[shwnwllms]

What is your opinion on oversampling to DSD? Should everything be oversampled to the same rate? I.E, 44.1 and 48khz both to 2,822,400 hz? Or should 48khz be oversampled to (I believe my math is correct) 3,072,000 hz?


I would think the even integer oversampling would be the best way to go, and is how I have my system setup, but considering how many DACs actually do not support DSD at any base rates other than 44.1khz, I wonder sometimes. Perhaps it is easier therefore cheaper to just oversample both to the same sample rate?

If your DAC supports it check adaptive output rate and 48 kHz so you stay within the rate family which will be less resource intensive

 

[MLGrado]

If your DAC supports it check adaptive output rate and 48 kHz so you stay within the rate family which will be less resource intensive

yeah, that is how I have it setup. I just wonder why DACs would not support DSD at 48khz base rate. Less resource intensive and I would presume higher quality too.

 

[shwnwllms]

yeah, that is how I have it setup. I just wonder why DACs would not support DSD at 48khz base rate. Less resource intensive and I would presume higher quality too.

Plenty of DACs only support 44.1 kHz (it’s cheaper to manufacture that way as you have to have a separate pathway for 48 kHz so it cuts down on components etc)

 

[MLGrado]

Plenty of DACs only support 44.1 kHz (it’s cheaper to manufacture that way as you have to have a separate pathway for 48 kHz so it cuts down on components etc)

yes that was my speculation. Cost saving.

Thanks so much

 

[dericchan1]

What is your opinion on oversampling to DSD? Should everything be oversampled to the same rate? I.E, 44.1 and 48khz both to 2,822,400 hz? Or should 48khz be oversampled to (I believe my math is correct) 3,072,000 hz?


I would think the even integer oversampling would be the best way to go, and is how I have my system setup, but considering how many DACs actually do not support DSD at any base rates other than 44.1khz, I wonder sometimes. Perhaps it is easier therefore cheaper to just oversample both to the same sample rate?

There really is no sound quality difference with upsampling to 44 or 48 base, it just uses more resources

 

[EMINENT]

Hallo.

I report.

After testing with Dac Correction now correctly applied, as of the current 5.7.0 version, my current pc hardware is incapable of rendering stutter free playback with DC at 1024 rates. Necessitating audition with DC at 512 rates instead for trial.
Upon further review, I was not able to find appreciable sonic or bodily engagement with DC at 512 rates in comparison to my previous working 1024 settings.

I shall conclude inconclusive results for now until suitable hardware upgrade or future revisions are capable of allowing those auditions.

 

[bogi]

I just wonder why DACs would not support DSD at 48khz base rate.

It's simple. For most people incl. DAC developers DSD is DSD64 (since only such spec exists) or DSD is the content one can download or stream from streaming services. No official music content at 48k based DSD rates exists since no such standard exists. DSD production software (Pyramix, Saracon...) most probably does not support 48k DSD rates at all. Furthermore, not only DAC developers, but also USB driver developers (XMOS, ...) don't take the playability of 48k DSD rates into account. Companies like iFi or Project who made 48k DSD working needed to fix USB driver code which was delivered to them from Thesycon. So it brings additional efforts and hassle for DAC developers to support 48k DSD rates. When 48k DSD rates work with usual DAC chip, it is most probably only a coincidence, not an intention. Except of very few companies which state 48k DSD support in specs of their DACs. But such companies are then more oriented to audiophile world (iFi, Holo etc).

How many Topping DAC users are doing PCM to DSD? Maybe 0.5% ? I don't have an idea, but statistically not much. Companies like Topping want to sell large amount of devices for the lowest price. They are oriented to usual customers. They don't invest significant efforts into drivers and firmware area. They simply use what was delivered to them (manufacturer specs, development boards, datasheets with usage examples) with all the bugs and incompleteness and they follow known standards to make something working. As soon they have something working they want to sell it in large amounts and not to go through unknown hassles because of 0.5% of (potential) users. In the previous years Topping, SMSL and other similar companies showed that instead of fixing old device bugs they bring new devices with new bugs, more times in a year...

 

[MLGrado]

It's simple. For most people incl. DAC developers DSD is DSD64 (since only such spec exists) or DSD is the content one can download or stream from streaming services. No official music content at 48k based DSD rates exists since no such standard exists. DSD production software (Pyramix, Saracon...) most probably does not support 48k DSD rates at all. Furthermore, not only DAC developers, but also USB driver developers (XMOS, ...) don't take the playability of 48k DSD rates into account. Companies like iFi or Project who made 48k DSD working needed to fix USB driver code which was delivered to them from Thesycon. So it brings additional efforts and hassle for DAC developers to support 48k DSD rates. When 48k DSD rates work with usual DAC chip, it is most probably only a coincidence, not an intention. Except of very few companies which state 48k DSD support in specs of their DACs. But such companies are then more oriented to audiophile world (iFi, Holo etc).

How many Topping DAC users are doing PCM to DSD? Maybe 0.5% ? I don't have an idea, but statistically not much. Companies like Topping want to sell large amount of devices for the lowest price. They are oriented to usual customers. They don't invest significant efforts into drivers and firmware area. They simply use what was delivered to them (manufacturer specs, development boards, datasheets with usage examples) with all the bugs and incompleteness and they follow known standards to make something working. As soon they have something working they want to sell it in large amounts and not to go through unknown hassles because of 0.5% of (potential) users. In the previous years Topping, SMSL and other similar companies showed that instead of fixing old device bugs they bring new devices with new bugs, more times in a year...

But didn’t you know topping is making DAC’s with SINAD approaching 125db and dynamic range greater than 130db, and near perfect linearity to -120db? They are measurement kings which we all know means they are obviously better than anything iFi or Holo could possibly make

Lol of course I’m being sarcastic and I’m sure most people will know exactly what I’m talking about.

The two Topping DACs I have had in my ‘lab’ for inspection were both under 500 dollars. One ESS and one AKM. Neither of them bothered to offer any DSD filter options even though they were listed in manual, of course ESS has no direct DSD mode, but the AKM does but it was not available. All the while some very big players in the review game confused fixed output mode with DSD bypass mode and I guess the team at Topping were oblivious to this or perhaps if it helped sell product whether it was true or not, so be it.

Oh, and the lowest scoring DACs for sound quality and enjoyment in my reviews are some of the best measuring. Almost as if they are engineered to perform well on the most common industry tests, such as 1khz distortion FFT more so than being designed to play music.


But at the same time I don’t want to be too hard on them over everything because not every product falls into that stereotype. They also make some fine stuff. But as always buyers beware at all times

 

[shwnwllms]

Hallo.

I report.

After testing with Dac Correction now correctly applied, as of the current 5.7.0 version, my current pc hardware is incapable of rendering stutter free playback with DC at 1024 rates. Necessitating audition with DC at 512 rates instead for trial.
Upon further review, I was not able to find appreciable sonic or bodily engagement with DC at 512 rates in comparison to my previous working 1024 settings.

I shall conclude inconclusive results for now until suitable hardware upgrade or future revisions are capable of allowing those auditions.

Change DSP pipelines to 8

 

[plumpudding2]

Change DSP pipelines to 8

dac correction GPU usage went down by 8gb.. the manual mentions it affects resource consumption I just have a hard time understanding why this is a setting and not dynamically adjusted depending on your output configuration.

 

[shwnwllms]

dac correction GPU usage went down my 8gb.. the manual mentions it affects resource consumption I just have a hard time understanding why this is a setting and not dynamically adjusted depending on your output configuration.

It’s a bug, it will be fixed in a future update. This is the workaround until then

 

[EMINENT]

Change DSP pipelines to 8

Hollo.

I have had set to 8 during test along with all trials with Default, Filter, Pool and every block amount combination each.

 

[OhHeyItsLou]

Stopping by to share my journey into HQPlayer. I've spent the majority of my time trying to optimize for DSD512 playback according to my taste. Since putting this together, I've already stumbled upon setting "Bits" to "20" for my Spring 3. Hopefully this can help some MacOS users. I've included my Hardware info so people can compare apples to apples. As far as DSD goes for most tracks I've tested, this sounds amazing. I'm still chasing the magic combination for the realism that you get in NOS on the Spring 3.

 

[Failed Engineer]

Anyone ever heard of this DSD converter?

http://www.rtaudiodesign.com/w/products/dacs/pure-dsd-converter/

Seems to have been around since 2016 or so.

 

[bogi]

Anyone ever heard of this DSD converter?

The author is using incorrect information and terminology.
According to him all DAC chips convert DSD to (pseudo) PCM format. With direct DSD capable DAC chips it is not true. Some BB amd ROHM chips don't allow any other mode than direct DSD mode (and thus volume control is not possible in DSD mode with them).
The formulation "Pure DSD converter without DAC" is itself a nonsense. Without a DAC chip it would bring sense.
But he continues: "In our converter the DSD signal can be directly used to reconstruct the original analog signal with an analog filter.".
In delta sigma design, analog low pass filter acts as DAC (D/A stage) in narrower sense.

There are many DACs which do DSD to analog conversion without a DAC chip involved. Holo DACs, T+A DACs, some Maranz and TEAC offerings, Gustard R26, Denafrips, Exasound, Jussis open source DSC1 and others.