Upsampling to anything other than your DAC's internal conversion rate

[YashN]

This is going to be a very unpopular statement around here. Let me quailfy it by saying that my NAD M51 DAC does not do DSD. However, it upsamples all PCM input to 844kHz PWM.

 

If the size of each segment (width) is 1 level, it equates to PDM which is the nature of DSD

[ericuco]

I get what you're saying but beware of confusing native streaming with native DSD.

 

For example, my iFi DAC still does native DSD on Mac, despite the streaming method to its interface being DoP.

 

In other words, "native" doesn't equate to "non DOP".

 

I know member TedB who is connected to NativeDSD, etc. considers native DSD to include DoP so I am never sure how to say this.

 

In my case, I am not using DoP.

[YashN]

I know member TedB who is connected to NativeDSD, etc. considers native DSD to include DoP so I am never sure how to say this.

 

In my case, I am not using DoP.

 

 

I know it's not often people take the time to make the distinction. I use it this way:

 

native streaming vs DoP streaming (that's until the USB interface).

 

native DSD DAC or non-native DSD DAC (DAC-level).

[gldgate]

Another possible PCM exception - Schiit Yggy (and perhaps their other multibit offerings). I have tried upsampling in HQP and JRMC but prefer sticking with native file and Schiit's internal filters. Not sure if I am in the majority or not. I have another DAC that does DSD (Eastern Electric Minimax Supreme) and prefer using HQP when using that.

[Jud]

I know it's not often people take the time to make the distinction. I use it this way:

 

native streaming vs DoP streaming (that's until the USB interface).

 

native DSD DAC or non-native DSD DAC (DAC-level).

 

What do you classify as native vs. non-native DSD DACs?

[Miska]

Another DAC to list as "forget DSD and use PCM at 352.8/384k" is the new Prism Callia.

[hornytoad]

Another possible PCM exception - Schiit Yggy (and perhaps their other multibit offerings). I have tried upsampling in HQP and JRMC but prefer sticking with native file and Schiit's internal filters. Not sure if I am in the majority or not. I have another DAC that does DSD (Eastern Electric Minimax Supreme) and prefer using HQP when using that.

 

Upsampling to Dsd is a negative using the Yggy.

 

It colors the sound too much

[Miska]

Upsampling to Dsd is a negative using the Yggy.

 

It colors the sound too much

 

Sure, because it doesn't support DSD in first place at all...

[ElviaCaprice]

Maybe a thread, updated with a list of all DAC's tested with HQP, optimum SQ upsampled or not bitrate and format, preferred HQP settings.

You could call it, HQP for Dummies.

[hornytoad]

Sure, because it doesn't support DSD in first place at all...

 

Neither does my Chord 2 qute .

 

It sounds better not upsampling

[alubis]

I have tried upsampling using hq player both for Chord Hugo and Hugo TT, and prefer not to upsample as well.

 

Sent from my SM-G930F using Computer Audiophile mobile app

[徐中銳]

Recalling Low's comment :

The digital HiRes initiative is further confusing this issue. While high-resolution files can offer significant audio advantages, the most important thing they offer is that when properly processed, there are less audio band artifacts, less misinformation to corrupt our music. Unfortunately, higher speed processors are noisier, so a 24/192 file sometimes sounds better when processed at 24/96.

Had reason to recall what WL said at

From a 2 hours interview at recent High End Munich 2016' date=' edited down to 53 minutes :

[/i']

[ted_b]

Neither does my Chord 2 qute .

 

It sounds better not upsampling

What are you saying? The Yiggy does not play back DSD, the Chord 2Qute absolutely does. This is Miska's point; the Yiggy cannot possibly accept any format upsampled to DSD; the 2Qute of course can. Whether it sounds good is a debate, of course, but I don't understand your "2Qute doesn't support DSD" comment.

[One and a half]

To avoid sample rate switching delays in DACs, it's easy to set on sample rate like 192 or DSD128 and let the DAC run.

Well, on the other side the coin is the DAC inventor, that would have fed the DAC with native sample rates of 192, 174.6 etc and judged the DAC's performance on those rates.

Considering the amount of filtering available these days, there is no chance in Hades of selecting the right filter 'to make the sound better' when rarely does anybody (I'm talking about the DAC user) know what the topology is in the DAC to begin with, let alone to apply an external filter characteristic or even consider increasing the sample rate.

 

I don't recall seeing, oh, the DAC has ESS9018 and 5th order butterworth filters and the optimum rate for this DAC is 2x upsampling, plus Izotope a,b,c,d, e settings, or polysinc..with dither..? This is all too hard!

 

After changing to A+ a month ago after about 18 months on HQPlayer, the sound for me is better with no upsampling, leaving the filtering bog standard and play the file in its native format, whether 96, 192, DSD64 whatever. There is a delay time when switching sample rates, but I play albums and 70% of the library content is redbook, so the sample rate switching delays occurs, but it's not bothersome.

 

Initially I came round to Miska's way of thinking about using filtering and upsampling, but certain filters behave with certain music, and I couldn't be bothered changing the settings to suit, all got a bit hard. So now I just accept what the DAC designer does and work on the sound that goes in and comes out.

[YashN]

What do you classify as native vs. non-native DSD DACs?

 

Native: the D/A is done on the DSD content without intermediate processing in PCM domain

 

Further down the rabbit hole with yet another distinction or rather terminology (direct) by people who know far more than me (Jesus, Miska): The Nativity of Native DSD:

 

So, for those reasons I'd like to define two terms in the DAC world: Native DSD processing and its smaller subset, Native Direct DSD processing (aka Direct). In both cases the DSD signal is not converted to PCM and is therefore native. In the case of Direct the signal is not even processed digitally beyond being handled as a one bit signal. In my world both of these technologies have a good chance of producing beautiful DSD playback, given good power supplies and analog stages.

 

 

Examples of non-direct Native processing are when a DAC uses a SABRE 90XX chipset. The DSD signal is digitally processed, usually to 6 or 8 bits [This treatment of DSD is called "multi-bit," and is not at all the same as PCM. – Editor], but remains DSD and is treated with DSD filters (to remove noise shaping artifacts) and then sent to the analog side. One could argue that FPGA processing like PS Audio's Directstream, where one bit DSD is converted to 30 bit, 30Mhz, then processed "down" to DSD128, is another example of non-direct Native DSD processing. Meitner's upsampling to DSD128 falls in this category for me, too.

 

 

Examples of Direct Native DSD processing is when any of the TI/Burr Brown chipsets like 1792, 1793 and 1795 that support DSD processing are used in a DAC. They may also do PCM processing, but have a direct conversion process for DSD. Strangely, they may be labeled PCM1793 or DSD1793, and the only difference is the pin setups.

 

[sdolezalek]

Ok, after a slow start, this has gotten quite a bit more lively. Let me start by accepting the comment made by "One and a half" that trying to figure out exactly which setting, filters, etc. to use on which types of music might just be too much work for most. I have no issue with that and it makes life after the original DAC buying decision much easier.

 

For the rest of us, who like the freedom that front end software with its own upsampling algorithms and filtering choices provides, this thread was intended to provide some guidance as to where to start exploring what you like and what you don't. What I have heard so far is that there are some DACs that do seem to prefer DSD and others that seem to prefer PCM. I think Miska has made a very coherent case for why doing the filtering at much higher sampling rates is capable of reducing noise artifacts in the audible range (see post #19 above). So, based on the responses submitted so far let me take a cut at a DSD pro and PCM pro list (and please feel free to correct my mistakes and I'll republish the chart:

 

Best with DSD input: iFi, Oppo, T+A, Teac, exaSound, Fostex, Sony and most other ESS Sabre-based DACS

 

Best with PCM input: Chord DACS, NAD, Schitt (Yggy), Prism Callila

 

After this list gets a bit further along, I'll try to take a shot at also including some specific input rates for each DAC (knowing that risks much greater disagreement).

[semente]

I know it's not often people take the time to make the distinction. I use it this way:

 

native streaming vs DoP streaming (that's until the USB interface).

 

native DSD DAC or non-native DSD DAC (DAC-level).

 

 

Can we feed a non-native DSD DAC with a DSD stream?

DAC chip is PCM1794.

 

R

[YashN]

Can we feed a non-native DSD DAC with a DSD stream?

DAC chip is PCM1794.

 

No, won't work with that chip. That's a PCM DAC chip, which is confusing the matter even further (please don't call a PCM DAC Chip a non-native DSD DAC chip).

 

With this one you can: http://www.ti.com/lit/ds/symlink/dsd1793.pdf

 

So, if you want to listen to DSD with that chip, you will have to pre-convert the DSD to PCM before sending it to the DAC, for instance real-time with Audirvana+ or HQ Player.

[Miska]

Considering the amount of filtering available these days, there is no chance in Hades of selecting the right filter 'to make the sound better' when rarely does anybody (I'm talking about the DAC user) know what the topology is in the DAC to begin with, let alone to apply an external filter characteristic or even consider increasing the sample rate.

 

I don't recall seeing, oh, the DAC has ESS9018 and 5th order butterworth filters and the optimum rate for this DAC is 2x upsampling, plus Izotope a,b,c,d, e settings, or polysinc..with dither..? This is all too hard!

 

It doesn't matter... The filter choice depends on the source material and preferences of the listener, not on the DAC.

 

Dither algorithms for PCM are up to your preferences, if you don't want to bother, just select TPDF and you are fine.

 

For modulator, it is also largely preference based, but for simple lowpass filter DSD DACs (Lampizator) it may be technically better to choose lower order modulator (5th order) and for the ones with more complex lowpass filter (DAC chip based and T+A) it is better to choose higher order modulator (7th order).

 

After changing to A+ a month ago after about 18 months on HQPlayer, the sound for me is better with no upsampling, leaving the filtering bog standard and play the file in its native format, whether 96, 192, DSD64 whatever.

 

The DAC cannot know if for example 192 kHz data is result of upsampling or happens to be hires source originally created at 192 kHz. For the DAC it is just data sampled at 192k.

 

So now I just accept what the DAC designer does and work on the sound that goes in and comes out.

 

You could also accept my defaults or recommendations and stick to those without playing with the settings. This is no different from sticking to DAC designer's choices. Noting here that "DAC designer" is usually the engineer at the company who manufactures the DAC chip and he's under strict resource and thermal budget, so has to make many compromises on his design so that it

1) doesn't increase the chip price above the set price point

2) as a result of (1) can fit within the computational resource constraints of the chip

3) doesn't exceed set power consumption limits - meaning that the DAC chip cannot require heat sink or active cooling

4) fits in the chip manufacturing process

And once the design is set, it is manufactured at least for 10 or more likely for 20+ years without changes or possibility for updates/improvements. The improvements then come as a new product at some point.

[semente]

No, won't work with that chip. That's a PCM DAC chip, which is confusing the matter even further (please don't call a PCM DAC Chip a non-native DSD DAC chip).

 

With this one you can: http://www.ti.com/lit/ds/symlink/dsd1793.pdf

 

So, if you want to listen to DSD with that chip, you will have to pre-convert the DSD to PCM before sending it to the DAC, for instance real-time with Audirvana+ or HQ Player.

Thanks.

So if I understand correctly there are DSD-able chips and those which do PCM only.

 

R

[YashN]

Thanks.

So if I understand correctly there are DSD-able chips and those which do PCM only.

 

and those who do both...

 

... and there's chip-less DSD as well

[Miska]

So if I understand correctly there are DSD-able chips and those which do PCM only.

 

Yes, and for the TI/BB models the PCM path is pretty much the same with some differences in digital filters and converter output stages. But the overall architecture is the same. They've omitted DSD capability from the parts that are pin-controlled ("H/W control" in their terminology) instead of I2C/SPI software control. This is because the number of pins that can be used for controlling the DAC is limited. With software control over communication bus (I2C and SPI) there are no such limitations.

 

There are two otherwise identical models, PCM1794A and PCM1792A, where 1794 is pin-controlled and 1792 is software controlled. And 1792 supports DSD.

 

The latest flagship model PCM1795 also supports both PCM and DSD.

 

Some manufacturers don't want to bother to have a microcontroller (and the associated programming needed) in the DAC and want to use easier pin-control DAC chip models. In the past I wrote the necessary control software for various S/PDIF receivers and DAC chips for a small microcontroller. It took a week or so to get everything done and working, but I have strong software development background. The MCU (microcontroller) itself costs just $1/piece, so it is not so much of a cost issue.

[hornytoad]

What are you saying? The Yiggy does not play back DSD, the Chord 2Qute absolutely does. This is Miska's point; the Yiggy cannot possibly accept any format upsampled to DSD; the 2Qute of course can. Whether it sounds good is a debate, of course, but I don't understand your "2Qute doesn't support DSD" comment.

 

Yggy upsamples , not to Dsd you are correct .

 

What I meant to say is the Yggy sounds better without software doing any of the upsampling and letting their filter do the work .

[semente]

Yggy upsamples , not to Dsd you are correct .

 

What I meant to say is the Yggy sounds better without software doing any of the upsampling and letting their filter do the work .

 

But if you feed the Yggy with upsampled or higher sampling material it'll still use its filter won't it?

 

R

[hornytoad]

But if you feed the Yggy with upsampled or higher sampling material it'll still use its filter won't it?

 

R

 

Yeah as far as I know . I could be wrong but I think the Yggy upsamples all files to 24/192 .