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

[sdolezalek]

If your DAC chip internally converts musical input at DSD512 or at PCM384, should you be upsampling to anything other than that and if so, aren't you just adding more conversion artificats?

 

I'm currently upconverting everything to DSD128 in HQPlayer because that is the highest DSD rate my Oppo BDP105D accepts. To me, it sounds great. But am I fooling myself?

 

I actually don't know what I would have to feed this DAC in order to forgo any further internal resampling, nor do I know if it will natively accept that format, but it seems that for those who know what that rate is and are feeding their DAC that rate and format, they should be getting the maximum benefit from HQPlayer.

 

For those of us, feeding it something that still requires further internal resamplings/conversions, I wonder just how much we are actually improving the sound vis-a-vis just feeding the DAC the original format and letting it apply its own upsampling and filtering (even if that is inferior to HQPlayers upsampling and filtering).

 

Of course, it would also really help if we could collectively come up with a chart that shows the actual native conversion rate for popular DACs.

[Norton]

If your DAC chip internally converts musical input at DSD512 or at PCM384, should you be upsampling to anything other than that and if so, aren't you just adding more conversion artificats?

 

I wonder just how much we are actually improving the sound vis-a-vis just feeding the DAC the original format and letting it apply its own upsampling and filtering.

 

Unless using a DAC that's purpose designed for such use or with upsampling that can be turned off, can we be certain that any software upsampling/filtering is not making things worse than better? How do you know that by software upsampling to the DAC's max rate you are somehow bypassing the DACs own processing - maybe its just doing the job twice and making things worse as a result?

 

From my listening "feeding the DAC the original format and letting it apply its own upsampling and filtering" gets the best results.

[sdolezalek]

Unless using a DAC that's purpose designed for such use or with upsampling that can be turned off, can we be certain that any software upsampling/filtering is not making things worse than better? How do you know that by software upsampling to the DAC's max rate you are somehow bypassing the DACs own processing - maybe its just doing the job twice and making things worse as a result?

 

From my listening "feeding the DAC the original format and letting it apply its own upsampling and filtering" gets the best results.

 

Norton: I think there are a lot of folks here who would disagree with that last statement. Just read the thread on the T+A DAC8 DSD to see how many people hear the benefit of feeding it 512DSD rather than native format. The same is true of the iFi Nano and Micro DACs. I have heard that many of the Chord DACs prefer high rate PCM. So my question was not about original format versus the DAC native format, but about halfway measures.

 

I also don't see how feeding a DAC DSD512 when it uses DSD512 internally could cause it to "do the job twice." Are you suggesting it would take DSD512 convert it back to PCM 16/44 and then back up to DSD512?

[Panelhead]

I have a Dac that sample rate converted everything to feed a Room Perfect chip. It sounded much better when using Pure Music to do this versus the internal async sample rate converter.

[Miska]

If your DAC chip internally converts musical input at DSD512 or at PCM384, should you be upsampling to anything other than that and if so, aren't you just adding more conversion artificats?

 

I'm currently upconverting everything to DSD128 in HQPlayer because that is the highest DSD rate my Oppo BDP105D accepts. To me, it sounds great. But am I fooling myself?

 

AFAIK, Oppo uses ESS Sabre. With Sabre chips, the DSD path through the DAC is much simpler and more straightforward than the PCM path. So you can happily send it any DSD. Regardless of the DSD rate it behaves the same way, but higher rates just improve performance.

 

Even if a DAC supports only up to 96k PCM, it is still worth upsampling RedBook to 88.2/96k before sending to the DAC. The first step up is sonically most critical. And doing 2x conversion would already put 22.05 kHz wide space between DAC's filter and the original content, so the DACs built-in filter is already having much lesser impact. In addition, with typical DAC chip, every doubling of input rate drops out one cascade digital filter section from the built-in filter chain.

 

In addition, many devices doing digital room correction or other DSP (AVRs and such) typically resample everything with no so great filters to either 96k or 192k to apply the DSP at fixed rate. Sending these already good quality resampled data will typically also improve performance.

 

Sometimes it may be useful to choose between PCM and DSD output though. In particular with Chord DACs it is better to use highest PCM rate, since they convert DSD inputs to that same one anyway it is better to send PCM out.

 

So no worries...

[semente]

I wonder why, with all the positive feedback of audiophiles using software upsampling and filtering, there aren't more manufacturers producing NOS filterless DACs...

 

Does it somehow disturb the status quo?

 

R

[sdolezalek]

Even if a DAC supports only up to 96k PCM, it is still worth upsampling RedBook to 88.2/96k before sending to the DAC. The first step up is sonically most critical. And doing 2x conversion would already put 22.05 kHz wide space between DAC's filter and the original content, so the DACs built-in filter is already having much lesser impact. In addition, with typical DAC chip, every doubling of input rate drops out one cascade digital filter section from the built-in filter chain.

 

In addition, many devices doing digital room correction or other DSP (AVRs and such) typically resample everything with no so great filters to either 96k or 192k to apply the DSP at fixed rate. Sending these already good quality resampled data will typically also improve performance.

 

Sometimes it may be useful to choose between PCM and DSD output though. In particular with Chord DACs it is better to use highest PCM rate, since they convert DSD inputs to that same one anyway it is better to send PCM out.

 

So no worries...

 

Jussi: Thank you, I was hoping you would reply. I had forgotten that most DAC chips do this by repeated doublings from the input starting point ( i.e. 44>88>192>384) till their internal conversion rate. I assume that means they are also applying their own filters at each step? So every time I substitute a preferred HQP filter for the DAC Mfg's filter, I benefit at each doubling?

 

My other request, not of you, but of the community here, was to begin putting together a list we might all share as to what the "optimal" or "native" input rates are for at least the most popular DACs. Do you think that would be a worthwhile exercise for CA users in order to help newer HQP customers maximize their benefit from upsampling?

 

Thanks, Stephan

[ElviaCaprice]

I would disagree about Chord Dac's sounding better by upsampling PCM. From my experience with the Hugo this is not the case. Best to feed the Hugo with given bit perfect feed. In fact, I think Redbook sounds better streamed as is, not upsampled.

 

I don't disagree with upsampling with HQP in general for most DAC's, just not the Chords.

[Jud]

I wonder why, with all the positive feedback of audiophiles using software upsampling and filtering, there aren't more manufacturers producing NOS filterless DACs...

 

Does it somehow disturb the status quo?

 

R

 

More expensive parts than using a commodity DAC chip, and puts you on your own as to how to implement the circuit (thus requiring engineering time and again more expensive). And to reach what fraction of an already small market?

[semente]

More expensive parts than using a commodity DAC chip, and puts you on your own as to how to implement the circuit (thus requiring engineering time and again more expensive). And to reach what fraction of an already small market?

I guess you're right.

Still, someone could take advantage of the tiny niche...

 

R

[hornytoad]

I would disagree about Chord Dac's sounding better by upsampling PCM. From my experience with the Hugo this is not the case. Best to feed the Hugo with given bit perfect feed. In fact, I think Redbook sounds better streamed as is, not upsampled.

 

I don't disagree with upsampling with HQP in general for most DAC's, just not the Chords.

+1

[Miska]

I assume that means they are also applying their own filters at each step? So every time I substitute a preferred HQP filter for the DAC Mfg's filter, I benefit at each doubling?

 

Yes. From this perspective, less you have filters left at the DAC side, the better. In many cases if you can reach 352.8/384k there are no filters left. That is also sort of down side, and that's why in many cases DSD is better option.

 

My other request, not of you, but of the community here, was to begin putting together a list we might all share as to what the "optimal" or "native" input rates are for at least the most popular DACs. Do you think that would be a worthwhile exercise for CA users in order to help newer HQP customers maximize their benefit from upsampling?

 

Yes, although the easiest way may be to list "exceptions to the rule". Like; "with Chord DACs, use highest PCM rate available".

[wwaldmanfan]

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.

 

I have experimented extensively with Audirvana's iZotope settings, as well as HQPlayer, which I also own. If there is any audible benefit with software upsampling, I'm just not hearing it. It always sounds veiled or muted. Maybe I'm just not using the right upsampling settings. Or, perhaps this is because my system and/or my ears are not sensitive enough, but the question that newbies and others here like myself want explained is how does upsampling add any enhancement to music from its native format? You can't add detail that's not there, and if we are talking about ultra-high frequency noise or artifacts, who can hear that, anyway? Is there inherent distortion with PCM at Nyquist frequencies when played "straight"? Miska has explained it in theory, but, is it audible? if so, I would LOVE for someone to suggest some specific Audirvana or HQPlayer PCM upsampling setting for my particular DAC that will enhance the SQ.

[whoozwah]

in my own experience my DAC will upconvert to either PCM176 or 192 depending on the the starting sample rate. if it's fed one of those sample rates at the outset it goes into NOS mode. I don't use HQplayer but I have fed the same track in 24/48 and 24/192 and I wanna say they sounded the same to my ears. That led me to believe that the upsampler in my DAC is good enough to warrant me not really caring what the sample rate of the source file is.

[Miska]

I would disagree about Chord Dac's sounding better by upsampling PCM. From my experience with the Hugo this is not the case. Best to feed the Hugo with given bit perfect feed. In fact, I think Redbook sounds better streamed as is, not upsampled.

 

I don't disagree with upsampling with HQP in general for most DAC's, just not the Chords.

 

I only have Mojo, have not tried Hugo. But Mojo clearly performs best when it's volume control is disabled and it is fed with 705.6/768k PCM data. Volume can then be controlled for example form HQPlayer as necessary.

 

Of course if you prefer the sound/performance of the Chord's WTA filter, then you prefer it and there's nothing to argue. The most similar sonic signature you get from HQPlayer with poly-sinc-hb or closed-form filters.

 

For me, the second biggest improvement with Mojo was to disable it's volume control and use external headphone amp (Schiit) instead.

[Jud]

Generally, to use internal DAC filters/modulators least, feed the highest DSD rate the DAC will accept. For Chord and DACs that don't accept DSD, feed the highest PCM rate they'll accept.

 

Now for the exceptions...

[Jud]

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.

 

I have experimented extensively with Audirvana's iZotope settings, as well as HQPlayer, which I also own. If there is any audible benefit with software upsampling, I'm just not hearing it. It always sounds veiled or muted. Maybe I'm just not using the right upsampling settings. Or, perhaps this is because my system and/or my ears are not sensitive enough, but the question that newbies and others here like myself want explained is how does upsampling add any enhancement to music from its native format? You can't add detail that's not there, and if we are talking about ultra-high frequency noise or artifacts, who can hear that, anyway? Is there inherent distortion with PCM at Nyquist frequencies when played "straight"? Miska has explained it in theory, but, is it audible? if so, I would LOVE for someone to suggest some specific Audirvana or HQPlayer PCM upsampling setting for my particular DAC that will enhance the SQ.

 

So what external sample rate will it accept?

[ElviaCaprice]

Of course if you prefer the sound/performance of the Chord's WTA filter, then you prefer it and there's nothing to argue.

 

I definitely do prefer the Chord's WTA filter, otherwise I wouldn't own a Chord. I also prefer a media player that gives me the most media format versatility and library functions for media playback. Thus JRiver for HQ video with lossless audio. Currently the best bang for the buck.

 

Meanwhile there are far more bigger system wide issues that I find more pressing in SQ than media software for playback (other than format needs met). But if one was inclined to go down the HQP route (it fulfills your format needs), then one should start with the proper DAC and streamer to get the most out of HQP benefits.

 

Thus back to the intent of this thread. Which ones?

[Miska]

but the question that newbies and others here like myself want explained is how does upsampling add any enhancement to music from its native format?

 

The native format is just essential set of "cues" needed to reconstruct the analog waveform. It is a bit like kids "connect the dots" drawing, where you have numbered points and need to draw a curve through the points to get to a full picture. If you add more very carefully calculated dots between the original ones, drawing the curves becomes easier and more accurate. Analog filter at the DAC output is the one drawing the final curves, everything is just dots before that.

 

You can't add detail that's not there

 

It is not about adding detail that doesn't exist, it is about making it easier for the analog conversion section for to accurately reconstruct the waveform. Data is not the signal, the signal needs to be built from the data. If there is more data, it helps getting an accurate end result.

 

and if we are talking about ultra-high frequency noise or artifacts, who can hear that, anyway?

 

Through intermodulation mostly. If you have two high frequency artifacts, at 100 kHz and 101 kHz -> you get 101 - 100 = 1 kHz difference tone which falls into audible range.

 

Is there inherent distortion with PCM at Nyquist frequencies when played "straight"?

 

Yes, this is best demonstrated with a NOS DAC.

 

Let's first take a look at 1 kHz tone played through a DAC without any upsampling, you can see that there's 1 kHz tone around every multiple of the 44.1 kHz sampling rate up to over 1 MHz frequencies. These are called "images" that are supposed to be removed by the analog filter at DAC's output. Those are essentially distortion - jaggedness in the output waveform.

 

Then the same source data, but now upsampled to 384 kHz sampling rate, you can now see that there's anymore tone left repeating around multiples of the 352 kHz rate and much lower in level.

 

This is particularly clearly visible when you make a 0 - 22.05 kHz frequency sweep.

 

Here at the original 44.1 kHz sampling rate, you can see the entire sweep spectrum repeating around every multiple of the 44.1 kHz sampling rate, the lowest frequencies being the strongest ones (since the DAC's output low-pass filter has not yet cut into those as much).

 

And here the same source upsampled to 384 kHz sampling rate, now you can see much lower level images which are repeating around multiples of 384 kHz and as such the number of those is also lower (four pairs, so total eight).

 

Since these components are fully correlated with the source signal, also the intermodulation products are fully correlated.

 

I would LOVE for someone to suggest some specific Audirvana or HQPlayer PCM upsampling setting for my particular DAC that will enhance the SQ.

 

Unfortunately I don't have specific good values for the NAD since I don't have it and it uses it's own approach. I would just feed it with 176.4/192 kHz PCM data and try different filters. If that doesn't sound good, then hires recordings at the same rate are likely not going to sound great either.

 

But it certainly has it's own set of digital filters / upsampling.

[Miska]

But if one was inclined to go down the HQP route (it fulfills your format needs), then one should start with the proper DAC and streamer to get the most out of HQP benefits.

 

Thus back to the intent of this thread. Which ones?

 

Well, I though this wasn't about HQPlayer or any other player in first place, but more about finding which particular input format makes best out of the DAC. There are even cases where for example 44.1k and 48k base rates have different jitter performance and the other performs better. Many DACs have one certain "sweet spot" - one input format that gives the best performance.

 

Frequently there are are also settings on the DAC that are involved in the end result.

 

For HQPlayer, there are bunch of good DACs. Just some examples being T+A DAC8 DSD, iFi DACs, TEAC UD-501/UD-503/NT-503, exaSound, Fostex, Sony, etc.

[Miska]

Unless using a DAC that's purpose designed for such use or with upsampling that can be turned off, can we be certain that any software upsampling/filtering is not making things worse than better? How do you know that by software upsampling to the DAC's max rate you are somehow bypassing the DACs own processing - maybe its just doing the job twice and making things worse as a result?

 

If the DAC is based on a DAC chip from the typical set of manufacturers, then it is usually pretty clear cut. If the DAC uses chips from TI/BB, AD, CL (incl Wolfson), AKM or ESS for example, then the behavior is known.

 

I have a collection of DACs that keeps growing that I measure and test.

 

If upsampling makes things worse, then also playing equivalent hires content would be making things worse. Pretty simple. I wouldn't buy such a DAC in first place.

[ericuco]

If the DAC is based on a DAC chip from the typical set of manufacturers, then it is usually pretty clear cut. If the DAC uses chips from TI/BB, AD, CL (incl Wolfson), AKM or ESS for example, then the behavior is known.

 

I have a collection of DACs that keeps growing that I measure and test.

 

If upsampling makes things worse, then also playing equivalent hires content would be making things worse. I wouldn't buy such a DAC in first place.

 

Miska - over on the Twisted Pear forum on DIYAudio I asked about this a few weeks ago, specifically what is the highest DSD rate supported by the ESS chip (ES9018) but really didn't get an answer. Do you know? In line with this thread topic, if I am feeding my DAC the highest rate possible then the less processing required by the DAC, supposedly a good thing.

 

Currently, I am using HQP to send DSD256 data to my DAC, the limit being the JLSounds USB board supporting native (non-DoP) DSD256 on Linux.

[wwaldmanfan]

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.

 

I have experimented extensively with audirvana's izotope settings, as well as hqplayer, which i also own. If there is any audible benefit with software upsampling, i'm just not hearing it. It always sounds veiled or muddy. Maybe i'm just not using the right upsampling settings. Or, perhaps this is because my system and/or my ears are not sensitive enough, but the question that newbies and others here like myself want explained is how does upsampling add any enhancement to music from its native format? You can't add detail that's not there, and if we are talking about ultra-high frequency noise or artifacts, who can hear that, anyway? Is there inherent distortion with pcm at nyquist frequencies when played "straight"? Miska has explained it in theory, but, is it audible? If so, i would love for someone to suggest some specific audirvana or hqplayer pcm upsampling setting for my particular dac that will enhance the sq.

 

so what external sample rate will it accept?

 

24/192 pcm

[Miska]

Miska - over on the Twisted Pear forum on DIYAudio I asked about this a few weeks ago, specifically what is the highest DSD rate supported by the ESS chip (ES9018) but really didn't get an answer. Do you know? In line with this thread topic, if I am feeding my DAC the highest rate possible then the less processing required by the DAC, supposedly a good thing.

 

I should be reading it more frequently than I am...

 

Note! I don't know details about the latest series ESS has put out! exaSound is first I've seen to use a chip from that series.

 

Based on the datasheet it should work up to DSD512 and there are some implementations. But it doesn't seem to be reliable.

 

The only requirement they have is that MCLK must be at least 3x the DSD BCLK and MCLK must be <= 100 MHz. I personally think it should work if one would use exactly the 3x clock. This is a bit hard of course with normal audio clocks. But if someone wants to experiment, someone could come up with some frequency like 90 MHz for MCLK and then I could add 30 MHz output to HQPlayer. That would of course require a bit of customization to the USB firmware too.

 

Currently, I am using HQP to send DSD256 data to my DAC, the limit being the JLSounds USB board supporting native (non-DoP) DSD256 on Linux.

 

That is, IMO, pretty good option.

[YashN]

the limit being the JLSounds USB board supporting native (non-DoP) DSD256 on Linux.

 

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".