I’m interested in the Chord Hugo M-Scaler, but due to unavoidable bad room acoustics I rely heavily on room-correction convolution filters in Roon. Chord’s Robb Watts has said that only bit-perfect input should be sent to the M-Scaler, and obviously you wouldn’t use Roon’s upsampling with it. However, I’m wondering if anyone with an M-Scaler (or Blu MKII) has tested any of the other Roon DSP settings (convolution, EQ, etc.) and found a negative impact on the efficacy of Watts’ filters as a result. Thank you!
If it was me I would want to do convolution and EQing after upsampling.
Yeah, sounds reasonable, but…
Most DACs these days do upsampling, and that is by definition last in the chain.
In my main room, I do convolution for room correction, but no upsampling, in Roon. (I have tried with upsampling in Roon, up to the maximum 192k the Meridian 818 accepts, no difference.)
Then, after all this, the DSP8000 active speakers upsample to 705/768.
I woudn’t worry too much about this. The PCM output from Roon is bit-perfect and lossless even if EQ and convolution is used on the signal. Imagine that you apply the same EQ and convolution to a file and then play it in Roon with DSP disabled. Roon would display ‘Lossless’ (i.e. bit-perfect) in the signal path and it would sound the same as the original file played through DSP. Of course the converted file is changed compared to the original, but it has not become lossy (like mp3) in any way. An album (cd or high-res) is always treated with EQ and various processing before release anyway, but pure DSD is different of course…
What type of outputs do you use? If you use active speakers, many brands have built-in room correction capability (Audyssey, Dirac,…). Same goes for higher end amplifiers and most AVRs.
Room treatment is preferable, but if this is unfeasable, audio room correction at source level is considered a last resort. Better to do the room correction at amp or speaker level. That way the correction is applied to all your sources, not just to your Roon output.
Hello @tc4972,
I actually asked Rob Watts about this at the NYC CanJam event last weekend. He said that as long as the filters are done correctly and no sample rate conversion is being performed, using FIR convolution for room correction before sending to the M-Scaler will still allow for the full benefits of the M-Scaler to be realized. Hope this helps!
-John
3 Likes
@john , thank you for thinking to ask this of Rob and thank you very much indeed for sharing his response with us. This is just what I needed to know and this is great to hear.
Thanks @ogs Sounds like your hunch was correct, based on what Rob Watts has said in john’s report. And thanks @Frank_Daman. I’ve been able to do a bit of acoustic treatment but for now, I must take into account the necessities of a small family apartment layout that falls far short of ideal, and my equipment does not include the possibility of any processing in amp or speakers. Perhaps someday!
You might look in to building yourself an acoustic diffuser. This page offers some insight: https://fullenglish.co/blog/2017/2/21/how-to-build-an-acoustic-diffuser
It shows a frequency attenuation table, but honestly, just make a cork board with different sized cubes and beams and this will make a huge difference and it looks kinda cool as well.
2 Likes