HQPlayer. This is a very nice player and provides some pretty advanced 4x-upsampling filters (and also noise-shaping schemes). As such, it's a great alternative to XX for NOS1 users, providing just enough filtering before the totally filterless NOS1 DAC. IMO, HQPlayer (with WASAPI) using one of its advanced minimum phase filters is better than XX with Engine#3 - here, AP is simply too much and AI just kills the sound.

BUT... XX with Engine#4 and OAP easily beats any other player/upsampling-scheme through the NOS1.

Mani.

Marcin, sorry I totally forgot to reply to this.

In much the same vain as I've posted in the JPlay thread, using the NOS1 with no upfront upsampling is not a good idea. To give you an example, I'm one of those people who has never had a headache - I mean literally, I've never had a headache. I just don't know what one is or feels like (for which I count myself very fortunate, BTW). BUT... I'm pretty certain that if I listened to the NOS1 at 16/44.1 I'd manage to induce a headache eventually. There's something very uncomfortable about it - it's as if your conscious mind can't tell, but your unconscious mind is picking up all the distortion.


Thanks for sharing. I'll look into this once I've loaded W7 SP1.

Mani.

I respectfully disagree. My 24/176.4 recordings played back on the Model Two sound much closer to the vinyl source than my 16/44.1 recordings upsampled by XX and played back either on the Model Two (with QAP) or the NOS1 (with OAP).

Mani.