24/352.8 vs. 16/44.1 - an attempt at a comparison

[mansr]

Here's a quick analysis. To start with, look at the spectrum of the DXD source:

Just above 22.05 kHz there is a very small amount of analogue noise. This is followed by a fairly flat noise floor up to 44.1 kHz. Beyond this, we have increasingly strong modulator noise from the ADC, reaching -100 dB/Hz at top end. It is safe to say that this music would have been captured perfectly at 88.2 kHz. Dropping the sample rate to 44.1 kHz loses nothing of value.

 

Now add the spectrum of the 16/44.1 source file along with 16-bit TPDF dither for reference:

The curves are clearly diverging above the centre, the 16/44.1 version being higher in level. Plotting the power ratio shows this more clearly:

 

Moving on to the captures, their spectra show the same diverging trend, albeit to a smaller extent:

And the power ratio of the two captures:

 

Finally, the power ratio of (what I assume is) the DXD capture to the source:

The bulk of the music content is about 0.2 dB lower in the capture, while at high frequencies the noise is much stronger.

 

The only anomalous thing here is the discrepancy between the DXD original and the 16/44.1 version. A plain downsample shouldn't differ that much. I wonder if perhaps they intentionally boosted the highs by a few dB for some reason.

 

As for the captures, the ADC has possibly been a limiting factor here. I would have used a 24-bit ADC at a higher sample rate to make certain nothing was missed.

[mansr]

7 hours ago, PeterSt said:

[impossibility of sample alignment]

 

Watch out please: I am not saying or suggesting that this explains the differences mansr pointed out.

The graphs I posted were based on the average power spectrum and thus don't rely on sample alignment.

[mansr]

33 minutes ago, manisandher said:

13 hours ago, mansr said:

he only anomalous thing here is the discrepancy between the DXD original and the 16/44.1 version. A plain downsample shouldn't differ that much. I wonder if perhaps they intentionally boosted the highs by a few dB for some reason.

How strange. Why on earth would they do this deliberately? I'm tempted to do my own decimation from 24/352.8 to 16/44.1 to see if I could do a more accurate job.

Never mind. It's the 16-bit dither that causes the discrepancy. At the high frequencies, the signal is so low in level that the dither becomes significant in comparison.

[mansr]

Hey goldenears, how does this one sound compared to the original DXD and 2L's down-conversion?

2l-092-1644.flac

[mansr]

16 minutes ago, sandyk said:

2 hours ago, mansr said:

Hey goldenears, how does this one sound compared to the original DXD and 2L's down-conversion?

2l-092-1644.flac 3.44 MB · 2 downloads

Message from Foobar 2000

Quote

Could not load info (Unsupported file format) from:
"K:\2L-139_01_stereo_01.flac.flac24"

So you have a broken file. What does that have to do with the one I posted?

[mansr]

7 minutes ago, sandyk said:

Perhaps the naming of the file as shown has caused the problem ?

 I will try renaming it.

 

 O.K.  Getting rid of the ".flac24" in the name  worked.

Still not clear why you're mentioning a 2L-139 track.

 

7 minutes ago, sandyk said:

 Do you have the link to the version you want this to be compared with ?

DXD original: http://lindberg.no/hires/test/2L-092/2L-092_stereo-DXD_01.flac

16/44.1 conversion: http://lindberg.no/hires/test/2L-092_01_44kHz_16b.flac

[mansr]

2 minutes ago, sandyk said:

 I was referring to your post 32, which perhaps you meant for Mani's consumption only ?

That file is for anyone to play (with). Note that it is named 2l-092-1644.flac whereas the one you complained about was 2L-139_01_stereo_01.flac.flac24. Either you have a habit of randomly renaming downloads, or you got the files mixed up.

[mansr]

6 minutes ago, sandyk said:

Tom

As I said, it's not worth stuffing around with. I never have such problems with any of the large numbers of .flac files that I download from elsewhere. With the other file I was able to successfully rename it, but not this one.

I also had to muck around renaming the other files from Mani.

Maybe you should try downloading them again using a better power supply.

[mansr]

44 minutes ago, manisandher said:

Just compared your version to 2L's DXD and 16/44.1.

 

The DXD, played back with no upsampling, filtering or SDM, remains the clear winner for me. It simply sounds the nicest and the easiest to listen to.

 

Against 2L's 16/44.1, yours sounds softer, slightly fuller, almost as though the transients have been flattened a little. Not sure which I'd take given a choice. But they do sound audibly different to me.

Thanks for checking. I used shaped dither, lowering the noise by 10 dB below 15 kHz. Like this:

[mansr]

Just now, manisandher said:

Did you apply this to the DXD during decimation, or directly onto the 16/44.1?

I downsampled the DXD to 44.1 kHz using 32-bit precision, then applied the shaped dither at 16 bits before saving the result. That's how these things are always done.

 

The 16/44.1 file provided by 2L doesn't appear to use shaped dither.

[mansr]

2 hours ago, manisandher said:

I could (and still can) hear a clear difference between the original DXD file (played back with no further upsampling, filtering or SDM, which 99.9% of DACs on the market will perform, to varying degrees), and 2L's down-converted 16/44.1 file (played back with 8x upsampling and filtering, but still no SDM).

 

If there were any point to this thread, it might be to figure how the two could sound different, considering there's very little signal above 10kHz. I decided to share a couple of 16/44.1 captures to help.

Do you suppose it could be the bit depth difference that's relevant here? Have you tried a 24/44.1 downconversion?

[mansr]

18 minutes ago, manisandher said:

You comparisons of the two captures look incredibly similar:

If they didn't, at least one of us would be doing something wrong.

 

18 minutes ago, manisandher said:

I know @pkane2001 is using his newly developed DeltaWave software, but what about you @mansr? 

I'm using Octave.

[mansr]

7 hours ago, Rexp said:

Does your 16/44 down conversion sound the same as the DXD? 

Listen and decide for yourself.