First, I wouldn't worry too much about playability on a mobile phone. It's definitely a consideration, but usually, if you get the mix right for a good system, it will work out okay on a phone, and the listener will be used to their phone's deficiencies.

Before I could say much about the balance, I'd need to address some tonal issues. The vocal and piano have too much midrange energy. The piano seems to have nothing going on below about 300Hz. The vocal seems to be too strong around 800-1000Hz and lacking in clarity in the highs. A 2 octave 6dB cut at 850Hz and a 2 octave 6dB boost at 8kHz helped the vocals, and didn't hurt the piano, but it didn't address the lack of lows from the piano.

I suspect that once you address the tonal issues, the piano can come up a bit. You might need to automate the bridge section to keep the levels a bit more even.

Another thing to look at is the overall level. On my LUFS meter, your maximum True Peak level is -5.7dB. That means you could just bring the overall volume up 4.7dB without having to do any compression or limiting. Since the LUFS integrated level is 21.9, that would get you up to 17.2 LUFS. With a bit of judicious limiting, and perhaps that automation in the break, you should be able to get it up to a more standard 14 LUFS without much loss of dynamics.

Attached files

Fabulous.... (y) just what i needed, will get on it tomorrow :D and come back with version 4.5

You're a star really appreciate it.

Note that the eq changes will make all that about LUFS and True Peak levels change, but the principle is the same. You left a lot of volume on the table. You reclaim that without giving up much in the way of dynamics.

bouldersound, post: 467261, member: 38959 wrote:
Another thing to look at is the overall level. On my LUFS meter, your maximum True Peak level is -5.7dB. That means you could just bring the overall volume up 4.7dB without having to do any compression or limiting. Since the LUFS integrated level is 21.9, that would get you up to 17.2 LUFS. With a bit of judicious limiting, and perhaps that automation in the break, you should be able to get it up to a more standard 14 LUFS without much loss of dynamics.

True Peak being max signal level where an increase of 4.7 would have him -1 at max peak.....? I get that

Integrated LUFS --- at -21.9 means average signal - and assuming its linear then raising the ceiling to -1 by 4.7 you get -17.2 average signal -- then compressing or limiting the peaks he could boost the signal more till it averages out (though this boost with limiting and compression is not linear in the same way... am i understanding this correctly?

DogsoverLava, post: 467271, member: 48175 wrote:
True Peak being max signal level where an increase of 4.7 would have him -1 at max peak.....? I get that

Integrated LUFS --- at -21.9 means average signal - and assuming its linear then raising the ceiling to -1 by 4.7 you get -17.2 average signal -- then compressing or limiting the peaks he could boost the signal more till it averages out (though this boost with limiting and compression is not linear in the same way... am i understanding this correctly?

Yep. There's no sonic magic about 14 LUFS, but that's kind of the standard (with variations) for streaming platforms.

I think your track sounds fabulous! Maybe a bit too much reverb for my taste but really good.
Listening on my iPhone.

audiokid, post: 467273, member: 1 wrote:
I think your track sounds fabulous! Maybe a bit too much reverb for my taste but really good.
Listening on my iPhone.

I agree. I think it's more about the reverb times and the predelay. Maybe a shorter dampening on the piano

Thanks guys for your comments....

Boldersounds - As I mentioned earlier very helpful to get some specifics on EQ and levels - hopefully you'll notice a difference - good ears to pick up those things.. impressed!

Dogsoverlava - Thanks for that... I'm just gonna say I'm not technically minded to be able to picked up on the math.. especially on LUFS especially integrated ones, but I do have a plugin which helps me get near to the 14LUFS target... appreciate your comments (its a work-on for me)

Audiokid - Thanks for that good to get that kinda feedback too I didn't pick up on your reverb comment till id been ready to print off n back to here.. I did mess with reverb but whether it needs more reduction i dunno .. but good to get you thoughts.. on it now

Davedog - Thanks again... picked up your point after id printed off for the day so might go back to that tomorrow - I think there's a piano damping option I can look at

So what I've done today..

Taken on board Boldersounds technical adjustments, I was able to hear the difference which has been very helpful...

1. Went back to the mix - Found the dynamic EQ I used to avoid the VOX masking the piano (Just dipping certain freq) needed adjusting- released below 300Hz bass part of piano came back in (at least I think so)
2. EQ Adjusting the 800-1000K and 8khz levels etc...made a massive difference to the Vox - so I went back to the mix and made my adjustments there too.
3. Started looking at VOX transients and used some automation to in the mix to flatten them out
4. Dropped the mix reverb back a tad - (Might need to revisit this)
5. Brought the piano up a tad
6. Mastering- Used Izotope Insight to help me get closer to that 14LUFS
7. Went back n forward with all -bit of an iterative process the way I work but much happier with this version.. Still happy to get more feedback I've probs solved one problem and caused a few others

   Version 4.5

    
   Attached files
   

I just discovered that my default channel configuration had the pan law set to -3dB, so all my numbers above are 3dB off. I had just changed the default a couple of days ago and it tripped me up.

I think this version is better. Maybe dial back on the vocal HF boost and mid cut a bit. In parts the piano seemed to step on the vocal.

The loudness numbers came out to 14.8 LUFS and -0.9 dB True Peak. That's probably close enough. Worst case, a streaming platform limits it down 0.9 dB or so. One of the reasons to leave some headroom is that converting to a lossy compressed format can raise the peaks slightly. So if you mastered it with -1 dB True Peak, going to mp3 might have shifted it up.

Hey Thanks for the prompt feedback

Your a star...

Will take another look tomorrow.

Pleased it sounds better.. I tend to lose perspective after a while. I dont do enough mixing to have the ears you guys seem to have.

I have a few piano ballads on the go so am hoping will be able to use this one as a reference track for the rest so am happy to keep at it till its right..

I did master it at -1db, do you think I should leave more room... say -1.4db ?

Cheers

Well now...Thats so much better separation. I'm still hearing a reverb clash. It's minor. Are you using a global verb on the mix or are you using separate verbs on the piano and the voice? Other than that its really good.

Mach13, post: 467278, member: 52472 wrote:
I did master it at -1db, do you think I should leave more room... say -1.4db ?

If the wave file is coming out at -1 dB True Peak, I wouldn't worry about it. Leaving that dB of headroom accommodates the slight increase you often get with compressed formats, which is probably what happened here. I think it all worked out just the way it's supposed to.

Davedog, post: 467279, member: 4495 wrote:
Well now...Thats so much better separation. I'm still hearing a reverb clash. It's minor. Are you using a global verb on the mix or are you using separate verbs on the piano and the voice? Other than that its really good.

Ha erm well ☺️ agaim an impressed... thinking this is another thing im going to learn on here, ( feels like a genuine master class ) but i have to admit:
vox and indtrument each have their own verb, the mix track has one and there was a slight touch of one on the master erm... so i will be adressing that today too..

Is it better to have one global one on the mix?

What do you suggest?

KUDOS to you guys for your help, im learning loads ☺️

The traditional way to apply reverb is by way of a parallel effects bus (as opposed to a series mix bus). How you implement that varies with the software you're using, but it generally involves a post fader send from each track to a shared reverb which is set to 100% effect (no unaffected audio passing through). Then you can adjust the sends from tracks to have more or less reverb on each. It's less load on the CPU, which with only two tracks is probably not a major consideration. If you want different styles of reverb on different tracks, either put the reverb right on the track or make more effects buses.

bouldersound, post: 467285, member: 38959 wrote:
The traditional way to apply reverb is by way of a parallel effects bus (as opposed to a series mix bus). How you implement that varies with the software you're using, but it generally involves a post fader send from each track to a shared reverb which is set to 100% effect (no unaffected audio passing through). Then you can adjust the sends from tracks to have more or less reverb on each. It's less load on the CPU, which with only two tracks is probably not a major consideration. If you want different styles of reverb on different tracks, either put the reverb right on the track or make more effects buses.

Agree with this 110%.

Class 101-reverb.

Reverb is time. Adding time signatures at different points of a mix like you have done is a very tricky balancing act.
Reverb is space. Defining an area the recording has been captured in in order to give the instruments a sense of dimension inside a geometrical figure is the goal.
Reverb is seasoning. Just like the dry rub for the wings, too much IS too much. How much is too much? ThAt much.

The absolute hardest recordings to add the right touch of reverb to is solo instrument with vocal.

You may ask..."Why Uncle Davedog is this so?" And here's why.......
When you use a global single verb on this type of capture, you risk having the reverb tail being too long for the sustain being used on the instrument (in this case the damping pedal on piano and holding the notes) and creating an odd feeling of disconnect for the vocal to the instrumental. But you WANT a type of verb to create the SPACE that the recording sounds like it was made in as one complete performance. For this task you will want to look to Convolution types of reverbs to establish the size and depth of the overall space you want the recording to sit in. Careful. It won't take much and the time stamp being used is going to determine the side-chained individual verbs of the instrument and voice as described by 'bouldersound'.
Having a side bus control of these does two things specifically to the sound.
1. And this is true for any chain of effects on a track....driving a time-based device directly in-line on a track will affect its triggering for the effect due to the differing amounts of input from the instrument or vocal. This is a good rule to follow when choosing the order of your plug-ins in ANY case.
2. The separate bus allows the time-based effect to be entered into the mix at whatever point you choose. And it gives you much more finite control of how much and at what time it shows up. Experimentation with this will open a large garage door in your search.

So. In summary and from my experience with these sort of mixes....I feel like you DO need a global verb inserted before the master slider and controlled through a submaster sidechain. This should be the convolution type of verb that creates and defines a "space" for the entire mix.
You DO need a separate and defined verb for the piano. A nice plate with HPF and LPF, diffusion, and depth controls plus amount will work here.
You DO need something for the voice that is time based but is complimentary to the voice as an instrument yet isn't competing for space with the room verb or the plate on the piano. So a delay of some sort set in time to the meter of the song. And just used as a suggestion of size for the voice.

Have fun. You are in a laboratory. Break a few glass beakers.

I echo the encouragement to experiment, and I use special reverbs on occasion. But my approach is to resist adding complexity until I have a compelling reason. So I'll work with a single shared reverb as long as possible, then resort to special reverbs for specific tracks when that just doesn't work. Even then, I generally put the reverb on an effects bus rather than on the track directly, which allows things like eq and compression in series with the reverb.

Ha that's been an 'interesting' day :X3: So glad I try to keep to a strict versioning process... Changed the verbs in the mix and it played havoc with whatever I thought I had. The mix completely got away from me... haha..

So the upshot is I'm gonna quit whilst I have a reasonable print and say that I when I start something new I'll rethink my verb strategy at that point. Barring starting from scratch (which is what I think id have to do, I'm gonna save this project as is and maybe come back to it one day sometime when I'm not so sick of listening to it any more :ROFLMAO:;)

Thanks for your help folks- You've all been incredibly helpful... I feel I learnt more in the past few days than weeks of bimbling about on my own would have ever produced... and at least I have a reference to work with that will help me in the future. (y) (y) (y) (y) :giggle:

Would you be open to letting one or more of us have a shot at mixing this? I could do it in Reaper and send you the project file so you could see exactly what I did.

Wow thats an amazing offer.. Sure what do you need?

And how do I do that?

sounds great, decent song. In The second version the voice cuts better, but is still feels like it is missing a bit of body.

I dont use Reaper tho.. (I use Sonar with Izotope Ozone and Neutron plugins) theres a bandl labs free version of the one I'd bought from Cakewalk.. So not sure how to see the Reaper project file.... But if I could see the EQing that would help loads...

Thanks again...amazing offer to do that

It will depend a bit on what software you're using and how you recorded the tracks. Did you record this live straight through or did you record it in parts? Voice and piano together or piano first then voice? I suspect the piano at least is a single take since I didn't hear anything that sounded like an edit, but I could be wrong.

The goal is to end up with two files, piano and voice, that start from the same point on the timeline and have little or no processing on them. If it's a live take of both voice and piano, you should be able to find the source audio files to send. Even if you recorded them separately but started the second track all the way through from the same spot, that should work.

If you, for example, recorded the voice in parts, you'll have to generate new files that start together. Probably the simplest thing to do is to save the project with a new name, then strip off all the effects, put the faders to zero, then mute the voice and export the song. After that, unmute the voice and mute the piano and export the song from the same point in the timeline. Ideally you'll export at the same setting (bit depth/sample rate) as you recorded.

Then use a cloud transfer deal like Dropbox or Google Drive to send them to me. I can send back a project file. If you put that in the folder with your copies of the files you sent me, it should open them up with my mix settings. You would have to install Reaper, which you can use for free for a limited time (after which you can still use it, but it has a nag screen). Then you can inspect everything I did on the mix: eq, routing, effects settings etc.

[Edit] I don't know if there's someone here who knows Sonar, but it may have features like Pro Tools' Consolidate and Export. That would be a bit simpler than all the stuff I just explained.

Oopsie! I see we've had a reset n lost some of the chat on here, thankfully I've noted most of it so that's cool :cool:

Bouldersounds - You are such a generous guy to do that work on the stems for me... To miss quote an old Jonny Nash lyric.. 'I can hear clearly now the crud has gone..' :giggle:

I'm gonna have a go an emulating your settings from Reaper and see (or hear) what comes.. I'll post it later when I'm done (y)

Thanks for your help with this...

Here's the version I've now been able come up with (Bouldersounds mix eq's etc has been very helpful)

I've remixed from scratch whilst referencing Bouldersound's version

I'm gonna put them all here for ease of referencing the versions

First one Version 4.4
 
Second Version 4-5
 
Most recent Version 4-6
 

I'm not expecting it to be without the need for more tweaks and your comments are most welcome.

Always appreciated

Last update

Davedog made a good point about my lack of 'enunciation' on 4.6 and of course he was spot on with that, so I went back and sorted what I could... And of course fiddled some more on the piano whilst I was there

So this is the last version :X3: Version 4-7
No it really is... been a good learning curve for me on here thanks...
 

Attached files Marks Thank You All BSG mix.mp3 (6.9 MB)  Thank You All (Covid 19) Forum Fix V 4-5.mp3 (6.8 MB)  Thank You All (Covid 19) V 4-4.mp3 (6.9 MB)  Thank You All Marks BSG Mix V4.7.mp3 (6.8 MB) 

bouldersound, post: 467338, member: 38959 wrote:
I don't know if there's someone here who knows Sonar, but it may have features like Pro Tools' Consolidate and Export. That would be a bit simpler than all the stuff I just explained.

It does and the missing post from the reset explains that.

I thought the piano sounded 'better' in the second version...4-5....

The vocal is better also throughout but still some 'murkiness' in the defining of the lyric. But thats just me. I like to hear the words.

I no longer hear the verbs colliding around and washing out the clarity. Good job

Davedog, post: 467362, member: 4495 wrote:
It does and the missing post from the reset explains that.

I thought the piano sounded 'better' in the second version...4-5....

The vocal is better also throughout but still some 'murkiness' in the defining of the lyric. But thats just me. I like to hear the words.

I no longer hear the verbs colliding around and washing out the clarity. Good job

Thanks Davedog you're abs right of course.... I went back to try n fix the enunciation with Version 4.7 (above).. (y)

Now that we are coming to an end with this nightmare, how is everyone feeling?