[Sonic77]

Maybe off-topic, but at the Montreal Audio Show this weekend I listened to the BluDave for a long stretch and had them switch the mid/high M-Scaler upsampling switch on Blu2 and found I preferred the mid (500k taps). Less detail, yes, but also less screechy ...at least to my ears although other people just nodded in approval when 'told' which was supposed to sound better. Just my guess, but this was probably an issue with the increased RF on the dual coax caused by the FPGA working harder for the more taps. Do any bludave owners notice this?
Thanks

Yes

 

[yakaway]

Has anyone compared shunyata's sigma reference power cable to clearer audio's reference power cable? With the Dave and or Blu2 of course. One is copper and the other is silver I believe .

 

[Jawed]

Maybe off-topic, but at the Montreal Audio Show this weekend I listened to the BluDave for a long stretch and had them switch the mid/high M-Scaler upsampling switch on Blu2 and found I preferred the mid (500k taps). Less detail, yes, but also less screechy

Yes

Hmm, new test for good BluDAVE setup?

I wonder if @Crgreen finds the same with his BluDAVE.

Now playing: Michelle Shocked - Winter Wheat

 

[etnt]

Bludave at chord booth, canjam sg.

Finally the wife doesn’t think I’m crazy buying ferrites

 

[Rob Watts]

A few points.

500K against the full 1M does certainly does not normally change the tonal quality - so something is bizarre there. The FPGA always runs at 1M taps, it's just the OP that is changed, so RF levels are identical. Going from 500k to 1M opens up the soundstage massively; and if anything it's a tad warmer whenever I have done it, unless my memory is deceiving me.

192 to 44.1 is not good, as asynchronous sample rate conversion is used, which always increase THD and noise. Integer conversion (192 to 48) in principle can be achieved with zero THD and noise (excepting dither noise), and so innately no loss in sound quality (assuming perfect reconstruction and the required decimation bandwidth limiting having no effect).

As far as the M scaler interpolation goes, let's consider 48k or 192k. Both instances use identical coefficients from the WTA filter; so this means both have sinc function accuracy of better than 16 bits (or same as ideal sinc to 16 bits). So is there benefit with 192 against 48? Yes - as the transient timing error area without any interpolation is 16 times better with 192 against 48k; but will this actually allow better performance after the WTA filter? That I am not sure about, but it may be reasonable to expect so.

I will have some measurements soon, that will quantify the size of the transient timing error after the WTA filter, and this is one of the areas I wish to explore...

Rob

 

[flyte3333]

Deleted - too controversial !

 

[flyte3333]

Deleted - too controversial.

 

[Triode User]

For sh!t$ and giggles, would you consider trying this shielding method proposed by John Swenson. I've done it for analogue interconnects (with good results) but haven't tried for digital cables yet.

In summary, connect 24 AWG unshielded insulated silicone rubber wire to the main cables shield, at each end of main cable and run it parallel along the main cable - or just a few turns around the main cable if you really need to, for a 2m cable. In other words, only connect each end of the cable shield with the unshielded silicone wire.

https://www.computeraudiophile.com/...c-power-cables/?do=findComment&comment=659092

Note, that thread title mentions DC power cables but if you read the entire thread (if you're bored one weekend) that post above applies to shielding of digital and analogue cables - not DC and AC power cables though, where John says prefers unshielded star quad configuration....

Thanks for that link. It is a load of b*llocks (excuse my french). All my self made power cables are shielded and earthed at ONE end only. This is intentional and is to stop a loop. The faraday cage effect works perfectly well by grounding at one end and does not require grounding at both ends. I previously posted a link to a youtube clip showing that my power cables have no leakage of EMI and here it is again.

"sh!t$ and giggles" indeed as you quite rightly say but as long as that refers to the post in your link!

Also, that link quotes "at audio frequencies and power supply frequencies, 60Hz etc, it (the shield) is totally ineffective". I think my clip proves otherwise.

Maybe continue this in PM so we don't take the thread off topic!

 

[flyte3333]

LOL eep!

How can I delete my posts. Apologies for bringing it up.

 

[flyte3333]

Also, that link quotes "at audio frequencies and power supply frequencies, 60Hz etc, it (the shield) is totally ineffective". I think my clip proves otherwise.

Let's take the rest offline but that post does not apply to DC and AC power cables, I should re-state.

 

[marcmccalmont]

Deleted - too controversial.

No it’s not too controversial this is an open forum you should feel free to post

 

[Triode User]

Deleted - too controversial.

No it wasn't controversial. I just suggested you and I go to PM to save going too far off topic. Anyway, your post lives on in my quote!!

 

[flyte3333]

No it wasn't controversial. I just suggested you and I go to PM to save going too far off topic. Anyway, your post lives on in my quote!!

Ha no worries mate.

If anyone reads that post of John, for the 3rd time please do NOT think about AC power cables as you read it, as it doesn't apply to AC power cables, as I mentioned in my post.

 

[spotforscott]

Blu2Dave owners that use this combo for both 2 channel and headphones, how much of the m-scaler magic comes through on your headphones vs your 2 channel system? Trying to understand the degree to which a TOTL headphone scales with Blu2 relative to hi-end 2 channel system... thx

 

[Jawed]

Does anybody use their Blu2 with DVD concerts through TV? Is there any way of dealing with lip sync issue?

On the back of Blu 2, the dither switch can be set to on. For USB and BNC inputs only, instead of applying dither, it changes Blu 2 processing to use 666,000 taps but with only 0.1 second of delay.

0.1s is still troublesome in my view, but it's much better than 0.6s.