Anyone upsampling to Naim streamers or DACs?

I tried upsampling using Asset and it made the music really bland, like it had been artificially smoothed. When you upsample, the files are padded out with lots of zeros. The suggestion from those more knowledgeable than me was that Naim streamers 'expect' files in a certain format, so when they are modified with this padding they are not as intended and therefore sound worse. Nice but dull. 

Not what I'm hearing but then that's the beauty of this hobby. I also think that Roon / audirvana/ HQPLAYER upsampling is most likely more advanced than what asset uses. 

It really seems to depend on whose company one keeps. Over with my Roon mates or on CA you get the impression nearly everyone upsamples, I was wondering if many here do. 

Judging from the white hot heat of replies, it does not seem to be in the radar!

.sjb

Interesting subject;  I understand the Roon is a DSP process that genuinely does upsample.   Not to be confused with ripping a 16/44 CD or converting a low bit/sample rate file to (e.g.) 24/192 which is not much more than padding out the file size.     It would be good if someone who understands the Roon & other DSP up/down sample processes can contribute.   

Sloop John B posted:

I'm using the Roon upsampling to 24/192 smooth minimum phase 7th order CLANS. The idea being that a PC can do upsampling more effectively as it has more power to do so and the DSP/DAC skips this upsampling step  

SJB ... it doesn't skip the upsampling though as the DSP still upsamples / oversampled up to the maximum the DAC chip can accept.

Thats not to say the "sound" can't be different, but you aren't avoiding the DSP upsampling.

Not an answer to the question as I haven't been tempted to try and in any case my DAC isn't a Naim. However I, too, have noticed this in CA and Audirvana forums.

If I understand correctly some DACs automatically upsample internally to whatever is its maximum rate if presented with a lower bitrate stream, and the theory of using the (presumed) greater computing power of the rendering computer to do the upsampling so minimising load on the DAC to 'concentrate' on the core process seems to make sense. The big question in that is whether or not the DAC in question does do that upsampling - if not then doing beforehand would seem to me to have the potential to make tge sound worse as it would be presenting the DAC with more data to decode (but no more information), so more workload.

I suppose it's academic as you've tried and found that with what you're using it sounds better, but perhaps its a question worth putting to Naim, as they will know whether their DACs upsample internally.

Naim DAC's do oversample,  explanation can be found in NDX & DAC white papers