Pro-Ject Pre Box S2 Digital: MQA HW decoding at reasonable cost

[asdf1000]

8 minutes ago, GUTB said:

 

DACs typically take around 7-10 days. I’ll be checking back frequently though to observe changes.

 

I found it depends. My DirectStream DAC did sound noticeably different (better) after a couple weeks.

 

My Chord Hugo2 never changed in how it sounded from the moment it was un-boxed.
 

My iFi iDAC2 maybe needed 48 hours.

 

Looking forward to reading your impressions.

[asdf1000]

1 minute ago, left channel said:

 

@Em2016 Pro-Ject is now considering a Mac version of the firmware update. Wow!

 

 

Thanks so much for asking! Their consideration is a great start and this would be great indeed for those without easy access to a Windows PC.

 

[asdf1000]

7 minutes ago, fritzg said:

 

What other one?

 

This one :-) 

 

The two threads were already merged by Chris, into this one.

[asdf1000]

22 minutes ago, left channel said:

This DAC requires 1000mA (= 1A)

 

I'm not sure this is correct. The DAC can be fully powered by a USB 2.0 port (0.5Amp max) only, without external power.

 

24 minutes ago, left channel said:

maximum USB OTG current is 500mA, and actual output from various models may be far lower.

 

I think this might be the issue. See below link:

 

"OTG devices can provide up to 500mA, but in realistic terms, handheld portable electronics don't have 500mA to spare for external loads. 100mA is a commonly accepted realistic maximum."

 

https://www.eetimes.com/document.asp?doc_id=1226476

 

[asdf1000]

16 minutes ago, left channel said:

That must be it, thanks. 100mA is just too teeny.

 

I know so much about this DAC now but still patiently waiting for it's official release here in Australia - sigh.

 

Just for easier local warranty support, otherwise I'd have ordered from eBay long ago.

 

Apparently the S2 range of components hits shores here in January, so it's getting close.

 

[asdf1000]

9 hours ago, Miska said:

The "best" is really the crappiest setting. You get good performance with sharp roll-off filter and distortion compensation enabled.

 

So you mean the linear phase & fast roll off filter , performs best if you don’t do PC side upsampling, ie bit perfect playback?

 

with DSD512 and PCM768k upsampling arent the S2 DAC’s filters effectively bypassed?

 

 

[asdf1000]

58 minutes ago, Miska said:

 

Yes, I consider it good compromise and I have that selected for playing content from something like YouTube and such. There are other filter options too, like "minimum phase fast" or "linear apodizing".

 

 

Yes, certainly. I think 352.8k upwards already, although I didn't do much measurements at 352.8/384k rate.

 

 

Cool, thanks. I have the S2 DAC arriving this week. Looking forward to playing with it.

 

[asdf1000]

On 21/01/2018 at 6:14 AM, Miska said:

The "best" is really the crappiest setting. You get good performance with sharp roll-off filter and distortion compensation enabled.

 

What is the "optimal transients" filter doing, according to your guesstimates? It sounds like it's very different to "linear phase & fast rolloff", since you say the latter performs best.

 

Or is it too difficult to determine/guesstimate what this filter is, from your measurements?

[asdf1000]

9 minutes ago, Miska said:

 

It is very short, like the MQA filters, leaking a lot of images. It doesn't do much of anything.

 

Ah right. I wonder if this is really Pro-Ject's own filter (they say it's Pro-Ject preferred) or MQA's own filter.

 

Similar to the MQA filter mode on the Mytek Brooklyn, where even non-MQA files pass through this filter unless you disable MQA decoding?

 

[asdf1000]

2 minutes ago, Miska said:

 

No, I think those are standard set of ESS filters. You can see some of those in the Manhattan review:

https://www.stereophile.com/content/mytek-hifi-manhattan-ii-da-preamplifier-headphone-amplifier-measurements

 

 

Different one, but also leaky...

 

Ok cool. What is distortion compensation actually doing? I saw you mentioned you prefer it enabled all the time.

 

[asdf1000]

4 minutes ago, Miska said:

 

It is some ESS DAC chip feature. But based on my measurements it does precisely that - reduces distortion significantly. And I cannot see any negative effects on it either.

 

 

Nice. Thanks for this info.

 

[asdf1000]

2 hours ago, GUTB said:

2. Lack of dynamic power.

 

Did you power it via the USB audio input or use external power also?

 

[asdf1000]

3 minutes ago, Nihilnisibene said:

Their website and downloading is obnoxiously amateurish though. 

True. It seems they have responded to feedback on Facebook at least once, so perhaps leave some feedback there if you can.

 

I don't think they're around these forums much. Feedback can be valuable with improving things.

 

[asdf1000]

5 hours ago, left channel said:

@Nihilnisibene if you need the firmware and other files PM me. I have them in a download folder, no login required.

 

@Em2016 yes amateurish. I believe "support" is just one guy who is actually the PR manager and seems to have no understanding of technical issues, or of customer service really. I doubt there was any good reason to take down the 2.12 update while awaiting a fix, but even if there was he could have done a better job informing customers that have reported the bug.  They seem to think they can rely on their dealers to provide support. But the dealers I've contacted in the US and Germany were never informed of any bugs or updates, and are mostly just online resellers moving boxes.

 

Their testing is also amateurish. I'm beginning to think their manufacturing partner doesn't even have a testing lab. Also the manuals are copy-paste hack jobs. I could go on... 

 

 

 

Eep! While I’m sure we’re all able to separate John Westlake from this mess, I’m almost sure it must annoy him to be reading these comments, as the contracted designer 

[asdf1000]

22 minutes ago, Miska said:

Any modern thing will have bit of firmware/software that will need maintenance and they need to be prepared for that.

 

Maybe not forever but initially anyway. Taking the iDSD micro range as a recent example. It’s pretty rock solid now and hasn’t needed any major bug fixes since 2016. But there was an initial 20 months of firmware updates.

 

https://ifi-audio.com/micro-idsd-ifi-xmos-firmware/

 

Maybe some of these issues were fixed with the newer XMOS hardware that Pro-Ject is using but they should have budgeted for at least 12 months firmware support. Especially with a new unknown feature , MQA.

 

[asdf1000]

Hi @left channel and anyone else using this DAC

 

Am I correct in my understanding that the MQA bug ONLY affects the optimal transients filter used?

 

So if one uses the linear phase + fast rolloff filter (as recommended by @Miska ) then this MQA bug doesn't appear? And the unit is relatively bug free overall?

 

How it sounds is something else completely but that's always subjective of course. Just wanted to ask about bugs and reliability.

 

Cheers

 

[asdf1000]

2 hours ago, left channel said:

I had been splitting the difference between John and @Miska by using Optimal Transient with Distortion Compensation on, but I'll stick with Fast Rolloff now.) I haven't tested the other filters. There are a few other minor bugs/behaviors like the mushy volume control, but otherwise you're good-to-go. 

 

Nice! I wonder after some days of listening to linear phase + fast rolloff (and not experiencing any show stopping bugs) and getting used to this filter, if you end up liking the unit again? Do share your thoughts.

 

[asdf1000]

On 21/01/2018 at 10:27 PM, Miska said:

 

Yes, I consider it good compromise and I have that selected for playing content from something like YouTube and such. There are other filter options too, like "minimum phase fast" or "linear apodizing".

 

 

Yes, certainly. I think 352.8k upwards already, although I didn't do much measurements at 352.8/384k rate.

 

 

Do you prefer linear phase + fast rolloff with BOTH DSD512 and PCM768k up-sampling?

 

I know Rob Watts likes linear phase and fast rolloff for his FPGA PCM up-sampling.

 

But for DSD up-sampling I read Ted Smith likes linear phase and slow rolloff:

 

“What’s a proper reconstruction filter? An infinitely long filter with linear phase. Non linear phase will mess with the phase, well, non-linearly. Since filters can’t be infinitely long, what are the best approximations to one? Opinions differ, but I agree with the folks at Schiit Audio and Chord that the filter should be as long as possible and shouldn’t mess with the phase – we all use as long of a filter as is practical (well, we all use slightly impractical filters )”

 

and

 

"There are multiple things going on here:

I knew that DSD could sound correct and great and that it “only” needed a simple passive low pass filter for the analog output – You can’t have a simple passive low pass filter as the only analog output hardware for a PCM DAC implementation.

To simplify the passive output filter (which should lead to better sound) the filter is optimized for one frequency. I chose double rate DSD since single rate DSD needs analog output filters steeper than I’d like.

To get single rate DSD to double rate DSD there’s no choice but to implement an upsampling filter.

It turns out that I could get PCM upsampled to that same output rate by using the same math, just stuffing more zeros.

In my opinion the feature of DSD that make the biggest difference in resultant sound is slow rolloff filtering (everywhere). The slower the better. Slow rolloff filters allow you to keep the timing information more correct: you can more easily have the output waveform look like the input waveform – transients aren’t distorted. Most PCM DACs have filters that mess with transients: they often have preringing for example. You can’t have (only) slow rolloff filters in a PCM DAC, you need steep filters somewhere, either digital or analog (or more likely both).

To get slow rolloff filters you need a high sample rate. So I consider the most important features of a DSD DAC to be a high sample rate and using only slowly rolling off filters. Too many people think that DSD is defined by a high sample rate and a single bit per sample. The number of bits in a sample isn’t important: one bit is sufficient for the final output but not necessary, more don’t hurt.

Since one bit output is sufficient for DSD and a one bit DAC is much easier to implement well with simple analog hardware I chose a one bit output from the FPGA – but the real “goodness” comes from a high sample rate and simple slow rolloff filters."

 

Sources:

 

http://www.psaudio.com/forum/directstream-all-about-it/questions-for-ted-about-upsampling-and-fpga/

 

and

 

http://www.psaudio.com/forum/directstream-all-about-it/mconnect-control-app/page-4/

 

[asdf1000]

On 11/02/2018 at 8:30 AM, Miska said:

I have to admit I didn't encounter that problem ever until I specifically tested it out. Mostly because of my use cases:

 

1) Play YouTube from browser -> 44.1/48k PCM

2) Play music from HQPlayer -> DSD512 (or sometimes to 768/32 for testing)

 

So I never had quick switch from PCM to DSD64 until I tried it out.

 

 

 

How do you find the digital volume control work with this DAC's 2 x ESS chips used?

 

I read Archimago wrote this, regarding the Oppo Sonica DAC:

 

"These days, with 32-bit DACs, a -30dB reduction to hit a comfortable listening level would only amount to 5-bits worth of attenuation any way. With a high resolution DAC like this, I'm happy to stick with digital volume control."

 

http://archimago.blogspot.com.au/2017/06/measurements-oppo-sonica-dac-ess-sabre.html

 

I know it's advertised as being a 32 bit input chip, so I assume it's 32 bit digital volume control?

 

[asdf1000]

5 hours ago, Miska said:

 

Yes, it has 32-bit pipeline. It is not a problem, adverse effects are well below the analog noise floor. 32-bit has dynamic range of 192 dB while the entire DAC in question with it's analog stages has 120 dB dynamic range.  So you have 72 dB (12-bit worth) of digital headroom there.

 

Alternatively you can use software digital volume control in the playback software instead.

 

 

Nice, thanks. Software digital volume control (even if 64 bit like Roon) will be no different to using the S2’s digital volume control will it? As the limiting factor is still the S2’s analogue noise floor? 

 

Or can 64 bit software volume control still be better (technically) with this DAC? 

[asdf1000]

3 hours ago, left channel said:

Are all digital volume controls weird, or just the one on this product, or even just my unit? I've mentioned this before but nobody else has, so I'm wondering. Sometimes when I turn the volume knob, nothing happens for a few notches (the little clicks or detents you feel when you turn it). No change in sound or display for at least three of those. After that, now that it is "awake", the same few clicks will immediately change the volume. But turning back the other way may require extra clicks. And after not being turned for a while, it's back to requiring at least three extra clicks in either direction. It's really strange.

 

I noticed the same, even when adjusting the knob to enable distortion compensation. Maybe it’s a “feature” to prevent accidental change in volume? Worth a question to Pro-Ject though in case it’s a bug. I can’t imagine they haven’t noticed it.

[asdf1000]

8 minutes ago, Miska said:

I've got cleanest output from the DAC with it's volume control set to 0 dB and DSD512 from computer.

 

Nice, I'll play with DSD512 + set the DAC to 0db volume + use software control. For software volume control, maybe turn on auto volume levelling in the software, which means less adjusting when shuffling music.

[asdf1000]

On 13/02/2018 at 1:12 AM, Miska said:

 

Yes, it has 32-bit pipeline. It is not a problem, adverse effects are well below the analog noise floor. 32-bit has dynamic range of 192 dB while the entire DAC in question with it's analog stages has 120 dB dynamic range.  So you have 72 dB (12-bit worth) of digital headroom there.

 

Alternatively you can use software digital volume control in the playback software instead.

 

 

What are your thoughts on incoming jitter rejection with the latest ESS chips, like used in this S2? Like for the TOSlink input for this S2 DAC.

 

ESS marketing have long said incoming jitter is eliminated, even with a high jitter source like TOSLink. I guess a good listening test for this is the DAC's TOSLink input vs a mobile or isolated USB source. 

 

[asdf1000]

50 minutes ago, Miska said:

I'm not sure if the ESS chip receiver is used in this case due to the dual-chip design. I suspect the S/PDIF input is implemented in the XMOS controller.

 

Ah I see. My gut instinct would prefer the ESS chip receiver over XMOS controller, to do the incoming jitter reduction. I’ll see if Pro-Ject can help clarify.

 

[asdf1000]

@Miska some technical questions about the 2 x ESS chips used in this S2 DAC, if you have the info:

 

1. when doing DSD up-sampling to the S2 DAC, are the 2 x ESS chips doing any kind of digital filtering? Or only analogue filtering?

 

2. For analogue filtering, where does the low pass filtering start? Is it the same for DSD128, DSD256 and DSD512 or different for each?

 

3. I read from Ted Smith and others that going higher in DSD rates can result in an increase in clock phase noise. Do you see see with your measurements of DSD128 vs DSD256 vs DSD512?

 

Cheers