Hi! I analyzed with sonible truebalance some current top 100 tracks, but also in ozone or SPAN I get the same measurements.
It immediately jumps out at me this rolloff from 10khz and up that is always present, and I actually have to say that by trying it on my tracks I immediately get a cleaner and more defined sound in the small headphones or cubes.
Is this an intervention that you are used to doing? Is there any specific reason for it?


I attach some examples for you, thank you!

(All the files are in FLAC, i tried with WAV too. It's not the typical mp3 or youtube rip cut)

 

Last edited: Jul 10, 2024

For 80% of ears and reproduction systems out there, yeah, probably, but I'd never voluntarily do that. 20Khz, sure. I expect a fun thread anyway.

 

At first glance I thought those were graphs of my last hearing test.

 

Ouch

 

It is relevant. First, you are looking at Rolloff, not a "cut". You do not have all that many instruments up there in the first place.

Google this "good eq curve for mastering pop music" and hit the images filter. Look at the frequencies of the actual waveforms in the images you see of EQ displays. Don't even look at the nodes, just the real frequencies displayed before they are even EQ'd. It's the same thing, because there is generally not all that much information in that range without changing anything at all.

Also, you are looking at the sum of all the individual tracks, not individual tracks. Your individual channel will show information up there; but they are not summing to become a considerable portion of your amplitude. That's always bass and mids anyway.

 

hmm, interesting... it seems it is like that, im following a dj/producer of my taste and he do something similar in his videos or recommendations lets say... never though of this!!

 

today i know this eh eh....but why ? is cutting steep messing up phase ? is it psychoacustic ? is it mild air ?

 

I mean, some of the tracks might naturally have a roll-off there

 

I remember Jaycen Joshua tell once he would roll off the top end of the snares and only unmute and balance the hihats untill he had the top end of his vocal done. I mean, you shouldn't annihilate the top end of a song with a filter on the stereo buss, but as @clone said, it's not an area full of information, so you should pick well whats going up there, because too much tends to fatigate your ear, especially if you are young.

Spoiler: bonus

Aliasing comes from intermodulation around nyquist and this scaries a lot of folks

 

I agree, if rolling off at 10k somehow makes a mix better or clearer than there are probably things going on somewhere on individual instruments that need to be rolled off / dealt with individually that should be addressed. But don't nuke it all. If you're using an acoustic drum kit, say, or facsimile thereof chances are you can leave that high-end sparkle in your overheads but kill it on dang near everything else on the kit...

 

Last edited: Jul 10, 2024

where did you get these flac or WAV files?

 

It's mostly hiss, air, random "Extra transients", and other noise. How many of the instruments ranges shown up there on a frequency chart are often used in Top 100 songs? The more transparent portions of frequency ranges show where there is potentially information per instrument. It's not generally where there is any melodic content fundamentals. But it's not going to be junk generated by stuff you are doing at your groups, buses, or master. It's from your recordings, samples, earlier plugin use, from before your buses.

He also makes mention of using high frequencies like this to create depth or layers, and other psychoacoustic tricks and things. But one thing he does make a point of is "I use the 2 bus to make it sound good". There is no way he's putting a high cut at 10K on any project, never mind all of them.

 

Last edited: Jul 10, 2024

If you have a bunch of instruments and do nothing to them at all you get this "roll off", because it is natural. in the range there are "only" the harmonics of the instruments, and harmonics are lower in volume than the fundamentals of the instruments. These "roll offs" are no roll offs, they are the natural appearance of a "normal" received instruments.

My, what times. Not everything in audio is deliberately "constructed". Have a listen to 70 or 80 year old recordings. You will find the exact same things in the frequency response.

Urgent: get iZotope's Tonal Balance, make some profiles of complete albums (!!! complete albums !!!), put Tonal Balance on your monitor bus (Reaper) and mix according to the profile you like the most. Forget about that referencing of one (!!!) track that most of the times doesn't even fit and has completely different sounds and approach overall, you make yourself a madness hell. Check your mix against Tonal Balance. And never (NEVER!!!) against a single track. Always (ALWAYS!!!) against an at least whole album. (Yeah, an album is an album is ancient entity, consists mostly of a bunch of tracks, that, properly mastered, sound coherent but not equal. In ancient time we made this way sure that the listening experience lasted longer than 10 seconds and was enjoyable.)

Beside the fact that this frequency response game gets you 90% out of all loudness problems, because your tracks fit the natural human hearing.

 

Well yeah, when recording "old school" instruments the problem has a tendency to solve itself (or not be a problem, depending on how you look at it). As clone mentions, with modern sampled reproduction of the same instruments though you can wind up with a lot of arbitrary hiss and crud that's in the sampling that really adds up... like, Superior Drummer, as much as I love it, can be a nightmare for that kind of stuff. A certain amount of that can give a mix that old school feel but it can get out of control.

 

Thin construction and with concessions to loudness and normalization.

Even the best recorded cymbals and high hats sound like shit if you push shit and just having the meat of the song without the extended high end makes it easier to limit and control ISP.

Just take those lossless tracks and push up with some high shelf filters. Listen to what you are boosting and what it sounds like.

If you just have to have -8LUFS to - 6LUFS then this is just a tool to get you there.

 

I was going to say as we're close to the same age that you're doing better than I am
This chart helps if in doubt but I still use my ears

 

Last edited by a moderator: Jul 10, 2024

It's not that surprising if you know how sound reacts to our ears

 

It may be natural, but it is also a behavior of a filter or EQ. It is attenuation. What it is not, is a "cut". If you have a better term for it someone will understand, then please deploy it. For some reason I "missed that part". The OP is about "top 100" tracks. Are you trying to suggest the frequency range in question did not pass through an EQ? I'm sure they do that all the time on popular music.

Enjoy some reading: https://www.digikey.com/en/resource...sion-calculator-low-pass-and-high-pass-filter

 

Last edited: Jul 10, 2024

"top 100 tracks" but coming from...? ripped from a compilation CD? downloaded from a tracker?

can we really judge those frequency analysis without knowing the exact origin of the files?

I ask x2 : )

 

Let me guess, you downloaded the mp3s of these songs and "analyzed" those?
This rolloff is a 128kbps .mp3 file.
And if you try and mix the song without 10kHz and above - the same streaming encoding will eat up even more of the highs.
And if you recode said mp3 to a WAV - the junky high stays.

 

Last edited: Jul 10, 2024