What's new

ELI5: Mixes in headset sounds great, in friends monitor sounds shit - why?

Yeah - you'll need to listen on those headphones a lot to other music and also practice translating your mixes. I have the dt880s. They're overly bright and, to my ears, almost require correction (I use sonarworks) AND more recently I've also added Can opener by goodhertz as well in the monitoring chain before sonarworks to better approximate speakers. Others can't stand any software and think it's bad or that mixing on headphones is evil and will never turn out well.

Can opener and Sonarworks - gotcha! Thank you for kind advice.
 
+1 on this. the natural compression effect of human hearing really kicks into high gear with headphones, and its harder to sense loudness differences - this is a great thing for hearing details, but it makes it hard to know if you got the right levels - you might not know when one thing is to loud compared to another because you can hear them both just fine with headphones. This human hearing compression effect also totally screws with the perception of reverb decays. And on top of that, combined with the "in your head" stereo image (as opposed to "in front of you" phantom center on speakers), it really gives you a different sense of space than speakers. I feel it's been easier to get mixes to translate from speakers to headphones rather than from headphones to speakers.

I'm not an expert, but have learned of several other gems of advice thath have worked for me, and I think these are essential:
  • Listen at slightly lower-than-comfortable volume levels. It will help force you to mix the song so that important elements are all audible without the human hearing compression effect.
  • Try to make levels sound good in both mono and stereo, favoring mono slightly. Between lack of acoustic treatment and casual listener setups, mono might actually be a more representative baseline of what people hear when they don't use headphones. An awesome stereo image is only a bonus.
  • Try checking heavily high passed and low passed versions of your mix - it should still sound decent. For example, if you high pass at 400-800 Hz, or even as high as 1000 Hz with a 12db/octave slope, it would be ideal to still feel like the low frequency stuff is at least present and audible... Like the click of the kick drum, or the rich harmonics of a string bass. Obviously there are limits to this - a subbass with no harmonics and no transient isn't expected to be audible after a high pass...
  • I find it necessary to test multiple speakers/ headphones
  • mix referencing in combination with all the above tricks is pretty essential too. It's the only way to overcome the human psychology biases of judging your own mixes
Looks like dt770 did have a bit of a bass bump and a dip around 4-5 kHz - https://www.innerfidelity.com/headphone-measurements

Knowing that can also help you decide how to mix. No long term experience with sonarworks myself but had a demo from them before, and it was definitely great for ironing out the frequency response part. But that's still only part of making mixes translate well.

Cheers!

Wow, overwhelmed by the amount of information in this reply - thanks a lot!
1. Lower volumes when mixing - gotcha.
2. Several here is mentioning mono-mixing. Seems like I have to check that out more.
3. Checking high/low-passed versions - done.
4. Yes, Sonarworks keeps popping up in this thread - downloading the trial-version now.
 
I've been mixing in headphones a lot lately , and I think it's always a bit of a gamble. Having several different headphones/monitors to alternate between is helpful, as is constantly checking the mono mix. I've been using Sonarworks lately, but that sure hasn't stopped me from making bad decisions, particularly in the bass.

I'm considering iZotope's Tonal Balance Control, as a reality check. Visually comparing your mix to one you know and love on a spectrograph or analyzer can point out things your ears may have missed too.

And most importantly, if your mix sounds even better than the reference to you in your monitors, something is probably horribly wrong. I make this mistake every single time.

Thank you for mentioning the Izotope-plugin! Will check that out, sounds really handy.
 
I haven’t seen it mentioned yet, but mixing in mono will reveal quite a lot. Mix mono first, then stereo, then back to mono, etc. It is surprising how bad your mono mix will sound when learning how to mixing on headphones. Seriously. It happens to the best of us.

Waves NX is a great tool used in conjunction with Sonarworks. But, being honest about it, I find I strongly dislike the artifacts added by both pieces of software and find myself returning to a clean signal straight to headphones (HD650s btw). But I do use the plugins on every headphone mix before I return to hating them again. Necessary evil I suppose...

You’ll need to make sure you have a great headphone preamp with great DA converters too. Even if you don’t think you’ll hear the difference, you’ll be mixing to an inadequate signal chain that you will subconsciously mix to.

Edit: And as Avery mentioned just before my post, Tonal Balance is an excellent tool! I use it all the time.

Yes, mixing in mono is something new for me - really have to look more into that. Just tried my newest mix in plain mono, and it was very hard to separate the instruments from one and another. Think this can be a good way to create a wider listening experience / stereo-field.

Thanks for mentioning Waves NX - will check it out!

And yes, guilty as charged - my portable soundcard is both old and with a weak DA, so need more power in that departement too.
 
Another great plug in that works great is Mono maker by BX Digital.
Not sure if it’s in the BX Digital Suite as I use their ancient DSP Plug Ins.
But it’s a mid/side mastering plug in, it can go from mono to stereo, or mid /side, and Mono maker keeps everything below a selected sub range frequency in Mono.

This is very useful as a stereo mix where low freqs get dual signals is no bueno.
Cancellation is destructive to a mix.

Mono-plugings keep being mentioned in this post, thanks for sharing the advice!
 
This is true, but it's a little bit more complicated. When wavefronts from both speakers combine in the air, they are not well correlated, wich gives you +3dB compared to if only one speaker was working. When you add signals electrically (or digitally), their correlation is perfect, which gives you a +6dB gain. This means whatever is in the center will get a +3dB rise when downmixing to mono, compared to what's in the sides or very wide.
So, the wider your mix, the more perturbed the balance will be when switching from mono to stereo or the other way around (unless it's so wide there's nothing in the center ;)).
That is one reason (among a few others) why mix vocals and percussions in or near the center.


Have you ever tried slightly distorting the whole master (or a bus grouping the bass with something else), in order to hear the effect of the bass modulating the higher frequencies ? It has limits too, obviously, but it's always worth trying.

I 100% agree with everything else you've said.

This one is a bit too advanced for me at the moment, but I'm glad to see pro's commenting and helping out others in the same situation as myself. Cheers!
 
Undermixing the bass is surely one of my shortcomings - any advice on how to deal with that? My BD770 feels a tad soft and thin in the lower frequencies, so I have applied a headset-adjuster-plugin to flat'en the responsecurve a bit. Is that advisable to use on the master channel?
If you want to use EQ correction, why not EQ-correct your monitoring (the opposite way you would your master), instead of your master ?

Personally I would reserve that solution for extreme cases, however (like when I mixed student films for optical 16mm - the loss of HF was so drastic, I just couldn't compensate by ear), partly because it's approximative.

Alternatives are: get used to your BD770's sound (by listening to a bunch of other music you like in it)... and if you can't get used to it, then replace it.

Just as it is a good idea to check your work on other systems, it is also a good idea to listen to works of the quality you aspire to on your own system, to get a reference. As others have suggested in this thread.

As for replacement... I don't know your budget, but the Senheiser HD25 is a reliable headphones set, a de-facto standard, and still affordable enough. Or save a little bit for decent monitors. Here's a thread about affordable ones: https://vi-control.net/community/threads/best-studio-monitors-for-price-quality.79727/ - but before you open your wallet, know this: if your room is small and rectangular (or even worse, a square), new monitors won't help you much in the bass department - not without serious acoustic treatment.

This one is a bit too advanced for me at the moment, but I'm glad to see pro's commenting and helping out others in the same situation as myself. Cheers!
The thing to remember is that what's in the center will get a slight boost in mono.
Since mono sounds more crowded than stereo, compensate for that by placing what needs to stand out in the center. That's usually the soloist or vocals. And the percussions, if you make percussive music.

We also like to place the bass in the center, but for another reason: since the human ear can't locate low frequencies well anyway, this reduces the risk of clipping one channel, by spreading the high energy of the bass equally across both channels.

[EDIT: except in classical music, where one convention is to pan the instruments like a real orchestra, placing the bass to the right. But there's much less energy in bass string instruments than in an electric or synthetic bass.]

I know this is more than you asked for, but it's free :)
 
Last edited:
If you want to use EQ correction, why not EQ-correct your monitoring (the opposite way you would your master), instead of your master ?

Personally I would reserve that solution for extreme cases, however (like when I mixed student films for optical 16mm - the loss of HF was so drastic, I just couldn't compensate by ear), partly because it's approximative.

Other alternatives are: get used to your BD770's sound (by listening to a bunch of other music you like in it)... and if you can't get used to it, then replace it.

Just as it is a good idea to check your work on other systems, it is also a good idea to listen to works of the quality you aspire to on your own system, to get a reference. As others have suggested in this thread.

As for replacement... I don't know your budget, but the Senheiser HD25 is a reliable headphones set, a de-facto standard, and still affordable enough. Or save a little bit for decent monitors. Here's a thread about affordable ones: https://vi-control.net/community/threads/best-studio-monitors-for-price-quality.79727/ - but before you open your wallet, know this: if your room is small and rectangular (or even worse, a square), new monitors won't help you much in the bass department - not without serious acoustic treatment.


The thing to remember is that what's in the center will get a slight boost in mono.
Since mono sounds more crowded than stereo, compensate for it by placing what needs to stand out in the center. That's usually the soloist or vocals. And the percussions, if you make percussive music.

We also like to place the bass in the center, but for another reason: since the human ear can't locate low frequencies well anyway, this reduces the risk of clipping one channel, by spreading the high energy of the bass equally across both channels.

I know this is more than you asked for, but it's free :)
Great, great advice. Really appreciate it, man! Also, thanks for the link for the monitors.
 
So, I made a mix after following some of your suggestions:


Can I ask you what you think about the sound in general? To much hiss, high frequencies etc? What about the lows? I'm struggling to make my mixes sound the same across headphones and monitors. This piece was mixed on Beyerdynamic DT700 pro with Sonarworks-plugin + Goodhertz Can Opener, after checking in mono and mixing on fairly low volume.
 
An alternative to the CanOpener software mentioned earlier is Audified's MixChecker. This allows you to simulate playback on various consumer devices.
I'll check it out too :)

So, I made a mix after following some of your suggestions:


Can I ask you what you think about the sound in general? To much hiss, high frequencies etc? What about the lows? I'm struggling to make my mixes sound the same across headphones and monitors. This piece was mixed on Beyerdynamic DT700 pro with Sonarworks-plugin + Goodhertz Can Opener, after checking in mono and mixing on fairly low volume.


I am by far no expert in orchestral music, but as the spectral balance goes, I would say: there are too much low mids, and not enough highs. So it sounds kinda 'closed'.

Maybe correcting the low mids will automatically balance the highs... Start by trimming the former before boosting the latter.

Since the third quater of the song (from 2:00 to 3:00) is better balanced spectrally (it sounds more open, with enough highs), I would advise EQinq tracks individually before trying to EQ the whole master. There are good tutorials for that, e.g. on Youtube.

It's hard to judge the low freqs, because there's not a lot of them in the composition to begin with.
(If anyone wants to check them out, the lowest notes are around 1:30.)

I find the song pleasing btw :)
 
Last edited:
Thank you for mentioning the Izotope-plugin! Will check that out, sounds really handy.

I've been demoing it for the past few days, and it's actually really handy. I'm not using any of the integrated features with Ozone or Neutron, but it's a great guide. Might be worth upgrading Ozone for? Maybe.
 
I'll check it out too :)



I am by far no expert in orchestral music, but as the spectral balance goes, I would say: there are too much low mids, and not enough highs. So it sounds kinda 'closed'.

Maybe correcting the low mids will automatically balance the highs... Start by trimming the former before boosting the latter.

Since the third quater of the song (from 2:00 to 3:00) is better balanced spectrally (it sounds more open, with enough highs), I would advise EQinq tracks individually before trying to EQ the whole master. There are good tutorials for that, e.g. on Youtube.

It's hard to judge the low freqs, because there's not a lot of them in the composition to begin with.
(If anyone wants to check them out, the lowest notes are around 1:30.)

I find the song pleasing btw :)
Great! Will take you up on these advices, and try to mix it again, instrument by instrument. Will try to post the new version sometime tonight or tomorrow. Glad you found it pleasing!
 
So, hold on - I feel the opposite about @Fredeke 's "low mid" feedback =) But maybe not in a contradictory way.

First off the music itself and performances of the instruments are beautiful. Nice orchestration too. the words "provocative tranquility" are all I can think of to describe it.

For the mix - to me there's an entire octave of richness that is missing from low frequencies. For example - I can barely hear some pizzicato basses at around 0:45. They do exist, right? or am I imagining it? Bringing those out a LOT will help greatly. I feel that the "bed of strings" is not rich enough and needs more fullness. I can't tell if it's an composition/orchestration question or a mixing one, maybe those notes don't yet exist and I'm just yearning to hear them here.

The cymbal swell is another example that's easier to verbalize what I feel is wrong - it doesn't feel like it's place correctly in a stereo image. It sounds panned, but then the natural reverb I would expect to hear is not filling the entire room. Are you by chance panning your post-reverb signals? Are you using virtual instruments, and you can pan/position them before putting them through reverb? and also, are you using a true-stereo reverb (i.e. each input channel is processed twice, one for each output channel)? I'd recommend trying that for at least some of the instruments, or maybe overall.

Are you using dry virtual instruments? Another thing that helps magically make instruments melt into place is to EQ them to simulate the correct distance that I think you had intended. When I do this for myself, I find that the EQ curve usually looks like roughly straight-line sloped so that lower freqencies are reduced and the slope contines all the way past 600-800 Hz. I know this sounds a little bit contradictory to my complaint that there's not enough low frequency present, but it's not - the difference is that you can EQ individual instruments to have a certain tone, but you may still need to adjust the levels of some of the low bass instruments and mid-low instruments, too.
 
Just to add, so that you don't overdo it based on our feedback - a lot of the mix is already lovely and spot on. My suggestions about EQ and levels might only apply to the strings. The piano and woodwinds and bell like layers seemed to work really well together already.
 
For the mix - to me there's an entire octave of richness that is missing from low frequencies. For example - I can barely hear some pizzicato basses at around 0:45. They do exist, right? or am I imagining it? Bringing those out a LOT will help greatly. I feel that the "bed of strings" is not rich enough and needs more fullness. I can't tell if it's an composition/orchestration question or a mixing one, maybe those notes don't yet exist and I'm just yearning to hear them here.
We might indeed not contradict each other. I was not talking from a composition/arrangement point of view, and I wasn't talking about balance (I didn't pay much attention to that). I was merely talking about EQing, because I thought that was the question.

the difference is that you can EQ individual instruments to have a certain tone
Yes. I didn't want to get too much into that, but it's a good idea to reserve a specific band of the spectrum for each instrument - but less caricaturally than I just made it sound...

What I meant is I think several instruments deserve a dip in the low freqs and low mids (or higher lows ;)) : certainly some strings, maybe the piano, and maybe some other stuff too...

But it actually sounds like you (@beyd770) didn't EQ anything (I may be completely wrong, but that's my first impression), and if so, the result is a tribute to the quality of your instruments. Nevertheless, you have to EQ each track. You fashion the sonic landscape as much with EQs as with levels or panning.
In fact, when each instrument's EQed timbre fits into the others harmoniously (mainly by not stepping on each other's spectral toes), the right level balance is much easier to find - it practically finds itself.

don't overdo it based on our feedback
Indeed.

I wish I could compose a piece like that.
 
Last edited:
So, hold on - I feel the opposite about @Fredeke 's "low mid" feedback =) But maybe not in a contradictory way.

First off the music itself and performances of the instruments are beautiful. Nice orchestration too. the words "provocative tranquility" are all I can think of to describe it.

For the mix - to me there's an entire octave of richness that is missing from low frequencies. For example - I can barely hear some pizzicato basses at around 0:45. They do exist, right? or am I imagining it? Bringing those out a LOT will help greatly. I feel that the "bed of strings" is not rich enough and needs more fullness. I can't tell if it's an composition/orchestration question or a mixing one, maybe those notes don't yet exist and I'm just yearning to hear them here.

The cymbal swell is another example that's easier to verbalize what I feel is wrong - it doesn't feel like it's place correctly in a stereo image. It sounds panned, but then the natural reverb I would expect to hear is not filling the entire room. Are you by chance panning your post-reverb signals? Are you using virtual instruments, and you can pan/position them before putting them through reverb? and also, are you using a true-stereo reverb (i.e. each input channel is processed twice, one for each output channel)? I'd recommend trying that for at least some of the instruments, or maybe overall.

Are you using dry virtual instruments? Another thing that helps magically make instruments melt into place is to EQ them to simulate the correct distance that I think you had intended. When I do this for myself, I find that the EQ curve usually looks like roughly straight-line sloped so that lower freqencies are reduced and the slope contines all the way past 600-800 Hz. I know this sounds a little bit contradictory to my complaint that there's not enough low frequency present, but it's not - the difference is that you can EQ individual instruments to have a certain tone, but you may still need to adjust the levels of some of the low bass instruments and mid-low instruments, too.
First off - provocative tranquility - I can _really_ live with that description, thanks a bunch!

Correct, there are indeed pizz bass going on in the background, but sparsely. On my mix they are already quite loud, so something is certainly a bit off with the mix. And it's true that a lot of the tones in the lower register is not there, since I've only included bass strings, and tuba / trombones in small parts of the track. Will really look into how to make that stand out a bit more.

Interesting with the observation on the cymbals, I struggled quite a lot with them. I have used SF Percussion, but the swells are Loops de la Creme, and they are recorded smack in the middle, so I used Viritual Sound Stage to place them according to how the Spitfire-cymbals are positioned. Will try to remove this artificial placing, and see how it affects the overall sound.

No completely dry viritual instruments, but some lack the proper mic-options, and had to be treated with some extra reverb to sound more back on the room. But I'm still very new til all this, so might have screwed up the soundstage while trying to place them correctly.

Thanks again for taking your time to write such good feedback!
 
We might indeed not contradict each other. I was not talking from a composition/arrangement point of view, and I wasn't talking about balance (I didn't pay much attention to that). I was merely talking about EQing, because I thought that was the question.


Yes. I didn't want to get too much into that, but it's a good idea to reserve a specific band of the spectrum for each instrument - but less caricaturally than I just made it sound...

What I meant is I think several instruments deserve a dip in the low freqs and low mids (or higher lows ;)) : certainly some strings, maybe the piano, and maybe some other stuff too...

But it actually sounds like you (@beyd770) didn't EQ anything (I may be completely wrong, but that's my first impression), and if so, the result is a tribute to the quality of your instruments. Nevertheless, you have to EQ each track. You fashion the sonic landscape as much with EQs as with levels or panning.
In fact, when each instrument's EQed timbre fits into the others harmoniously (mainly by not stepping on each other's spectral toes), the right level balance is much easier to find - it practically finds itself.


Indeed.

I wish I could compose a piece like that.
Duly noted, I already am hearing something related to strings and piano, so will try to adress that with some EQ and volume-tweaking.

But it's true - the original is not treated with EQ at all. I am very new to this, and usually when I try to scope out the "good stuff", it often sounds to muddy or to nasaly, so I end up with reverting to no EQ, which also is the case with this track.

I have also noticed that there are a lot of collision from instruments in the same mid-low range in the part we already talked about, so I have decided to alter some of the orchestration as a start, to see if that can deal with some of the sonic aspects of it, before I start cutting like a madman :P

Thanks again for write such nice comments, I'm learning a ton!
 
Top Bottom