Which HQP Filter are you using? [2023]

No problem for my 9900K

Maybe it is your overclocking?

I use stock BIOS settings

The upsampling works fine at 44k DSD.

My question was the following : do you degrade sound quality by changing family rate (i.e. 96 or 192 pcm to 44k DSD)?

Noā€¦ It takes more computing resources, but thatā€™s all.

1 Like

Thanks! Is there any other reason to use 48k then?

1 Like

If you want to use a filter that cannot convert between the rate families, then it can be useful.

2 Likes

Yes! Now that you mentioned I brought my laptop to test the Holo Spring 3 with Wildism Audio before purchasing it. I was using my same filters I have on ifi NEO iDSD ASDMSECv2 poly-sinc-gauss / poly-sinc-high-res-mp. The amazing listening experience was the final yes.

Now at home Iā€™m stuck with NS9 PCM 1.538Mhz poly-sinc-gauss / poly-sinc-high-res-mp :rofl: on Spring 3 and havenā€™t used the REW EQ updates yet!!.

But I totally enjoy as @KML comments experimenting new variants! Is like getting a new DAC :notes::tada:

1 Like

So how does Spring3 compare with the Neo iDSD

When you compare same HQP settings of course.

Quite similar?

1 Like

Iā€™ll be happy to a similar test . Right now is not fair because I have the ifi Neo IDSD with Roon connected with my Mac Mini, is used now for Movies as it can connect easy with Bluetooth with Apple TV.

If I do PCM on Roon with ifi Neo IDSD and Holo Spring HQplayer, the sound reproduction is more ā€œAnalogā€. Overall the sound is more like ā€œlive stageā€

While I was at Wildism Audio It also had different speakers and I using Holo Red NAA. I used HQplayer directly with Qobuz and some of my local files. My laptop was able to do mostly DSD256 and I felt like DSD512.

So in a quiet time I will test both at DSD256 with same filters on HQplayer. Only difference is Spring is on XLR and NEO on RCA on same amplifier

3 Likes

Nice! Looking forward to the DSD256 battle! Donā€™t forget to make sure volume is equal in both.

You can play a 500 Hz tone and use an SPL meter app to make sure both playing at same SPL.

Otherwise the louder DAC will always win :smile:

3 Likes

So Iā€™m I using REW for this ? SPL app recommendations? Connect the Umik microphone to iPhone like HouseCurve?

1 Like

I guess you can use your UMIK-1 and REW to measure the amp volume knob gives matching SPL for each DAC.

Or use a free SPL meter phone app.

Try both - both free :grin:

1 Like

That contradicts my findings:

  1. I clearly hear a degradation when I go PCM192 to DSD256x44 instead of DSD256x48. The latter is more clear, textured and sharp.

  2. ā€œMatch rate familyā€ also takes less computing power (which seems to be logical since there is less conversion involved).

The only problem is that on certain DACs (including the A26) there is a slight pop when changing rate families).

There is just no reason for this from mathematical perspective. IOW, of what kind of data goes to the DAC.

Amount of calculations per sample is equal, but the RAM access hits larger memory portion when going between rate families. In fact, going to 48k multiple involves a bit more calculations because the output rate is higher (N samples per second need to be calculated). But usually the load seems higher when converting between rate families, because of the memory access wait times. Unless everything happens to fit in the cache. Which depends on amount of CPU cache and the filter in question. So far the AMD 7950X3D has biggest cache of whopping 144 MB.

Of course, then when the output data is ready, running modulator at 48k multiple takes 1.0884x more CPU time than running it at 44.1k multiple.

2 Likes

There might be a difference with your DAC but Itā€™s possible you are actually enjoying worse technical performance with the 48k output rate, compared with 44.1k output rate

What DAC ?

Would be interesting to measure the output

This would be the only reasonable explanation, adding to what Jussi wrote about

1 Like

Both Gustard A26 and Musetec MH-DA005.

If you have a 48k capable DAC, and if you are willing to do the listening test, Iā€™d be curious to know how you hear things.

1 Like

My comment wasnā€™t general. I have a properly designed DAC that works at DSD512 x 48k (ProJect Pre Box S2), one of many DACs i have

My comment was specific about analogue output of your DACs. It might be technically worse that you say sounds better, but only way to know is to measure

I have a good Cosmos A to D and measured a few DACs outputs

With the Project there is no difference in SQ at these 2 output rates - it is properly designed DAC

2 Likes

I have a 48k capable DAC, the Holo Audio May KTE, and I donā€™t hear any sound quality differences when using 48k x 256 or 44.1k x 256 when listening to any contentā€¦16/44.1ā€¦24/96ā€¦DSD64ā€¦24/192ā€¦DSD128ā€¦

7 Likes

anyone with a rockna Wavedream? any suggested settings for this dac?

Wow! I donā€™t know how I could have overlooked DSD upsampling for so long. :sweat_smile: Iā€™m loving the sound. Thanks Jussi, and everyone!

3 Likes

It works now? I thought you had stuttering?

What did you change?

1 Like