Help - Search - Members - Calendar
Full Version: Winamp equalizer
Hydrogenaudio Forums > Hydrogenaudio Forum > General Audio
vladimirovich
Well. This is quote from winamp 2.90 history:

Winamp 2.90:
* updated PCM equalizer with code from 4Front Technologies/George Yohng. Quality is a lot better.


This is test settings:

user posted image

Generally, those settings shouldn't change sound at all. But, what we see in freq graph?

user posted image

So, "a lot better" doesn't mean perfect at all. Well, if you want you could do this tests by yourself.
siokaos
Good post. Just curious... What freq. analysis program are you using? I've toyed with labview a bit... but it seems there's much easier software out there. Can anyone help?

I'm also looking for a good realtime FFT analysis, perhaps for all DirectSound, but a winamp plugin will do... I really hate 'DSP Spectrum Tool' for winamp, and I'm a fb2ker now, anyway =)


Thnx.
Lev
What I've never quite understood is why does that 'equalizer' (a bad name for it if ever I heard one) react badly with MP3's and absolutely terribly with MPC's? A long time ago, I left the equalizer on and promptly scoffed in the face of Musepac....
vladimirovich
QUOTE
What freq. analysis program are you using?


Cooledit pro. But it not the best in therms of frequisy analysis.

QUOTE
why does that 'equalizer' react badly with MP3's and absolutely terribly with MPC's?


Because in case of MPC and MP3 equalizer is applyed on file decoding stage, not afterwards.
Garf
The test in the first post is done with extreme settings (+12dB preamp and all sliders down 12dB), IMHO this is not really fair, as it it totally at the extreme end (or maybe not depending on how it works, but in any case these are not typical settings).

Can you do a test with preamp at 0dB and sliders at 0dB (is it activated then?) or for example a single 6dB boost at a single band and look at what comes out.

Also, what was the reference signal?
budgie
QUOTE(vladimirovich @ Apr 9 2003 - 10:38 PM)
Generally, those settings shouldn't change sound at all.

I admire your self-confidence... rolleyes.gif
KurtCocain
So what EQ settings do u recommend as good in WinAmp, or is there a EQ plu-in that would make the sound better. I personally think that MMJ, in terms of EQ, does the best job. Dont like MMJ, though cuz eats too much smile.gif In WinAmp i use the Full bass n treble, plus the ENHANCER plug-in. that kinda works best for me...
Krest
To do such a test you need pink noise as a sound reference (all frequencies at the same level) or you need to compare winamp's output with the original wav without eq...
kotrtim
QUOTE(KurtCocain @ Apr 10 2003 - 05:23 AM)
I personally think that MMJ, in terms of EQ, does the best job.

MMJ is too hissy if you put the high-end to the max! (my point of view)

Try J.River MJ's EQ,
That's the EQ I love,
It just apply min. changes to the audio smile.gif
If you look through to frequency analysis.
no changes at all COOL
budgie
QUOTE(KurtCocain @ Apr 10 2003 - 05:23 AM)
In WinAmp i use the Full bass n treble, plus the ENHANCER plug-in. that kinda works best for me...

OMFG... It must sound horribly; for metal and rock maybe acceptable.
I never use anything, no EQ, no filters, no replay gain. Just digital output from my soundcard into amplifier... or using headphones. That's all...
Lev
QUOTE
In WinAmp i use the Full bass n treble, plus the ENHANCER plug-in. that kinda works best for me...


blink.gif

Something tells me these must be "The Kazaa / WinMX Crowd's" creations, if that sort of equalisation is needed wink.gif
Case
QUOTE(Garf @ Apr 10 2003 - 01:42 PM)
Can you do a test with preamp at 0dB and sliders at 0dB (is it activated then?) or for example a single 6dB boost at a single band and look at what comes out.

Winamp doesn't activate EQ when settings are at 0. When 2.90 betas came out I tested the EQ with white noise signal by setting all levels down minimum possible amount (0.6dB). The result wasn't as bad as in vladimirovich's test but it wasn't flat either. But sure the EQ has improved since earlier Winamp's, those actually started the HF rolloff at around 3 kHz.
Case
QUOTE(Lev @ Apr 10 2003 - 10:53 AM)
What I've never quite understood is why does that 'equalizer' (a bad name for it if ever I heard one) react badly with MP3's and absolutely terribly with MPC's?  A long time ago, I left the equalizer on and promptly scoffed in the face of Musepac....

Winamp's equalizer doesn't know what the source signal is, what are you meaning by "react terribly with MPC"?
vladimirovich
QUOTE
To do such a test you need pink noise as a sound reference


Yes, but I call this "while noise". Freq sweep from 0 to 22050 could fit fine to.

QUOTE
The test in the first post is done with extreme settings (+12dB preamp and all sliders down 12dB),


You call this extreme? Equalizer in CoolEditPro do this job fine.

QUOTE
Winamp's equalizer doesn't know what the source signal is


No, it DOES. Input plugin could equlize sound by itself.
Lev
QUOTE
what are you meaning by "react terribly with MPC"?

--Ages ago, my first MPC was created, and, with the equaliser on in winamp, I rated it way worse than my --r3mix mp3 of the same song.

Could be I am making a complete gibbon out of myself here, but I definately recall WA's EQ worsening MPC more so than it did the MP3
Case
QUOTE
No, it DOES. Input plugin could equlize sound by itself.

No, it does not know the format the signal is coming from. All the EQ knows is whether to be on or off, input plugins can turn it off when they provide their own equalizing. By default for example MPC plugin doesn't use Winamp's EQ and I would like to know how it "reacts terribly" when it is actually activated and source is MPC.
Lev
I recall this thread as well, a certain alien had a similar problem

http://forums.afterdawn.com/thread_view.cfm/1945
KurtCocain
QUOTE(budgie @ Apr 10 2003 - 06:40 AM)

I never use anything, no EQ, no filters, no replay gain. Just digital output from my soundcard into amplifier... or using headphones. That's all...

If you want a better sound, esp. if u r stuck w/ a cheak ass laptop sound card, I think it is better to use EQ and DSP smile.gif Without any EQ or DSP it sounds like $SHIT$
budgie
QUOTE(KurtCocain @ Apr 10 2003 - 12:31 PM)
If you want a better sound, esp. if u r stuck w/ a cheak ass laptop sound card, I think it is better to use EQ and DSP smile.gif Without any EQ or DSP it sounds like $SHIT$

Depends a lot on the source you're listening to... Any "cheap" EQ (and the built-in in WA is no exception) makes things only worse. Learn yourself to listen without EQ... maybe you will "suffer" a few songs at first, but human ear is a great instrument to fake easily rolleyes.gif
picmixer
Sorry for double posting this here. But maybe I should not have posted it in a years old thred in the first place. Didn't realize it. Anyway I'll just paste it here then:


All right have to jump in here quickly as well. Maybe it's a bit off topic, but I seem to have found the right people for my question.

What about actually just using the Equalizer in the Sound Card Control Panel and leaving out all software eq's?

Have tried that out on my system for a while now and am actually quite happy with the sound I have now.

This post has been edited by picmixer on Apr 11 2003 - 01:22 AM
KurtCocain
QUOTE(budgie @ Apr 10 2003 - 11:11 PM)

Depends a lot on the source you're listening to... Any "cheap" EQ (and the built-in in WA is no exception) makes things only worse. Learn yourself to listen without EQ... maybe you will "suffer" a few songs at first, but human ear is a great instrument to fake easily  rolleyes.gif

I dont know dude. I tried it. Turned off the equalizer.. i am tellin you -- it sounds like S H I T, seriously. And I tried it on headphones and conn my pc output to my stereo. the sound is much better w/ the EQ on. Also, much, much better with Enhancer plug-in. The DSP really gives it that feeling that u r listening to a 5.1 surr sound smile.gif. Well, it manages to approach a liitle bit that kind of sound ... Without EQ on, my mp3s (most of them 320, and many lame encoded), sound like somebody is hitting on a metal can smile.gif). And eqiualizer in WinAMp -- am talking only about the "Rock" and "Full Bass n Treble", and "Techno" settings. All the rest are terrible. Also u guys talk about enabling 3d, bass or what ever from your sound card. Well, I guess then u dont need a WA EQ. But as I said before. I do not have such options in my sound card....
Wish
From new EQ's developer, George Yohng:

QUOTE

<Yohng> EQ is psychoacoustically tuned with different bandwidths on different frequencies, and if there is a person who wants old Justin's EQ back, he's probably deaf.
<Yohng> Also new EQ works for non-MP3s only
<Yohng> and that probably some newer winamp versions will have MP3 support smile.gif

<Yohng> well - EQ10 is not complementary
<Yohng> (band-complementary)
<Yohng> that's why it sounds good

<Yohng> if guys want linear response - they should set EQ bands to zero or turn EQ off
<Yohng> there is a "preamp" slider to control preamplification, btw
<Yohng> but Winamp2 has no "FAT" and "MP3 restore" options, so any feedback is welcome


Don't know if this is enough to satisfy you people on the new EQ..... laugh.gif
amano
according to the quotation above, it works for non-mp3s only. why doesn't it work with mp3???

Please tell me.
AstralStorm
Winamp's new EQ is inferior both to my external and Foobar SSRC. (
But I don't need any of these, my equipment has nearly flat frequency response, (with headphones)
and I like it that way.
(There is still a tiny amount of reverb with speakers,
even though I've changed room layout - it's much less than before anyway)

We shouldn't dicuss tastes here.
Pio2001
QUOTE
QUOTE
To do such a test you need pink noise as a sound reference


Yes, but I call this "while noise". Freq sweep from 0 to 22050 could fit fine to.


It depends on the spectrum analyser. A true audio spectrum analyser (logarithmic power/bandwidth) would need a pink noise in order to return a flat response, a white noise would not be flat. But this is quite non-existent on PCs. I tried all Winamp plugins two or three years ago, plus SoundForge, CoolEdit, Wavelab, Samplitude, and an expensive dedicated spectrum analysis software, all were flat with white noise (constant power/bandwidth scaling).
The first spectrum analyser with the logarithmic power/bandwidth option that I saw was Magix Samplitude 6's one. Maybe SoundForge, CoolEdit etc have included this option in their latest versions too.
Without it, the spectrum is very far from what we hear. We see only bass.

White noise : noise whose spectrum has a constant level over frequency.
Pink noise : noise with the same power in each octave, that is the same power in each audio band. Each octave being twice as wide as the previous, its true spectrum is decreasing 3 db each octave.

Note that both scalings return a flat response with a sweep tone.
amano
means:

you can't judge a eq's quality with a white noise???
Differenciam
1) Whatdaya expect??? You put the left bar all the way up, you're supposed to use the presets.

2) I am the winamp EQ presets b*tch user posted image biggrin.gif. When I listen on a good computer and decent headphones, there is no need for it. When I listen on my laptop and cheap sonys, it is absolutely necessary, or the music is almost unenjoyable.
Pio2001
QUOTE(amano @ Apr 24 2003 - 09:35 PM)
you can't judge a eq's quality with a white noise???

Sorry for being confusing.

In short :

Pink noise <> White noise

With a PC software, use white noise
With a standalone audio spectrum analyzer, use pink noise.
kennedyb4
QUOTE(Krest @ Apr 10 2003 - 08:38 AM)
To do such a test you need pink noise as a sound reference (all frequencies at the same level) or you need to compare winamp's output with the original wav without eq...

That's what I used when I was fooling around with the native winamp 2.8 eq and supereq. Its in some thread deep, deep, deep in the archives.

I was finding very audible changes in high frquencies when the low frequency sliders were adjusted. I even tried to upload the files to ff123 but I got a little nervous about some ftp protocol I needed to download.

Nobody seemed to care too much anyway so I let it go. But its easy enough to do.Create about 60 seconds of pink noise, play with the bass and listen. The problems are easy enough to hear. No abx seemed necessary.

I switched to the Waves Linear Phase EQ using the adapt-x interface.

Its a very transparant sounding eq, but between this eq and the ssrc resampler, playing an mp3 uses over 50% of my resources on my old P3 733. The docs with the Waves bundle state that low frequency equalization has to be handled differently from highs but did not elaborate too much.
Frank Klemm
QUOTE(Garf @ Apr 10 2003 - 12:42 PM)
The test in the first post is done with extreme settings (+12dB preamp and all sliders down 12dB), IMHO this is not really fair, as it it totally at the extreme end (or maybe not depending on how it works, but in any case these are not typical settings).

Can you do a test with preamp at 0dB and sliders at 0dB (is it activated then?) or for example a single 6dB boost at a single band and look at what comes out.

Also, what was the reference signal?

The absolute setting of the sliders is uninteresting.
For the EQ behaviour only the sum of global gain + frequency gain is interesting.

Problem looks like the K-501 equalization is switched on.
Garf
QUOTE(Frank Klemm @ May 15 2003 - 10:23 AM)
The absolute setting of the sliders is uninteresting.
For the EQ behaviour only the sum of global gain + frequency gain is interesting.

You're assuming it's well designed. If it does something stupid like applying preamp to 16 bit fixed point data before EQ, then these settings will generate a large amount of clipping. There can be other ways to mess up things that we don't know since we don't know the internals.

I see no reason to presume that an extreme combination of settings that would theorethically give the same results will actually always do so in practise.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.