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: foobar2000's sound worse than mpc-hc's? (Read 5271 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

foobar2000's sound worse than mpc-hc's?

Since many years I watch my videos with mpc-hc (updated fork)
https://en.wikipedia.org/wiki/Media_Player_Classic#Updated_fork
And due to comfort I took it also for audios.
Although no (dsp-) expert, music is quite important to me
and so I use a good soundcard and headphone
(ASUS Xonar Essence STX, Sennheiser HD600).

Recently I tried foobar2000 and was astonished that
it sounds quite different. I would describe it as
"bath tube" equalized (more bass and trebles).
MPC-HC is more "harmonic" to me.
Of course I know that personal habituation may be
of importance... So maybe I got used to a "strange sound"?

I had to have a closer look; so made a technical test:
Used a flac 44.1/16, neutralized all settings
at driver, mpc-hc and foobar2000.
Played the audios and recorded it with Audition.
Then compared the diagrams (spectrum, freq.response,
spectr. balance and spectr. phasis) to the original.

The spectr. resonses are all about the same
(amplitude diff. < 0.01 dB @ 4kHz).
But (strange enough) the detailed spectrums show differences
to the original. Kind of "noise" in the higher frequ.
In which foobar2000 has more "noise" added than mpc-hc.
As the overall spectral responses are about the same
(amplitude diff. < 0.01 dB @ 4kHz)
it means moreover that it "smears" the frequencies.
The spectral balances and phasis especially of foobar2000
also differ clearly from original (mpc-hc not much).
If you save the enclosed images and switch between them
you see the difference.

Interpretation:

All in all foobar2000 indeed outputs the orig signal (at least for flac)
much worse than mpc-hc.
But can this be?
At last foobar2000 is said to be the most "honest"
audioplayer.
Many features of foobar2000 I like very much
(it fits to me being an "audiophil" perhaps? ;) )
But the most important thing is the sound, right?
And I have to admit I don't like its sound.
Now it turns out it's because it's disturbed...?

What is going on here?

Re: foobar2000's sound worse than mpc-hc's?

Reply #1 –
Addendum: the (optical) diff diagrams for the spectral balances...

Re: foobar2000's sound worse than mpc-hc's?

Reply #2 –
When using wasapi exclusive output, no DSPs, max volume, both players will output exactly the same data.

Description on how and what you actually recorded is vague. With such description, results of your experiment are useless. Extraordinary claims require extraordinary proofs. So you should to describe your experiment much more detail and precise if you want to someone to take it seriously
 

Re: foobar2000's sound worse than mpc-hc's?

Reply #3 –
Well, I used a pretty "normal" use case, the one I consider
most foobar2000's user will operate at.
Otherwise I would have said so.

- No wasapi exclusive output
- No dsp
- About middle volume

As you get identical data with your config, foobar2000's error
(and it is foobar2000 that diverges much more from orig)
must have to do something with my typical use case.

For my circumstances I provided much info for that.
If you're not concerned about my "useless" results,
then leave it as it is.
(Beside your reply is pretty unpolite).

Re: foobar2000's sound worse than mpc-hc's?

Reply #4 –
What's the samplerate of the soundcard in Windows?
If it's not the same as the source audio, resampling occurs. Foobar handles resampling on its own, whereas many other programs leave it to Windows. That's a possible reason for the difference. Otherwise there shouldn't be any (assuming everything is set up the same).

edit: Both recordings look quite noisy. Did you record them through an analog path? Because that can be affected by other things.
You can perform such tests digitally, with a loopback function.

Re: foobar2000's sound worse than mpc-hc's?

Reply #6 –
@Brand:
The samplerate of the soundcard is 44.1 kHz - same as the source audio.

"Analog path"
As I said, I'm no real expert on these things.
I tried to describe my test settings as best as I could
hoping that everything missing can be derived from that.
As I wrote in my first post:
"Played the audios and recorded it with Audition."
So this seems an analog path to me.
BUT... I might think that this a typical use case, isn't it?
And - as I already said - I did this with both players identical.
So analog or not: foobar2000's audio is much worse than mpc-hc's
in that config.

@mudlord:
You read my description and saw the attachments, didn't you?
My "subject opinion" was just the cause to have a
following closer "scientific" look.
(Very strange all this...)

Re: foobar2000's sound worse than mpc-hc's?

Reply #7 –
Well the moment you said

Quote
MPC-HC is more "harmonic" to me.
Of course I know that personal habituation may be
of importance... So maybe I got used to a "strange sound"?

rang alarm bells. I admit I didn't even glance at the diagrams as a result of that.

Re: foobar2000's sound worse than mpc-hc's?

Reply #8 –
I just tried to describe what kind of feeling the difference was to me.
Actually I'm very bad in naming what actual the difference is
(jangling, reduced or enhanced of frequencys, fine resolution etc.).
A friend of mine can tell to the point - but not me.
I just feel something is different.
OK.
But as I said - that was just the cause for further investigation.
I was astonished to be confirmed by them so obviously.

At last the point is:
If I hear music with foobar2000 with my config it actually is worse
than (even) mpc-hc. In a way that I even hear ("feel"  ;)  ) it.

I wanted to help with my feedback, and didn't expect to be smacked.

Re: foobar2000's sound worse than mpc-hc's?

Reply #9 –
I recommend you record them with a digital loopback instead. Audacity has that option (maybe Audition too). That should at least give you some info about what's going on.

Re: foobar2000's sound worse than mpc-hc's?

Reply #10 –
Could you post the samples you are using to compare?

I just tried this using Audacity (loopback) to record the 2 players at 44,1kHz 32bit float then exported as 44.1 16bit and then diffmaker to compare. Both players were at 100% volume and the original wave was a FLAC file converted to wave using Foobar2000.

I compared the original wave to the Foobar2000 output saving the difference. Then doing the same with the original wave and MPC-HC (with SaneAR audio renderer enabled). Then compared the Foobar2000 difference to MPC-HC difference and got this below using Spek.

Apart from the glitch at the start and the end there isn't really anything notable.

How were you volume matching the 2 players BTW?

Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

Re: foobar2000's sound worse than mpc-hc's?

Reply #11 –
Quote from: guju2022 on
- About middle volume
That's what your graphs look like too. The players use different volume scale. foobar2000 has a logarithmic scale that follows human hearing, MPC-HC has just linear scale. The output from foobar2000 is much quieter at the middle level and thus noisier in your graphs.

Quote from: mudlord on
Please provide proof in the form of ABX tests. Otherwise I have no choice then to not believe you. Scientific fact matters, not subjective opinions.
Please do not. This is not about lossy audio and subjective quality assesment. Proper measurements are the way to prove correct output capabilities. But the measurements have to be done correctly.

Re: foobar2000's sound worse than mpc-hc's?

Reply #12 –
The diffeence between our tests:

- I used a flac 44.1/24 as input (sorry for the mistake before (16) !)
- I didn't record in loopback
- I recorded in 16bit (but both cases however)
- my volumes were about middle (mpc-hc), lower quarter (foobar2000)
  (I matched the volume exactly (1dB) in Audition after recording, before comparing)

I forget to mention: My "system volume" is set to 49%. (For both cases however)

But... I got a "little" nervous now... ;) so I repeated my test.
Very carefully.
The result: the same as before (PHEW...!)

I suggest that you try with my config.
If you get similar results as I you can try to trap the cause expertly.
If you still get different results than I (For example near identical
outputs), than there is something strange going on at my PC
and the problem is with me only.

A note on the diagrams:
They are only qualitative of course (arbitrary colors/ lightness),
and not quantitative. But still they reveal a remarkable larger difference
to original if played by foobar2000 instead of mpc-hc.
And they confirm my hearing impression ("bath tube" freq. resp.).
I don't hear noise effects (but my ears my be insensitive for that).

Please try to reproduce with my config.
As this seems the more typical one anyhow (most users won't listen
to the music with volume set to 100%);
and moreover I'm not able to track down the cause expertly.

Edit (see above):
- I recorded in 16bit (but both cases however)

Re: foobar2000's sound worse than mpc-hc's?

Reply #13 –
I attach the audios (as splitted 7zip) - maybe it's of any use...

Re: foobar2000's sound worse than mpc-hc's?

Reply #14 –
Annotation to my spectrograms (1st post):

The frequencies above 20kHz are of no significance
and therefore shouldn't be taken into account.

I used linear depiction for the frequency;
better (more "natural") is probably log,
which provides interesting results.

Maybe the audio is not the best for test,
as to the middle part (9s to 28s) is too "mixed".

Re: foobar2000's sound worse than mpc-hc's?

Reply #15 –
Foobar and MPC-HC are supposed to sound exactly the same (excluding resampling).
If this isn't the case for you, there are pretty much only two options:
A) something is broken or not configured properly
B) you're testing it wrong

Before venturing into any further speculation, I would again suggest that you perform the digital loopback test. That's going to eliminate the soundcard as a variable and make troubleshooting easier.
I would also strongly recommend to keep both players at the exact same volume for testing, even if that's 100% in your case because lower values are harder to match.

Re: foobar2000's sound worse than mpc-hc's?

Reply #16 –
If I get you right...

It is no bug if foobar2000 gives worse results (than e.g. mpc-hc)
when played with volumes < 100%.

You are stating there is something wrong with me,
although no one (!) has even tried my typical config
- to confirm or to reject.
A bug in foobar2000 is impossible as a fact,
as it is audiophile, right.


I spent much effort in this (more for explaining
than for testing), but now it's enough.
I got the impression your effort is mostly to not
see the problem.

I'm sorry.

Re: foobar2000's sound worse than mpc-hc's?

Reply #17 –
I''ll try and replicate your results tonight.

So both players set to 100% volume and windows volume set to 49%?
Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

Re: foobar2000's sound worse than mpc-hc's?

Reply #18 –
I've had a quick play with your sample,

Setting both players to 100% volume and then setting the Windows volume slider to 49% shows that there is a massive volume difference between the players (as Case said).

for the MPC-HC recording the overall RMS = -13.2866 dB.
for the Foobar2000 recording the overall RMS = -21.289 dB.

After getting both recordings between -20dB to -22dB before normalising this is the result after letting diffmaker works it's magic.


Once again nothing of notable difference except the glitch that MPC-HC adds at the start of playback.

Maybe it's a case of louder is better?
Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

Re: foobar2000's sound worse than mpc-hc's?

Reply #19 –
Last bunch of tests by me.

I've tested the following players using the original file is the 44.1/24bit file provided by guju2022 in reply 13. Digital loop back was used in Audacity and saving the recordings as 44.1kHz 24bit wave (my previous tests were saved as 44.1kHz 16bit). All players were set to 100% volume as well as the windows volume was set to 100%.

Foobar2000 1.6.2 (WASAPI Shared)
MPC-HC 1.9.23 (WASAPI Shared)
Musicbee 3.4.8033 P (WASAPI Shared)
VLC 3.0.17.4 (Wave Out)
Winamp 5.666 (Wave Out)

Foobar2000 vs Original.

MPC-HC vs Original.

Musicbee vs Original

VLC vs Original

Winamp vs Original


Looks like the glitch that i blamed on MPC-HC is was incorrect, it was part of the recording process.
Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

Re: foobar2000's sound worse than mpc-hc's?

Reply #20 –
Quote from: guju2022 on
A bug in foobar2000 is impossible as a fact,
It certainly is possible. That's what I meant with "something is broken". But a difference in sound quality as you describe it is highly unlikely.
I am only suggesting that you be more thorough when testing.

Quote from: A_Man_Eating_Duck on
Setting both players to 100% volume and then setting the Windows volume slider to 49% shows that there is a massive volume difference between the players
There shouldn't be any difference. Make sure you don't have ReplayGain enabled. Also check the Windows volume mixer. There was a bug at some point where Foobar was stuck at a lower level.
And OP did not test the players at 100% volume, but at some approximate medium level (which is different between the two, because they don't use the same scale, as Case pointed out).

Re: foobar2000's sound worse than mpc-hc's?

Reply #21 –
This is what i see when i set both players to 100% volume and then decrease the windows volume down to 50%.

Replaygain is disabled and no DSPs active.

Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

Re: foobar2000's sound worse than mpc-hc's?

Reply #22 –
That's interesting. I tried to reproduce it, but I couldn't. They always get recorded at the same volume here.
Side note: Windows volume control in some cases affects the loopback recording volume while in other cases it doesn't. But that depends on your soundcard driver and not on the software player — at least I think it shouldn't depend on the player, assuming both are set up the same way.

I can think of two possible reasons:
1. The Foobar volume bug is still there somehow, hidden behind the scenes. I forgot the details of how that bug worked, but I would try again with a separate clean portable Foobar installation.
2. Something in MPC-HC is configured differently in your case. Try disabling "Enable built-in audio switcher" and "SaneAR Audio Renderer". It actually makes no difference for me, but you can try anyway just to check.

EDIT: You can also try switching to a different audio device in Windows.

Re: foobar2000's sound worse than mpc-hc's?

Reply #23 –
MPC-HC sound related settings in the Option pages:

Playback > Output > Audio Renderer:
"MPC Audio Renderer" and "SaneAR Audio Renderer" may utilize the built-in resampler (possibly Secret Rabbit Code / libsamplerate), WaveOut and DirectSound may use Windows resampler.

Everything under "Internal Filters":
Under "Audio Renderer", "MPC Audio Renderer Settings", one can set WASAPI shared / exclusive mode. Pay attention that MPC-HC won't show error if any of the exclusive settings fails, it may silently fallback to shared mode. Play audio in other software (e.g. browsers) to verify if exclusive mode really works or not.

Check Internal Filters > button "Audio decoder" too.

Some combinations of the above settings may trigger Windows limiter, while some other settings may clip the audio (lossy and float formats) before audio data entering Windows mixer / audio driver.

It is also not unusual that mono and multichannel (i.e non-stereo) contents exhibit differences in different software due to mixing behaviour.

In my case, I use "MPC Audio Renderer" and disabled "SaneAR Audio Renderer", with "MPC Audio Renderer" set to shared mode and "Audio Switcher" disabled. I am a 2-channel user.

Re: foobar2000's sound worse than mpc-hc's?

Reply #24 –
I followed the test procedure to the best of my abilities. Windows volume was set to 49%, MPC-HC volume was at 50% and foobar2000 volume was at -6.00 dB. I recorded the sample provided by OP in digital domain using Audacity's loopback option. Over here MPC-HC was quieter than foobar2000. I trimmed and normalized the recordings and after that the spectal frequency Adobe Audition shows for the tracks was identical.

I can't replicate foobar2000 outputting huge amounts of noise.

@guju2022 - are you sure you are not recording foobar2000 output with ridiculously low volume to 16-bits and then amplify it to match a louder recording? That would be one explanation for the extra noise your graphs show.