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: Which audio track is better and why? (Read 2780 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Which audio track is better and why?

So I've converted two dsd files to flacs with foobar2000 and they seem different and i was wondering which one is better and which one should i keep? I've also noticed that the left one has more noise at the high frequencies. How does that affect sound quality and also how do the subtle lines on the right one affect it. Thanks for the responses!!!

Re: Which audio track is better and why?

Reply #1 –
Choose the one that sounds better!!!  ;)

Or, you may not hear a difference at all...   Just because you an see a difference doesn't mean you can hear a difference.   Listen with your ears and if you don't hear any difference don't worry about it.

I wouldn't worry too much about the "radio frequency" information, except sometimes  if it's too strong and not filtered-out by the DAC filter, it could "cause problems" with your amplifier.  

Re: Which audio track is better and why?

Reply #2 –
Do you actually have a 352kHZ DAC?  If not, all that ultrasound will be cleaned up on playback anyway and can be ignored aside from the fact that it is probably ruining your compression ratio in FLAC.  If you do have a 352 kHZ DAC, and it isn't removing that noise, it might be a bad idea to leave it in since you're going to be heating up your tweeter and maybe introducing additional distortion. 

Re: Which audio track is better and why?

Reply #3 –
From everything I've read, those graphs make it look like the first one is missing some of the standard filtering that gets used in DSD playback.
I'm no expert, so I may be wrong.
Processed audio in java and python.

Re: Which audio track is better and why?

Reply #4 –
Test your hearing: Online Tone Generator  ;)

My battered old raver ears get ~14000kHz.  ???

And as has alreayd been said: always worth using your ears to check if there is any difference. If not, then I personally would go with the smaller file (unless you're going to use the audio for other purposes, like mastering, editing etc).

Re: Which audio track is better and why?

Reply #5 –
The thing is - If these are different performances/recordings, different mixes, or different masters, etc., you may prefer one over the other regardless of what the spectrum looks like.

Re: Which audio track is better and why?

Reply #6 –
If you mean that those are two different dsd files from the same song, which you have converted to flac, then it seems that the first one is a true dsd file and the second one is a post-processed file.

First one looks quite correct, but as explained, the output has not been filtered.
The file is as-is-on-dsd, so it contains high-intensity noise on the second half of the spectrum. This has to be filtered.
Resampling the flac file to 96Khz with a good resampler ( so max frequency is 48Khz) would probably be the best option.  (Resampling it to 192Khz would still leave noise at 70Khz, although at that frequency is below -100dB so it is also a possibility).

I would not keep the one from the right.  Look at the range from 0 to 0.10 around the 20Khz.  The one at the right has some sort of duplication of the content there.  (somehow visible at the whole track around that frequency)

Aside of that, the lines that go up to 40Khz or more could indicate some kind of clipping or distortion happening at that point on the whole spectrum, not just where you see them.

Re: Which audio track is better and why?

Reply #7 –
I'd say that DSD's increased resolution of low frequencies is more of a merit than its high sampling rate. After decoding, that high freq content is probably DSD's inherit dithering which supposedly increases bit resolution. Why do you care about DSD? Because of low resolution high frequencies which you can't hear? Or because of probably inaudibly increased resolution of low frequencies? Fit decoded parameters accordingly. Though as already said such strong HF content may be dangerous.

Re: Which audio track is better and why?

Reply #8 –
Inherent to DSD is the substantial amount of quantization noise.
That is why the standard says to use a low pass filter.
If one play DSD, the gear used will recognize it and apply the filter.
If you convert to PCM manually without applying the filter,  all quantization noise will be in the PCM and of course, your gear won’t filter this as it sees PCM.
If the first file is the unfiltered one and the second the filtered one, 2 is your best bet.
TheWellTemperedComputer.com

Re: Which audio track is better and why?

Reply #9 –
Quote from: Roseval on
Inherent to DSD is the substantial amount of quantization noise.
Which works like dithering. I tend to think that DSD is one big dithering...

Re: Which audio track is better and why?

Reply #10 –
It doesn't
Dither decorrelates the LSB
The noise of DSD is something completely different


TheWellTemperedComputer.com

Re: Which audio track is better and why?

Reply #11 –
What do you mean by LSB, least significant bit? Which one is it in 1-bit DSD? ;)

Re: Which audio track is better and why?

Reply #12 –
Talking PCM of course.
DSD is a non editable format, you can't apply any kind of DSP to it, including dithering.
TheWellTemperedComputer.com

Re: Which audio track is better and why?

Reply #13 –
Quote from: rutra80 on
Quote from: Roseval on
Inherent to DSD is the substantial amount of quantization noise.
Which works like dithering. I tend to think that DSD is one big dithering...

I think the term you're looking for is "noise shaping" (causing the spectrum of the quantization noise to be non-white).   There is no dithering present here. 

Re: Which audio track is better and why?

Reply #14 –
Nope, I mean dithering. And that when converting from DSD to PCM, there is inherent noise, which works like dithering.

Of course there is no technically applied proper dithering anywhere. That noise is a natural byproduct of conversion from DSD, which by its nature is "one big dithering".

Re: Which audio track is better and why?

Reply #15 –
Quote from: rutra80 on
And that when converting from DSD to PCM, there is inherent noise, which works like dithering.

No. The noise there is called quantization noise, and it does not work like dither since it is correlated. This is one the downsides of DSD.

Re: Which audio track is better and why?

Reply #16 –
Delta-sigma modulation is related to dithering in that it utilizes a form of noise to retain audible dynamic range in a higher bit rate, and as a result typically yields similar results to a shaped dithered signal when downsampled. But it is not dithering in that DSD is so correlated that a dithered signal trips up the feedback loop and distortion can be heard under certain conditions—something only alleviated by attenuation prior to A/D or increasing the bit depth. It's not a format I would recommend using.