Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Fred H]

Running 31.0.43 (64 bit) on a fully patched Win 10 PC. The playback device is JRiver Media Center as I'm running the WDM Driver. MC's Audio device is Headphones (2- TX-384Hz Hifi Type-C Audio)[WASAPI], connected by USB. It's currently set to 32-384, although when I tried 24-384 it made no difference. I tried it without WDM and had the same problem.

Edited to add: It works fine for greater than 88.2, e.g. 24-96 and 24-192 both work fine.

The physical device is a Loxjie-A30 desktop DAC/amp, which accepts up to 32-384. MC's output format is "No change" to the A30's limit and either 352.8KHz or 384KHz above that.

Neither 88.2 format will play on my PC, message is "Automatically configure output (will close in 10 seconds)", defaulting to 48KHz. 24-88.2 played in the past (before 31, 16-88.2 was created for testing only). Our IdPi (running 30.0.96) plays 24-88.2 without any problem to two different linked DACs.

[Fred H]

Bump. I'm wondering if this is a bug; if so, i think I posted it in the wrong place.

To find out, would someone please play a 24 or 16 bit 88.2KHz flac file in 31.0.43 and tell me if it works for you? If you don't have one and are willing to test, please let me know and I'll post one for download.

TIA, Fred

[bob]

Tried with the same setup, just a different DAC. Worked fine.
Make sure output format is off and Auto Configure... (Under audio advanced) is set to no.

[Fred H]

I disabled output format and set Auto Configure to no. Still didn't work. Closed and reopened. Same result.

24-192 plays ok; just not 88.2. Should I reinstall?

[bob]

I disabled output format and set Auto Configure to no. Still didn't work. Closed and reopened. Same result.

24-192 plays ok; just not 88.2. Should I reinstall?

That shouldn't make any difference.
Do you have exclusive mode set? (audio->settings)?
Is the WDM driver out of the picture (that changes things)?

[whoareyou]

If you look in the windows sound panel, does the driver support 88.2? 

[Fred H]

Exclusive mode is set.
Same behavior with and without the WDM driver.
Windows sound panel supports up to 32-384.

[bob]

Exclusive mode is set.
Same behavior with and without the WDM driver.
Windows sound panel supports up to 32-384.

Up to doesn't tell the story.
There are plenty of devices that don't support 88.2

[Fred H]

Re: "Up to doesn't tell the store.
       There are plenty of devices that don't support 88.2"
I didn't know that; thank you.
The driver lists: 44.1, 48, 96, 192, and 384.
In Output Format I set all multiples of 44.1 to 44.1, and now 88.2 plays (as 44.1).
Thank you again.
As my driver is dated 2019 I'll look for a newer one and see if it offers more choice. But I'm happy with this outcome.

[Awesome Donkey]

For 88.2 kHz in Output Format, it might be better to set it to resample the sample rate to the closest one that's supported for your device, which for 88.2 kHz would be 96 kHz versus downsampling to 48 kHz or 44.1 kHz.

[blgentry]

For 88.2 kHz in Output Format, it might be better to set it to resample the sample rate to the closest one that's supported for your device, which for 88.2 kHz would be 96 kHz versus downsampling to 48 kHz or 44.1 kHz.

What is your reasoning on that?

Brian.

[bob]

I usually just set unsupported sample rates to the closest lower power of 2.
Just intuition but I'd be interested in seeing other reasoning.

[Awesome Donkey]

What is your reasoning on that?

Forewarning, this is a personal opinion and I know it likely isn't going to matter either way. For me it's about maintaining (at least as much of the) audio data by upsampling instead of downsampling and potentially losing any audio data. My reasoning for this is, while (for example) it's unlikely for most digital downloads of "hi-res" music to have any viable worthwhile audio data above 20 kHz (not to mention we can't hear it), on the off-chance there actually is something there I'd rather upsample and maintain all of it or as much of the audio data than downsample and potentially lose data... again, assuming there's actually anything there other than inaudible noise... which that's all there likely is. That's why I favor upsampling than downsampling in such scenarios.

999 times out of 1,000 it probably isn't going to matter. Probably 9,999 out of 10,000 actually.

[Awesome Donkey]

I usually just set unsupported sample rates to the closest lower power of 2.

Yep. For upsampling 88.2 kHz it'd probably be better to go 176.4 kHz instead of 96 kHz (though IMO it's a waste). And to be fair downsampling to 44.1 kHz is just as good a solution too. I guess it's an OCD thing with upsampling and wanting to avoid any sort of data loss... at least as much as possible. You likely can't go wrong and it doesn't matter either way though.

[blgentry]

I can't say I agree with your reasoning, but now I understand where you are coming from.

88.2 is unusual, but I think most of these files are designed to be "higher quality" than the standard 44.1 offering.  I would personally want to play them on a DAC that supports 88.2. If not that, then either at 176.4, or at 44.1. 

The biggest advantage of 176.4 here is that the brick wall filter in the DAC is way above human hearing so is should not introduce phase shift in the audible range like a brick wall filter in a 44.1 DAC would.  On the other hand, the majority of DACs today are Delta Sigma so these considerations probably don't make any difference any more.  My main DAC at home is one of the very few that are not Delta Sigma.

Brian.

[JimH]

Why would 2.0001 x be worse than 2.0 x?  Math is math.  Isn't it?

Try the listening test we made a while back:  https://yabb.jriver.com/interact/index.php?topic=107924.0

[bob]

Since the numbers are stored as integers the math is really simple dividing or multiplying by a power of 2. You either take every other sample or double each sample.
His device only supported 44.1k not any other multiples of that so he can't up-sample to 176.4k

[JimH]

I see the logic, but computers don't struggle with math.

[blgentry]

I see the logic, but computers don't struggle with math.

It's not a matter of speed or anything.  It's more a matter of accuracy.  Extremely badly done sample rate conversions produce easily audible artifacts.  If you've ever heard unusual tinkling or rushing sounds in the background of some sound sources, you've heard bad sample rate conversion.  I've heard this on the radio dozens of times.  Not on songs.  Usually on interview footage or something else from a digital source that got poorly converted.

MC doesn't do bad conversions.  It does high quality conversions.  But these kind of sample errors are what you are trying to eliminate by doing integer based conversions.

Brian.

[JimH]

With a 64 bit audio path, there's precision to burn.