Topic: Problems with setting sampling-rate, external clock and DSD recording

I love this device with my master clock. Internal it gets a bit muddy!!!
(But I don´t like this Forum software, for I have to rewrite all of this, cause I did hit the preview button and everything was gone...)

But back to the topic.

When plugged into the computer (Mac) the sample rate handling gets a bit confusing. The Driver does not recognize changes from the external clock (or show them correctly) and somehow does not properly communicate with the unit. And I´m a bit puzzled for it´s not possible to change the clock setting in the device while connected.
Therefore, if I work with the external clock, its not possible to get into double speed mode.

In standalone mode I can get up to 384kHz or DSD128 with the external clock signal of 192kHz (or 176.4kHz), which sounds great. When I first read about DSD I was very sceptical, but it´s amazing how the Stage opens up with it, compared to PCM. All I have to do is set the sample rate source to an external clock and adjust the sample rate in the ADI-2 Pro to the next higher position.
Now I want to record that phantastic sound, which is not possible, because, as soon as I plug in the USB, it jumps back to the originally provided rate.
If I set a higher value in the Driver, nothing happens at all. And as I have mentioned, it´s impossible to operate the sampling rate slider in the unit, while connected to my Mac.

Therefore I am only able to record up to 192kHz PCM and DSD64, which is a shame. DSD128 with 150kHz filter sounds a bit more transparent than DSD64, as well as 348kHz compared to 192kHz PCM. It could be useful for analogue mastering purposes and the final recording.

Is this a bug? Or am I missing something?

Well, and it´s not of any help to record with internal clock, for the converters sound better on 48kHz external clock than on 768kHz internal !!!!

Thx for any helpful Information.
Boris

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

2 (edited by KaiS 2020-12-07 14:15:41)

Re: Problems with setting sampling-rate, external clock and DSD recording

Isn't it that you have to set the Sample Rate in the recording DAW?

At least here in my studio that's the only way to change sample rates on any USB converter I use.

I'm not sure or tested how that works with ADI-2 Pro in the combination external clock / USB, but on my other USB interfaces it simply does, with external 32 CH of MADI signal.



A sceptical remark according your sound observations:
For "the stage to open up" it would need to be congested before by some technical losses, something I've never experienced with (quality) conventional recording formats.

When I'm recording in the studio I have a permanent A/B comparison between the analog original and the replayed recording, all day long.
I have the best possible test signals available: musicians playing live.

My RME converters do sound perfectly transparent, no difference between analog original and recorded (AD-DA-ed) signal is noticeable.



If a sound would "open up" I would suspect something is level misaligned (louder sounds better) or broken, in the direction of added distortions or frequency response changes.
For example, if converter and recorder work unsynced, with different sample rates, a permanent pseude-sample-rate-conversion would happen (skipped samples) that adds a lot of distortions.

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

Isn't it that you have to set the Sample Rate in the recording DAW?

First of all, there is no such setting option in Vinylstudio (wich is the only DSD recording software for Mac I know of). It has no setting dialog for picking certain devices or even requesting sample rates. It just takes the settings of the system. And in Reaper, it´s the same as I described.

KaiS wrote:

At least here in my studio that's the only way to change sample rates on any USB converter I use.

Not with the ADI-2 driver. There it is possible to do it in the Audio-Midi-Setup.




KaiS wrote:

A sceptical remark according your sound observations:
For "the stage to open up" it would need to be congested before by some technical losses, something I've never experienced with (quality) conventional recording formats.
...
If a sound would "open up" I would suspect something is level misaligned (louder sounds better) or broken, in the direction of added distortions or frequency response changes.
For example, if converter and recorder work unsynced, with different sample rates, a permanent pseude-sample-rate-conversion would happen (skipped samples) that adds a lot of distortions.

Yes your right. I got the unit used and it sounded good enough - listening to Spotify - to trick me. But now I tried my old ADI-2 against it, with a decent Signal, and it sounded way better. Have to send it back or get it fixed.
Thx for open my ears on that.

Let´s see until I get a flawless unit...

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

4 (edited by KaiS 2020-12-07 19:21:22)

Re: Problems with setting sampling-rate, external clock and DSD recording

I don't think your unit is flawed.

I'm not too much into Mac (although I do use both, Mac and PC), but as far as I can see Vinylstudio supports ASIO, which is a much more potent interface then Mac's system drivers/settings.

So your solution might be a few mouseclicks away installing RME's ASIO driver.

This might even fix the sound difference, that shouldn't exist.
Don't forget - for any valid sound comparison levels need to be very tightly matched.
When the race is close 0.05 dB make THE difference.


Regarding DSD:
In the recording studio we don't use this format, because not the simplest processing is possible, not even volume changes.

5 (edited by borispietsch 2020-12-07 20:57:38)

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

I don't think your unit is flawed.

Your right. Much of confusion, because something hocked up in the unit. (Which was after discovering the effect.)
When reseted, everything was fine again with the Sound.
Puhh...

KaiS wrote:

I'm not too much into Mac (although I do use both, Mac and PC), but as far as I can see Vinylstudio supports ASIO, which is a much more potent interface then Mac's system drivers/settings.

So your solution might be a few mouseclicks away installing RME's ASIO driver.

I will try that...
But the behaviour does´nt occur just in Vinylstudio. It´s also there, when you use the ADI-2 Pro just as the system preference.

KaiS wrote:

This might even fix the sound difference, that shouldn't exist.
Don't forget - for any valid sound comparison levels need to be very tightly matched.
When the race is close 0.05 dB make THE difference.

The sound observation was done with ADI-2 in Standalone mode.
When Did with another converter for PCM and the ADI-2 for PCM the levels can freely be adjusted.

Try the following:
(Use a good Organ and Choir recording)
Go out of your iPhone-Headphone-Jack, right into your Adi-2. The converter of the Phone ads some exiter to the Signal which can sound nice, if you like it.
If you switch between DSD and PCM now you will experience, that DSD is much more capable of capturing this exiter-effect.
If I compare the Signal, send via DI Box into the excellent pres and converters of the Sounddevices 702, it sound exactly the same as unbalancend into the ADI-2 converted into PCM.
DSD however preserves the somehow bigger Room. Its more like being in the Scene. Compared, PCM seems to be just a bit flat. Puts the center together again. While on DSD it stays a bit confused, like if you listen to the iPhone output just analogue. It´s no matter of volume. If you start do decide which sounds better, then, in most cases the higher volume wins.

This comparision I did in DSD64 and 176.4kHz for the Sounddevices handles only up to 192kHz.

If you disable SRC you can do the test just with the ADI-2 on DSD128 and 352.8kHz PCM with an external 192kHz clock, but without the possibility to rematch the levels. The effect is about the same, just both signals are a bit more brillant.

Never thought that there is an audible difference with this high sampling rates. But it seems, that with very good equipment, the inaudible frequencies are influencing the ones u can hear for more liveliness.

Its probably very good to experience with this odd effect of the iPhone converter.

I wonder, how bad the clock of the ADI-2 is, compared to the sound of the converters. It sounds even better with the O.K. clock of the Sounddevices 702 which already sounds better with Ambient Recordings "CLOCKIT".


KaiS wrote:

Regarding DSD:
In the recording studio we don't use this format, because not the simplest processing is possible, not even volume changes.

Yes, its only usable for analogue mastering purposes. Like a tape-machine. Unless you invest in very expensive Software...

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

So your solution might be a few mouseclicks away installing RME's ASIO driver.

Well, there is no ASIO driver for Mac OS!

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

7 (edited by KaiS 2020-12-08 00:56:39)

Re: Problems with setting sampling-rate, external clock and DSD recording

Have you seen this short video from RME about Mac driver installation:
https://youtu.be/qmrcKYf6TaI

Re: Problems with setting sampling-rate, external clock and DSD recording

Well, yes, I did this with my Fireface 800 Driver. But for the ADI-2 Pro there is no driver from RME!

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

9 (edited by KaiS 2020-12-08 01:42:15)

Re: Problems with setting sampling-rate, external clock and DSD recording

Did you read manual page 58?
ADI-2 Pro needs to be in Stereo-, not in Multichannel Mode for higher samplerates (page 39, 2nd passage)

The comprehensive Mode switching (page 36 ff) is quite a unique feature.
But can be a stumbling stone too.
Needed some trials here find the one that fits best to my demands.

10 (edited by borispietsch 2020-12-08 02:05:15)

Re: Problems with setting sampling-rate, external clock and DSD recording

Well, it´s set to stereo and everything works fine. Just when connected to the Computer I can not use the higher sample rates with a 192kHz external clock. In Standalone it works well. (Up to 384kHz PCM / DSD128).

(Off Topic: Scheinst ja ein schönes Studio in Karlsruhe zu haben. Im Januar nehme ich in Kaiserslautern eine CD auf und erweitere dafür gerade mein Equipment.)

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

11

Re: Problems with setting sampling-rate, external clock and DSD recording

The ADI uses the OS driver under Mac, and therefore can't change what that driver does and want. Clocking works as Apple wants clocking to work.

A different topic is that external sync is limited to the sample rates of SPDIF/AES. Multiples of these sampe rates are not supported, and not needed in real-world applications.

The description in post one seems inaccurate. Double Speed mode works, it is even stated that Quad Speed mode works. But one can not use Quad Speed input signal to use the unit in even higher sample rate modes, then externally clocked. Nobody needs such sync. If you think you do then I predict your application setup is flawed - like in post 1, where the ADI is supposed to do things it can't and wasn't built for.

Regards
Matthias Carstens
RME

12 (edited by borispietsch 2020-12-08 12:00:19)

Re: Problems with setting sampling-rate, external clock and DSD recording

MC wrote:

The ADI uses the OS driver under Mac, and therefore can't change what that driver does and want. Clocking works as Apple wants clocking to work.

So with an external clock I always have to set this external clock and the drivers clock to get it work? Let´s see wat happens when I implement it in a main device.

MC wrote:

A different topic is that external sync is limited to the sample rates of SPDIF/AES. Multiples of these sampe rates are not supported, and not needed in real-world applications.

Sorry, but that is simply not true.

1. Try it yourself. Disconnect the ADi-2 Pro from USB. Connect an external clock thru, lets say SPDIF-coax. Make shure that "SPDIF in" is set to coax or auto and "SRC" is set to OFF. Set your external clock to 192kHz.
Now go to clock options. Set "clocksource": SPDIF, "sample rate": 768kHz. You will see that the samplingrate jumps back to 384kHz. So you can choose between 192/384kHz internally. (With 176.4 it will be 176.4/352.8kHz)
If you have a good clock, the difference is well audible, for the internal clock of the device is not good. Maybe the FS version is better.

2. For the bad clock in the device itself it is needed to have an external clock, to do high quality archivation of analogue recordings. (MasterTape/Vinyl)
(As mentioned, the ADI-2 Pro with an exellent external clock in 48kHz sounds better than on 768kHz internally!)
So, if I want to use it´s full potential (or at least the ability to record DSD128 with Vinylstudio or HQPlayer4Pro, or PCM 384kHz to any capable program, I will need to get the behaviour of the device, explained in point 1., while connected to the Mac. - And if I want to play it back from the computer to record it on Tape (or lacquer cut, wich I don´t do at home) I will also need this feature.
(If the "sample rate Slider" in the device was´nt blocked on USB connection it could work.)

I think this is a real-world application. What do you think?

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

13

Re: Problems with setting sampling-rate, external clock and DSD recording

If I write what I think here I will get sued. There is some basic understanding of digital clocking as well as how AD/DA conversion uses such clocks missing on your part.

And you missed the chapter in the manual explaining the reason for the special 384 kHz single channel mode with 192 kHz SPDIF (14.1.2 Clock). The fact that you didn't even notice that it's no longer stereo speaks for itself.

Regards
Matthias Carstens
RME

Re: Problems with setting sampling-rate, external clock and DSD recording

Hey man, no need for getting offensive.
I just share my observations.

Yes I can´t get a digital Stereosignal out of the device by AES/SPDIF higher than 192kHz. But if I use it as a Preamp, the AD and the DA converter inside the unit seem to be able to use the 192kHz signal to work in 384kHz. Am I wrong here?

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

15

Re: Problems with setting sampling-rate, external clock and DSD recording

In Preamp mode the sample rate shown in the Status bar jumps back to 192 kHz when one sets the Sample Rate in the Clock menu to higher values as the incoming 192 kHz SPDIF/AES.

Edit: not true, I had ADAT as input signal. With SPDIF/AES the 384 kHz option is there for the single channel measurement mode.

Regards
Matthias Carstens
RME

16 (edited by borispietsch 2020-12-08 18:24:50)

Re: Problems with setting sampling-rate, external clock and DSD recording

If you happen to be in Munich once, I would be glad to show you how it sounds on my equipment when I use this settings.

Do´nt get me wrong. Using AD/DA conversion in the ADI-2 Pro is superb when driven by an excellent clock. I love this device. Also for its EQ capability. I have some slight room modes (around 98Hz to 178Hz) here. With the precise EQ settings the ADI-2 Pro as an "analogue" Preamp sounds better here, than a pure analogue signal without proper treatment.

I just tried it again, not to fool myselfe.
If you do the procedure described in post 12 there is one way to catch the difference from external clocked 384kHz to internally clocked. It´s hard to decide, for the difference is subtle of course and you can´t easily switch back and forth. It´s just a one time switch for direct comparison. After that, you have to do two steps to get back to the external clock setting.

If you are in 384kHz mode, driven by SPIF external clock (192kHz) just set "clocksource" back to "internal" (and it could be that there is a slight drop of clarity).
It´s very hard to tell right now because my very good clock just does 48kHz. I get an updated device this days which does any imaginable rate up to 192kHz. It´s even not that expensive. It´s from Ambient Recording, a Munich based company wich does or did special timecodemoduls for SoundDevices wich are very precise for Film and Sound synchronisation. The name of the device is "CLOCKIT". But the new devices do just 48kHz again. Despite of this, some older "CLOCKITs" can handle up to 192kHz and are sold around 250,- to 400,- € (depends on the model).
https://ambient.de

To come to the conclusion that 48kHz sounds better with this clock than 768kHz internally clocked, I used slow filter and the EQ compensation you presented in this post: https://forum.rme-audio.de/viewtopic.php?id=26952

I´m not doing statements just randomly. I do my research, experiments and most of all listen very carefully.

It´s not about bashing or some Voodo-Theories. It´s just about listening and trying to make things better... ...for fine recordings.

(I listen on STST Speakers by Stefan Strohmetz a manufacturer of very fine eqipment in Allgäu Germany, driven by Naim NAP 140.)
STST
http://www.stst-hifi.de (Sadly does´nt build Speakers any more.)
NAP 140
https://www.stereophile.com/content/nai … 2BUs7r9.97

Boris

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

17 (edited by borispietsch 2020-12-08 14:23:59)

Re: Problems with setting sampling-rate, external clock and DSD recording

MC wrote:

In Preamp mode the sample rate shown in the Status bar jumps back to 192 kHz when one sets the Sample Rate in the Clock menu to higher values as the incoming 192 kHz SPDIF/AES.

Thats not true with my device...

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

Re: Problems with setting sampling-rate, external clock and DSD recording

Here are some Pictures:
https://www.dropbox.com/sh/i4w8ohl6did3 … NQOaa?dl=0

The Sound is clearly opener on 384kHz!
Why ever that is, if the ADI-2 Pro does not support it and just shows false information on my unit...

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

19

Re: Problems with setting sampling-rate, external clock and DSD recording

That is the 384 kHz state which shows in the Global Level Meters that SPDIF and AES I/O are disabled, but send out one channel of data in SMUX mode.

So far I seem to have missed a clear statement what exactly you do here. Analog input is source and you listen via Phones or the rear analog outputs? In this case there is no audible difference between Internal clock 192 kHz, 384 kHz or using SPDIF/AES as clock source. You can claim that, but you can't proove it. Not to mention that with statements aka 'external clocks sound different' you won't find a big audience here. Technically this is indeed nonsense. Our SteadyClock FS completely (!) removes any influence the external 'very good' clock could have.

A possible reason for sound differences would be the fact that on 384 kHz some DSP effects will be switched off. If you overlooked that and have B/T, PEQ, Crossfeed, Loudness etc active, then the switch from 192 to 384 could sound different because one of these gets disabled. Details are in the manual, chapter 8.8.

Regards
Matthias Carstens
RME

20 (edited by borispietsch 2020-12-08 18:30:36)

Re: Problems with setting sampling-rate, external clock and DSD recording

MC wrote:

That is the 384 kHz state which shows in the Global Level Meters that SPDIF and AES I/O are disabled, but send out one channel of data in SMUX mode.

Yes!

MC wrote:

So far I seem to have missed a clear statement what exactly you do here. Analog input is source and you listen via Phones or the rear analog outputs?

Yes! Yes, rear! You seem to have missed my post N#16.
Just talking about stand alone mode. Tape > A/D ADI-2 Pro D/A > Amp > Speaker.

MC wrote:

In this case there is no audible difference between Internal clock 192 kHz, 384 kHz or using SPDIF/AES as clock source. You can claim that, but you can't proove it.

I don´t want to argue about subjective sensations. It could be individually proofed by a blind test.
Just wan´t to know my gear inside out for the next recording session. An Album with two acoustic Artists.

I wonder myself that there is an audible difference between this high rates. I just guess that even if I´m not able to notice a sine wave higher than 16-18kHz (don´t remember exactly), that I´m able to perceive some interference, that the higher frequencies produce with the ones in my hearing range.
In my last test I used a tape recording. 384kHz has a more transparent room impression to me than 192kHz.
Still have to evaluate my impressions on DSD, for in this mode the EQ and probably most other settings don´t work, even in DSD A/D to PCM to D/A. So I have to check that first.
And why did you implement 768kHz if already 384kHz is no improvement. (Well, maybe just for test-equipment?)

MC wrote:

Not to mention that with statements aka 'external clocks sound different' you won't find a big audience here. Technically this is indeed nonsense. Our SteadyClock FS completely (!) removes any influence the external 'very good' clock could have.

Well, I don´t have the FS Version. And my different external clocks sound indeed different, each, with the ADI-2 Pro (not FS).
Just a subjective observation so far, indeed. For a real blind-test I need somebody else to operate the unit. But closing your eyes while listening helps a lot. ;)

RME wrote:

Bei SteadyClock FS lag der Fokus auf einer Reduzierung des Eigenjitters auf niedrigste Werte. Dies gelang durch Verbesserung der zweiten analogen PLL-Schaltung, und das Referenzieren der Direct Digital Synthesis als auch der PLL zu einem Quarz mit niedrigerem Phasenrauschen. Der über eine DA-Wandlung messbare Jitter erreicht nun Werte, die normalerweise nur im Master-Quarzmodus verfügbar sind, während SteadyClock weiterhin grundsätzlich im PLL-Modus läuft - egal ob interner oder externer Takt, der Klang ist (wieder) exakt identisch. Der für die aktualisierte Schaltung genutzte Low Phase Noise Quarz besitzt eine Jitter-Spezifikation unterhalb einer Pikosekunde (ps), ein Bereich der FemtoSecond genannt wird. Daher SteadyClock FS.

https://www.rme-audio.de/de_steadyclock-fs.html

So you did improve the SteadyClock and gave it a better Quartz. Why would you do that if the old version was perfect?

MC wrote:

A possible reason for sound differences would be the fact that on 384 kHz some DSP effects will be switched off. If you overlooked that and have B/T, PEQ, Crossfeed, Loudness etc active, then the switch from 192 to 384 could sound different because one of these gets disabled. Details are in the manual, chapter 8.8.

AT 348kHz, external clock:
PEQ works fine
With works fine
Phase invert as well as Mono, Dim etc.
Loudness (not working, but was switched off anyway.)
Crossfeed (Seems to work. Not listening with headphones so far - have to try on my AKG K812. Was switched off)
B/T (not working. Was switched off)

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

21 (edited by KaiS 2020-12-08 23:25:09)

Re: Problems with setting sampling-rate, external clock and DSD recording

It's really a lot of effort to do a proper and valid A/B or better A/B/X test.
Then - just to find out, after all: there is no difference.
This happened (not only) here:
http://www.acousticplan.de/PCM_vs_DSD.pdf
I just read through this 99 pages of a "Diplomarbeit"  (University diploma dissertation).
They compare DSD vs PCM, under highly controlled and optimized conditions, with 148 participiants.

Result: nada, no audible difference!


A tip from producer/sound engineer to sound engineer, where to spend your energy:

Working intensively with the artists gains worlds of improvement even on the sound.

• Great artist's performance is the base for every great recording.
Let them play the best they ever did in their life.
• Making good arrangement decisions helps the songs.
• Next, microphone selection and placement is the key to catch this performance the best way possible.
• Catching (or not killing!) the vibe with a good mix is the icing on the cake.

• Everything else has really minor influence.

If you record with terra-blop technology or on basic stuff doesn't make THE difference, as long as the equipment doesn't come in the way of the music.
But - you have to know your stuff to concentrate on the musical performance.

Like Joseph Beuys said: "Jeder Griff muss sitzen":
https://www.edition-staeck.de/produkt/p … ss-sitzen/

Save your energy on that.
Of course get to know your equipment.

And one last tip: don't use a recording format that ties your hands on the back, preventing you from doing things you would normally do.

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

http://www.acousticplan.de/PCM_vs_DSD.pdf
I just read through this 99 pages of a "Diplomarbeit"  (University diploma dissertation).
They compare DSD vs PCM, under highly controlled and optimized conditions, with 148 participiants.

Result: nada, no audible difference!

Thank you. Very interesting Paper. So far, I read half of it. I will continue later.

So do you think that these different formats or different rates can sound eagual to each other in a perfectly treatet environment while on the other hand indeed could sound different on less precise recordings, listened to in an untreated room, because they create different impulse responses? Not talking about overall levels here.

And is it possible that the Adi-2 Pro produces different levels at different rates and formats/filters? That indeed would make it quite difficult to compare between them.

KaiS wrote:

Like Joseph Beuys said: "Jeder Griff muss sitzen":
https://www.edition-staeck.de/produkt/p … ss-sitzen/

Interesting that you mention Joseph Beuys, who was a very fine percieving human being in my eyes, with a great heart and an uncompromisingly clear mind, paired with, let's say, some punk rock. I adore his work.
Yes. feder Griff muss sitzen... That's why I do all of this investigation right now. I want to forget about it in the creative process without being negatively surprised afterwards by technical decisions I have made.
Maybe I should visit you in your studio, while I'm in Kaiserslautern.

But back to my very first question which got a little bit lost in all this emotionally loaded questions of believe, if there are perceivable differences in the spectrum above 20kHz, or about the flawlessness of the stady clock technology.

Is it somehow possible to record with a rate of more than 192kHz via USB, if the Adi-2 Pro is controlled by an external clock.

It could be easily achieved if I would be able to operate the sample rate switch on the device while being connected to USB, or if the device would listen to the driver settings while in external clock mode.

So, one simple question: Is there a way to make it possible?

Thx,
Boris

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

23

Re: Problems with setting sampling-rate, external clock and DSD recording

Answered already: this is not possible and will not be possible in the future.

Regards
Matthias Carstens
RME

24 (edited by KaiS 2020-12-09 09:40:30)

Re: Problems with setting sampling-rate, external clock and DSD recording

borispietsch wrote:
KaiS wrote:

http://www.acousticplan.de/PCM_vs_DSD.pdf
I just read through this 99 pages of a "Diplomarbeit"  (University diploma dissertation).
They compare DSD vs PCM, under highly controlled and optimized conditions, with 148 participiants.

Result: nada, no audible difference!

Thank you. Very interesting Paper. So far, I read half of it. I will continue later.

So do you think that these different formats or different rates can sound eagual to each other in a perfectly treatet environment while on the other hand indeed could sound different on less precise recordings, listened to in an untreated room, because they create different impulse responses? Not talking about overall levels here.

No - same stays same.


But, they used power amps with very broad HF power bandwidth - 200kHz.
Few power amps can do this.


Minor amps (or other electronics in the signal path) with limited ultrasonic power capabilities will produce intermodulation distortions with the ultrasonic noise above 20kHz DSD produces.
These IMDs would "fold" the ultrasonics into the audio range and compromise audio quality.

More distortions can sound like "more ambience", remember the famous Aphex Aural Exiter which ads distortion.


I would go that far: If it sounds different something goes wrong.
Remember, this paper is examining exactly your case:
High sample rate PCM vs DSD - they found no audible difference.

My bet is on a level difference, but distortions in your audio chain aren't too unlikely.
Once you have actively heard how these type of IMDs sound you will always recognize them.
They're most obvious with sibilants or single, high frequency percussions.
Recording a triangle is one of the best tests.


Use STAX headphones with STAX solid state amp, directly connected to ADI-2, to exclude faults in your speaker signal path.


borispietsch wrote:

And is it possible that the Adi-2 Pro produces different levels at different rates and formats/filters? That indeed would make it quite difficult to compare between them.

Not too unlikely, as much is different between PCM and DSD signal processing, and simple to check:

Use an AC voltmeter of high enough resolution, a sinewave, and measure it.
While in that, check for difference in the treble frequency response up to 20kHz.

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

Use STAX headphones with STAX solid state amp, directly connected to ADI-2, to exclude faults in your speaker signal path.

Thanx, but for now I will stay with my AKG K812. (They state to handle 5 - 54000 Hz. Should´nt it work as well if I plug them into the ADI-2 Pro?) My budget is limited and I think this would not be the most important investment right now.


KaiS wrote:
borispietsch wrote:

And is it possible that the Adi-2 Pro produces different levels at different rates and formats/filters? That indeed would make it quite difficult to compare between them.

Not too unlikely, as much is different between PCM and DSD signal processing, and simple to check:

Use an AC voltmeter of high enough resolution, a sinewave, and measure it.
While in that, check for difference in the treble frequency response up to 20kHz.

I might do this sometimes but as you mentioned, there are more important things to take care of first.

Would be great though if Matthias Carstens could answer this question.
But I think he does not like me. I´m very sorry for any offence, but I just tried to share my observations as precise as possible. Not meant to bash RME at any time or to have the final answer on open questions. We are all learning at least...

Boris

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

26 (edited by KaiS 2020-12-09 14:53:32)

Re: Problems with setting sampling-rate, external clock and DSD recording

The AKG K812 if a nice headphone too, I have it and would say it has a good resolution.
And yes, plugging it directly to the ADI-2 could do the job if you add a passive volume control.
With DSD there's non, and for a valid comparison you should run PCM on 0 dB Volume setting anyway.


Here's a graph of the ultrasonic noise DSD contains, and some explanations what creates this exorbitant noise level:
https://digital-audio-systems.com/pcm-i … ch-zu-dsd/

You can see I'm not talking about a bit of noise floor at the bottom, it's really a lot of noise energy outside the audio band.
I guess a lot of power amps struggle handling this.

I wouldn't want a signal like this in my mixing console, besides causing all kinds of distortion it's capable to crosstalk into other channels and mixing busses too.
And that's a point: with DSD you're forced to work in the analog domain, but only specially prepared electronics is capable to work with such ultrasonic polluted signals.


I can't speak for Matthias, but he aswered your questions.

27 (edited by borispietsch 2020-12-09 14:52:31)

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

I can't speak for Matthias, but he aswered your questions.

Maybe you are right and i´m over interpreting things. Maybe it was just in post N#13, where he got me wrong - thinking I was not able to notice the difference between mono and stereo - that he sounded a bit offensive. It´s hard to tell, communicating just by written text...

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

28 (edited by KaiS 2020-12-09 15:18:48)

Re: Problems with setting sampling-rate, external clock and DSD recording

Please re-read my posting above, I edited some.

The whole clocking issue is complicated in practise once external clocks and USB meet. They're not the best friends.
Bringing x2 and x4 into play doesn't make things easier.

Without step by step rebuilding what you did (which I can't, as I neither use MAC any more, nor Vinyl Studio) impossible for me to predict results.

So post #13 can be seen as a hint to read manual page 28, 14.1.2 too.


As you're new to ADI-2 Pro I'd strongly suggest to read the manual, this device is VERY different to anything on the market, much more versatile even than it looks from the outside.
I did, and needed some time to become familiar.

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

As you're new to ADI-2 Pro I'd strongly suggest to read the manual, this device is VERY different to anything on the market, much more versatile even than it looks from the outside.
I did, and needed some time to become familiar.

Yes, I am reading the manual. I always do both, reading the manual and playing around to discover what happens...

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

http://www.acousticplan.de/PCM_vs_DSD.pdf
I just read through this 99 pages of a "Diplomarbeit"  (University diploma dissertation).
They compare DSD vs PCM, under highly controlled and optimized conditions, with 148 participiants.

This is indeed a very interesting paper.

And it should be noted that there where 4 persons who seemed to consciously perceive a scientific significant difference. I know that they try to explain it with a flaw in their experimental layout, but this is just another hypothesis.

I have the experience, that if I notice a subtle difference in a signal, and another person is not - just stating that it sounds equal - there is a way to sharpen her awareness, so that this person is enabled to lift her perception into consciousness.
I do this by describing as precise as possible, how this different sound effects me. And often tell the one I want to be able to decide for himself, which sound he prefers: "Do´nt try to catch the difference of what you hear, but be aware of what you feel when you hear it." "When do you feel it to be more alive?"
In most cases they are able then to tell exactly when I change the sound to which setting, even if they do´nt know when and to what setting I do change it. When they once catch it, they also can very precisely describe what difference they do perceive. And what setting they do prefer.

Probably they where effected by the difference before. They just could not get it into their consciousness.

Perception, awareness and consciousness are very tricky things. Peoples feelings are often effected by sound even if they do´nt recognize that they hear it. Is´nt that what the magic of sound is about...
I do´nt know if statistic is helping here.


(I often do this with myself when I want to decide on a subtle change. Close my eyes and hit the bypass button a few times to disable myself of knowing in which state it is and then press it as often I like, to catch the difference and to decide for the one or the other. Still eyes closed.)

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

31 (edited by KaiS 2020-12-10 00:18:28)

Re: Problems with setting sampling-rate, external clock and DSD recording

borispietsch wrote:

... Perception, awareness and consciousness are very tricky things. Peoples feelings are often effected by sound even if they do´nt recognize that they hear it. Is´nt that what the magic of sound is about...
I do´nt know if statistic is helping here.


(I often do this with myself when I want to decide on a subtle change. Close my eyes and hit the bypass button a few times to disable myself of knowing in which state it is and then press it as often I like, to catch the difference and to decide for the one or the other. Still eyes closed.)

Blind A/B - the best one can do when working on a project.
I do this quite often.
Did you notice, you have to sit absolutly still when switching.
A tiny move and the sound changes.

More often I do simple A/B.
A/B/C/D/... - decisions have to be done all the time working in our job. When recording, producing, mixing, all the time things have to be decided to create the "sound".
Every action incorporates such a decision, so I am used to decide - not only if something's different, but what's better.
The difference is, there is the real difference - if you move a microphone, if you choose another recorded take, set up an EQ or compressor...

If the difference, in case it exists, is so minor and everything else I do as a producer is multi-magnitudes more effective I put my energy into this more rewarding.

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

Blind A/B - the best one can do when working on a project.
I do this quite often.
Did you notice, you have to sit absolutly still when switching.
A tiny move and the sound changes.

Probably doing it. But I was´nt conscious about it... :))


KaiS wrote:

If the difference, in case it exists, is so minor and everything else I do as a producer is multi-magnitudes more effective I put my energy into this more rewarding.

Yeah, you´re right.
But do you know the point, when everything is done and you recognise that this last 1% is missing and you try to make it happen. That is so much more effort than doing it right in the beginning of the process. Therefore I do all of this stuff... Getting better gear, finding out in what format to record best, or even on Tape maybe, sometimes... In the end it is always about finding the best possible compromise for sound and workflow. (Not to forget the inspiration of beauty around good old gear and by doing things with your hands.)

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

33 (edited by borispietsch 2020-12-18 16:03:23)

Re: Problems with setting sampling-rate, external clock and DSD recording

borispietsch wrote:
MC wrote:

Not to mention that with statements aka 'external clocks sound different' you won't find a big audience here. Technically this is indeed nonsense. Our SteadyClock FS completely (!) removes any influence the external 'very good' clock could have.

Well, I don´t have the FS Version. And my different external clocks sound indeed different, each, with the ADI-2 Pro (not FS).
Just a subjective observation so far, indeed. For a real blind-test I need somebody else to operate the unit. But closing your eyes while listening helps a lot. ;)

Now I got an FS Version ADI-2 Pro here. Indeed it is more detailed in sound than the non FS Version. Still it sounds even a bit more clear with my external clock (even if this is technically impossible Edit: (It works only if SRC is off. So I suggest not to use SRC if you have a very good master clock.)), but now the highest internal sample rate delivers a finer sound than 192kHz externally clocked.


If you want to try yourself, catching this subtle differences between high sample rates and different clock sources, listen to the crackle of an analogue Record, or the smack sounds of mouth and tongue while singing or speaking in a very good microphone. Both are very short transient and frequency rich signals, which have a great liveliness to them. (For my last test I used "Hallelujah - Live in Studio" by Lola Christine (3:10 - 3:40)). The higher the sample rate gets and the more precise the clock works, the more detail you hear on this kind of sounds and the more natural and alive these come across. They are a very good reference because even cranking up the volume a bit does not really help here to make them more alive or detailed. They just get annoying. (Of course very good speakers or headphones help a lot or are even necessary.)

borispietsch wrote:
MC wrote:

In Preamp mode the sample rate shown in the Status bar jumps back to 192 kHz when one sets the Sample Rate in the Clock menu to higher values as the incoming 192 kHz SPDIF/AES..

Thats not true with my device...

With the FS Version this seems to be true...

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

Re: Problems with setting sampling-rate, external clock and DSD recording

O.K. hre´s more on my DSD/PCM comparison experiment.

I took everything in consideration, what we have discussed so far. Especially the following:

KaiS wrote:

Minor amps (or other electronics in the signal path) with limited ultrasonic power capabilities will produce intermodulation distortions with the ultrasonic noise above 20kHz DSD produces.
These IMDs would "fold" the ultrasonics into the audio range and compromise audio quality.

More distortions can sound like "more ambience", remember the famous Aphex Aural Exiter which ads distortion.


I would go that far: If it sounds different something goes wrong.
Remember, this paper is examining exactly your case:
High sample rate PCM vs DSD - they found no audible difference.

My bet is on a level difference, but distortions in your audio chain aren't too unlikely.
Once you have actively heard how these type of IMDs sound you will always recognize them.
They're most obvious with sibilants or single, high frequency percussions.
Recording a triangle is one of the best tests.


Use STAX headphones with STAX solid state amp, directly connected to ADI-2, to exclude faults in your speaker signal path.

First I tried my AKG K812 headphones.
So just analogue into the ADI-2 Pro FS, conversion to DSD or PCM, then back to analouge out of the headphone preamp of the ADI (wich is only possible by converting DSD back to PCM) into the phones.

After that, done the same (but could stay with DSD all the way, because direct DSD is possible on rear outs), just going into my amp and speakers (see post N# 16).

I tried it with different PCM filters. (Slow I like the most.)

1. 176kHz PCM vs. DSD64
(Same formats as in your paper.)

The difference is subtle but obvious.

2.  705.6kHz vs. DSD256

The difference is even more obvious.
It is the room impression that is changing. With DSD it seems to be a bit more "spacy" and might be closer to the analogue source-material (hard to tell with all the switching und plugging in between). I was reminded of my first impressions with good headphones, listening to the stereophonic experiments of Pink Floyd. Room-wise it took me into another Dimension. With DSD I was more getting there...
With PCM everything seems to be a bit more tidy.

It´s audible on both, phones and speakers.

What is better? I think it´s just another media, like Vinyl, Tape or ADAT etc., with it´s own sound. So now one can record on Tape, ADAT, PCM, or DSD and decide when to use which for what purpose. (I have even seen a Video about a recording session directly to lacquer cut...)

So my conclusion is, that, if I want to do analogue processing I consider using DSD as well.

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

35 (edited by KaiS 2020-12-18 00:06:55)

Re: Problems with setting sampling-rate, external clock and DSD recording

I'm out!

Matthias - when switching PCM - DSD in the AD-DA chain, do we get the same level?

Or is something broken, software-wise, like totally different frequency responses?
This would be the only explanation for such huge differences like described.

I'm too lazy this evening for doing a measurement.


Boris, it's my job all day long:
Doing A/B between the analog original, live played from musicians directly into high class microphones (real, not from recordings), and these signal monitored live through PCM.
I'm completely satisfied with standard sample rates.

36 (edited by borispietsch 2020-12-18 00:31:29)

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

Matthias - when switching PCM - DSD in the AD-DA chain, do we get the same level?

I also tried slightly different levels in both direction. (It´s only possible with the rear outs, switching between DSD direct (fixed level) and PCM.)
Hard to tell but seems to be no matter of levels.

KaiS wrote:

Or is something broken, software-wise, like totally different frequency responses?
This would be the only explanation for such huge differences like described.

Experienced it with two different devices, one is running V.97, the other V.98.

KaiS wrote:

I'm too lazy this evening for doing a measurement.

Yeah, you´re right. More important things to do first.
But curious I am...

KaiS wrote:

I'm completely satisfied with standard sample rates.

For most stuff I also work on 48/24 and it sounds good.
But when digitalizing Vinyl I always found 192kHz keeps more of the Spirit, especially when it comes to the sound of the crackles or very high flute tones for example. That´s why I was curious about even higher rates and the DSD option on my new device...
(Did´nt even know about it when I bought it. Thought 192kHz PCM is top of the line.)

If we are mastering for CD or regular streaming, it´s probably totally O.K. and not even noticeable if we stick to the conventional rates.

Despite of all of this I´m pleased to have met you online.
(Maybe we really get in touch when I´m in Kaiserslautern.)

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

37 (edited by borispietsch 2020-12-18 12:48:54)

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

Boris, it's my job all day long:
Doing A/B between the analog original, live played from musicians directly into high class microphones (real, not from recordings), and these signal monitored live through PCM.

Well, one last thought about this.
Maybe all of these differences become more obvious, if reprocessing already compressed material. As every recording results in one way or the other in compressing the source, wether it´s a digital or analogue process.
I´m just starting to work with analogue recording gear. Until now, each of my preamps did only work, by also digitalize the signal. Every processing of the material I did digitally. I wait on my analogue mixing console and already have a very good tape-machine which waits to be calibrated. With the console I will be able to compare the real deal. (Analogue microphone signals of a very high quality with their digitalization.)

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

38 (edited by KaiS 2020-12-18 13:28:21)

Re: Problems with setting sampling-rate, external clock and DSD recording

Everything is possible - but is it likely that degraded (be it ever so slightly) audio is more prone to further degradation of the same type than the original?


A hypothesis isn't more than that, unless it's proven by practical tests under controlled conditions, excluding other factors.

Confirmed by such a test, hypothesis becomes theory, one step below confirmed theory, until others reproduce your results.
Then, confirmed more and more, over time it becomes common knowledge or acknowledged fact, but still not written into stone forever.

That's science, everything else is ideas, that might or might not lead to somewhere.

I acknowledge you do hear something, but the reason for this needs to be seriously investigated.
First step is to exclude other factors when switching between PCM and DSD.
ADI-2 is not made to be a comparison machine, someone has to do some measurements before going on.


Sidenote about compression:
When I record classsical music compression is not used.
Even simple relative level changes are used only occasionally and sparsely, because they very much tend to make the result unnatural.

Re: Problems with setting sampling-rate, external clock and DSD recording

KaiS wrote:

Everything is possible - but is it likely that degraded (be it ever so slightly) audio is more prone to further degradation of the same type than the original?


A hypothesis isn't more than that, unless it's proven by practical tests under controlled conditions, excluding other factors.

Confirmed by such a test, hypothesis becomes theory, one step below confirmed theory, until others reproduce your results.
Then, confirmed more and more, over time it becomes common knowledge or acknowledged fact, but still not written into stone forever.

That's science, everything else is ideas, that might or might not lead to somewhere.

Yes, that is about common knowledge. In the end I will always have to refer to what I perceive with my individual senses and judge on them in any specific situation.
As a creative human being in sound processing I have to trust my ears more than technical data and their visualization, even if it´s helpful to take all of that into consideration.

KaiS wrote:

I acknowledge you do hear something, but the reason for this needs to be seriously investigated.
First step is to exclude other factors when switching between PCM and DSD.
ADI-2 is not made to be a comparison machine, someone has to do some measurements before going on.

Yes, even if they state, that now we are able to compare. (Manual 19.5)

KaiS wrote:

Sidenote about compression:
When I record classsical music compression is not used.
Even simple relative level changes are used only occasionally and sparsely, because they very much tend to make the result unnatural.

borispietsch wrote:

Maybe all of these differences become more obvious, if reprocessing already compressed material. As every recording results in one way or the other in compressing the source, wether it´s a digital or analogue process.

When I mentioned compression here it´s probably a bit missleading. I´m not talking about level-compression, but the compression of the original data in the process of recording it on a media.

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung

Re: Problems with setting sampling-rate, external clock and DSD recording

borispietsch wrote:
borispietsch wrote:
MC wrote:

Not to mention that with statements aka 'external clocks sound different' you won't find a big audience here. Technically this is indeed nonsense. Our SteadyClock FS completely (!) removes any influence the external 'very good' clock could have.

Well, I don´t have the FS Version. And my different external clocks sound indeed different, each, with the ADI-2 Pro (not FS).
Just a subjective observation so far, indeed. For a real blind-test I need somebody else to operate the unit. But closing your eyes while listening helps a lot. wink

Now I got an FS Version ADI-2 Pro here. Indeed it is more detailed in sound than the non FS Version. Still it sounds even a bit more clear with my external clock (even if this is technically impossible),

Edit: (It works only if SRC is off. So I suggest not to use SRC if you have a very good master clock.)

MANUFAKTUR DER KLÄNGE - Hörstücke | Musik | Filmvertonung