[MetalAudio]

I'm not surprised that the Denafrips sounds different than the Bryston in your system, it's always about synergy.

But you are comparing apples and oranges. The Stereophile review says the Bryston uses dual AKM chips whereas the Denafrips is a R2R NOS (non-oversampling) DAC.
R2R NOS DACs slowly roll off the highs starting earlier than the AKM chip.

The Stereophile Denafrips Terminator measurements are a good example:

Denafrips Terminator, NOS mode, frequency response at –12dBFS into 100k ohms with data sampled at: 44.1kHz (left channel green, right gray), 96kHz (left cyan, right magenta), and 192kHz (left blue, right red) (0.5dB/vertical div.).


The Stereophile Bryston BDA-3 measurements show an extended high frequency response with the 44.1kHz brick wall filter

Bryston BDA-3, frequency response at –12dBFS into 100k ohms with data sampled at: 44.1kHz (left channel green, right gray), 96kHz (left cyan, right magenta), 192kHz (left blue, right red), 384kHz (left green, right gray) (0.5dB/vertical div.).


The measurements by Golden Sound raise doubt that the Denafrips Terminator Plus is even a NOS DAC, his analysis concludes the Denafrips does linear interpolation oversampling, that doesn't mean it doesn't sound great, it just isn't a true NOS DAC.
https://goldensound.audio/2021/10/07/denafrips-terminator-plus-with-gaia-measurements/


Another reason for the difference in sound is how the digital signal is filtered. Different filter choices change how a DAC sounds.

JA's Bryston BDA-3 measurements state:
"The impulse response with data sampled at 44.1kHz (fig.1) revealed that the digital reconstruction filter used by LG was a minimum-phase type, with all ringing occurring after the single sample at 0dBFS."

Sound Stage measured the Denafrips Terminator Plus and found "the  NOS mode also yielded a symmetrical impulse response, but with less pre- and post-ringing effects. Of note here is that this is not the shape of NOS DAC impulse response (it should be a square), but that of a typical sinc filter."

In oversampling mode the Denafrips has a Sharp and Slow filter which have more pre- and post-ringing effects than the NOS filter.
https://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=2612:denafrips-terminator-plus-digital-to-analog-converter-measurements&catid=434&Itemid=577


I completely eliminated the "which DAC sounds best" quandary by buying a true NOS DAC and using HQPlayer to upsample PCM to DSD256. Why DSD? The simplest answer is it just sounds better. The Bryston does DSD so it is easy to prove it to yourself, don't believe me, believe your ears.

Start with the HQPlayer poly-sinc-gauss-long filter and ASDM7EC-light to upsample to DSD256. The combo has been optimized for all types of music. The learning curve can be steep, the Audiophile Style HQPlayer thread is dauntingly long but the rewards are worth it. Hint: instead of reading the entire thread, read the last 5 pages and plug in the filters everyone else is using as a starter. You may never need to change.

Other filters can closely approximate other manufacturer's DAC sound characteristics.

[Chord DAC] "Chord-style filters sinc-L group and sinc-short/medium/long the length also affects stop-band attenuation."

[In the latest HQPlayer 5.2.0] "...find new linear phase halfband filters.
"I found the xs one to be very interesting. IMO it shows some similarity to xtr-short."

Miska (HQPlayer's developer) replied:
"It is something I actually made because someone asked what would be closest to Mola-Mola filter, and there weren't any. So this one fairly closely matches the one used by Mola-Mola. And it is very similar to most DAC chip filters (ESS etc) in terms of response as well. Relatively slow roll-off and fairly low attenuation.
https://www.audiocircle.com/index.php?topic=179643.msg1963005#msg1963005


Interesting stuff but most audiophiles would rather buy a new DAC every couple of years to get a new sound. When I talk about HQPlayer and it's 165 filters, dithers, noise shapers, and Delta-Sigma (DSD) converters to the guys in our audio club I can watch their eyes roll.

Which DAC did you end up buying?

[WGH]

Which DAC did you end up buying?

I wanted to geek out with HQPlayer so needed a DAC that is basically a blank slate, the HoloAudio May KTE.

My original 10 year old music server died so I built a faster server so I can upsample PCM and DSD.
Building an Affordable High-End Silent Music Server
https://www.audiocircle.com/index.php?topic=177495.msg1867916#msg1867916

I was doing too much reading and found out a little knowledge is a dangerous thing so I wrote a primer so I wouldn't forget what I learned
An Upsampling Primer or Why Make More Bits?
https://www.audiocircle.com/index.php?topic=183465.msg1926276#msg1926276

The HoloAudio DAC and HQPlayer combo became a review
HoloAudio May KTE DAC with HQPlayer Review
https://www.audiocircle.com/index.php?topic=182724.msg1919415#msg1919415


Everything in the above articles can be applied to the Bryston BDA-3. The Bryston looks like the perfect HQPlayer companion.

"Not only did Bryston enable DSD in the BDA-3, but it realized that PCM and DSD have different requirements. The company went the extra mile by designing independent signal paths to the DAC chips for DSD and PCM. According to Bryston, "Audio is processed in their native format with no conversion ensuring each song is totally bit perfect — an exact replica of the master recording."
https://audiophilestyle.com/ca/reviews/bryston-bda-3-dac-review/



I would be careful with the filter type because the BDA-3 uses a minimum-phase filter and would pick either 384kHz or DSD256 (my preference) in HQPlayer.

    On 2/7/2022 at 1:47 PM, mfsoa said:
        "... can you please describe what happens when we take HQP data that has been filtered w/ HQP, and then send it to a DAC that has its own filtering?  If I send an HQP minimum phase filtered signal to a DAC with a linear phase filter - What happens?  A 50:50 split? Who wins? Are we really hearing what HQP can do, or are the HQP goodies being subsequently mangled by the DACs own filter?"

Miska (HQPlayers developer) replies:
 No, it is not mangled by DAC's filter, because you have plenty of "empty space" at end of upsampled frequency spectrum. You should always use linear phase filter at DAC side in such cases where you end up chaining filters like that. If you use minimum-phase filter at DAC side too, it may modify the phase response. Linear phase filter there will leave things untouched.

But in most cases when you reach 352.8k/384k input rate to a DAC, it's own digital filters are already bypassed. This is the case for AKM chips for example.

Unless the DAC is NOS R2R, it will still likely do basic sample-and-hold (S/H aka ZOH aka zero-order-hold) oversampling and run it's modulators to produce SDM type output (similar to DSD). Just like it would after it's own digital filters too.

 On 2/7/2022 at 1:47 PM, mfsoa said:
    Is this the advantage of a DSD dac that does native DSD - No additional filtering beyond HQP?
 
Miska replied:
One of those yes. You also get to skip the poor sample-and-hold or linear interpolation stage that is used due to lack of DSP power to do any better. And instead you get full proper digital filters run to final modulator rate, such as 256x, 512x or 1024x of the source rate.

... AKM chips have possibility to process DSD inputs or bypass processing and just send the data straight to the actual D/A conversion.
https://audiophilestyle.com/forums/topic/19715-hq-player/page/962/#comment-1182886


HQPlayer's filter choices are subtle, if a system isn't resolving or speakers are open baffle like GR Research or Magnepan with half the energy reflected off the front wall then the differences will be obscured. But even though the differences are subtle, they are still there and can make the difference between good and great sound.


A-B tests can reveal differences but my ears and brain gets fatigued and then confused. I prefer to listen for a week or two, then switch for another week or two, then switch again. Differences will then jump out and a choice can be made. In the big scheme of things there are no wrong choices, our local audio group taught me that.

[MetalAudio]

That's supposed to be a good DAC - people often compare it with the Terminator or the Terminator Plus. Enjoy!

[WGH]

I'm as attracted to the next shiny new thing as much as anyone. The Bryston BDA-3 fell off my radar years ago, this can be a problem for mature products that have achieved technical excellence and don't need constant upgrades. Bryston products still get great reviews in the consumer press, both in magazines and on the web. The BDA-3, not so much anymore. I don't mean it doesn't get great reviews, it just doesn't get mentioned at all. The most recent mention was as part of an all Bryston system at the 2023 Montreal Audio Fest.
https://www.stereophile.com/content/bryston




The BDA-3 was reviewed in the Stereophile November 2016 issue by Larry Greenhill, who wrote "I recommend that the BDA-3 be listed in Class A+ of Stereophile's "Recommended Components."
https://www.stereophile.com/content/bryston-bda-3-da-processor

The BDA-3 achieved an A+ listing in Stereophile Recommended Components in 2017, 2018, and 2019. Stereophile has a policy that if a product isn't reviewed in 3 years then it is removed from the Recommended Components listing. An unfortunate circumstance for products that don't need to be upgraded every 3 years. As technologies mature Stereophile may have to extend how long a product is listed.

There are always new DACs appearing so not many reviewers are going to review a 7 year old DAC, let alone say it still sounds better than the latest and greatest.

I sold my 10 year old Van Alstine Fet Valve Hybrid DAC, not because it sounded bad, but because I was curious and it didn't do hi-res. Digital technology surely must have advanced in 10 years. Right?
Not so much as I found out, the technology pretty much matured 10 years ago (and the Wolfson 8742 chip is still highly regarded). Improvements can still be squeezed out by using massive amounts of processing power

Some products try, while keeping the price within reach, the Chord Electronics Hugo M Scaler* upsampling digital processor uses the Xilinx XC7A200T FPGA but will never be able to do what HQPlayer does, there too many calculations. Jussi Laako, HQPlayer's developer, wrote:

"Sometimes companies ask what it would take to run my algorithms on FPGA, when I tell them that it would need Xilinx biggest Virtex FPGA's, they quickly forget about it. Such big FPGA's cost way more than CPU/GPU and you are likely not going to find such chips in DACs."

Dedicated music servers with stand alone programs like HQPlayer are at the bleeding edge of digital reproduction but not many audiophile are willing to add that much complexity to their hobby.



* For comparison the Chord Electronics Hugo M Scaler costs $4750 and upscales standard 44.1kHz digital audio up to 705.6kHz (16x CD’s 44.1kHz native resolution).

The Hollis Audio Labs HAL-MS6 for $450 (an AC member) will probably upsample to 705.6kHz although I haven't tried. I have found that 384kHz sounds better, there are trade-offs upsampling to 705.6/768kHz in PCM anyway. The HAL-MS6 plus HQPlayer will cost $740.

A higher powered music server that will upsample to DSD will start at $1300.