• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Playback of High Rez files

babadono

Active Member
Joined
Jul 8, 2021
Messages
173
Likes
120
I am ignorant on this subject but not stupid. I just recently got satellite internet installed so I can stream music. I installed a trial version of Tidal. I am trying to go out of my laptop over USB to a S.M.S.L SU-10 DAC. I am doing something wrong or have not got everything set up correctly. It appears I can only get 44.1k out of my laptop from Tidal. What do I need to check?
 
I am doing something wrong or have not got everything set up correctly. It appears I can only get 44.1k out of my laptop from Tidal.
Windows itself is set in audio settings to downsample everything to 44.1kHz... best to install;


Set ASIO4ALL as the default device in the sound settings and also in the Tidal app, or other media players, maybe control panel>devices for the DAC as well if needed.


JSmith
 
I am ignorant on this subject but not stupid. I just recently got satellite internet installed so I can stream music. I installed a trial version of Tidal. I am trying to go out of my laptop over USB to a S.M.S.L SU-10 DAC. I am doing something wrong or have not got everything set up correctly. It appears I can only get 44.1k out of my laptop from Tidal. What do I need to check?
The Tidal Desktop App has an exclusive mode, which will bypass Windows and send all tracks at their native sample rate. No ASIO4All required.

Two things to keep in mind:

1. Tidal is still full of messed up MQA files: https://youtu.be/48IPHc43M1k
Use other services for high fidelity streaming.

2. Sample rate does not matter: https://people.xiph.org/~xiphmont/demo/neil-young.html
 
Last edited:
ASIO4ALL will probably be a fix, but I agree with staticV3 - I would prefer to get the Windows WASAPI driver working properly.
I don't use TIDAL myself, but Google tells me you should do this -
Go to the TIDAL audio output settings, choose your DAC from the list of output devices, click on "more options," and then enable Exclusive Mode and Force Volume.
 
How can ASIO4ALL help at all? Last time I checked, which admittedly has been a while ago, all it did was to emulate ASIO interface on top of a WDM driver. In the old days when the WASAPI exclusive mode was unavailable it was a way to bypass Windows mixer resampling for DACs lacking ASIO drivers, but one had to use a player that could write to the ASIO interface. Does Tidal support ASIO? I don’t think so. WASAPI exclusive mode is the way to go with the Tidal desktop app, don’t bother with ASIO4ALL.
 
Wow.......thank you gents..I told you I was ignorant.....
 
Tidal audio output settings? where?
1737346936210.png



JSmith
 
Go to the TIDAL audio output settings, choose your DAC from the list of output devices, click on "more options," and then enable Exclusive Mode and Force Volume.
Where? nonesuch in the Tidal trial I downloaded 2 days ago
 
This is what Youtube shows me -
Launch the TIDAL app. Click on your profile initials in the top left corner > Settings > Streaming > Audio Output
 
All righty I finally found the settings that JSmith was pointing me to. I can select either use exclusive OR force volume but not both. When I chose exclusive I was able to cue up a track that my DAC showed was 48K and not standard CD 44.1K. But no output...then output from speakers in laptop (WTF?) then on laptop I toggled the USB SMSL sound device and then sound switched from laptop speakers to DAC and system speakers (WTF?) Criky mates this is clunky. Hopefully this switching to/from laptop speakers will go away and was just a first set up anomaly.
 
The Tidal Desktop App has an exclusive mode, which will bypass Windows and send all tracks at their native sample rate. No ASIO4All required.

Two things to keep in mind:

1. Tidal is still full of messed up MQA files: https://youtu.be/48IPHc43M1k
Use other services for high fidelity streaming.

2. Sample rate does not matter: https://people.xiph.org/~xiphmont/demo/neil-young.html
The argument put forth in "Sample rate does not matter" seems to be rather elaborately constructed around the notion that Hi-Res files contain audio recording and playback frequencies that directly correspond to the sample rate. The sample rate refers to the number of times per second the audio signal is sampled, not the audio frequency response.
 
The argument put forth in "Sample rate does not matter" seems to be rather elaborately constructed around the notion that Hi-Res files contain audio recording and playback frequencies that directly correspond to the sample rate. The sample rate refers to the number of times per second the audio signal is sampled, not the audio frequency response.
You might want to reread the sampling theorem proofs one more time…
 
If you are still on the trial version of Tidal, also try Qobuz for free. I had both and liked the music selection better on Qobuz, all CD quality or higher and no MQA. Tidal does have a connect feature but Qobuz says they are about to launch one too.
 
The argument put forth in "Sample rate does not matter" seems to be rather elaborately constructed around the notion that Hi-Res files contain audio recording and playback frequencies that directly correspond to the sample rate. The sample rate refers to the number of times per second the audio signal is sampled, not the audio frequency response.

The argument put forth is correct - sample rate directly translates to bandwidth. The frequency range that can be accommodated (perfectly) by any sample rate is exactly half that sample rate. Technically <1/2, so call it exactly half less 1Hz. :)

So 44.1kHz sample rate can reproduce frequencies up to 22.05 khz.
48kHz can go up to 24khz
96kHz can go up to 48kHz.

That is how digital audio works.


EDIT - I admit this is counterintuitive, and if you don't have the maths capability (as I don't) to follow the sampling theorem proofs, is difficult to grasp how this can be. I didn't "get it" (despite "knowing" it for decades) until watching the (so called) Monty video - a favourite round here.

If you're interested in getting a much better grasp - without any complex maths - about how sample rate and bit depth influence audio - I can post the link for you.
 
Last edited:
The argument put forth is correct - sample rate directly translates to bandwidth. The frequency range that can be accommodated (perfectly) by any sample rate is exactly half that sample rate. Technically <1/2, so call it exactly half less 1Hz. :)

So 44.1kHz sample rate can reproduce frequencies up to 22.05 khz.
48kHz can go up to 24khz
96kHz can go up to 48kHz.

That is how digital audio works.


EDIT - I admit this is counterintuitive, and if you don't have the maths capability (as I don't) to follow the sampling theorem proofs, is difficult to grasp how this can be. I didn't "get it" (despite "knowing" it for decades) until watching the (so called) Monty video - a favourite round here.

If you're interested in getting a much better grasp - without any complex maths - about how sample rate and bit depth influence audio - I can post the link for you.
If you accept that any signal is a sum of sines by Fourier transformation, it is fairly easy to grasp the bandwidth restriction. Any sampled sine could have several frequencies other than its actual frequency due to its cyclical nature. If the lowest of those is f and the sampling frequency is fs, than it can have frequencies f, fs-f, fs+f, 2fs-f, 2fs+f etc. At a frequency of fs/2, the sampled values are always the same pair, which makes amplitude and phase ambiguous. To be able to reproduce the sampled signal, the bandwidth has to be restricted such that only one sine at one frequency, at one amplitude and with one phase is generated per sampled sine, which is done by limiting bandwidth to below fs/2.
 
If you accept that any signal is a sum of sines by Fourier transformation, it is fairly easy to grasp the bandwidth restriction. Any sampled sine could have several frequencies other than its actual frequency due to its cyclical nature. If the lowest of those is f and the sampling frequency is fs, than it can have frequencies f, fs-f, fs+f, 2fs-f, 2fs+f etc. At a frequency of fs/2, the sampled values are always the same pair, which makes amplitude and phase ambiguous. To be able to reproduce the sampled signal, the bandwidth has to be restricted such that only one sine at one frequency, at one amplitude and with one phase is generated per sampled sine, which is done by limiting bandwidth to below fs/2.
It is relatively easy to see how bandwidth is limited to no more than fs/2 - it is less intuitive to grasp how a signal just below fs/2 can be perfectly reproduced with only 1 sample per half cycle.
 
It is relatively easy to see how bandwidth is limited to no more than fs/2 - it is less intuitive to grasp how a signal just below fs/2 can be perfectly reproduced with only 1 sample per half cycle.
Because there are more than one samples per half cycle if only ever so slightly. This allows the sampled value to vary ever so slightly across every two samples removing any ambiguity.
 
Because there are more than one samples per half cycle if only ever so slightly. This allows the sampled value to vary ever so slightly across every two samples removing any ambiguity.
Of course. I get that is how it works. I've seen it demonstrated, and have demonstrated it myself. I can understand that there is only one wave form within the band limit that can fit. I can analogise it to drawing a perfect circle with only two points.

But to the layman : Everyone who has been told how digital audio works at the very basic level has (at best**) been told that that you interpolate to join up the dots. How you do that with just 1 point above the line and one point below the line, and get to a perfect sinewave is, as I stated, counterintuitive.


** At worst, they've simply been told "straight lines" or even worse "stair steps"
 
Last edited:
Back
Top Bottom