Expert opinion on my mix needed!

[Marcello]

hi guys!

i really need an opinion from some expert on my mix cause I’m mixing at home with headphones on my own and I don’t have external feedbacks on my work.

I’m quite a newbie and this is my song that I did with programmed drums and guitars amp sims. 
I then used LANDR AI automatic mastering platform online, you just drop your song and comes out the mastered version, I used a reference track from Mogwai since I play post rock stuff.

do you think the mix version is better than the master? Someone from a forum told me this, the master is too compressed and too much in you face, also the drums disappear in the master. He also claims that is confusing and I should EQ more to make space for different instrument frequencies.

also around minute 4, the instruments are a bit out of time.

what do you think? Is he right? Do you confirm all these things?

I don’t pretend to have a professional mix cause I’m doing everything at home, but I just want it good enough.

 

Mastered

https://drive.google.com/file/d/1IZN1eH2jW2s5eJMiFKTG7XLs0AShSLdN/view?usp=sharing

MIX only

https://drive.google.com/file/d/12l7wh6Nrken01P-DP8AdZ-PzrxO_Yn2p/view?usp=sharing

[Wookiee]

I don't claim to be an expert but I have had a listen which leaves me with some questions.

What levels are you recording at? i.e. how are you gain staging this.

Is the acoustic guitar intentionally distorted?

Looking at the wave file in Soundforge Pro 15 the overall levels when the whole band kicks in looks about right for pre mastering levels. Prior to the the signal is very low.

Examining the frequency spread/content of the wave using Span shows that is little below about 80hZ with a very steep roll off and equally very little above 5kHz with an equally steep roll off.

[Bajan Blue]

Marcello

The mix seems to be a wav file while the mastered version is MP3 - if so you can't really compare them in my opinion.

The mix version sounds OK to me

The mastered version sounds very brittle to me - very top end - but as I think I said before, I wouldn't use LANDR anyway!

Hope that helps

Nigel

 

[Marcello]

26 minutes ago, Wookiee said:

I don't claim to be an expert but I have had a listen which leaves me with some questions.

What levels are you recording at? i.e. how are you gain staging this.

Is the acoustic guitar intentionally distorted?

Looking at the wave file in Soundforge Pro 15 the overall levels when the whole band kicks in looks about right for pre mastering levels. Prior to the the signal is very low.

Examining the frequency spread/content of the wave using Span shows that is little below about 80hZ with a very steep roll off and equally very little above 5kHz with an equally steep roll off.

So The acoustic guitar is intentionally distorted a bit cause I like it that way,  otherwise would sound too thin, and weak personal taste.

What you are saying is that the solo guitar part at the beginning is too quiet (in volume) compared to the part where the band starts? I tried to give some dynamics there, maybe too much?

About the frequencies, did you check the mix or the master?

Is the master too compressed compared to the mix?

The mix to me sounds a bit muddier than the master, I agree with Bayan though  tthat the master might be too bright? or maybe is just too compressed by LANDR?

Edited May 22, 2021 by Marcello

[Wookiee]

1 hour ago, Marcello said:

About the frequencies, did you check the mix or the master?

As I said I do not consider myself anything but a Wookiee Jedi 😎

I checked the original wave track as I thought this is the same track s you showed the other day plus I doubt the master can add that much to the frequency range.

An MP3 is going to be frequency reduced by default but a quick look again in SPAN shows that there is little below about 80hZ with a very steep roll off and equally very little above 5kHz with an equally steep roll off.

The MP3 master is also showing a 3 to 5dB difference between left and right hand signal level with the left being higher.

As Nigel says comparing a wave file with a "mastered by AI computer algorithm" is not comparing like for like.

See the output from Youlean pro below the red line at the top tells me there is clipping with the true peak hitting +0.2 means your are hitting the endstops and over running by a small amount. On poor playback gear that would sound bad.  Commercial streaming sites are going to crush it back to -14 LUFS not the -8 LUFS

Ideally I always aim for a max of -1.0 dB True Peak for streaming and an integrated level of around -14 to -16dB not the -8.0 you are hitting.