[Giogio]

  Can't help you with the resampler, but for crossfeed, there's a whole thread on page 3 on the subject:
http://www.head-fi.org/t/202365/best-crossfeed-plugin-for-foobar
 
ADDENDUM:
You piqued my curiosity so I did some searching.  Ended up installing SoX resampler

And, which resampler do you prefer now? DO you really notice any difference in SOUND, between the stock Foobar one and any other resampler?
Would the difference be noticeable in Mp3 too?
 
Thanks for the other link. I was wandering, if I use crossfeed, I have less stereo effect, right? Will this not affect negatively the soundstage? Should I, and is it even possible and meaningful, use at same time a crossfeed and a dolby or whatever else surround plugin (not yet found a simple, good, effective one, and I hate when they just add a cheap metallic reverb).

 

[castleofargh]

So you mean that Wasabi give better sound than Asio4all? I was just wondering about that, as I do not have Asio4all...  
About the buffer, I try to keep it as low as possible and generally even with a BT headphone I can even keep it at 50 actually, but I leave it on 300 mostly.
I put it on the max buffer possible when I have other CPU sucking programs running in background, like the MP3Gain (which correct me if I am wrong but it is still the best software to normalize MP3), which even in idle mode is terribly heavy. A huge buffer, and putting Foobar in high priority, solve the unwanted problems.

asio4all is like an adapter to make stuff that like to talk to asio, able to talk to other protocol(I think I read somewhere it was kernel streaming but I'm not sure). so the rational thing to do when you have a choice, would be to use the real asio drivers provided for your DAC if they happen to exist. or go for wasapi or even directly to KS if that works for you.
but there is no actual situation where it really make sense to use asio4all for a basic listening. so more than saying "why not use it?", I would say "why use it? " ^_^.
 
 
Quote:

  And, which resampler do you prefer now? DO you really notice any difference in SOUND, between the stock Foobar one and any other resampler?
Would the difference be noticeable in Mp3 too?
 
Thanks for the other link. I was wandering, if I use crossfeed, I have less stereo effect, right? Will this not affect negatively the soundstage? Should I, and is it even possible and meaningful, use at same time a crossfeed and a dolby or whatever else surround plugin (not yet found a simple, good, effective one, and I hate when they just add a cheap metallic reverb).

sox is known to be pretty good, can work fine with basic settings but you can add some magic spells if you're into making your own stuff. it's fine and versatile, known to measure pretty well, so there really is no reason not to use it IMO.
 
 
crossfeed will put some frequencies that were on the left, and copy them on the right, usually attenuated by a value and delayed by a value. the aim is to simulate in part what would happen to a real sound coming from the left, you would get it in your left ear, then a little latter and a little quieter, into the right ear.  no sound would ever come only into the left ear, that doesn't exist in real life. but it exists in headphones. so unless the album was mastered for headphones(99.9% aren't), you end up with panning that is too strong and unnatural.
it is my experience that you get used to whatever panning you get after half an hour or less(the brain is amazing), but most crossfeeds feel less fatiguing to me in the long run. at first it is less impressive, and often at least the bass suffers a little from it(voices too for some IMO). so some people just never get into it.
crossfeed isn't speakers and is far from a perfect simulation, that must be clear for everyone. it's just slightly more of a realistic experience than music mastered on speakers and used on headphone.
 
 
ps: I was desapoint by the naked picture. I expecting a cool juggling panda or something.

 

[Giogio]

  sox is known to be pretty good, can work fine with basic settings but you can add some magic spells if you're into making your own stuff. it's fine and versatile, known to measure pretty well, so there really is no reason not to use it IMO.
 
 
crossfeed will put some frequencies that were on the left, and copy them on the right, usually attenuated by a value and delayed by a value.
often at least the bass suffers a little from it(voices too for some IMO).
 
 
ps: I was desapoint by the naked picture. I expecting a cool juggling panda or something.

Well, you can still watch Kung Fu Panda, one of the best films ever, and get some of that stuff
 
The reason why not to use SoX (or the other one, SS something, I forgot how it is called, which soma people think it is better, for what I have read in the first page of the sox thread on hydrogenaudio), is that Foobar comes with a resampler already, which does not have to be configured. Now, if I have to use something more complicated where I have to set parameters, given how obsessive and perfectionist I am I may lose ages to configure it. SO I would only get into that hassle if that resampler really give a noticeable better audio quality than the stock one (also with MP3, which I listen a lot, because most of my music I got it from people met during travels).
So... Does it, sound noticeably better?

 
About crossfeed, I had that in mind (Dmitry of Neutron explained that once to me, confused by the fact that I was using the Surround instead of the Crossfead effect with Headphones. But I like to have a spacious sound, and some BT headphones do not have it... Unfortunately you cannot use both effects on Neutron).
But thanks! One never knows.
What I am trying to find out is which Crossfeed plugin is better on Foobar. Apparently I am not the only one!
The stock "crossfader" does nothing to my ears, I do not even know what is it supposed to do.
I have downloaded the naive, which I still do not know how to configure (no help file, no faq, nothing, no idea what those three parameters really do).
I have noticed the slight bad effect on bass on some tracks, which anyway does not affect me too badly because I have Real Bass Exciter doing a very good job.
But on other tracks it was even better, because some tracks have a panned bass (never pan bass!), which with crossfeed becomes more central. Just in these days I was listening to the wonderful Electro Swing track "True Love Sweet Georgia Brown" of Ecklektic Mick (unfortunately not available online), where the punch/bass is almost all on the right. Very annoying.
Crossfeed made it better.
Btw, do you put it at the beginning or end of the chain? I have it just after the resampler, as second plugin, so that the Real Bass Exciter will excite the centred bass, not accentuating eventual panning of the recording.
 
In the while I have other questions:
what the Advanced Limiter and the Hard -6 Limiter do?
I know, more or less, what a limiter does (stopping all sounds which passes a certain peak level, right?), so I though I could use it at the end of the chain, or bewteen (hmmm, I swear it was not a porn thought, just a typo) Real Bass Exciter and Graphic EQ, to minimize some eventual distortion (sometimes I do not reduce so much the gain in the EQ, to have more loudness. But apparently bass and gain do not understand each other perfectly well).
I did not notice any effect with the Advanced, but I have noticed that the Hard even creates more distortion sometimes.
Do you use them? Which? How?
Does it makes sense to combine them?

 

[Bob A (SD)]

  And, which resampler do you prefer now? DO you really notice any difference in SOUND, between the stock Foobar one and any other resampler?
Would the difference be noticeable in Mp3 too?
 
Thanks for the other link. I was wandering, if I use crossfeed, I have less stereo effect, right? Will this not affect negatively the soundstage? Should I, and is it even possible and meaningful, use at same time a crossfeed and a dolby or whatever else surround plugin (not yet found a simple, good, effective one, and I hate when they just add a cheap metallic reverb).

I enjoy using crossfeed algorigthms and have for years (since Tyll introduced his in the Supreme headphone amp in 1993).  They simply take what is there and mix the channels ever so slightly to lessen the "in your head" sound of headphones/IEMs.   Subjectively it seems to broaden the soundstage.
 
Your reverb and other special effects, colors the music in my view distorting what the artist has created.  I avoid them.
 
I don't use dolby multichannel playback unless the tracks were originally recorded that way and I have the equipment to properly play them (e.g. dolby 5.1 music tracks on a home theater system).
 
A resampler simplistically can reduce digital distortion.  And no, I have no heard any substantive differences with the SoX resampler engaged.  But as castleofargh stated there is no reason not to use one and SoX is touted as the best of what is available for foobar2000.
 
So as far as DSP in my foobar2000 setup, all I use is foo_dsp_xfeed and foo_dsp_resampler.   The graphic equilizer foo_dsp_xgeq is active  but I have yet found a reason to use it.

 

[castleofargh]

Originally Posted by Giogio /img/forum/go_quote.gif
 
Well, you can still watch Kung Fu Panda, one of the best films ever, and get some of that stuff
 
The reason why not to use SoX (or the other one, SS something, I forgot how it is called, which soma people think it is better, for what I have read in the first page of the sox thread on hydrogenaudio), is that Foobar comes with a resampler already, which does not have to be configured. Now, if I have to use something more complicated where I have to set parameters, given how obsessive and perfectionist I am I may lose ages to configure it. SO I would only get into that hassle if that resampler really give a noticeable better audio quality than the stock one (also with MP3, which I listen a lot, because most of my music I got it from people met during travels).
So... Does it, sound noticeably better?

 
About crossfeed, I had that in mind (Dmitry of Neutron explained that once to me, confused by the fact that I was using the Surround instead of the Crossfead effect with Headphones. But I like to have a spacious sound, and some BT headphones do not have it... Unfortunately you cannot use both effects on Neutron).
But thanks! One never knows.
What I am trying to find out is which Crossfeed plugin is better on Foobar. Apparently I am not the only one!
The stock "crossfader" does nothing to my ears, I do not even know what is it supposed to do.
I have downloaded the naive, which I still do not know how to configure (no help file, no faq, nothing, no idea what those three parameters really do).
I have noticed the slight bad effect on bass on some tracks, which anyway does not affect me too badly because I have Real Bass Exciter doing a very good job.
But on other tracks it was even better, because some tracks have a panned bass (never pan bass!), which with crossfeed becomes more central. Just in these days I was listening to the wonderful Electro Swing track "True Love Sweet Georgia Brown" of Ecklektic Mick (unfortunately not available online), where the punch/bass is almost all on the right. Very annoying.
Crossfeed made it better.
Btw, do you put it at the beginning or end of the chain? I have it just after the resampler, as second plugin, so that the Real Bass Exciter will excite the centred bass, not accentuating eventual panning of the recording.
 
In the while I have other questions:
what the Advanced Limiter and the Hard -6 Limiter do?
I know, more or less, what a limiter does (stopping all sounds which passes a certain peak level, right?), so I though I could use it at the end of the chain, or bewteen (hmmm, I swear it was not a porn thought, just a typo) Real Bass Exciter and Graphic EQ, to minimize some eventual distortion (sometimes I do not reduce so much the gain in the EQ, to have more loudness. But apparently bass and gain do not understand each other perfectly well).
I did not notice any effect with the Advanced, but I have noticed that the Hard even creates more distortion sometimes.
Do you use them? Which? How?
Does it makes sense to combine them?

a lot of questions.
does it sound noticeably better, I don't think so, but then again I'm one of those deaf objectivists that doesn't hear much of anything ^_^. I do use the default foobar converter sometimes and it didn't ruin my life. but I always use sox for real time downsampling.
at least one of the reasons why I believe everybody doesn't agree on a preferred crossfeed, could be simply that we don't all have the same distance between our ears. meaning that we don't all need the same delay to get something similar to real life delays. xnor's crossfeed let you set the delay yourself, so you could try and measure your head, and estimate the value with the speed of sound. or you can just move the slider around until you feel that it's right for you. anyway the changes aren't dramatic(I actually struggle a lot to notice a change unless I change the values a lot) and even if we're a little off, again our brain usually takes care of what's wrong.
 
crossfade isn't a crossfeed ^_^.
 
limiters are used to set a maximum volume level. on the digital side it can been seen as an anti clipping tool. I'm pretty conservative with my gain values+ I use replay gain and my EQ gets mad when it clips so I know about it. so I don't feel like I need the limiter. I always wondered if the "prevent clipping" option when choosing what replay gain to use was not doing the exact same thing? anyway, in general the limiter will do nothing as long as the signal doesn't clip, so there is nothing wrong with using it as a safety measure. you can put it at the end of the DSP chain and forget about it. but it should not serve as an excuse not to correctly set the gain on your EQ to go with whatever boost you applied. a limiter is an improvement over hard clipping, but it isn't something you want to activate too often, it could sound really bad too if your signal is pushed up too much too often.
and no using 2 limiters only means the lower one may do something, the higher threshold one will never get the opportunity to do a thing.

 

[Giogio]

   
Your reverb and other special effects, colors the music in my view distorting what the artist has created.  I avoid them.  
I don't use dolby multichannel playback unless the tracks were originally recorded that way and I have the equipment to properly play them (e.g. dolby 5.1 music tracks on a home theater system).
 
A resampler simplistically can reduce digital distortion.  And no, I have no heard any substantive differences with the SoX resampler engaged.  But as castleofargh stated there is no reason not to use one and SoX is touted as the best of what is available for foobar2000.
 
So as far as DSP in my foobar2000 setup, all I use is foo_dsp_xfeed and foo_dsp_resampler.   The graphic equilizer foo_dsp_xgeq is active  but I have yet found a reason to use it.

I meant if you or anybody ever heard a difference between the different resamplers. I am dangerous when I start with something new, so before I open a pandora box I want to be sure what could be the advantage respect to the stock foobar resampler.
 
I do not use reverb and I understand what you mean but I do not agree and you may read why in the section "my sound ideology" of the opening post in the first link of my signature.
Beside what written there, which is mostly objective, I also have other more personal reasons why I like to customize sound: I do not consider music like a mystical ipse dixit which must be venerated as it is.
If I listen to Dubstep I doubt the "artist" would ever possibly care less if I push the bass or add this or that. And same I would say of most musicians, given that I seriously doubt that in pop and rock music each single little detail of the composition, and of the recording and mastering and mixing, is thoroughly thought (apart Pink Floyd, Alan Parsons Project, and guys like that). I think more artists do some jam, find something cool, work on it, and fix it when it is good enough. I have talked to musicians and producers of electronic music, and I have done some myself, and the creation of music is not that mystical in the 90% of cases. Artists know that they could stay ages working and reworking on something without ever feeling "oh, now it is ready".
So, I do not see all this mystical need to respect the "sound as it was meant to be".
Besides, when they record and mix, they have THOSE equipment, THOSE monitors, THOSE headphones. They listen to their own record through limited technology (nothing is uncoloured) and have no idea how that music will sound on your devices, nor you will ever be sure that how your devices make it sound is exactly how it sounded in the studio.
For me it is like cooking. You can follow a recipe in each little detail, but, do you have the SAME ingredients? Same quality? Same origin? And do you have same mouth, do you perceive tastes in the same way? Are you sure that if you follow the recipe as it is, you will taste the same taste the creator of the recipe tasted?
So, I change things. It is an active listening, where I participate in the creation process, in a dialogue with the artist.
Unless I am facing an absolute masterpiece of music, a real ultimate work of art where each single little sound has a mystical reason the way it is, and where the recording was made state of the art with extremely sofisticated technology. In that case I would definitely do my best to organize a setup which would let me hear the recording as it was supposed to sound (just tweaking what I need to, to adapt the sound to the limitations of my devices and of my ears).
 
But to be honest, mostly what I do is just adding salt and pepper. I can eq a bit, mostly one single preset to adjust the basic sound of my headphones overcoming their limitations/colour and tuning them to my ears and psychoacoustic traits.
I can push a bit the bass because I like so and I could not care less if Sade would agree with me.
I can apply crossfeed.
And I can try to find something which can improve the soundstage, which in my opinion is NOT altering the creation of the artist but overcoming the limitations of the equipment I have (I do not have 2000$ open back planar magnetic with 2000$ top class amp). On a little AKG Y45BT, if you do not apply something, you have almost no soundstage, and I will bet my life each second that that is not what the artist wanted.
If I was sure that the artist wanted a direct sound with no soundstage and I respect and admire the artist more than I like spatious sound, I would definitely avoid effects.
 
But I avoid them now already, as I did not find ANY, at least, not any free, which can improve soundstage without sounding like I am in a case underground (as said, I do not like reverb. Although I would argue that artist do use reverb, but, whatever).
 
I felt in writing mode tonight.
I will try the xfeed as soon as hydrogenaudio is back online.

 

[Giogio]

Originally Posted by castleofargh /img/forum/go_quote.gif
 
 I always use sox for real time downsampling.
 
at least one of the reasons why I believe everybody doesn't agree on a preferred crossfeed, could be simply that we don't all have the same distance between our ears.
 
crossfade isn't a crossfeed ^_^.
 
limiters are used to set a maximum volume level. on the digital side it can been seen as an anti clipping tool. I'm pretty conservative with my gain values+ I use replay gain and my EQ gets ma when it clips
and no using 2 limiters only means the lower one may do something, the higher threshold one will never get the opportunity to do a thing.

Thanks
 
I use a resampler because some of my Mp3 are 44, some are 48. Both my intel soundcard and my bt headphones can get to 48, so I put the resampler to 48. So, it is mostly upsampling.
Would that make a difference?
I do not even know if it would make a sense setting it at 96 or so, which with the output as "direct sound" it does not crash foobar (with wasabi I must set 48 if my device can only 48).
Anyway, I will try sox
 
So crossfade is like, a track ends and another begin and the two mix? I hate that. Just used it for a party at home with Musicbee once.
 
I have all my Mp3 set at 88,8 with mp3gain and option "do not clip" (so some are lower than 88,8).
I also select the apply replay gain and do not clip in foobar.
As EQ I use the graphic EQ, and it does not ma to me but I did not understand what you mean What do you use? I like that because it goes deep till 20 and I can save presets.
I do set the gain a bit lower if I am pushing the bass a lot (which I do in the presets I use for Techno, and even more for Dubstep).
But I have no idea how much should I reduce the gain to avoid clipping, and despite replay gain there are tracks which clip and distort, and some which does not, although the perceived volume is the same.
One which distorts very easily is Balloons (club mix) of Nils Hoffmann. I use it like a standard to set the gain!
An autogain option like in Neutron would be cool. Is there any such plugin? It reduces the gain of x% when a peak is detected. Each new track it resets the gain at 0 (or the value you want).
The graphic eq have a fake autogain, it does not analize the music real time like neutron does, it just set the gain as many db down as you have pushed some frequency up. This is way too much.
So, basically I use the limiter "just in case". I do want power and try to reduce the gain as few as possible, but I definitely do not want distortion and clip.
 
Using the two limiters would be in different moments of the chains. One after the Real bass Exciter, and one after the EQ. Because I push both till their limits.
And I was just wondering how the two limiters differ.
 
I like what you say of crossfeed, it sounds exactly like what I wrote in "my sound ideology" on my thread (feel free to read and feedback if you agree).
I find it much more clever offering a customizable crossfeed than 20 different ones where people must try each to see if by chance it come close to what their ears need.
I did not know this one you mention. Ah, ok, it must probably be the xfeed, right?
I will try that soon.
 
EDIT: as we are here, do you know if it is better 16 bit plus dither, or 24 bit without dither?

 

[castleofargh]

it was supposed to read "my EQ gets mad when it clips" sorry about that, depending on the keyboard I use I tend to eat up some letters ^_^. I just meant there is a warning for clipping.
for EQ there is an endless supply of VSTs that will work with foobar as long as you also add a vst adapter. I would always recommend parametric EQs as they tend to have cleaner impact and better controls. for years I've used electri-Q, great but crashes way too often for my taste. so I moved on to easyQ that offers a little less but also crashes less. now I've moved on to some paid EQ(DMG equilibrium), but both easyQ and electriQ were really cool ones and I never regretted using them and learning on them.
 
what I meant about setting the gain correctly on the EQ, if you boost 10db, you should also move the gain to -10db at the same time. a digital signal is written with one value per sample, that value gives an amplitude and goes from 0db to -96 or -144db(16 or 24bit) the loudest sound being 0db. so anytime you boost something that was close to 0db with the EQ, you risk pushing the value above 0db which is impossible as there is no number for that, so the value that will be used instead is still 0. that's clipping. all the stuff that would go above ends up being 0db. at this point if you use nothing it clips, if you use the limiter, it will reduce the value of the signal by a given value for a given time and sound more or less natural depnding on the kind of limiter, but it's not a transparent process. so whatever boost you had decided upon will not be applied anyway if the music was already close to the max, one way or another, clip or limiter you will still be stuck to 0db(of course that's in the digital domain, on the analog side any amp can be pushed to be louder).
so the proper method to avoid having to alter the music in an undesired way, is to simply make sure the sample values will always stay below 0db. it's easy, as long as you do nothing that will boost the signal you will not risk degrading said signal. that means trying to EQ by removing values instead of boosting. or if you boost, lowering the main gain setting in the EQ by the same value you used on the biggest boost. if your maximum boost is 20db, then you need to set the global gain of the EQ to -20db. the result will be that the loudest frequency will stay where it was and never risk clipping, while the rest will go lower.
as far as digital go it is the right method. if your only concern is that you can't get a 20db boost in the bass, I was talking about it on another topic a few days ago, with easyQ you can make up to 48db boosts!!!!!!!!! ^_^

here it is, let's say I want to apply this made up curve with my massive up to +48db bass boost. I can't leave it like this, instead I will lower the global gain by the value of the maximum boost. that way the loudest signal getting out of the EQ cannot try to be higher than 0db no matter what was recorded on the music. and the EQ I will use is in fact this one:

this one cannot clip the music and won't need a limiter to try and save the day. I'm not using the EQ to boost the bass, I'm using it to lower everything else!!!!
 
 
 
 
the only reason direct sound seems to work with 96khz when your devices can only do up to 48 is because windows must resample back down to 44 or 48 in the end. wasapi bypasses windows mixer so it doesn't resample down and your device says "no speako el 96khz". in both cases you do no end up with 96khz.
 
now on the other hand setting the output to 24bit is very fine if your DAC can use 24bit(be it connected DAC or bluetooth DAC inside the headphone(does the streaming protocol accept 24bit????).
if the DAC part is ok with 24bit then do use it, that gives some headroom for those EQ things we're doing and for some more volume setting by foobar or windows. if the DAC doesn't accept 24bit(or the bluetooth protocol), well then obviously select the 16bit output. ^_^
 
and yes it's xfeed I was talking about for xnor's crossfeed, although I suppose several .dll are called the same.

 

[Giogio]

So, well, thanks again.
I had read about subtractive EQ and I had used for a while but got sick of having to drag down all the other sliders (I mostly push the bass, so with subtractive I must drag down all other sliders on the Graphic EQ).
I did not realize, well, I had partially realized or suspected that decreasing the main gain was the same as doing like the images you have posted.
 
With MP3gain I have seen (analysing tracks) that most modern music has a value of 94, +/-.
But older music is less. So MP3 gain suggests 89. As I do that, I have a lower 0, and I can push the EQ more without decreasing the gain so much, or this is what I thought till now.
Also, decreasing the gain of 6 if you push one freq 6 up, it is only preventive, just in case, but we do not know if that freq was near to 0, so, maybe it war -3, and I just have to reduce the gain -3.
The other problem would be that as I only push the bass, the overall sound will lose lot of loudness if I reduce a lot the main gain. With Dubstep I push the bass even more. More or less, 3db the 40-80hz with Techno, 6db with Dubstep. Plus the Real Bass Exciter, this is lot of bass (which the ATH-WS99BT do very well).
I go by ear, I know which of my tracks have the most devastating (near 0) bass, and I use them to test my EQ and gain setting, reducing the gain just the min needed.
Lately I am addicted to the highs of Dirac, which I use on the ATH although I only have the filters made to compensate FR and IR of the XTZ Headphone Divine.
Dirac has got a clipping led so I can see what my ears do not hear.
 
So, I wonder, does any of your EQ have both a clipping led and an automatic gain protection (like Neutron)? I think it is the best solution. I leave the EQ gain at 0, and the EQ automatically sees the peaks and decreases the gain as much as needed.
I think MusicBee had a function, "dynamically normalize tracks". I wonder if it is the same.
And would that be the same as Peak Normalization, but on the fly?
Although Neutron as said does something different, it does not constantly increase and decrease the gain for every peak, it just reduces it a bit when the 0 is reached, and it leaves it like that till the end of the track also if there are no other peaks.

 

[castleofargh]

even with the basic foobar equalizer, once you're done with your EQ you just click "auto level" and boom it will move the loudest boost to zero. unless you're failing to go loud enough with your sound system, I don't really get why you wouldn't just do it as it should?
+ if you really use something that reduces the loudness for a lot of potential clipping that you would have created yourself, in the end you're ruining the job that was done with mp3gain to make all our songs feel as though they were at the same loudness.
 
 
 
if you insist on using the EQ wrongly, you could always try and add something like this https://www.foobar2000.org/components/view/foo_r128norm  at the end just before the limiter, it would most likely do something close to what you're asking for (didn't try it so IDK if it's any good).  and it would make mp3gain redundant.

 

[Giogio]

My system are BT Headphones. Which means, I do not have speakers, nor wired headphone, nor anything I can push in loudness with an analogical Amp.
And although my favourite BT Headphone is the probably loudest of all BT headphones or among the loudest ones, if I do an EQ preset for Dubstep with 6db push on the sub-bass plus other hardcore push on Real Bass Exciter (for that I also need to get headroom by lowering the gain on the EQ, as the Exciter does not have any levelling option), and as said I only push the bass, and I have to reduce the main gain something like 9 to 12 db, I do not have much power left to enjoy music loud enough.
Another story is with the setting for Techno and Electro Swing, where I push bass 3db and the Exciter mush less than with Dubstep (the bass of Dubstep is so tricky, it needs lot of push to be really juicy, fat and dense).
In this last case I normally see that -6db main gain are ok and I can live with that. That is anyway the "border value" of my loudest headphone, meaning that already with that I am just "loud enough". Which is indeed louder than many people would like to listen to music for many hours (me included) but which is not loud enough to me for occasional moments when a special track make me wish to jump around my room doing club moves.
 
That's why I try to find a way to avoid reducing the main gain too much.
 
So, you gave me one. I will look at what it does, but if you say it nullify Replay Gain then it must probably constantly do real time peak normalization? Does that not reduce dynamics? Besides, I could never understand why the creator of MP3Gain says that peak normalization is not the best way to make tracks sound at same volume.
 
Another option would be maybe (I really have no idea if it would work, maybe you can confirm) using an external DAC or Soundcard, something I can connect to my notebook either via USB or with the 3.5 jack, and which would give a much higher loudness than the one I can get now.
I see that some Soundforge have USB ports where you can stick an USB BT Dongle like the Creative BT-W2, otherwise I could use the 3.5 Avantree Priva 2 adapter.
But I am not sure if once into the BT adapter the sound of the external card, even if originally louder than in my notebook, would be levelled digitally and have same exact loudness than those BT adapters would give me when connected to my notebook...

 

[castleofargh]

ok so you're a special case with a special problem. anybody else should avoid clipping and get a more powerful amp, but that can't be done with a bluetooth headphone as the amp is in the headphone. if the aim is really to feel louder, then the best known option is a dynamic compressor. like has been done for so many years on CD and on the radio (loudness war). it does alter your music but it's a proved method to feel loud.
 
about peak limit vs mp3gain: a limiter or peak normalizer will make sure the maximum signal in the song doesn't get over a fixed value. while mp3 gain set a preferred target for how loud the song will feel! not how loud the peaks will go will change depending on the dynamic of the song. some pop song will have little dynamic range so the peaks are actually very close to the main music. if you do peak normalization, then that song will sound very loud. if you do the same for some classical music, the music will sound way quieter(probably hard to hear compared to the pop song).  if you use mp3gain or replaygain or some R128 stuff, your music is set to feel like all musics are as loud psycho acoustically. not electrically .so it will take the pop song down a lot so that it feels as loud as the high dynamic classical song. by doing that mp3gain tends to make most music quieter than what they were before as a mean to make them all sound about the same in perceived loudness.
so each tool has it's own purpose.
 
 
 
 
 
and no adding an external whatever wouldn't change a thing as you're most likely already pushing things to the max with your EQ+limiter. as I said, with digital music the maximum loudness can't be more than 0db. and the only part that can boost the analog domain is your bluetooth headphone in your particular situation.
 
 
should I mention that listening to music loud for long periods is bad for your ears? ^_^

 

[Joe Bloggs]

My system are BT Headphones. Which means, I do not have speakers, nor wired headphone, nor anything I can push in loudness with an analogical Amp.
And although my favourite BT Headphone is the probably loudest of all BT headphones or among the loudest ones, if I do an EQ preset for Dubstep with 6db push on the sub-bass plus other hardcore push on Real Bass Exciter (for that I also need to get headroom by lowering the gain on the EQ, as the Exciter does not have any levelling option), and as said I only push the bass, and I have to reduce the main gain something like 9 to 12 db, I do not have much power left to enjoy music loud enough.
Another story is with the setting for Techno and Electro Swing, where I push bass 3db and the Exciter mush less than with Dubstep (the bass of Dubstep is so tricky, it needs lot of push to be really juicy, fat and dense).
In this last case I normally see that -6db main gain are ok and I can live with that. That is anyway the "border value" of my loudest headphone, meaning that already with that I am just "loud enough". Which is indeed louder than many people would like to listen to music for many hours (me included) but which is not loud enough to me for occasional moments when a special track make me wish to jump around my room doing club moves.

That's why I try to find a way to avoid reducing the main gain too much.

So, you gave me one. I will look at what it does, but if you say it nullify Replay Gain then it must probably constantly do real time peak normalization? Does that not reduce dynamics? Besides, I could never understand why the creator of MP3Gain says that peak normalization is not the best way to make tracks sound at same volume.

Another option would be maybe (I really have no idea if it would work, maybe you can confirm) using an external DAC or Soundcard, something I can connect to my notebook either via USB or with the 3.5 jack, and which would give a much higher loudness than the one I can get now.
I see that some Soundforge have USB ports where you can stick an USB BT Dongle like the Creative BT-W2, otherwise I could use the 3.5 Avantree Priva 2 adapter.
But I am not sure if once into the BT adapter the sound of the external card, even if originally louder than in my notebook, would be levelled digitally and have same exact loudness than those BT adapters would give me when connected to my notebook...

Why not just gauge the volume using the foobar peak meter? It indicates clipping whenever it happens. So just set the EQ gain however you like and turn it down if you see / hear clipping. I agree that you don't need to strictly follow the "everything below zero" rule, especially if Replaygain has been applied, but it's anybody's guess what a safe setting is for all the stuff you listen to. I don't listen loud, listen to wired earphones and have an amp that goes crazy loud, but if I were in your shoes I'd either watch the meter manually, or add the Advanced Limiter to the end of the DSP chain and call it a day.

Also, you do know there's volume buttons on the bluetooth headphones right?

 

[castleofargh]

Why not just gauge the volume using the foobar peak meter? It indicates clipping whenever it happens. So just set the EQ gain however you like and turn it down if you see / hear clipping. I agree that you don't need to strictly follow the "everything below zero" rule, especially if Replaygain has been applied, but it's anybody's guess what a safe setting is for all the stuff you listen to. I don't listen loud, listen to wired earphones and have an amp that goes crazy loud, but if I were in your shoes I'd either watch the meter manually, or add the Advanced Limiter to the end of the DSP chain and call it a day.

Also, you do know there's volume buttons on the bluetooth headphones right?

it's also the first thing I thought about, but as you say it's a great song by song method, but you can't tell much as a general purpose. I have songs where I could add +30db at 40hz without clipping, but I also have songs where 40hz almost reaches the max level of the entire song. so I didn't mention it. but of course it's a perfect song by song control. 
 
second point, when people know what they are doing and understand the signal chain, I'm with you that the everything below zero in the EQ is rarely needed. but for anybody who isn't yet confident with what he's doing, basic rules are how you learn how to do it right IMO. maybe we would have less EQ haters on headfi if they had cared for the overall gain.

 

 

[Giogio]

  if the aim is really to feel louder, then the best known option is a dynamic compressor. like has been done for so many years on CD and on the radio (loudness war). it does alter your music but it's a proved method to feel loud.
 
about peak limit vs mp3gain
 
and no adding an external whatever wouldn't change a thing as you're most likely already pushing things to the max with your EQ+limiter. as I said, with digital music the maximum loudness can't be more than 0db. and the only part that can boost the analog domain is your bluetooth headphone in your particular situation.
 
should I mention that listening to music loud for long periods is bad for your ears? ^_^

I could not hear your last sentence, can you repeat?
Anyway, you deserved THIS (you know the original but maybe not this, so eventually watch till the end).
 
Aim 1: quality
Aim 2: bass
Aim 3: loudness
I have tried the Dynamic Compressor of Foobar, it makes some tracks louder, some other strangely quieter, and it mostly ruins the dynamics. All feels constipated. Compressed, exactly.
No, that's not for me.
 
I understood your explanation of peak vs gain. Thanks. So, a Techno track with a constant loud low kick pushed near 0 would be treated from mp3gain like the classical music or like the pop one in your example? I would say like the pop, as these peaks are constant, so I am not sure how could mp3gain understand that they are just the low kick and give priority to the main melody. But if mp3gain would do that, then it would push the low kick above the 0 (which is not possible but, you know what I mean), right? And to avoid this, there is the option "avoid clipping". Did I understand all right?
 
Now, as I have already asked and nobody proposed any option, I must suppose there are none, but, I ask again just in case: are there any plugin (usable in Foobar2000) which would analyse the peaks in real time (probably before time, thanks to the buffer) and quickly reduce the gain as much as needed to avoid clipping?
You may argue that this is what the option "apply replay gain and avoid clipping according to peak" is doing, but I would argue back that it is not doing it well. It is mostly wishing to do it than really doing it.
 
About the 0 thing, the notebook must have some DAC to make the audio come out of the 3.5, right? And you said that with analog we can go above 0.
In this case there may be an external DAC which is louder than my notebook. Right?
The real question is, would the greater higher loudness of the new DAC result in a higher loudness on my headphones?
I do not know if the ADC of the BT adapter, which can just go till 0, would make the sound equally loud no matter how loud was when it entered in the ADC.
Do you?

Why not just gauge the volume using the foobar peak meter? It indicates clipping whenever it happens.

or add the Advanced Limiter to the end of the DSP chain and call it a day.

Also, you do know there's volume buttons on the bluetooth headphones right?

I have just tried the Peak Meter, it does NOT shows clipping. There is no red, no led for clipping. Or am I doing something wrong?
Anyway I can at least see when I am getting close to 0. It can be useful at times.
The advancer limiter is always there. And it definitely does not avoid clipping. I can notice it from the bad distortion when I am on red zone and from the clip led on Dirac flashing.
I think it is not a very good limiter. But the Hard -6 is not better.
 
Which buttons? Ah, yes, you  mean, that one with a + on it, which is used to turn headphone on and off?
(joke explained for dumb: all or nothing. loud or off.)
Nah, do not build a bad impression of me, I only listen loud to energetic music. And only when I feel like energizing myself.