[camrector]

Thats incorrect. You need to be a bit careful making a habit of coming out with definitive statements that are just wrong and can mislead beginners.

For anyone wanting to look into this check out the data. Linearity plots and 16bit to 24bit FFTs show the resolving power of a R2R.

Dan Lavry many decades ago worked out how to compensate for resistor drift so R2R are not limited by resistor specs.

I think bringing up a DACs ability to resolve LSB is going to confuse beginners more.
And thank you for thinking that my thoughts are definitive statements.
I can assure you that they are not.
I think the key word is output.
Is a R2R dac outputting 16 bits of actual information or is it padding MSB/LSB with 0s and really outputting 14bits?You wouldn’t hear the difference if was.
ENoB has the R26 output at 15.65bits. The Pontus 2 at 14.6bits and the Terminator plus at 15.1bits

No commercially available DAC (R2R or chip) is outputting beyond 22 bits(I’m doubtful it’s that high too) Mostly we have the thermal effects of Johnson–Nyquist noise to thank for that. Companies saying a DAC is 24bit or 32bit is true in the digital domain but not in the analog output.

 

[cobrabucket]

Quick update in regards to my experiences with the LHY OCK-1.
It's a shame the lack of transparency and the seeming "silicon lottery" aspect to this product.
That being said, I can't properly articulate exactly what I am hearing but I absolutely prefer the sound of the R26 with this external clock over the internal clock.
After the recent discussions about measurements etc. on the OCK-1, I decided to give the Sine wave output a try over the Square Wave I was previously using.
I also added one of the recommended low-pass filters. It makes for an enjoyable listen.
As always, this is in no way scientific as I didn't A/B against the Square wave, but FWIW I think I prefer the Sine Wave + LPF.

 

[Deleeh]

I think bringing up a DACs ability to resolve LSB is going to confuse beginners more.
And thank you for thinking that my thoughts are definitive statements.
I can assure you that they are not.
I think the key word is output.
Is a R2R dac outputting 16 bits of actual information or is it padding MSB/LSB with 0s and really outputting 14bits?You wouldn’t hear the difference if was.
ENoB has the R26 output at 15.3bits. The Pontus 2 at 14.6bits and the Terminator plus at 15.1bits

No commercially available DAC (R2R or chip) is outputting beyond 22 bits(I’m doubtful it’s that high too) Mostly we have the thermal effects of Johnson–Nyquist noise to thank for that. Companies saying a DAC is 24bit or 32bit is true in the digital domain but not in the analog output.

Hello,
How do I have to imagine this when I set 32/96 on the PC as well as on my Su2,plus Roon?
Does it digitally upscale it, but at the R26 only 16/96 actually arrive and is it also output at the output?

With the Ank Kit Dacs I know that their Dacs can play 24 bit, in fact the 24 bit supposedly work real with 22 bit with Cd but on the stream it is supposedly not as good as with 16 bit.
So 22 bit is supposed to be worse than 16 bit in the stream.

 

[camrector]

Hello,
How do I have to imagine this when I set 32/96 on the PC as well as on my Su2,plus Roon?
Does it digitally upscale it, but at the R26 only 16/96 actually arrive and is it also output at the output?

With the Ank Kit Dacs I know that their Dacs can play 24 bit, in fact the 24 bit supposedly work real with 22 bit with Cd but on the stream it is supposedly not as good as with 16 bit.
So 22 bit is supposed to be worse than 16 bit in the stream.

This is just a digital process. You can input 16, 24 or 32bits to the R26 no problem. It just isnt going to output that bit depth in an analog signal to be amplified.
If you send a 32bit/96khz track to the R26 it’s going to internally oversample it (unless you have NOS on) to a higher rate most likely 768kHz (8xrate), apply digital filters, then downsample back to the original rate (truncate and noise shape the extra bits) then reconstruct and low phase filter to output.

 

[sajunky]

If you send a 32bit/96khz track to the R26 it’s going to internally oversample it (unless you have NOS on) to a higher rate most likely 768kHz (8xrate), apply digital filters, then downsample back to the original rate (truncate and noise shape the extra bits) then reconstruct and low phase filter to output.

Not sure a DAC will use additional processing for truncating. Most likely it will truncate at the beginning to the 24-bits, process data at integer 24-bits (not a floating point), so there is no truncation, no dithering afterwards, letting a noise on the LSB's do dithering work naturaly.

It is why common ladders have more bits than required for matching accuracy of the ladder on one side and upsampling on the PC to the maximum supported frequency improve SQ in most cases on the other. It is bypassing the process of inferior upsampling in a DAC. NOS mode also allows to select the optimum sampling rate of the ladder, as settling time is limited and transition errors bring more noise at a higher frequency.

 

[Forgisound]

New R26 DAC body firmware is ready for trial

https://www.mediafire.com/file/qwq9ih7azboh55y/Gustard_R26_CT_m2.1_via_type-C.zip/file

 

[Alenotta]

New R26 DAC body firmware is ready for trial

https://www.mediafire.com/file/qwq9ih7azboh55y/Gustard_R26_CT_m2.1_via_type-C.zip/file

Anything good like Tidal Connect?

 

[Nada]

ENoB has the R26 output at 15.3bits. The Pontus 2 at 14.6bits and the Terminator plus at 15.1bits

Making repeated definitive statements like this based on erroneous reasoning doesn't do your reputation any good. Just saying.

 

[Forgisound]

Anything good like Tidal Connect?

I don't know yet. But that isn't streamer software.

 

[domiji]

New R26 DAC body firmware is ready for trial

https://www.mediafire.com/file/qwq9ih7azboh55y/Gustard_R26_CT_m2.1_via_type-C.zip/file

Thank you very much!

The new firmware has a really nice new menu - great work Gustard!

 

[Odezra]

Thank you very much!

The new firmware has a really nice new menu - great work Gustard!

Any other nice new features @domiji ? Is the update process the same as the streamer software?

 

[Odezra]

Making repeated definitive statements like this based on erroneous reasoning doesn't do your reputation any good. Just saying.

Rather than berating people for getting things wrong (with no data / evidence) - can you try and educate us on how you view those enobs as being erroneous? That way we all learn. It’s a genuine question as I am curious. I am a relative beginner and there isn’t a definitive wiki/source of truth for audio on stuff like this - I plugged the sinad numbers I had on hand a few months ago for the r26 (and some of those dacs Cam mentioned) and got back similar numbers. I understand from your previous posts that the FFT should enable a higher upsampling rate but the precise way to get there I am not sure on. Jussi himself from HQ player recommended 16 bits for the r26 for pcm upsampling (based on xp, not real world testing as he does not have the dac yet). Cams posts seem to suggest that regardless of the dac processing - the output stage can behave differently - how do we work that out / prove or disprove it?

With respect - the need to criticise and snipe really takes away from what has been one of the best online communities for an audio product I have seen. Would be great to keep the collaboration, positivity and education going.

 

[domiji]

Any other nice new features @domiji ? Is the update process the same as the streamer software?

Yes, now you can turn-off the display completely as well as the power button. I love this new feature!
Beside of that there are no new features.

The update process is a bit different but it is well documented.

Step 1
Turn off the DAC at the back (but leave the power cord connected )

Step 3
Start the firmware update program

Step 2
Connect the USB-C port to you pc

Step 4
Hit the "click here" button on the left corner

Step 5
Turn the DAC on at the back

Step 6
Wait a few seconds until the update progress is completed

 

[Currawong]

It seems to have fixed my one gripe, which was the menu being too small to read.

 

[camrector]

Making repeated definitive statements like this based on erroneous reasoning doesn't do your reputation any good. Just saying.

Lol I didn’t make the math
ENOB =SINAD - 1.76/6.02

https://www.everythingrf.com/rf-calculators/sinad-to-enob-calculator