1 (edited by ning 2019-06-09 11:36:51)

Topic: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

Hi, I was reading the manual of
- ADI-2 Pro FS
- ADI-2 Pro

In manual chapter 34.13. The new FS version seems to have worse THD+N and SNR numbers because of higher harmonics in AD conversion. why?

In addition, the FS version manual seems to reuse the old version's DA measurement graphics. Is this a mistake?  The ADI-2 DAC manual also borrows the same picture. With the latest SteadyClock FS I assume the new device should measures better than the old one?

Also, the FS AD spec numbers in chapter 33 are indeed worse than the old one. However on the ADI-2 Pro FS page from official website, the spec is using the old (better) numbers. Is this a mistake as well?

The DA specs seems to be exactly the same between ADI-2 Pro and ADI-2 Pro FS . Is this expected?

Would you please update the doc and website to reflect the actual spec? I know a few dbs difference in SNR and THD+N in low level is probably fine, but it's better to correct the documentation.

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

Also interesting that much cheaper ADI-2 DAC measures better than more advanced ADI-2 PRO:

https://www.audiosciencereview.com/foru … -dac.2682/

Sonicaly I like more the PRO version on my system.

3 (edited by ramses 2019-06-09 16:38:59)

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

Pitrs wrote:

Also interesting that much cheaper ADI-2 DAC measures better than more advanced ADI-2 PRO:

https://www.audiosciencereview.com/foru … -dac.2682/

Sonicaly I like more the PRO version on my system.

Sorry - you mix chicken with eggs.

Are you aware of, in what order the devices have been brought to the market ?
That ADI-2 Pro and ADI-2 Pro FS are two different products ?
That the test you refer to is between ADI-2 DAC and the older product ADI-2 Pro (not the current ADI-2 Pro FS) ?

Seems not wink

1. 04/2016 ADI-2 Pro has been released 1st

2. 12/2017 - over 1 1/2 y later - ADI-2 DAC arrived with FS clock

3. 09/2018 - 10 months later - ADI-2 Pro got FS clock as well, simply to be on par with measuring results of the ADI-2 DAC. This little circuit upgrade didn't bring audible changes, simply better measuring results.

I hope it's clear now.

So if you think that the slightly better measuring reasults bring your "bat ears" into sonic heaven, then sell your Pro and get the Pro FS. And if you like black cases more, then you can get even the ADI-2 Pro FS BE (case without mounting holes).

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

4 (edited by Pitrs 2019-06-09 16:57:31)

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

Sorry - I don't ;-) .

If I write about ADI-2 PRO I mean ADI-2 PRO which might be quite obvious, I (maybe wrongly) suppose :-) .

Hope it is clear now.

Regarding sonic diferrences, I was not writing about any thoughts or heaven, so any weird buying recomendations are neither needed nor requested.

Hope it is clear now :-D .

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

You wrote:

> Also interesting that much cheaper ADI-2 DAC measures better than more advanced ADI-2 PRO:

Again, the ADI-2 DAC came 1 1/2 years later with FS clock. There is nothing "interesting" or spectacular.
And the newer Model of the Pro, the ADI-2 Pro FS now also got FS clock.

We have now 2019 and if you compare devices, then please the actual ones ADI-2 DAC / ADI-2 Pro FS.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

> So if you think that the slightly better measuring reasults bring your "bat ears" into sonic heaven, then sell your Pro and get the Pro FS. And if you like black cases more, then you can get even the ADI-2 Pro FS BE (case without mounting holes).


Sorry, please do not start a war. What I said is the manual and official website should document the right numbers and graphs.
In the first post I already said the THD+N and SNR are so good that a few db difference doesn't matter.
The manual is a good read and I learned a lot from it. It would be even better if its technical accuracy is improved.
The FS manual just copies-and-pastes the DAC FFT graph from the old manual, and doesn't seems right.

> 3. 09/2018 - 10 months later - ADI-2 Pro got FS clock as well, simply to be on par with measuring results of the ADI-2 DAC. This little circuit upgrade didn't bring audible changes, simply better measuring results.

The strange thing is, contrary to what you believed,   the spec and graphs in the manuals (ADI-2 Pro and ADI-2 Pro FS) indicate the FS one has worse, not better, measuring results.  That's why I'm asking the question. I am interested to know what happened.

7 (edited by Pitrs 2019-06-09 17:25:56)

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

Ramses it is obvious that you made wrong assumptions and then carefully explained ME why YOUR assumptions are not correct :-D . Interesting. Anyways thanks for your meaningful contribution.

This topic is about weird situation when newer PRO FS measures worse than the older PRO version.
In the same manner it is surprising that much more advanced PRO targeted to more demanding users measures worse than much cheaper DAC.

I understand that DAC is newer and some tweaks were made, but for a long time these two models were sold together, both as current versions.

No need to create another unnecesarry assumptions and constructions as well as pointless nonrequested advices.

8 (edited by ramses 2019-06-09 17:48:03)

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

Now we all talk about different things.

One guy talks about a certain unclarity of graphs in the manuals.
Other guy talks about an old review in a blog.

Maybe you should 1st wait for RMEs answer in terms of graphs in the manual.

And my statement about chicken and eggs is correct the way I see it ... it was between two different devices, one without and one with FS technology, that came 1.5y later. Its simply kind of unfair to compare these two devices and start to complain about performance. You need to see also the history and evolution of the devices.

This is not a war .. simply some facts and a little patience to give RME maybe a little time ... its weekend isn't it ?!
And for those of you who are not from Germany .. its Pfingsten, long weekend with a public holiday on monday,
where most people spend all the time with the families.

RME worked already hard on the firmware upgrade I would guess, so be fair and give them a little time over Pfingsten.

Cheers.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

Regarding Pfingsten: nobody did complain about RME not answering so asking for patience is completely pointless with zero context in this thread.

10 (edited by jiw 2019-06-09 23:08:13)

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

ning wrote:

Hi, I was reading the manual of
- ADI-2 Pro FS
- ADI-2 Pro

In manual chapter 34.13. The new FS version seems to have worse THD+N and SNR numbers because of higher harmonics in AD conversion. why?

In addition, the FS version manual seems to reuse the old version's DA measurement graphics. Is this a mistake?  The ADI-2 DAC manual also borrows the same picture. With the latest SteadyClock FS I assume the new device should measures better than the old one?

Also, the FS AD spec numbers in chapter 33 are indeed worse than the old one. However on the ADI-2 Pro FS page from official website, the spec is using the old (better) numbers. Is this a mistake as well?

The DA specs seems to be exactly the same between ADI-2 Pro and ADI-2 Pro FS . Is this expected?

Would you please update the doc and website to reflect the actual spec? I know a few dbs difference in SNR and THD+N in low level is probably fine, but it's better to correct the documentation.

The levels of the harmonics in the Pro FS are different than in the Pro. The differences are that the 3rd harmonic in the Pro FS is about 5 dB higher than in the Pro while the 4th and 5th are 2dB lower, the 6th and 8th are 5 dB lower while the 2nd, 7th and 9th are at the same level and the 10th is below the displayed noise floor while it is 8 dB above it in the Pro. Further, the 13th and 14th are 2 dB higher while the the 16th is 1 dB higher and any higher harmonics are below the displayed noise floor while for the Pro the 17th and 21st are 5 dB above the displayed noise floor. Of all this, the most important is the differences for the 3rd and the 5th harmonic.

One psychoacoustic effect is that the closer one tone is to another tone that is higher in level, the higher is the level at which the first tone becomes audible. Conversely, the further one tone is from another tone that is higher in level, the lower is the level at which the first tone becomes audible. This is called auditory masking (https://en.wikipedia.org/wiki/Auditory_masking).

Thus, the Pro FS's THD is higher but its spectrum of the HD may be superior in terms of (lower) perceived distortion. So merely looking at the THD numbers does not give an accurate indication of which AD is worse.

Regarding the SNR (and in consequence THD+N), RME specs the SNR for the AD in the Pro FS the same as in the Pro (manual 33 p. 64).

Judging by the 'skirt' around the 1 kHz tone and the Pro FS manual 34.6 (p. 75), the DA figure in 34.13 seems to be for the DAC/Pro FS.


Also, 'Pfingsten' is the german name for Pentecost.

11 (edited by ning 2019-06-10 07:21:14)

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

>  RME specs the SNR for the AD in the Pro FS the same as in the Pro (manual 33 p. 64).

You are right. but the FFT spectrum said otherwise. on page 80 FS has the SNR = 114.18, the non-FS has SNR=116.988.

> Judging by the 'skirt' around the 1 kHz tone and the Pro FS manual 34.6 (p. 75), the DA figure in 34.13 seems to be for the DAC/Pro FS.

You made a good point.  However I have another theory. If you take a look at FS manual 34.13 (p80), The DA measurement were done on Nov 2016, which is the same date they did the non-FS Pro's AD FFT. At that time there was no DAC nor Pro FS yet. The Pro FS AD measurement was performed on Aug 2018.



I hope RME could correct their manual and online website, and document the actual numbers and graphs. Seems many are not recorded correctly.

12

Re: Why ADI-2 Pro FS measures worse than ADI-2 Pro?

ning wrote:

>  RME specs the SNR for the AD in the Pro FS the same as in the Pro (manual 33 p. 64).

You are right. but the FFT spectrum said otherwise. on page 80 FS has the SNR = 114.18, the non-FS has SNR=116.988.

RME might have measured their SNR-spec differently (eg. different bandwidth). Maybe for 33 they measured dynamic range (amplitude of 0 dBFS signal less noise without signal) rather than SNR (amplitude of 0 dBFS less noise with signal present).

The -1 dBFS specs in 33 also differ from the measurements in 34.13 and the math only checks out for 34.13 (unless signal and noise are significantly correlated for 33 but not for 34.13).

For the DA, the -1 dBFS specs are given as 33 vs. 34.13: THD is -112 dB vs. -114 dB, THD+N is -110 dB vs. -111 dB, SNR is 117 dB vs. 114 dB.
For two uncorrelated -114 dB signals (THD and N from 34.13) the RMS of the combined signal is -111 dB. Whereas for uncorrelated signals with amplitude of -112 dB (THD form 33) and -117 dB (N from 33) respectively, the RMS of the combined signal (THD+N) is -111 dB which is 1 dB below the spec from 33.
For a signal with RMS amplitude -110 dB (THD+N from 33) which has two uncorrelated parts of which one has RMS amplitude of -112 dB (THD from 33), the RMS amplitude of the remaining part (N from 33) is -114 dB which is 3 dB higher than the spec from 33.

For the AD of the Pro, the -1 dBFS specs are again given as 33 vs. 34.13: THD is -116 dB vs. -117 dB, THD+N is -112 dB vs. -114 dB, SNR is 120-121 dB vs. 117 dB.
For two uncorrelated -117 dB signals (THD and N from 34.13) the RMS of the combined signal is -114 dB. Whereas for uncorrelated signals with amplitude of -116 dB (THD form 33) and -120-121 dB (N from 33) respectively, the RMS of the combined signal (THD+N) is -115 dB which is 3 dB below the spec from 33.
For a signal with RMS amplitude -112 dB (THD+N from 33) which has two uncorrelated parts of which one has RMS amplitude of -116 dB (THD from 33), the RMS amplitude of the remaining part (N from 33) is -114 dB which is 6-7 dB higher than the spec from 33.

For the AD of the Pro FS, the -1 dBFS specs are again given as 33 vs. 34.13: THD is -113 dB vs. -113 dB, THD+N is -110.6 dB vs. -110.6 dB, SNR is 120-121 dB vs. 114 dB.
For two uncorrelated signals with amplitude of -113 dB (THD form 34.13) and -114 dB (N from 34.13) respectively, the RMS of the combined signal (THD+N) is -110.5 dB which is only 0.1 dB below the spec from 34.13 which is well within the errors from rounding.  Whereas for uncorrelated signals with amplitude of -113 dB (THD form 33) and -120-121 dB (N from 33) respectively, the RMS of the combined signal (THD+N) is -112.2-112.4 dB which is 1.6-1.8 dB below the spec from 33.
For a signal with RMS amplitude -110.6 dB (THD+N from 33) which has two uncorrelated parts of which one has RMS amplitude of -113 dB (THD from 33), the RMS amplitude of the remaining part (N from 33) is -114 dB which is 6-7 dB higher than the spec from 33.


ning wrote:

> Judging by the 'skirt' around the 1 kHz tone and the Pro FS manual 34.6 (p. 75), the DA figure in 34.13 seems to be for the DAC/Pro FS.

You made a good point.  However I have another theory. If you take a look at FS manual 34.13 (p80), The DA measurement were done on Nov 2016, which is the same date they did the non-FS Pro's AD FFT. At that time there was no DAC nor Pro FS yet. The Pro FS AD measurement was performed on Aug 2018.

I missed the date on the figures. Also, the title in the DA THD spectrum figure only says Pro (no FS). The DA THD spectrum figure for the DAC shows no date but identical performance to the Pro.


ning wrote:

I hope RME could correct their manual and online website, and document the actual numbers and graphs. Seems many are not recorded correctly.

RME seems to have derived the THD, THD+N and SNR specs for 33 differently than by the measurement shown in 34.13. I think it would suit them well to remove or explain the discrepancies.