The No.1 Website for Pro Audio
Peak Headroom
Old 6th August 2020
  #91
Lives for gear
 
Hippocratic Mastering's Avatar
Quote:
Originally Posted by Rukas View Post
Side note, is FabFilter L2 reaaalllly worth getting over Waves L2?
Is it really worth getting a Boeing 747 over The Wright Brothers’ 1903 double-winged prototype?
Old 6th August 2020
  #92
Motown legend
 
Bob Olhsson's Avatar
 

Verified Member
High-end D to As are way less susceptible to isp distortion than most consumer devices.
Old 6th August 2020
  #93
Quote:
Originally Posted by Bob Olhsson View Post
High-end D to As are way less susceptible to isp distortion than most consumer devices.
100% agreed. That's the reason why I used my iPhone/headphones to do the listening tests I described earlier in the thread.
Old 6th August 2020
  #94
Lives for gear
 

Verified Member
Quote:
Originally Posted by Rukas View Post
Side note, is FabFilter L2 reaaalllly worth getting over Waves L2?
well, there's something to the L2 that one may like or not like...

(many people have come to dislike it altogether)

...but it's the same as with every tool: if you want it to behave as the original, then better get the original unless there's an equivalent - not everything got copied/re-modeled/emulated (yet) though!
Old 7th August 2020
  #95
Gear Head
 
Bongo Spank's Avatar
 

I was just searching for an article I read a year or so ago that went into detail about the exact situations in which the headroom mattered. I vaguely remember something about a certain combination of factors in Spotify like if someone has the matching off, and is listening on a certain type of device, or something like that where the combination created an audible issue.

Can't find the article at the moment, though.

Launching a new artist after being off on other projects for a while, and revisiting this notion since I too have been doing a little digging lately into the headroom of some references. I did -0.3 for years, then was convinced to switch to -1 TP, but I don't think I've found a single one of the references I looked up that's doing the same. In fact, most are pretty darn close to 0dB on a PCM reading.

Basically, I've taken it pretty much on faith that the issue exists since I have yet to hear a particular issue that I can definitively trace to ISP's. Since this discussion sems to have wandered, I thought I'd invite anyone who might have done any experiments or have anything to add (or who can recall the article I'm trying to find).

I'm doing the typical one master thing, so I certainly don't want to take a dB hit in all those places that aren't matching if it's not actually helping, so I've got two basic questions:

1) Who has looked into this recently, and what have you settled on for headroom in pop related genres?

2) I don't have an iPhone to check, but last time I looked into it (I think it was a year or so ago), Apple Music was still single largest point of consumption for US customers, and had loudness matching disabled by default so the whole normalization conversation didn't even apply to the largest single slice of consumers in US. Is this still the case?
Old 7th August 2020
  #96
Yes Apple Music is the most widely used, followed by Spotify as a close second.

Yes Sound Check is off by default.

But loudness normalization revolves around LUFS etc. and has little to do with peak headroom because in the vast majority of cases tracks are normalized lower, so peak headroom is actually increased when the tracks are turned down. Loudness and peak headroom are 2 different discussions. You can make a -14 LUFS master that clips or you can make a -6 LUFS master that peaks at -1.
Old 7th August 2020
  #97
Gear Head
 
Bongo Spank's Avatar
 

Quote:
Originally Posted by Trakworx View Post
Yes Apple Music is the most widely used, followed by Spotify as a close second.

Yes Sound Check is off by default.

But loudness normalization revolves around LUFS etc. and has little to do with peak headroom because in the vast majority of cases tracks are normalized lower, so peak headroom is actually increased when the tracks are turned down. Loudness and peak headroom are 2 different discussions. You can make a -14 LUFS master that clips or you can make a -6 LUFS master that peaks at -1.
Yeah, I get that. My most recent ballpark has been -10 LUFS (integrated whole song), but that's actually a PLR of 9 with 1dB TP haadroom.

I've just been taking it on faith for a while now that the added headroom is a good thing despite growing evidence that no one outside of this forum cares, and not having heard the issue personally.

I appreciate that others also bother to take the time and effort to run controlled experiments, and tend to take their results on faith unless given a reason not to.

Relative loudness does have to do with headroom in the sense that without normalization engaged, that headroom translates to a direct drop in relative loudness. In fact, if there's a 1 dB of difference between the PCM value and the TP, then it's actually a 2dB penalty since my loudest peak is down at -1TP whereas I see lots of current pop masters at roughly 0 as a pcm reading.

If I were to actually follow the Spotify guidelines of -2TP, then that's potentially a 3dB penalty on Apple Music, for instance.

I'm not hung up either way. Like I said, I did -.3 for years, and have been doing -1TP for the last year and a half or so.

I'm not opposed to headroom that IF it really matters, or a significant number of people end up with distortion, but I'm just not seeing any evidence in the comparables I've dug up that any of those ME's cared.

I'm also not saying the issue doesn't exist... just that I've yet to personally hear a particular downstream issue I could specifically attribute to the ISP's.
Old 7th August 2020
  #98
Some confusing logic in there that I'll just let pass...


At the end of your post you're basically repeating the question that's the subject of this thread. Assuming you already read the thread, I guess you just want more recent info, because there's a lot of detailed discussion here from less than 2 years ago.
Old 7th August 2020
  #99
Gear Head
 
Bongo Spank's Avatar
 

Quote:
Originally Posted by Trakworx View Post
Some confusing logic in there that I'll just let pass...


At the end of your post you're basically repeating the question that's the subject of this thread. Assuming you already read the thread, I guess you just want more recent info, because there's a lot of detailed discussion here from less than 2 years ago.
Yes, I stated that. I've read the thread, but I'm not really seeing a particular resolution to the question... nor am I able at the moment to find the article I read a while back that goes into greater detail and gives essentially a recipe for the exact combination of factors that make the issue audible for a certain segment of listeners. I don't see it referenced here, and I'm still wondering if anyone involved has found a particular resolution since, or some specific scenerio I can test that will clearly demonstrate that the issue does exist in the real world so I can make a more informed decision than simply taking it on faith as I have been.

If I have a PCM reading at 0dB, and a TP reading a dB above that, then I turn down to match Spotify guidelines of -2TP, then that means my PCM now reads -3, and it plays 3dB lower on an un-normalized platform than the same track uploaded without the headroom.

Not sure which part you're referring to as being confusing.
Old 7th August 2020
  #100
LUFS and headroom are completely separately controllable. If you want to you can have 3dB headroom and -10LUFS or whatever by simply limiting more. Whether that's a good idea will come down to engineering skill and decision-making. The point is, you don't necessarily have to turn your track's loudness down to set the headroom. 2 different things.

To me it's pretty clear that if there were a significant distortion issue then all the major new releases would not be peaking at -0.1 or 0 with ISPs above 0. Basically none of the big name MEs are worrying about headroom or normalization loudness targets. That's easy to see as I think you said you already demonstrated for yourself. Just listen to the music on consumer systems. How does it sound? QED.
Old 7th August 2020
  #101
Motown legend
 
Bob Olhsson's Avatar
 

Verified Member
Bob Ludwig speaks of offering his clients a choice. Unfortunately, his clients often choose smashed.
Old 7th August 2020
  #102
Gear Head
 
Bongo Spank's Avatar
 

Quote:
Originally Posted by Trakworx View Post
LUFS and headroom are completely separately controllable. If you want to you can have 3dB headroom and -10LUFS or whatever by simply limiting more. Whether that's a good idea will come down to engineering skill and decision-making. The point is, you don't necessarily have to turn your track's loudness down to set the headroom. 2 different things.

To me it's pretty clear that if there were a significant distortion issue then all the major new releases would not be peaking at -0.1 or 0 with ISPs above 0. Basically none of the big name MEs are worrying about headroom or normalization loudness targets. That's easy to see as I think you said you already demonstrated for yourself. Just listen to the music on consumer systems. How does it sound? QED.
Sounds fine to me. I'm just not entirely sure there isn't some relatively common scenario in which it doesn't sound fine but under circumstances I didn't think to or have the capacity to test.

Yes, they're two different things... which is precisely why kept them separate. Seems much more confusing to muddle the two.

Clipping / limiting change aesthetics. If my aesthetics are already dialed, and I'm just trying to address the technical issue of how much headroom I need to avoid downstream issues, then my other option is to turn it down... in which case I may as well be reaching into the homes of any listeners not on normalized platforms and turning down their volume knob directly.
Old 7th August 2020
  #103
Probably overthinking it. All the majors are doing it the same way. What does that tell you? Losing 3dB over a maybe possible problem is something you won't find me ever doing...
Old 7th August 2020
  #104
Gear Head
 
Bongo Spank's Avatar
 

Quote:
Originally Posted by Trakworx View Post
Probably overthinking it. All the majors are doing it the same way. What does that tell you? Losing 3dB over a maybe possible problem is something you won't find me ever doing...
I tend to agree. Interesting. I was tarred and feathered a couple years back for suggesting the same thing after having run some tests. So much so that I assumed the "experts" must be attuned to something I'm missing, or there was some test I neglected to run.

Will look again for that article though. Can't remember the details, but it gave a very specific set of circumstances on Spotify in which it claims the the artifacts are obvious. For whatever reason at the time I read it, I wasn't able to test it.
Old 13th August 2020
  #105
Lives for gear
 
Jerry Tubb's Avatar
 

Verified Member
Loud & Proud for many years... leaving -0.3dB headroom as suggested by the late great Michael Gerzon.

But often I’ll go to -0.5dB headroom for cleanliness, but i can hear it change, a little.

If the project is headed to MFiT (ADM) I’ll go up to -1dB, to pass the Terminal afclip test.

That’s it. you can’t control what device people listen with, or the conditions.

cheers, JT
Old 13th August 2020
  #106
Gear Head
 

Quote:
Originally Posted by Jerry Tubb View Post
Loud & Proud for many years... leaving -0.3dB headroom as suggested by the late great Michael Gerzon.

But often I’ll go to -0.5dB headroom for cleanliness, but i can hear it change, a little.

If the project is headed to MFiT (ADM) I’ll go up to -1dB, to pass the Terminal afclip test.

That’s it. you can’t control what device people listen with, or the conditions.

cheers, JT
Hi Jerry!
Which Gerzon article are you referring to? Quite fancy reading this.
Old 13th August 2020
  #107
Lives for gear
 

Verified Member
I find these discussions (this, many threads on here and many many on MEW on FB) confusing.

Music isn't one thing? there's almost endlessly and constantly multiplying styles of music, production and instruments.. and formats are in a real wild west mode right now to make it even more choatic.

How can we quantify any of this?

I did a master for a client few years ago for CD up to -0.3 which was on a big playlist on Spotify, he's mortified because it clips every time. He insists we now test the codec and playback levels.

Should I tell him some other mastering engineer on some entirely unrelated music does something else and he should suck it up?

Odd discourse IMHO. A bit like asking if something should be delivered bright or dark.. dunno? what does it sound like?
Old 13th August 2020
  #108
Quote:
Originally Posted by Joe_caithness View Post
I find these discussions (this, many threads on here and many many on MEW on FB) confusing.

Music isn't one thing? there's almost endlessly and constantly multiplying styles of music, production and instruments.. and formats are in a real wild west mode right now to make it even more choatic.

How can we quantify any of this?
Peak headroom is perhaps the easiest thing to measure/quantify. I hope it's clear that we're talking about headroom on brick-walled PCM masters for digital before distribution, and CD. ISPs are ISPs, no matter the genre, though any discussion of different methods for different styles is welcome, and as I recall was already discussed in this thread a little.

Quote:
Originally Posted by Joe_caithness View Post
I did a master for a client few years ago for CD up to -0.3 which was on a big playlist on Spotify, he's mortified because it clips every time. He insists we now test the codec and playback levels.
That's interesting, and the first time someone has mentioned a specific instance of audible clipping when streamed. Was it a particularly loud master? What genre?

Quote:
Originally Posted by Joe_caithness View Post
Should I tell him some other mastering engineer on some entirely unrelated music does something else and he should suck it up?
Do you think that's what anyone is suggesting?

Quote:
Originally Posted by Joe_caithness View Post
Odd discourse IMHO. A bit like asking if something should be delivered bright or dark.. dunno? what does it sound like?
I think Peak headroom is much more of a technical matter than EQ decisions, don't you? Apples and oranges?


It's about getting the most punch/attack into a track at a given LUFS vs compromising for ISP concerns. IDK, seems straightforward enough to me...
Old 13th August 2020
  #109
Lives for gear
 

Verified Member
Quote:
Originally Posted by Trakworx View Post
Peak headroom is perhaps the easiest thing to measure/quantify. I hope it's clear that we're talking about headroom on brick-walled PCM masters for digital before distribution, and CD. ISPs are ISPs, no matter the genre, though any discussion of different methods for different styles is welcome, and as I recall was already discussed in this thread a little.

I meant qualify...my bad

You're correct that peak level is entirely quantifiable.

In the sense that it could be absolutely fine in some cases and an absolute disaster in others.

This was in relation to suggestion that either the opinion (paraphrasing) "ISPs are a technical fault and should be avoided" and "ISPs are fine" is somehow good mastering advice, when the program material, experience and listening. Neither is true.

In regards to genre, I can think of examples where ISPs could be essentially unheard and you might as well use that gain and I can think of examples where I got a sad email from a client regretting our joint decisions on peak headroom on a project.

Quote:
That's interesting, and the first time someone has mentioned a specific instance of audible clipping when streamed. Was it a particularly loud master? What genre?
Not that loud, modern classical, absolute crunch at one point played back on Spotify. Clean as a whistle played PCM or even 320kbps MP3 directly in a DAW or on Windows Media Player.
[/quote]

Quote:
Do you think that's what anyone is suggesting?
Hyperbolic reply from me yes, but,

"If Ted Jensen doesn't worry about clipping in codecs then maybe I shouldn't either."

I know your comment is purely philosophical and I'm not making a comment on how another ME conducts themself in practice, this is all in the spirit of healthy discourse, but I find the idea that because big ME does or doesn't do something means you should act against your instinct a bit odd.

I'm arguing that Ten Jensen may have made the decision to run that master on that job, but that doesn't really have any impact on whether or not I chose to implement a similar process on a completely different project with a completely different context.

Quote:
I think Peak headroom is much more of a technical matter than EQ decisions, don't you? Apples and oranges?

It's about getting the most punch/attack into a track at a given LUFS vs compromising for ISP concerns. IDK, seems straightforward enough to me...
Well both determine how the master plays back and how people perceive it.

ISPs aren't going to make you streamer or laptop crash, but they will affect how the audio sounds, same as EQ.

If it plays back as intended, it's not a technical fault is it? Some MEs decide that playing back with ISPs is fine, some don't, some do or don't depending on the project.

That decision, much like applying a high shelf or not, is on us.

I work on a lot of soundtracks and I make a point of those projects which are wildly dynamic and extreme in content at times complying to Spotify's suggestions...

Why? because if someone complains that it goes PRRRFT on one timpani hit I've failed my client.

Big loud rock with hundreds of snares and compressed stereo guitars fizzing away? maybe not
Old 13th August 2020
  #110
Gear Head
 
S_mask's Avatar
 

'The singularity' notwithstanding, today's music-delivery options, other than HD or vinyl, are atrocious. So, recommending a 96 kb/s stereo quattuordecimation (for mobile-streaming evaluation), as well as 160 kb/s (for the ūndecimation's slightly-less severe carnage-glimpse) during QC, prior to 'hand-off'.
Old 13th August 2020
  #111
Quote:
Originally Posted by Joe_caithness View Post
Big loud rock with hundreds of snares and compressed stereo guitars fizzing away? maybe not
It's true that I'm mainly talking about the meat and potatoes genres that I work with. Rock, Metal and Hip Hop where clients want loud aggressive masters and that's where the trade off between headroom and punch matters. Soundtracks, Classical, etc can and should require different treatments, and that discussion is not outside the bounds of this thread. 'Never meant to imply one size fits all. It's good that you brought that up, adding to the discussion and making it even more relevant to our work.

Although it's far from just Ted Jensen on one project. It's literally every track I've ever checked by a lot of major artists and MEs. 0 or -0.1. I did not check any soundtrack or Classical however, because that's outside of my purview.
Old 14th August 2020
  #112
Lives for gear
 

Verified Member
Quote:
Originally Posted by Trakworx View Post
It's true that I'm mainly talking about the meat and potatoes genres that I work with. Rock, Metal and Hip Hop where clients want loud aggressive masters and that's where the trade off between headroom and punch matters. Soundtracks, Classical, etc can and should require different treatments, and that discussion is not outside the bounds of this thread. 'Never meant to imply one size fits all. It's good that you brought that up, adding to the discussion and making it even more relevant to our work.

Although it's far from just Ted Jensen on one project. It's literally every track I've ever checked by a lot of major artists and MEs. 0 or -0.1. I did not check any soundtrack or Classical however, because that's outside of my purview.
Fair.

Also I think we're in a fairly confusing transitional stage with all this stuff right now.

I wouldn't be surprised if there was some audio engineering micro scandal ala Death Magnetic around the sounds of streaming with all the complications with after mastering processing in the next few years.

We shall wait and see..
Old 14th August 2020
  #113
Mastering Moderator
 
Riccardo's Avatar
 

Verified Member
2 Reviews written
Quote:
Originally Posted by Trakworx View Post
It's true that I'm mainly talking about the meat and potatoes genres that I work with. Rock, Metal and Hip Hop where clients want loud aggressive masters and that's where the trade off between headroom and punch matters. Soundtracks, Classical, etc can and should require different treatments, and that discussion is not outside the bounds of this thread. 'Never meant to imply one size fits all. It's good that you brought that up, adding to the discussion and making it even more relevant to our work.

Although it's far from just Ted Jensen on one project. It's literally every track I've ever checked by a lot of major artists and MEs. 0 or -0.1. I did not check any soundtrack or Classical however, because that's outside of my purview.
It depends on the producer in charge and label rep. Classic OST (Soundtracks ) here go out with no limiting bar very few exceptions where I have a sixties or early seventies master of a work for a medium to low budged movie, well written and played but with "time constrain" mixing issues.
A good percentage of these jobs are for all three formats: digital release (they go out at -1 dBfs checked for codecs and downsampling) either at 44/16 or 24/96 depending on label requests, CD release, DDP 44/16 at 0.3 dBfs and Vinyl prep. These measure louder compared to mid nineties releases at around -14 to -18 LUFS long term gated.


Classical is a different story and largely depends on producers and labels but yes quieter than Indie, Hip-hop, Trap, Pop and contemporary Rock. It could be anything between -20/21 and -16 LUFS long term gated.
Old 20th August 2020
  #114
Motown legend
 
Bob Olhsson's Avatar
 

Verified Member
You need to consider who the most important listener is. If a major label CD is headed for a broadcast programming meeting, you simply want it hot. If the goal is a word of mouth recommendations on Spotify, you want to optimize for that platform.
Old 20th August 2020
  #115
Gear Nut
 
fairchildren's Avatar
 

So for the guys who are using fab filter... are you more commonly using the TP setting on to prevent peaks down the line of conversion?

I haven't played around with doing masters with it off much, assuming it was better to just prevent those clips, but now that I know some of you turn it off for aesthetic reasons, it has me quite curious.
Old 4 weeks ago
  #116
Lives for gear
 
Jerry Tubb's Avatar
 

Verified Member
Quote:
Originally Posted by Rukas View Post
Side note, is FabFilter L2 reaaalllly worth getting over Waves L2?
Yes.

The Waves L2 is about 20 years old now. Great innovation at the time,

a huge upgrade from the original L1.

I still have the hardware version that i bought new, as well as the plug-in.

used it extensively for a decade, then i heard the PSP Xenon...

a quantum leap above & beyond the Waves L2. Leapfrogged over the L3.

also added the Fab L2

and most recently the Weiss Limiter plug-in, which i love.

but on a desert island, i would be totally happy with PSP Xenon.

cheers, jt

Last edited by Jerry Tubb; 4 weeks ago at 08:48 AM.. Reason: i love editing!
Old 2 weeks ago
  #117
Lives for gear
 

Verified Member
Can we write -.3dB TP when we mean true peak, just I am not sure after reading all this that we all see the difference between sample and True Peak. Makes me feel uneasy seeing just -.3dBs could be either TP or sample peak.

Dua Lipa track from spotify "Don't show up"

+.9 dB TP
0dB Sample peak

Maximum short term LUFS -6.0
Integrated LUFS -7.9

or there about, measured with auto normal off in real time with Insight

Sounds very loud with a wide response but with a moderate wide V dip and strangely quite dynamic sounding. I cant hear any audible distortions here in the studio.

Anybody know who mastered that?

What are you using to count ISPs?

The difference to me between something like that track, and one that comes in under TP is a feeling that when perceptually level matched by changing the volume the ones that are under TP have the feeling that you can really turn up the volume, the really loud -8LUFS stuff has a sort of scary feeling when you really crank it up. It seems to me that you hear more and more of the limiting as you increase listening volume.

I wonder if any of the top 20 most played list are under TP? Would that data change your targets if they were ALL over TP?
Old 2 weeks ago
  #118
What are data compression artifacts?
Old 2 weeks ago
  #119
Lives for gear
 
SmoothTone's Avatar
 

Verified Member
Quote:
Originally Posted by coreyspencer View Post
What are data compression artifacts?
Often referred to as 'space monkeys,' they sound similar to overdone digital noise reduction. Make a low bitrate mp3 and solo the side channel and you'll hear them.

Lots of great detail in this article:

https://www.soundonsound.com/techniq...oes-your-music
Old 2 weeks ago
  #120
Gear Addict
Quote:
Originally Posted by Rukas View Post
Side note, is FabFilter L2 reaaalllly worth getting over Waves L2?
Why do you compare Waves L2 and Fabfilter Pro-L2?
Pro-L2 is not an emulation of Waves L2. Don't mix up both. L is just for the word Limiter.
Also, Waves Q equalisers are not the same as Fabfilter Pro-Q equalisers.
📝 Reply

Similar Threads

Thread / Thread Starter Replies / Views Last Post
replies: 62 views: 22389
Avatar for MarkyGoldstein
MarkyGoldstein 30th December 2014
replies: 434 views: 117223
Avatar for Slug1
Slug1 25th January 2015
replies: 181 views: 34796
Avatar for ~ufo~
~ufo~ 3rd April 2020
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
🖨️ Show Printable Version
✉️ Email this Page
🔍 Search thread
🎙️ View mentioned gear
Forum Jump
Forum Jump