HQ Player

[tobes]

For a long time I just used the poly-sinc filter, then there were some things I liked about the xtr variant. I thought ext sounded good too.

At the moment, in my setup, I have a clear preference for the closed-form. Sounds natural without obvious artefacts and maintains good focus, layering and wide staging. Seems to sound good with all material (to me) and with both my speaker and headphone setups.

All systems/ears are different though.😉

[LoryWiv]

Installed HQ Player 4.03 and enjoying the improvements, including filter displayed within server as well as client, more manageable pixel size and scrollbar for settings dialogue, and in client, long album and song names intelligently begin a new line rather than encroaching into album art. The manual's inclusion of ratio compatibility info. and  checkbox for apodizing filters is very welcomed, and de-mystifies some things for me as well (EG - Closed-Form-16-M won't work on integer.)  This responsiveness to feedback, attention to detail and iterative improvements make HQ Player a very fair value over time IMHO....in addition to the excellence of it's audio engine, of course!Thank you, Miska. 

[LoryWiv]

Clarification requested, please: When default output mode is set to SDM at 48 x256, and I initiate playback of a DSD64-DSF track, does the oversampling filter apply or just the modulator? I ask because if I've last played an Nx PCM track (Closed-Form-16M), DSD playback won't start unless I first manually switch filter to the 1x setting (ply-sinc-xtr-mp-2s). I really enjoy Closed-Form-16M with Nx PCM but I would think that when tracklist contains a different file type the switch to settings selected in settings dialogue would occur automatically. Guidance appreciated, my settings are below:

[Miska]

1 hour ago, LoryWiv said:

Clarification requested, please: When default output mode is set to SDM at 48 x256, and I initiate playback of a DSD64-DSF track, does the oversampling filter apply or just the modulator? I ask because if I've last played an Nx PCM track (Closed-Form-16M), DSD playback won't start unless I first manually switch filter to the 1x setting (ply-sinc-xtr-mp-2s). I really enjoy Closed-Form-16M with Nx PCM but I would think that when tracklist contains a different file type the switch to settings selected in settings dialogue would occur automatically. Guidance appreciated, my settings are below:

 

Only modulator applies, and some of the settings from "DSD Sources" dialog. So the PCM-to-SDM filter restrictions shouldn't apply. I'll check this case that it is not unnecessarily applying the PCM filter limits.

 

[AnotherSpin]

3 hours ago, Le Concombre Masqué said:

 

I would upsample all PCM to 192 with Closed Form and shaped dithering ; I'd try none too and play with the dCS filters

 

 

+1 for combination of closed form and shaped dither.

[Miska]

1 hour ago, Miska said:

Only modulator applies, and some of the settings from "DSD Sources" dialog. So the PCM-to-SDM filter restrictions shouldn't apply. I'll check this case that it is not unnecessarily applying the PCM filter limits.

 

...fixed now for next release...

[evkatz]

Thanks so much to everyone for their help on my DCS Paganini question... One more question.  In HQ ver 4, what are the "Filter Nx" and the "Oversampling Nx" choices for?

[evkatz]

11 hours ago, Le Concombre Masqué said:

The idea is to best the DACs, their modulators, filters, processing power,  etc, by better ones provided by HQP and better run by a powerful computer ; however due to the prestige of dCS, I'd be interested to know what your ears prefer

 

I would let DSD 64 be DSD 64 and check Direct in DSD settings (but if you use convolution filters)

 

I would upsample all PCM to 192 with Closed Form and shaped dithering ; I'd try none too and play with the dCS filters

 

(nowadays) I don't like mp filters

 

 

Hi there.  When you refer to "closed form and shaped dithering" which filter/dither setting do you suggest?  And what, nowadays do you prefer over the mp filters?  thanks much!

[AnotherSpin]

45 minutes ago, evkatz said:

Hi there.  When you refer to "closed form and shaped dithering" which filter/dither setting do you suggest?  And what, nowadays do you prefer over the mp filters?  thanks much!

 

I would suggest you to read ten or so last pages of this thread, there are answers to your questions and lot of other useful information in length..)

[LoryWiv]

9 hours ago, Miska said:

 

...fixed now for next release...

Thank you!

[Sevenfeet]

4.0.3 seems to operate better for me, especially in DSD playback from PCM sources.  Previously it would not work sometimes or default back to PCM.  Now everything seems to work fine.

 

One mystery...when i first ran 4.0.3, it locked up on the splash screen.  A reinstall seems to have fixed the problem.  Not sure what happened (MacOS High Sierra).

[Confused]

14 hours ago, Miska said:

My theory (and ears) tell that filter choice tends to depend on source material.

 

Technically, especially newer source material needs apodizing filters to clean up some of the mess created by half-band digital decimation filters used in ADCs and some production software.

 

Other than that, subjectively I find minimum-phase filters sound good with things like older prog-rock (Pink Floyd etc) recordings and other such multi-track mix studio productions, also with some modern pop tracks. These don't contain any real acoustics at all, only little bit of artificial reverb. (curiously iPhone seems to use minimum phase filters for the headphone output)

 

While I find linear phase subjectively good sounding on recordings made with few microphones in real acoustic spaces.

 

Other than that, recently my preference has been poly-sinc-ext2 which is a linear-phase apodizing filter.

 

Would I be right that these comments apply more or less equally to upsampling to PCM and upsampling to DSD?

[freesteve]

thanks for suggesting all the different opinions on filters....

[Miska]

1 hour ago, Confused said:

Would I be right that these comments apply more or less equally to upsampling to PCM and upsampling to DSD?

 

Yes, regardless of output format. These reasonings are based on differences in the source content.

 

[Yviena]

@Miska with the closed-form-megs variants having more taps does that mean they are longer, and thus have more pre/post ringing than the normal variant? 

[Miska]

43 minutes ago, Yviena said:

@Miska with the closed-form-megs variants having more taps does that mean they are longer, and thus have more pre/post ringing than the normal variant? 

 

Yes, that's what it means. More taps/longer the filter is, longer it "rings".

 

[Yviena]

2 hours ago, Miska said:

 

Yes, that's what it means. More taps/longer the filter is, longer it "rings".

 

Ahh okay i was a little unsure about that as I thought longer filters  where independent of number of taps.

But then wouldn't longer filters like xtr/closed mega etc have worse time domain performance or does that somehow improve when upsampling to DSD?

[Miska]

3 hours ago, Yviena said:

Ahh okay i was a little unsure about that as I thought longer filters  where independent of number of taps.

But then wouldn't longer filters like xtr/closed mega etc have worse time domain performance or does that somehow improve when upsampling to DSD?

 

Using DSD is not really related to this much, it is more related to other aspects. Only notable time domain aspect of DSD is when the source content is also DSD, because then you don't really have much frequency domain band limiting which means that you avoid most of time domain effects of that frequency domain band limiting.

 

Longer impulse response is result of more taps, which gives you a have sharper cut-off in frequency domain. So yes time domain response suffers from length of "ringing" or "settling time" (which MQA calls "blur") perspective. It is same with analog filters too, there you just talk about order or Q of the filter instead of length/taps.

 

However, you cannot make that aspect of time domain performance better by making the filter shorter (less taps) forever, because at some point it begins to roll-off frequency response early which in turn slows down transients and thus making time domain response worse again (MQA being one prime example of such). Overdoing this "settling time" aspect also leads to more images/aliasing - less accurate reconstruction.

 

If we consider time domain performance as whole and talk about "transient response" or "step response" we note that it has two aspects.

1. Settling time (amount of ringing, related to Gibbs phenomenon)
2. Steepness/slope of the rise
3. Location/timing and shape of transient

If you try to improve (1) too much, then (2,3) suffers, or vice versa. Bound of (2) is ultimately defined by the sampling rate - fs/2 (Nyquist) frequency. We need to remember from the Gibbs/Fourier series that steepness of the step is defined by number of harmonics included in the series and accuracy of their relative levels. Lowest sampling rate involved defines how many harmonics can be included, but filter responses (both digital and analog) define correctness of the relative levels and possible limitations to the steepness/timing.

 

When MQA talks about time domain performance they talk about (1), while when Chord talks about time domain performance they talk about (2,3). And then they take things to the extremes.

 

Also if you overdo (1), the reconstruction accuracy also suffers due to severe images/aliasing and you instead just get a lot of distortion (MQA being example). If you overdo (2,3) you have transients that never settle.

 

"xtr" and "ext2" are still quite far from the "mega" filters. I've been putting a lot of effort in making filters that optimize all aspects, from all three time domain perspectives as well as frequency domain perspective, by using some nice math tricks. Then you can choose different weightings on these aspects with filter selection. But also all extreme approaches are supported (well, for DSD upsampling some "filter" choices like polynomial are not available at the moment).

 

 

I also want to strongly emphasize, that in any case, for recorded content (instead generated test signals), you always have effect of the ADC filters included in the source data. Or if it was recorded in hires and converted to RedBook for distribution in software, that software converter's digital filter fingerprint. Using non-apodizing (like closed-form) or all-pass digital "filter" just means that the source's digital filter fingerprint is passed through. So even if your playback filter wouldn't "ring", the source data already contains "ringing" from the production phase. To deal with this aspect of source data, apodizing filters were created. In recorded material, you won't find test signal -like pulses or transients because they have already passed some filters.

 

Also note that the filter "rings" only when it is limiting spectrum of the signal. If you have hires recording where sampling rate is high enough that no signal harmonics are reaching the beginning of filter's transition band (start of roll-off), the filter doesn't "ring". For example with MQA, it's filters begin to roll-off already around 30 kHz for 96 kHz sampling rate (48 kHz Nyquist), so their filters are much more likely to begin "ringing" than steeper filter with cut-off starting somewhere around 47 kHz. Because content is much more likely to have stronger harmonic at 30 kHz than at 47 kHz. In addition, it also means that slope/steepness of the rise is severely affected because of lower corner frequency (less harmonics included and their relative levels affected more).

 

[Yviena]

@Miska Thanks for the explanation  Miska.

 

 I think I remember reading somewhere though in this forum that higher signal speeds like sending/upsampling DSD128/256 improves time domain performance making it not matter so much at these rates, I'm unsure if it was in this thread or not, maybe it was you who wrote that can't remember.

 

On another note I'm wondering is it technically possible to create a filter that is apodizing.but still keeps original samples or is that impossible?

 

[Miska]

13 minutes ago, Yviena said:

 I think I remember reading somewhere though in this forum that higher signal speeds like sending/upsampling DSD128/256 improves time domain performance making it not matter so much at these rates, I'm unsure if it was in this thread or not, maybe it was you who wrote that can't remember.

 

Most of the limitations come from the lowest rate in the chain. Higher rates allow better construction accuracy though. I'm not completely sure of the context where that has been.

 

15 minutes ago, Yviena said:

On another note I'm wondering is it technically possible to create a filter that is apodizing.but still keeps original samples or is that impossible?

 

That is not possible because by definition the two are mutually exclusive. Intention of apodizing filters is to fix digital filter problems in the source content, and if nothing is changed, no fixing can happen.

 

[Yviena]

31 minutes ago, Miska said:

 

That is not possible because by definition the two are mutually exclusive. Intention of apodizing filters is to fix digital filter problems in the source content, and if nothing is changed, no fixing can happen.

 

Ah ok, is ADC ringing.really a bad thing though, as I still hear a larger sense of space, details etc with the closed form filter with my adi2/dac1541 so it can't really be so bad, unless it's meant for very badly mastered music.

[Miska]

5 hours ago, Yviena said:

Ah ok, is ADC ringing.really a bad thing though, as I still hear a larger sense of space, details etc with the closed form filter with my adi2/dac1541 so it can't really be so bad, unless it's meant for very badly mastered music.

 

All ADCs are oversampling and contain digital decimation filter to convert the DSD-like SDM data into lower rate PCM. Many ADCs like the AKM ones in ADI-2 Pro also have selectable digital filters just like DAC side.

 

Or if the material was converted from hires to RedBook for distribution, some software digital filter was likely used instead.

 

How it happens to be depends on what was used to create the source content.

 

Closed-form is interpolator, so it doesn't "filter" anything at all. Totally non-apodizing, you get all the good and bad through. So the results depend a lot on the source material.

 

I personally don't like the aliasing hash modern converters leave at top of the audio band, so I like to use apodizing filters to clean it up.

 

[PPk]

On 5/23/2019 at 4:34 AM, Le Concombre Masqué said:

In a word: clarity. 4 brings more micro-details 

I second it... Purchased the license straight away.. Can never go back to version 3 again.. 

[tboooe]

2 minutes ago, PPk said:

I second it... Purchased the license straight away.. Can never go back to version 3 again.. 

I've been waiting to purchase version 4 until all the bugs are figured out, specifically with Server 2016. Is anyone having problems using version 4 with Server 2016 and Roon?

[brightonjel]

Just tried 4.0.3 (on latest Mac OS .. 10.14.15) but have found that I can no longer adjust the volume of the DAC from HQPlayer (or from Room).  Everything else seems to work ok.  Anyone else seeing the same issue?  Am using NAA (microrendu) over Ethernet as the back end.