Any experiences with RME ADI-2 Pro DAC?

[asdf1000]

On 8/29/2017 at 7:54 AM, scan80269 said:

 

For best sound quality, I'd recommend leaving the ADI-2 Pro on continuously, just like with most DACs.  It does run a bit warm, and takes quite a while to stabilize in temperature after being powered on from cold.

 

 

Hi Scan, would really love your impressions if you ever get time, after burn in.

 

Cheers

[asdf1000]

On 1/12/2017 at 8:33 PM, Miska said:

If you use PCM input, just use the 705.6/768k input rate (and TPDF or Gauss1 dither) and your favorite filter.

 

Hi Jussi, the HQP manual states:

 

"Gauss1 - Gaussian Probability Density Function. High quality flat frequency dither recommended for rates at or below 96 kHz where noise-shaping is not suitable."

 

So is Gauss1 actually ok for 705/768 HQP PCM up-sampling?

 

[asdf1000]

On 1/11/2017 at 5:30 AM, Miska said:

I'm using mine configured to Direct DSD mode with 150 kHz analog filter (disables primary headphone output and volume control is disabled), sending DSD256 there.

 

Hi Jussi,

 

I read the infamous Andreas Koch article which says with DSD64 the noise shaping curve starts at ~20 KHz, DSD128 ~40kHz and DSD256 ~80kHz.

 

https://positive-feedback.com/audio-discourse/raising-the-sample-rate-of-dsd-is-there-a-sweet-spot/

 

So with the 150kHz analogue filter option of the RME DAC, does this mean there is a range of ~ 80kHz to 150kHz that is not analogue filtered at all?

 

I see there is a 50 kHz option but I'm guessing you tried it and don't like the sound of it - some higher res music content being filtered/affected?

 

I'm guessing ideally, you'd like the RME to have a 70kHz or 80kHz analogue filter option?

 

Cheers!

[asdf1000]

5 minutes ago, Miska said:

 

You can make that practically 25, 50 and 100 kHz.

 

Since the noise doesn't start suddenly, but has increasing slope, DSD converters usually don't begin to cut early. SACD (DSD64) by the spec starts cutting at 50 kHz. Also most "PCM" ADC chips operate like DSD128, so if you look at noise floor of 192k content, commonly you can notice increasing noise slope above 50 kHz.

 

Also note that the 50/150 kHz filter setting applies only to the AKM DAC chip's D/A conversion stage (switched capacitor filter), not to the final analog reconstruction filter that follows the DAC chip! Usually this final filter has corner frequency around 100 kHz.

 

 

All noted with thanks!

 

So for you it's really the final 100kHz analogue filter that you care about with your RME DAC and that's why you select the 150kHZ AKM DAC chip's filter with HQP DSD256 up-sampling?

 

In other words, you choose the 150khz filter so that the DAC chips switched capacitor filter does the least 'stuff' with the DSD256 passing through?

 

 

[asdf1000]

1 minute ago, Miska said:

 

Well, either selection is fine, most of the noise is above the corner point anyway. If one is more worried about remaining ultrasonic noise levels can select the 50k setting. With either setting, level of ultrasonic noise that comes out is lower than with PCM inputs at or below 384k. With 705.6/768k PCM inputs the level is roughly same. I would say that it is unlikely to hear much difference between the 50/150k settings and you can safely keep it at 50k too. If I listen very closely and switch between the settings, I believe I hear a little bit more air at 150k setting. But the difference is very small.

 

 

Thanks Jussi!

 

Does the "DSD Direct" function of the RME DAC enable the path circled in red below? 

 

Or does "DSD Direct" only disable volume control but it still goes through the "DATT Soft Module" and "DS Modulator" path?

 

[asdf1000]

Just now, Miska said:

Output of the DAC looks quite different with the DSD Direct on/off.

 

Output of the DAC is even different with volume at maximum with DSD Direct is "off"?

 

If so, is that because even with no volume control used (volume at max), the signal goes through a more complex path?

 

[asdf1000]

14 minutes ago, Miska said:

 

Yes, because with DSD Direct off, it goes through digital filter and remodulation scheme.

 

With DSD Direct enabled, you also get constant -3.5 dB volume drop. Partially reason being DSD spec and partially the chip's analog section implementation. A bit similar to different volume drops with different analog filters you get on BB/TI chips (on iFi DACs and TEAC UD-501 for example). To match levels, I keep the PCM volume setting at -3.5 dB too.

 

Note that you cannot use ADI-2's headphone outputs in DSD Direct mode due the lack of volume control. On ADI-2 Pro, the other DAC chip for the other headphone output cannot be switched to DSD Direct mode, so it is always available, but never in DSD Direct mode. Only the chip that is coupled to analog outputs and one of the headphone outputs can be switched to DSD Direct mode (which then mutes that headphone out).

 

 

Nice thanks.

 

How is the 8kHz packet noise with this DAC's USB input in your measurements? Is there no difference whether you connect to noisy PC (non DAC-UP port) and your microRendu for example?

 

 

[asdf1000]

Somebody here mentions the balanced headphones outputs measures better than the rear XLR outputs? Wouldn't work with DSD Direct though obviously.

 

https://www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-rme-adi-2-pro-comparison-to-adi-2-dac.2682/post-116299

 

[asdf1000]

4 hours ago, Miska said:

 

That is probably because with balanced headphone output mode you get one DAC chip per channel. While normally one of the chips drives analog outs and one headphone socket and another drives another headphone socket.

 

 

Ah that’s interesting. Is DSD Direct possible via this balanced headphones output , with XLR adapters, WITHOUT volume control? To connect to a preamp/headamp/integrated/powered speakers?

 

Or do they still block DSD Direct on headphones, regardless, To protect from accidental max. volume on actual balanced headphones?

 

 

[asdf1000]

3 hours ago, Miska said:

 

Headphone out with DSD Direct is muted to avoid accidents. And I can totally understand that. Software volume control would be an option, but I'm still hesitating to ask for such config option in the DAC.

 

Anyway, the performance is already very good with single chip. Better than what TEAC does with dual-chip (on 503). So I wouldn't worry too much. I still want to get the TEAC 505 though.

 

 

Agreed, too much risk of hearing damage on headphones.

 

What features of the Teac make you want to get it? Better in-built headphones section?

 

[asdf1000]

13 minutes ago, Montanari said:

Does this mean That if i Connect the headphone to the Jack  of the exasound 22 for example, upsampling to dsd256 with hqplayer i risk to damage My headphone? 

 

The risk being discussed above is more risk of serious damage to ears/hearing, not just headphones.

 

Not sure about the Exasound 22 though. Hopefully someone else can help.

 

[asdf1000]

47 minutes ago, Miska said:

 

It should be fairly well built and has one step newer AKM DAC chips capable of DSD512... So far I have UD-501 and NT-503, so it would be next step in the series...

 

 

We were discussing the DSD filter cut-offs with the RME DAC (50kHz and 150kHz). Similar for the 503.

 

The Teac UD-505's DSD filters look very different. Much much higher cut-offs with DSD256 and DSD512.

 

[asdf1000]

36 minutes ago, Miska said:

 

As you can see, conversion stage filters just scale in frequency, every time sampling rate is doubled, filter corner frequency is also moved 2x higher. This is just like the the behavior of my DSC1 design conversion stage (the following analog filter stays at fixed frequency of course). So with 505 the actual filter remains the same, just sampling rate changes and thus filter corner. There's still a fixed corner analog filter following the DAC chip.

 

With T+A DAC8 DSD this is different in a way that it has two alternate fixed frequency filters you can choose from and they both apply equally to PCM and DSD sources.

 

 

Maybe wait for the UD-507? With the new AK4499EQ?

 

Apparently not using the switching capacitor filter?

[asdf1000]

Hi @Miska

 

I'm trying to workout the output voltages (different gain settings) of the ADI-2 Pro.

 

Converting from dBu into V, are these the balanced XLR output voltages?

 

1.23V, 3.46V, 6.9V and 12.3V?

 

And unbalanced RCA output voltages are just half of the above?

 

I used this site. Hopefully it's correct:

 

http://www.sengpielaudio.com/calculator-db-volt.htm

 

Cheers

 

[asdf1000]

@Miska the measurements obviously speak for themselves (very very good and no jitter issues to worry about between digital inputs to analogue output) but just out of general interest only, how is the clocking in the RME ADI-2?

 

Does it have synchronous clocking, with dual clocks (e.g. ~22MHz and ~24 MHz family rates)?

 

Or just a single clock with ASRC?

 

 

 

 

 

[asdf1000]

11 minutes ago, Miska said:

 

AKM chips don't have ASRC, only ESS has. So they need matching separate clocks for the two rate families.

 

There's a clock generator for those.

 

Nice, thanks.

[asdf1000]

3 hours ago, barrows said:

Just to be clear, the ESS chips do not have to use the ASRC and a single rate XO for the DAC.

 

Yes Jussi has mentioned some Myteks (not the newest ones) do this and the Pro-Ject S2 DAC also. 

[asdf1000]

On 1/12/2017 at 8:33 PM, Miska said:

If you use PCM input, just use the 705.6/768k input rate (and TPDF or Gauss1 dither) and your favorite filter.

 

Hi Jussi, any technical reason you don't recommend NS5 here at 705/768k output rates?

 

Is it because at these PCM output rates specifically, there is very little difference in technical performance between TPDF, Gauss1 and NS5?

 

[asdf1000]

On 1/25/2019 at 4:12 AM, Miska said:

 

They are already translated for you in the analog input section of the manual...

 

 

So your conversion seems to match the RME's.

 

I'm using +13 dBu myself for consumer gear connections.

 

 

Hi @Miska, just to confirm, DSD Direct mode can be used with all XLR and RCA output gain settings of ADI-2 DAC?

 

Since all the gain settings are all done in the analogue domain ?

[asdf1000]

Just now, Miska said:

 

Yes, that's the case AFAIK.

 

 

Thx.

 

The manual says: 

 

“In DSD Direct mode there is no PCM conversion – and consequently no volume control anymore. After having activated DSD Direct in the ADI-2 DAC’s menu (SETUP - Options), the analog signal is available only at the rear outputs, with a coarse volume control via the analog output reference level control.”

 

What does this bold mean, in practical terms? 

 

There is analogue volume control?

 

[asdf1000]

7 hours ago, Miska said:

 

Yes, but only couple of steps, so I wouldn't call it "volume control" as such. Reason for the reference level control is just to provide means to match DAC's output level to input sensitivity of the following stage. Otherwise the preamp following the DAC could clip in worst case.

 

Thanks Jussi.

 

ADI-2 DAC supports both 1bit 11.2896 MHz and 12.288 MHz rates ?

[asdf1000]

6 hours ago, Em2016 said:

 

Thanks Jussi.

 

ADI-2 DAC supports both 1bit 11.2896 MHz and 12.288 MHz rates ?

 

It's ok @Miska. Found it earlier in the thread, DoP256x48k is supported.

 

I went through this thread - did you share measurements of your ADI-2 somewhere?

 

 

 

[asdf1000]

PCM44.1k vs DSD256 input, if you have them

[asdf1000]

56 minutes ago, Miska said:

DSD filter set to 50 kHz, ASDM5 modulator:

 

Thanks! Is this what you use with your ADI-2 Pro at DSD256? 

[asdf1000]

4 minutes ago, Miska said:

 

I think I have 150 kHz filter and ASDM7EC at the moment. That lets a bit more noise through. Amount of noise in the output is roughly the same with seventh order modulators + 50 kHz filter and fifth order modulator and 150 kHz filter. It is easy to compare sound of the two filter settings by just turning the knob between the two. The difference to me is quite small. If you want absolutely lowest amount of noise in the output and good performance, use ASDM5EC and 50 kHz filter setting. When I did the measurements, I didn't have the EC modulators yet. "50 kHz" setting has extra notch around 400 kHz.

 

 

Thanks again!

 

This probably got lost in the HQP thread but you previously mentioned:

 

On 7/14/2019 at 7:55 AM, Miska said:

keep the AKM chip in DSD Direct mode. What comes out with/without DSD Direct is vastly different.

 

Can you show this difference, with DSD256 into your ADI-2, with & without DSD Direct?

 

Would be great to visually see how advantageous it can be when avoiding all DAC chip based DSP & modulator.