Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: What is the the right sample rate and bit depth to play my music in my (Read 12460 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

What is the the right sample rate and bit depth to play my music in my

Hi,

I’m new to hi-definition PC audio and came here hoping to clear out some doubts about the right sample rate and bit depth to avoid issues that people with better knowledge describe as jitter, bit perfect, perfect timing while playing my music in my local library and Tidal.
I have three questions:
(1) Is it fine for a audio player software to lock the sample frequency of the DAC?
(2) Should I lock the sample frequency to the higher frequency available on my OS audio device (384 kHz)?
(3) Should I favoured my MAC instead of using my PC Windows to get the best possible audio quality, just because in my MAC I have Sampling frequencies up to 352.8 kHz and 384 kHz and Windows up to 192 kHz?

Some background:
I’m using Oppo Ha-1 DAC on Windows (installed the Oppo ASIO driver) and Mac (it doesn’t need driver). Below some specs listed in the Oppo Ha-1 website:
PCM Sampling Frequencies: 44.1 kHz, 48 kHz, 88.2 kHz, 96 kHz, 176.4 kHz, 192 kHz, 352.8 kHz, 384 kHz
PCM Word Length: 16-bit, 24-bit, 32-bit

Testing:
After installing the DAC, I started my tests on MAC and Tidal HIFI. The DAC showed the PCM Sampling Frequency of the audio file being played. Sometimes it showed PCM 352.8 kHz.

Then I installed a trial version of Amarra, and I noticed that it immediately locked the Sampling Frequency that shows into the DAC to 44.1 kHz. Now, I can’t go back to the variable frequency. In Amarra settings I can lock it on any sampling frequency up to 384 kHz.
This lead to my first two questions listed above: (1) is it fine for a program to lock the sample frequency? and (2) should I lock it to the higher frequency (384 kHz)?

Then, I did some testing on Windows. Here, the same Tidal songs that showed PCM 352.8 kHz on MAC were played as PCM 192 kHz. So I went to the Oppo playback device properties, and noticed that I can choose from 44.1 kHz up to 192 kHz, for both, 24 and 32 bits, but not 352.8 kHz, 384 kHz.
And this lead to my third question: (3) Should I favoured my MAC instead of using my PC Windows to get the best possible audio quality, just because in my MAC I have Sampling frequencies up to 352.8 kHz and 384 kHz and Windows up to 192 kHz

I hope I made sense (and I that haven’t broken any forum rules!)

Any advice will be appreciated,
Juan

What is the the right sample rate and bit depth to play my music in my

Reply #1 –
None of that makes any difference at all.

What is the the right sample rate and bit depth to play my music in my

Reply #2 –
Right. So why OS's allow choosing Sample Frequencies and audio software locking them?

What is the the right sample rate and bit depth to play my music in my

Reply #3 –
Quote from: blueoris on
Right. So why OS's allow choosing Sample Frequencies and audio software locking them?


Different applications, different reasons. 

In certain radio applications for example, anything below 176.4 kHz cannot capture a MPX encoded signal off an analog FM radio channel nor will it be able to feed it the transmitter without issues or some cases special equipment.  That's has to with some rather high frequencies such as the stereo subcarrier and RDS.  Please no one rip on me for this.
Do note that 32KHz @ 16-bits would be perfect for any signal captured off the air with a consumer FM radio, the stuff talked about above refers more to broadcasting where the PC is doing the work of otherwise more expensive equipment and broadcasting analysis software to check for certain things that may result in violation of your jurisdiction's local regulatory laws.

99.9% of the time it makes no difference in listening circumstances with any rate above 44.1 KHz.  Although some sound cards work better with multiples 48 KHz where others are not even affected.  44,100 Hz @ 16-bit is CD rate and is already better than human ears.

For recording 24-bit can be helpful as it allows you to record at a lower volume level without clipping the top and bottom off the waveform.  It's then possible to further process the audio more, before dropping it to 16-bits.  Depending on the noise floor of your sound card, 24-bits can be very useless if your sound card or onboard has a poor noise floor.

For recording, the highest frequency that can be recorded is about half the sample rate.  This is called nyquist frequency.  Anything above it when recording should be filtered out.  Anything resampling it should also filter those frequencies.  The original frequencies that were filtered out will be never recovered and you never heard them anyway assuming the nyquist frequency is above the upper lower limits of human hearing.  As far as playback is concerned there is nothing to filter out unless you're resampling the audio itself to a rate different then what the audio is recorded or stored at.

For playback sometimes a different device that is external usually a digital connection to a different device via SPDIF or HDMI only takes certain rates and depths.  An operating system that allows you to set this to what it is supported is more useful than one that does not.

Most modern hi-definition PC sound cards, do not allow you to go lower than 44.1 KHz @ 16-bit at least none that I've seen of on Windows 7.

Drivers or software may prevent changing of the sampling rate to keep things from going bad (crashing or stop working correctly, etc.) when any endpoint is in use by running applications on the computer itself.  Usually closing all applications that is using a certain endpoint may free it up.  If an application you installed doesn't allow any changes of this after closing it, then it maybe time to get rid of it, if you do not need it at all.  Finally, check to see if you anything is ticked or checked to allow listening of any inputs analog or digital through the computer itself, this may prevent changing of the sample rate.

Most modern operating systems allow to specify a common rate at which all audio is resampled to when shared with other applications, this is called shared mode in modern Windows.  Accessing the endpoint directly only allows one program to output sound.  This is often called under modern Windows as WASAPI Exclusive mode bypassing the settings for shared mode.

What is the the right sample rate and bit depth to play my music in my

Reply #4 –
Quote from: blueoris on
Right. So why OS's allow choosing Sample Frequencies and audio software locking them?


Lets say you have audio files ripped to CD resolution (16-bit/44.1kHz) and you want to output these through S/PDIF to your DAC which supports only 24-bit/48kHz/98kHz formats. What do you need to do to get the files played if OS doesn't support this feature?

What is the the right sample rate and bit depth to play my music in my

Reply #5 –
To me, anything above 16b/44.1 is nonsense. I cannot hear any difference.

Sometimes I get my hands on an album that is recorded in HD audio 24b/192. First thing I do is to convert it using Foobar/SoX to 16b/44.1 . There is absolutely no reason to use sample rates above the human hearing. The limits of human hearing are known for decades and the only thing that is sure about that is that it degrades every day you get older.

Young people today lose their ability to hear higher frequencies even faster because of the clubs/disco/earpods these days. Especially the last one. They are listening to loud music all day with in-ear monitors. That's absolutely killing!

Unfortunately modern CD's are recorded so badly.... People with good hearing prefer HD records. Not because HD is better, but it is (sometimes) not smashed down to a DR rate so low that you cannot even call it music anymore.

What is the the right sample rate and bit depth to play my music in my

Reply #6 –
Quote from: blueoris on
Some background:
I’m using Oppo Ha-1 DAC on Windows (installed the Oppo ASIO driver) and Mac (it doesn’t need driver). Below some specs listed in the Oppo Ha-1 website:


Quote from: blueoris on
Right. So why OS's allow choosing Sample Frequencies and audio software locking them?


Its not the OS, its the specific piece of software you installed that allows this. If you use directsound or a normal audio API, the OS will automatically handle all of this for you.

What is the the right sample rate and bit depth to play my music in my

Reply #7 –
Quote from: Fairy on
To me, anything above 16b/44.1 is nonsense. I cannot hear any difference.

I never could either.  But in some applications I have had to use higher rates.

Quote from: Fairy on
Sometimes I get my hands on an album that is recorded in HD audio 24b/192. First thing I do is to convert it using Foobar/SoX to 16b/44.1 . There is absolutely no reason to use sample rates above the human hearing. The limits of human hearing are known for decades and the only thing that is sure about that is that it degrades every day you get older.

I leave this as is.  On the computer I do real time resampling by either letting the operating system do it or the software program do it.  Any portable player will have 44.1 KHz in either 16 or 24-bit depending on the original bit depth.  If additional space is there I have the originals hidden on the portable player marked for computer use only.

Quote from: Fairy on
Young people today lose their ability to hear higher frequencies even faster because of the clubs/disco/earpods these days. Especially the last one. They are listening to loud music all day with in-ear monitors. That's absolutely killing!

Unfortunately modern CD's are recorded so badly.... People with good hearing prefer HD records. Not because HD is better, but it is (sometimes) not smashed down to a DR rate so low that you cannot even call it music anymore.

I'm young and I have good hearing.  I've notice the majority of CDs released are loud these days.  Sucks.  I've been able to find modern stuff that wasn't squashed on non-HD formats.  The average person just doesn't give a shit.  As long it works they enjoy their stretched images on HDTVs and horribly configured and hooked up sound systems.  Can't expect people to care or to be experts.  The lowest common dominator is the ones that expect stuff loud but don't care to understand the consequences of it anyway.  On the flip side there is some people that turn everything way down thinking they'll burst their eardrum or other such nonsense and refuse to turn it up when the material is clearly more quiet and dynamic and might even complain if the material is too dynamic, where others will ride the volume knob.

What is the the right sample rate and bit depth to play my music in my

Reply #8 –
Quote from: Fairy on
Young people today lose their ability to hear higher frequencies even faster because of the clubs/disco/earpods these days. Especially the last one. They are listening to loud music all day with in-ear monitors. That's absolutely killing!


Can't stress this enough. I can barely hear my headphones 2 meters away when they play music at my normal listening volume in my dead silent bedroom at night, yet these people are playing their Apple earbuds so loud that I can easily hear their noise 3 meters way with me wearing IEMs in an already noisy train! That's insane.

What is the the right sample rate and bit depth to play my music in my

Reply #9 –
Quote from: Palladium on
Quote from: Fairy on
Young people today lose their ability to hear higher frequencies even faster because of the clubs/disco/earpods these days. Especially the last one. They are listening to loud music all day with in-ear monitors. That's absolutely killing!


Can't stress this enough. I can barely hear my headphones 2 meters away when they play music at my normal listening volume in my dead silent bedroom at night, yet these people are playing their Apple earbuds so loud that I can easily hear their noise 3 meters way with me wearing IEMs in an already noisy train! That's insane.


The hipsters deserve it, too.  Oh right, because earbuds get super loud easy. 

I remember an idiot from high school that stole a pair and wasn't allowed back in the store and almost faced criminal charges, always thought he was better than everyone else, too. 

Sarcasm Alert: Because Apple makes such nice headphones, that no other brand can compete.

What is the the right sample rate and bit depth to play my music in my

Reply #10 –
While you have to give Apple credit for including some really decent stock earbuds, they'll always remain just that, earbuds - a concept that is just inherently flawed. Acoustic cancellation drives mids and bass distortion way up and makes for lousy bass response, and they isolate about as much as your average sieve. Both factors combined make for first-rate hearing killers in loud environments. Whenever I see (and potentially hear) someone with an Apple device and stock buds on a bus, I can't help but feel pity for them.

Sennheiser used to sell earbuds with silicone sleeves that were intended to provide an airtight seal (which would eliminate acoustic cancellation), not sure how well that worked. There's definitely a market for cans thare are not closed for sports use. With typical closed in-ears, amplification of internal noises would quickly drive you mad.

Anyway, yes, 16/44 technically is just fine as a distribution format. Frequency response up to at least 20 kHz (ample for most humans), effective dynamic range well in excess of 100 dB if shaped dither is applied. You'll probably want to output in 24 bit or at least enable dithering in order to avoid losing resolution when, say, applying ReplayGain and thus otten turning things down (which has the added benefit of usually providing ample headroom for any intersample-overs or lossy-codec-related overshoot), and if you don't trust the DAC's interpolation lowpass filter, you can always include a high-quality software resampler to whatever the highest supported sample rate is (the SoX resampler for Foobar seems to be well-optimized, even a years-old Core 2 Duo hardly shows any additional CPU load at all). But that's minutiae, having more to do with playing it safe and ensuring textbook-level performance under any circumstances.