DSD upsampling over original format

[kilroy]

With my modified W4S DAC-2 DSDse fed by JRiver 19, I greatly prefer low-res lossless files (44 to 192) to be upsampled in JR to DSD128. More body, more "organic" (?). When I go back to non-upsampled, the music feels almost "lightweight". Without the A-B comparison that can be made in seconds, I might still prefer non-upsampled. It's not bad, just not the same. I talked to my DAC mfr about DACs that upsample internally, they said that's a bad thing and it harms the music. Not sure. But I know what I like, although I'm not the best at putting these differences into words. Just better, to my ears.

[Miska]

With my modified W4S DAC-2 DSDse fed by JRiver 19, I greatly prefer low-res lossless files (44 to 192) to be upsampled in JR to DSD128. More body, more "organic" (?). When I go back to non-upsampled, the music feels almost "lightweight". Without the A-B comparison that can be made in seconds, I might still prefer non-upsampled. It's not bad, just not the same. I talked to my DAC mfr about DACs that upsample internally, they said that's a bad thing and it harms the music. Not sure. But I know what I like, although I'm not the best at putting these differences into words. Just better, to my ears.

 

Your DAC is also upsampling internally, inside the DAC chip...

[audiventory]

Kilroy,

 

 

You have complex systems:

 

file resolution - DSP in audio player - DSP into driver (except ASIO and WASAPI) - DSP in DAC.

 

First need provide driver without DSP.

 

Second detect "best sounding" mode (PCM/DSD + sample rate + bit depth).

 

Not fact what highest sample rate / bit depth is the best.

 

Also it can depend on available instance of DAC.

 

Best regards,

Yuri

[kilroy]

W4S told me the DAC does not upsample. Can you please explain? (and I'm not an engineer, so.... )

 

Your DAC is also upsampling internally, inside the DAC chip...

[Jud]

W4S told me the DAC does not upsample. Can you please explain? (and I'm not an engineer, so.... )

 

Unfortunately the subject is confused by marketing terms. W4S may say they "oversample" rather than "upsample." But Miska is right, nearly all DACs, including yours, do interpolation and accompanying filtering to raise the sample rate. As an alternative you can choose to do this in software before it gets to your DAC.

[kilroy]

Unfortunately the subject is confused by marketing terms. W4S may say they "oversample" rather than "upsample." But Miska is right, nearly all DACs, including yours, do interpolation and accompanying filtering to raise the sample rate. As an alternative you can choose to do this in software before it gets to your DAC.

 

 

Let's say I have a CD-ripped 16/44 file. I set JRiver to upsample to DSD 128. The DAC front readout shows "128".

 

Same file, played at original resolution in JR, reads "44" on the DAC. I hear a difference from the upsampled file.

 

But you are saying the DAC automatically upsamples. To what? You mean the 44 file is not playing at 44?

 

Anyway, does it even matter? I hear a difference when I upsample in JRiver. The point is not whether the DAC upsamples or not, but that I prefer the SQ of files upsampled in JR over their original resolution. That's all.

 

Curious to hear if others also prefer upsampled files.

[audiventory]

DAC indicate own input sample rate only.

 

Indication don't depend how processed audio into DAC.

[firedog]

Let's say I have a CD-ripped 16/44 file. I set JRiver to upsample to DSD 128. The DAC front readout shows "128".

 

Same file, played at original resolution in JR, reads "44" on the DAC. I hear a difference from the upsampled file.

 

But you are saying the DAC automatically upsamples. To what? You mean the 44 file is not playing at 44?

 

Anyway, does it even matter? I hear a difference when I upsample in JRiver. The point is not whether the DAC upsamples or not, but that I prefer the SQ of files upsampled in JR over their original resolution. That's all.

 

Curious to hear if others also prefer upsampled files.

 

The short answer is that almost all modern DACs take your 44.1 signal (or any PCM signal) and internally convert it to a form of DSD on the way to turning the original digital signal to analog.

That's even if your DAC doesn't have "upsampling" engaged. Your DAC still reads 44.1, b/c this isn't considered "upsampling" in DAC marketing speak, - so it's telling you it's received a 44.1 signal. In DAC marketing speak as long as it's not upsampling to another PCM family like 24/96 or 24/192, it isn't usampling, even if the DAC automatically changes all PCM to DSD internally, as part of the way the DAC does the DA conversion. Thus the comments about your DAC upsampling internally.

 

The reason Jud and Miska mention this is that many consider it advantageous to do this upsampling on a PC and not internally in your DAC chip. The idea is twofold: the more powerful PC can do a better job of conversion and also allow you - if you have the proper software - to use different filtering, modulating, and dithering that may make the sound more to your liking than what's built into your DAC chip. The other claimed benefit is that if you do this upsampling to DSD on your computer using software like Miska's (or with JRiver), then you skip that difficult conversion stage inside your resource constrained DAC, which should mean better SQ on the analog end.

 

But you are right, if you prefer the upsampling in JRiver to DSD 128, that's what matters. Your preference is anecdotal evidence that you get better sound by converting to DSD before the stream gets to your DAC chip, and not internally to the DAC chip.

 

I'm sure Jud will probably chime in with a better and more detailed explanation, but the above is a synopsis.

[kilroy]

Nicely put firedog, thank you.

 

Could I surmise then that the same file, at different PCM sample rates, would in theory sound the same if all were upsampled in JRiver (or other app) to the same DSD rate, as they are all converted to DSD in the DAC anyway?

 

Would this also mean that a native DSD file fed straight through JRiver is NOT upsampled in the DAC? And that different DSD file rates will all play at their native resolution?

 

 

 

The short answer is that almost all modern DACs take your 44.1 signal (or any PCM signal) and internally convert it to a form of DSD on the way to turning the original digital signal to analog.

That's even if your DAC doesn't have "upsampling" engaged. Your DAC still reads 44.1, b/c this isn't considered "upsampling" in DAC marketing speak, - so it's telling you it's received a 44.1 signal. In DAC marketing speak as long as it's not upsampling to another PCM family like 24/96 or 24/192, it isn't usampling, even if the DAC automatically changes all PCM to DSD internally, as part of the way the DAC does the DA conversion. Thus the comments about your DAC upsampling internally.

 

The reason Jud and Miska mention this is that many consider it advantageous to do this upsampling on a PC and not internally in your DAC chip. The idea is twofold: the more powerful PC can do a better job of conversion and also allow you - if you have the proper software - to use different filtering, modulating, and dithering that may make the sound more to your liking than what's built into your DAC chip. The other claimed benefit is that if you do this upsampling to DSD on your computer using software like Miska's (or with JRiver), then you skip that difficult conversion stage inside your resource constrained DAC, which should mean better SQ on the analog end.

 

But you are right, if you prefer the upsampling in JRiver to DSD 128, that's what matters. Your preference is anecdotal evidence that you get better sound by converting to DSD before the stream gets to your DAC chip, and not internally to the DAC chip.

 

I'm sure Jud will probably chime in with a better and more detailed explanation, but the above is a synopsis.

[Jud]

Nicely put firedog, thank you.

 

Could I surmise then that the same file, at different PCM sample rates, would in theory sound the same if all were upsampled in JRiver (or other app) to the same DSD rate, as they are all converted to DSD in the DAC anyway?

 

Would this also mean that a native DSD file fed straight through JRiver is NOT upsampled in the DAC? And that different DSD file rates will all play at their native resolution?

 

Depends on the chip (or whatever else is doing the conversion, such as an FPGA) in the DAC. In some DACs DSD128 may not be further processed. I don't know whether there would be DACs that convert, for example, DSD64 to DSD128, or DSD64 and DSD128 to DSD256, before final conversion to analog. With ESS SABRE DAC chips, everything coming in is converted to something around a 40-44.1MHz sample rate, if I remember correctly. (With regard to all of these items, someone like Miska would have better, more detailed knowledge.)

 

As for whether it matters, probably not beyond knowing there are various alternatives you can try to see what you like.

[kilroy]

Thank you Jud.

[YashN]

I don't know whether there would be DACs that convert, for example, DSD64 to DSD128, or DSD64 and DSD128 to DSD256, before final conversion to analog.

 

The PS Audio DirectStream does pre-convert to an intermediate high-rate format before doing DSD128 at output (if memory serves me well).

[kilroy]

I sent an email to W4S about their DAC chip based on what I'm learning here. I post that and the response from the marketing guy.

My questions:

I am told that the DAC chip in my DSDse oversamples/upsamples, via interpolation and accompanying filtering, to raise the sample rate. So a 16/44 file for example does not actually play back at this resolution, but by a higher one. Is it just your DAC that does this, or all DACs. They infer ALL DACs with this type of circuitry behave like this.

 

If that's the case, then the front panel reads out the resolution of the signal received, not the actual playback resolution. True?

 

It came up because I said that I prefer the sound of lower-res files upsampled to DSD in JRiver then played back via the DAC over playing at their original resolution. Their response was that all DACs upsample in the DAC chip.

 

 

The response:

 

It sounds like these ... people are getting caught up in the classic confusion between oversampling and upsampling. They are quite different. Upsampling is what you referred to in your email, i.e. taking a 44k sample rate and upsampling it to 88k or higher. Our chip does not do that, and whomever said "all DACs upsample in the DAC chip” is incorrect. Some do, some don't. Our DAC chip does oversample, however. What this does is to sample the signal at a faster rate than it is being received for increased accuracy and to reduce jitter. It doesn’t manipulate the signal (nor the sample rate). The front display shows the sample rate, which is the same input to output.

 

 

 

Oh boy. How does this correlate to earlier posts?

[audiventory]

Oversampling = upsampling = multiply input sample rate

 

DAC do upsampling/oversampling/[multiply input sample rate] and apply to upsampled/oversampled/multiplied input sample rate digital filter.

 

We can say that "all DACs upsample in the DAC chip” is incorrect.

 

"some DACs upsample in the DAC chip” is correct. As examples:

 

For upper sample rates (as example 352, 704, ... kHz) DAC can not-upsample but do digital filtration.

 

Also there are non-oversampling DAC - without upsampling/oversampling inside.

 

Same chip can used in different mode for certain DAC.

 

What inside DAC and how it use chip know developers only.

[Miska]

Same chip can used in different mode for certain DAC.

 

What inside DAC and how it use chip know developers only.

 

In this case it is ESS Sabre 9018 which has two distinct cascaded oversampling digital filters. It has also OSF bypass if you want to run external filters and for this case recommended is 352.8/384k and maximum supported is 1.536 MHz (32-bit integer input). Rest of the oversampling is done using SAH or similar function. Measuring the DAC output makes it nicely evident how it is being run, and you can also hook to the I2C configuration bus to see how the DAC chip is being configured (for example using a scope that can decode I2C).

 

It has 64-element D/A conversion array (64 1-bit DACs in parallel), equivalent of 6-bit resolution.

 

 

(running it in NOS mode would make the modulator undersampled and results similar to how iDSD Micro looks like in NOS mode: http://www.computeraudiophile.com/blogs/miska/ifi-idsd-micro-measurements-632/)

[Jud]

I sent an email to W4S about their DAC chip based on what I'm learning here. I post that and the response from the marketing guy.

The response:

 

It sounds like these ... people are getting caught up in the classic confusion between oversampling and upsampling. They are quite different. Upsampling is what you referred to in your email, i.e. taking a 44k sample rate and upsampling it to 88k or higher. Our chip does not do that, and whomever said "all DACs upsample in the DAC chip” is incorrect. Some do, some don't. Our DAC chip does oversample, however. What this does is to sample the signal at a faster rate than it is being received for increased accuracy and to reduce jitter. It doesn’t manipulate the signal (nor the sample rate). The front display shows the sample rate, which is the same input to output.

 

 

 

Oh boy. How does this correlate to earlier posts?

 

Correlates pretty exactly:

 

Unfortunately the subject is confused by marketing terms. W4S may say they "oversample" rather than "upsample." But Miska is right, nearly all DACs, including yours, do interpolation and accompanying filtering to raise the sample rate. As an alternative you can choose to do this in software before it gets to your DAC.

[YashN]

Correlates pretty exactly:

 

Yup yup, that was very funny!

[firedog]

Correlates pretty exactly:

 

to make it clear to kilroy: This means that his DAC does turn PCM into DSD on the way to analog?

[Jud]

to make it clear to kilroy: This means that his DAC does turn PCM into DSD on the way to analog?

 

Oh yeah. If I have time today or tomorrow I'll post some history of DAC innards which should help shed some light.

[kilroy]

I guess my original intention was not so much technical as listening experience directed, but what the hey it's all a learning experience. Not an electrical eng (real world I'm a mechanical guy) so a lot of Miska's and other's posts are beyond me.

 

As it has evolved my understading/questions now seem to be, PCM does convert to DSD before analog according to above posts, yet W4S says no if I interpret correctly; if in fact, as W4S says, the front panel readout is the output resolution, then it is true that it always reflects what the original file format is if I do not upsample or otherwise change these settings in JRiver, and when I do set DSP to upsample, it reflects that upsample. And if I do upsample, at some level the file reaches the point where the DAC will no longer upsample, and may even downsample, i.e. reaches its maximum output resolution.

[Jud]

I guess my original intention was not so much technical as listening experience directed, but what the hey it's all a learning experience. Not an electrical eng (real world I'm a mechanical guy) so a lot of Miska's and other's posts are beyond me.

 

As it has evolved my understading/questions now seem to be, PCM does convert to DSD before analog according to above posts, yet W4S says no if I interpret correctly; if in fact, as W4S says, the front panel readout is the output resolution,

 

No - they're saying they "oversample" instead of "upsample," so either way they're changing the input sample rate. The readout reflects the input resolution. Think about it for a moment and you'll see the readout can't possibly reflect the output resolution: It's a DAC (Digital to Analog Converter), so the output is analog (music) and has no resolution. (Resolution figures are bit depth and sample rate, of which the music coming out of your DAC has neither. )

 

then it is true that it always reflects what the original file format is if I do not upsample or otherwise change these settings in JRiver, and when I do set DSP to upsample, it reflects that upsample. And if I do upsample, at some level the file reaches the point where the DAC will no longer upsample, and may even downsample, i.e. reaches its maximum output resolution.

 

Nope. Sit tight for a bit and I'll explain this all thoroughly in another note, including why the explanations you tend to get from DAC manufacturers about this are often more confusing than helpful.

[k6davis]

I'm like you in that I don't go too far into the technical side of things, but my understanding is that most DACs have an optimum sampling rate and that all inputs are internally converted to that rate. This generally involves upsampling both Redbook/CD and DSD64 to a higher DSD sampling rate - particularly in the popular ESS DAC chips. The sample rate your DAC displays is the input rate and does not reflect what's happening internally in the DAC. It's a little analogous to an HDTV. If yours is 1080p, all video signals are internally upconverted to 1080p.

 

It makes sense to use your computer to do the upsampling instead as it has many times more processing power than the chip in most DACs and it's allows you to try different software/upsampling algorithms at will. I currently use HQPlayer to upsample everything to DSD128 or DSD256 and am getting great results.

 

There are DACs that explicitly do not oversample (NOS) and there's a new generation of DAC's that have chips which are far more powerful and configurable (FPGA) but presently, they are the exception.

 

Again, this is a greatly oversimplified and generalized explanation and the more technically inclined among us can provide a lot more detail.

[Jud]

If I have time today or tomorrow I'll post some history of DAC innards which should help shed some light.

 

So why, instead of just saying what goes on inside their (by all indications very fine) DAC, did W4S give Kilroy a reply so laden with marketing-speak that he wound up thinking they were saying the exact opposite of what they actually meant?

 

It all goes back to the first CD players. (I know, why do we need to look at something that happened 30 years ago? Just hang on, the context is actually helpful to understanding, and I'll try as much as possible to put it all in plain English.)

 

Many people thought the first CD players sounded pretty bad. Many engineers thought the reason for this was the filters used in those early players. Filters?

 

OK, digital audio is based on the Sampling Theorem (the name Nyquist gets mentioned a lot, though there were other guys too, like Whittaker and Shannon). Basically what the Sampling Theorem says is, in the world of mathematics, you can take samples of any signal (like music) and reconstruct it to mathematical perfection, as long as you take samples more than twice as fast as the signal changes (more than twice the highest "frequency of interest," which with music meant for us humans is the upper frequency limit of our hearing, ~20,000Hz). Thus the CD sample rate of 44,100Hz, a little more than double 20,000Hz. Once you've got the samples, you filter out the higher frequency half (the higher frequencies are stopped, the lower frequencies pass, so it's called a "low pass" filter), and you're left with the lower frequency stuff - music! Easy peasy, right? So why did folks think the filters might be at fault?

 

The Sampling Theorem requires a perfect filter, and no such animal exists outside of a mathematical proof. The perfect filter would allow all the frequencies we wanted to hear to pass, and perfectly block anything above that like a brick wall. For that reason and possibly due to the image of the response such a filter would show on a scope (a perfectly straight dropoff to silence, looking like a wall), these are known as "brick wall" filters. That's what the designers of the early CD players tried to get as close as possible to, the perfect "brick wall" filter. But because these filters were working in the real world and weren't perfect, they had a couple of problems.

 

One problem is called "aliasing." Any high frequency that's supposed to be blocked but gets through because the filter isn't perfect is mirrored or "aliased" around the filter cutoff frequency to create distortion in the audible frequency range. So let's say you've got an imperfect brickwall filter with a cutoff at 22,050Hz (half the 44,100Hz sample rate) and some noise gets through at 30,050Hz. No problem, you can't hear 30,050Hz. But that noise mirrors or flips around the 22,050Hz cutoff point to create an "alias" at 14,050Hz (30,050 is 8000Hz higher than the cutoff, so the alias is 8000Hz lower than the cutoff), which you very well might be able to hear. And of course there will be leaks through the imperfect brickwall filter at other frequencies, and these frequencies interact with each other to create yet more distortion/noise at audible frequencies.

 

The second problem is something called "ringing," which I won't go into too much except to say two things: (1) many people think ringing "smears" transients in the audible range; and (2) ringing is caused by a sharp filter transition - the sharper the cutoff the worse the ringing, so a brickwall filter is the poster child for causing ringing.

 

The CD and earliest DAC designers figured out a way around this, called "oversampling." (This was happening right around the time the earliest separate DACs were being created.) If you could first raise the sample rate way above 44,100Hz, and use a filter with a gradual cut starting at a higher frequency, that would accomplish two things: (1) by the time you got down around audible frequencies you wouldn't get much leakage, so it would help stop aliasing; and (2) the filter wouldn't need a sharp cutoff, so it wouldn't ring. The industry pretty rapidly settled on "8x oversampling" as a standard, meaning 44.1kHz input rates (CD) would be raised to 352.8kHz in the DAC, and 48kHz rates (DVD) would be raised to 384kHz in the DAC.

 

This oversampling is more properly called "interpolation." The thing is, when you do interpolation, you need filters for that, too. So it was kind of good news, bad news: no more bad sounding brick wall filters, but now the quality of the sound you got from the DAC's output depended on how good the interpolation filters were.

 

But wait a second, I thought the input to my DAC got turned into DSD along the way.

 

Yep, that's a little further along in DAC history. The 8x oversampling DACs mentioned above all used something called "Pulse Code Modulation," or PCM. Engineers found out you could do digital to analog conversion a lot cheaper using something called "Sigma Delta Modulation," or SDM. (DSD is a type of sigma-delta modulated signal.) Pretty soon the chips built into just about all DACs were *first* using interpolation filters for 8x oversampling to 352.8/384kHz PCM, *then* running the result (still in the chip) through a sigma-delta modulator to get a DSD-type signal, before finally converting that to the analog (music) output of the DAC.

 

This is still the standard way just about all DACs work internally today - interpolation, sigma-delta modulation, conversion to analog.

 

But hey, what about all this "upsampling" foofaraw W4S was telling me they just won't do, like it was being cruel to animals or something?

 

This involves another DAC problem you have likely read references to, called "jitter." Again I'm not going to go into any depth; just to note that jitter can cause distortion and DAC designers take steps to get rid of it. One thing DAC designers found out they could do to help minimize jitter is called "asynchronous sample rate conversion" - ASRC. This involves interpolation, which as you remember virtually all DACs these days do as standard processing anyway. But instead of interpolating to an integer (8x) multiple of the input rate, it's thought to be better for jitter minimization to interpolate to a non-integer multiple - for example, interpolating 44.1kHz input to 384kHz. This became known as "upsampling." Same exact type of mathematical operation - interpolation - just to a non-integer rather than an integer multiple. So why do W4S and a lot of other DAC manufacturers make the sign of the cross when you say "upsampling" instead of "oversampling"?

 

A few years after the advent of ASRC, a different jitter minimization technique with a similar sounding name was developed, called asynchronous USB input. Many people thought/think DACs with asynchronous USB input sound better than DACs using ASRC, and in any case async USB input was then (and is even to a fair extent these days) considered the "latest and greatest" DAC technology. So when the manufacturer of a DAC with async USB input tells you "Heavens no, we don't upsample!", they are most definitely *not* saying they don't interpolate in the DAC to high PCM sample rates and sigma-delta modulate the result before converting to analog. What they're saying is "We don't use none o' that stinkin' old-fashioned ASRC technology, nosirree Bob, nuthin' but the very latest async USB here!"

 

All right, long journey, but here it is in a nutshell: Yes, anything you feed to your DAC below 352.8/384kHz rates does get internally raised to 352.8/384kHz rates, then modulated into a DSD-type format before being converted to analog and sent to the DAC output. W4S wasn't denying that, they were telling you your DAC has what they consider the latest and greatest technology for jitter reduction and good sound.

[YashN]

foofaraw

 

Haha, never seen that before

 

Great explanation, Jud.

[Jud]

Haha, never seen that before

 

 

foofaraw: definition of foofaraw in Oxford dictionary (American English) (US)