Preparing a mix for a mastering engineer.

[newkid]

Originally Posted by Bigduggieface

I'm curious why you would need the 3-6dB of headroom. I assume it would be to accommodate any eq boosts in the digital domain....
In other words, why not submit 24/96 masters that peak at say -0.2 dBfs? Why leave the headroom?

from wikipedia:

...
Headroom in digital audio

In digital audio, headroom is defined as the amount by which digital full scale (FS) exceeds the permitted maximum level (PML) in dB (decibels). The European Broadcasting Union (EBU) specifies a PML of 9 dB below 0 dBFS (-9 dBFS), thus giving 9 dB of headroom. An alternative EBU recommendation allows 24 dB of headroom, which might be used for 24-bit master recordings where it is useful to allow more room for unexpected peaks during live recording.

Failure to provide adequate headroom can bring about clipping of brief, higher-level transients.
...

See the diagram included on the full headroom article
http://upload.wikimedia.org/wikipedi...6/Lindos10.png


...

Must be a reason so many standards require more headroom.

Do you know more than the pros? I don't.
More than the standards committees??


Might want to research headroom some more.
Read a couple of good audio recording mixing mastering books.
I doubt that the headroom was arbitrary to make your mix less loud.

[Danny Danzi]

Ah Dug, I know YOU wouldn't have done the overcompression thing....I'm sorry if I implied that. What I meant was, some guys just don't go for -0.2, they really hit it hard. Everyone has a different interpretation of what levels could or should be once they use a limiter and that's what always worries me and makes me a control freak. LOL!!

I also wouldn't complain about a -1dB peak but in the event there are some things that I have to really tweak, it doesn't leave me much of a window for it. What we as ME's hear are usually a bit different than what the engineers hear and most of this can be due to ear fatigue etc...you know how all that stuff works, I know. So for us, we may hear a problem area or see something that might be a bear to fix...it's nice to have a bit of headroom to do it, know what I mean? In a good mix, you don't need much. Especially if it's a mix that I am sort of in on with the client.

Part of my service is to listen to the mix ahead of time to make suggestions so we're all on the same page. I can't tell you how many times I've had guys boosting sub lows and boosting excessive mids then they expect me to give them a clear kick drum in all that mess. LOL! So for mixes I actually have a bit of say on, the headroom window doesn't have to be as strict and even -1dB would be perfectly acceptable.

Yeah I'm sorry Dug, the sampling thing you had mentioned lost me a bit so I wasn't sure how to reply there. I kinda thought you meant taking a recording and just upsampling it...which of course can be and has been done. I don't dither, change bit or convert sample rates until the very end so everything gets processed where it was recorded at.

If a 16/44 does need some help though, I will upsample it so it can be processed at a higher bit rate....but honest, for me it's been rare and some guys swear you should do this at all times and process in the highest realm possible. I just don't hear much of a difference in most things to marry the idea.

Now if someone decides to record from the start at 24/96 or above, that's ok too. I really don't think they need to though in all honesty due to the reasons I've mentioned before. I have a few gear head studios I master for that are hell bent on 24/96 and above. One of them who isn't as stubborn decided to take a test for me since he was a bit more approachable and a bit soft lol!

I told him to record a little project of about 8-12 bars at 16/44, then record the same thing at 24/48, then 24/96. His reply in the email to me was "why don't you go fook yourself for being right all the time!" LOL!!! He DID hear the difference between 16/44 and 24/48....but the only difference he could come up with on the 24/96 was it sounded a bit more crisp like I had mentioned.

He then let his wife play all the versions at random without looking at his screen and he couldn't even tell the 24/96 anymore...but he did pick out the 16/44. But again, I'm sure some guys will probably be able to tell....I'm not saying it's impossible...I just don't think it makes that much of a difference to where we have to all join hands and say "YES, this is the bit/sample rate to use for life! It makes such a huge difference I sound like Mutt Lange!" LOL!! In this subjective business, there really aren't any rules other than the ones we know make a huge difference for us. What may work for me may suck for someone else...what may work for them may work for me, but it may not be better than what I do...you know how it goes...the field is just too vast to rule anything out. That's part of the fun/challenge factor for me.

[Danny Danzi]

Originally Posted by newkid

Must be a reason so many standards require more headroom.

Do you know more than the pros?
More than the standards committees??


Might want to research headroom some more.
Read a couple of good audio recording mixing mastering books.

newkid, regardless of what you read, what pro producer or engineer says what or what the aesthetics may be, it doesn't always mean they are correct in this field. Honest when I tell you. One thing to keep in mind that I always say to people is you cannot always believe what you hear and what you read as factual until you yourself experiment and find out for yourself.

The reason being, if Andy Johns were to join this site right now and start posting information, how would some of it apply to us if we are not using the millions dollars worth of gear that he used? Some of the greatest recordings known to man....had engineers that were not concerned with all the digital hype and mathematics we read about that can totally deter someone from wanting to be involved in this career.

That said, I'm not saying headroom doesn't matter and I'm not disagreeing with your post. I'm just saying...sometimes the science behind this art as well as the pros that have something to say about it, just hype it up to where they try to make it out to be "law". And, we don't have the exact gear they have to apply some of these theories, know what I mean?

I can list 1000 things that people hold dear to their heart in this field to where I never keep them in mind and have no problems. Have I tried some of them? Sure....my place is a little lab most times. I'll try anything for as long as it takes from whoever decides to share something with me until I see if it will be beneficial or not. As decent as I feel I am in this field, I love to learn from everyone no matter what their level of expertise is. But some of it to me is just overkill in the math department...it's all about how it sounds...not what it looks like on a meter, whether you use the K System or not, what you dither to, whether you upsample, downsample, use 24/96, have loads of headroom or not....if it sounds good it is good really and all too often some of the tech heads forget this in order to try and intimidate us with their intelligence.

The only math "I" personally need in the audio field is the following:

Good instrumentation pre-print + clean signals + accurate monitors + knowing what to listen for and how to make it better = great mix. Great mix + good mastering engineer = Excellent end result. All the rest, the gear heads can have if it makes them feel better.

[newkid]

Originally Posted by Danny Danzi

newkid, regardless of what you read, what pro producer or engineer says what or what the aesthetics may be, it doesn't always mean they are correct in this field. Honest when I tell you. One thing to keep in mind that I always say to people is you cannot always believe what you hear and what you read as factual until you yourself experiment and find out for yourself.

The reason being, if Andy Johns were to join this site right now and start posting information, how would some of it apply to us if we are not using the millions dollars worth of gear that he used? Some of the greatest recordings known to man....had engineers that were not concerned with all the digital hype and mathematics we read about that can totally deter someone from wanting to be involved in this career.

That said, I'm not saying headroom doesn't matter and I'm not disagreeing with your post. I'm just saying...sometimes the science behind this art as well as the pros that have something to say about it, just hype it up to where they try to make it out to be "law". And, we don't have the exact gear they have to apply some of these theories, know what I mean?

I can list 1000 things that people hold dear to their heart in this field to where I never keep them in mind and have no problems. Have I tried some of them? Sure....my place is a little lab most times. I'll try anything for as long as it takes from whoever decides to share something with me until I see if it will be beneficial or not. As decent as I feel I am in this field, I love to learn from everyone no matter what their level of expertise is. But some of it to me is just overkill in the math department...it's all about how it sounds...not what it looks like on a meter, whether you use the K System or not, what you dither to, whether you upsample, downsample, use 24/96, have loads of headroom or not....if it sounds good it is good really and all too often some of the tech heads forget this in order to try and intimidate us with their intelligence.

The only math "I" personally need in the audio field is the following:

Good instrumentation pre-print + clean signals + accurate monitors + knowing what to listen for and how to make it better = great mix. Great mix + good mastering engineer = Excellent end result. All the rest, the gear heads can have if it makes them feel better.

Can't argue with you.
What I said was that a committee of experts set a standard (actually several of them) for headroom.

They think you need a significant amount.
Doubt taht they would risk a bad recording because of doing it.

So what benefit is there by bumping the ceiling as suggested?

Back in the day you wanted to minimise it so you didn't get noise on the low end. Now things are so quiet and have such huge range there is no reason at all to NOT have a lot of headroom.

Again,
What is the benefit in squeezing the headroom down to zip , zilch, bupkes, rien, nada, nil, squat?

[Bigduggieface]

Originally Posted by Danny Danzi

I also wouldn't complain about a -1dB peak but in the event there are some things that I have to really tweak, it doesn't leave me much of a window for it.

I don't mean to belabor this point, but I am still unsure of any evidence of why a mastering engineer would require a digital file that peaks well below 0dBfs.

In these modern days of 32 and 64-bit floating point digital workstations, clipping a signal in the digital domain seems unlikely. Even for a mastering engineer works his magic entirely in the analog domain, there seems to me to be no harm in delivering them a hot file.

Also, an unmastered file that has been mixed to -12dBfs does not utilize the entire bit-depth nor dynamic range of a file that has been mixed to peak closely to 0dB. The mastering engineer is now forced to boost portions (or the entirety) of the spectrum +12 dB in order to get a hot signal that approaches 0dB without clipping. This 12 dB boost also increases noise and inherent low-level digital distortion.

It seems counter-intuitive to me to lower my mix by -12dB or even -3dB only for the mastering engineer to ultimately boost it back up to zero again. It seems more likely (especially in modern rock and pop) that the engineer will take my hot file, and compress it. Thus negating the need for any headroom.

For example, I can imagine submitting a stereo mix that peaks at -4dB. The mastering engineer ultimately compresses it another 3dB. He then boosts the signal +7dB to reach 0dB. All of the noisy low level crap has now been unnecessarily boosted by +7dB.

I'm trying to investigate the rationale of the mastering engineer for requiring that a mix need such a "window" to improve a mix.

[venuestudio]

From the big boys of the ME world:

At what recording level should I print my digital mix?


"Digital yields such a low noise floor that now we don’t really have to worry about it anymore – if you print mixes on your 24 bit recorder at -48dBfs, you’ll still have a higher signal to noise ratio than is possible on analog tape. There’s no need to ‘get close to zero’, give your mixes room to breathe and you’ll be amazed how much better they sound when mastered." - Brad Blackwood

"Measuring the peak level of a mix tells us almost nothing about its quality. As long as your digitally-measured highest peak of the loudest song does not exceed, say, -3 dBFS, you are safe from the peak point of view. And if you are absolutely sure of your metering accuracy and you are using intersample peak meters (which are very rare), then you could conceivably exeed -3 dBFS on the highest peak and still be safe, but why play with fire when 24 bit recording permits low levels without losing signal to noise ratio or sound quality? Even if your recording never peaks above -10 dBFS it’s still going to be fine. It is the average level of your mix that determines more about its quality (measured on an averaging meter, such as a VU meter or a K-System meter). I recommend your mix should not exceed -14 dBFS RMS, preferably -20 dBFS RMS, if you want to have a dynamic (and loud) master. - Bob Katz

"There was a famous engineer some years ago that said, "I can mix by just looking at the meters." He was obviously an upstart wanker. If you stare at meters long enough, which is what I did for the first 15 years, you find they don't mean anything. It's what's in your soul. You hope that your ears are working with your soul along with your objectivity, but truly you can never be sure." - Andy Johns

On Compression on the two bus before mastering:

"Some mastering engineers will tell you not to compress your mixes, giving various reasons such as the quality of your compressors or your monitoring environment not being up to par, but I disagree. Mix compression can radically change the instrumental balances within the music; if left to the mastering engineer your mix will often come back sounding very different from what you sent. I recommend judicious use of compression, checking the results in various listening environments with varying levels (if needed) until you are sure it is beneficial." - Brad Blackwood


“Please don’t pre-master or compress in pro tools, as this leaves less room to make changes once the mastering engineer evaluates your mixes." -Bob Ludwig

"It is very difficult to undo compression and limiting after the fact. It's unnecessary to normalize, compress or limit your mix in order to boost level before it gets to the mastering stage. If you wish to use master bus compression for artistic reasons, then do so, but we recommend that you use compression judiciously, and that you print both versions of the mix, with and without the master bus compression." - Bob Olhsson

On compression you can see it is up to the engineer as to whether they believe you should compress ahead of mastering. I say this because compression for many people is a way to add volume, not just control transients or add color. These are world class MEs and even they differ a bit.

[Danny Danzi]

Originally Posted by Bigduggieface

I don't mean to belabor this point, but I am still unsure of any evidence of why a mastering engineer would require a digital file that peaks well below 0dBfs.

In these modern days of 32 and 64-bit floating point digital workstations, clipping a signal in the digital domain seems unlikely. Even for a mastering engineer works his magic entirely in the analog domain, there seems to me to be no harm in delivering them a hot file.

Also, an unmastered file that has been mixed to -12dBfs does not utilize the entire bit-depth nor dynamic range of a file that has been mixed to peak closely to 0dB. The mastering engineer is now forced to boost portions (or the entirety) of the spectrum +12 dB in order to get a hot signal that approaches 0dB without clipping. This 12 dB boost also increases noise and inherent low-level digital distortion.

It seems counter-intuitive to me to lower my mix by -12dB or even -3dB only for the mastering engineer to ultimately boost it back up to zero again. It seems more likely (especially in modern rock and pop) that the engineer will take my hot file, and compress it. Thus negating the need for any headroom.

For example, I can imagine submitting a stereo mix that peaks at -4dB. The mastering engineer ultimately compresses it another 3dB. He then boosts the signal +7dB to reach 0dB. All of the noisy low level crap has now been unnecessarily boosted by +7dB.

I'm trying to investigate the rationale of the mastering engineer for requiring that a mix need such a "window" to improve a mix.

Please "belabor" all you want...and thanks for teaching me that word!! LOL!! I love talking about this stuff and learning from you as well man, so thank you for taking part in this. All that you say is completely credible, Dug. I'll try to explain this to the best of my knowledge and ability and why it works for me. That's really all I can offer bro.

OK, Now I'm gonna get all mathematical on yo azz. LOL!! Eventhough 24 bit recording is the norm for most people, some engineers like to push the meters as high as possible and even sometimes hit 0. It's totally unnecessary and here's why. A 16 bit recording fits in the bottom 91dB of the 24 bit. This means you would have to lower the peak of a 24 bit recording by 48dB to get an effective 16 bit recording. LOL!!!

There is a lot of room at the bottom so going to -3dBFS or even as low as -10 like we've talked about briefly, and you'll end up with a much cleaner recording. Since distortion accumulates when a hot recording comes in to be mastered, I may have to attenuate the level to prevent processing the DAC from overloading. A digital mix that peaks at -3dBFS or lower just makes it easier to eq and process without needing an extra stage of attenuation when mastering.

Could I master at near clip levels? Sure but I'd be willing to bet I'd have to lower the over-all mix because it would just not leave me any headroom to boost frequencies IF I need to. And, most people boost it up using compression too much on the 2-buss and it sometimes really makes it hard for me to compress MY work. Another example...

If you send me a metal tune that's kicking all through the tune and then it has a really quiet, intimate passage, I'm going to split that part of the clip and work on it as an entity. The transients in quiet parts are sometimes nasty to begin with. If added compression the wrong way is in there and the mix is automated to make that part close to the level of the heavy part of the tune, my judgement has to be extremely selective.

In a -3dB tune with no compression on the 2-buss, this quiet part might peak at -10dB. I can eq it, compress it and boost it up so it keeps dynamics and doesn't sound as loud as the heavy part. We want a bit of a break in volume there but we don't want it too low. As I'm working on this part, I'm raising in increments and watching everything I do. When it's already hot, IF it happens to be hot, my abilities are more limited really. I can't un-compress something to where you hear the compressor kicking in. There is just no way for me to stop that.

This is why I always ask people not to do that on their mixes. They can send me a reference mix of what they would hope it would sound like when I'm done showing me the type of compressor they want, but for the most part, it shouldn't be something you really hear. I compress for the peaks and valleys I create when eqing. Most times, a 2:1 ratio or a 4:1 with a -12 threshold works for rock or metal...but this depends on the eq curve I have to draw. To me, the -3dB just allows for a bit more control to where I have a little more flexibility if that makes any sense?

I understand a guy like you would probably just give me the mix without compression raising the gain....but as soon as I raise something, there's a clipping chance. -3dBFS doesn't let it get close to that because most times, I will not have a compressor plus an eq that exceeds more than a gain of 0.5 to 1dB. I try my best to make sure there is no volume change when I apply these effects. I don't watch a meter though....I use my ears. When I have used a meter, 05dB to 1dB is how off I was...which isn't too bad. So realistically, if the mix came in at -3dB I've raised the mix to, at the most, with compressors, eq's and maybe a multi comp and an imager, -2dB for the top peak.

Another thing I notice...and this is really senseless to mention, but it's been a pet peeve of mine...is guys like Bob Ludwig seem to get perfect analysis numbers. Meaning, if you run something they've done that isn't in mp3 format through an editor and bring up the numbers they've achieved, min sample on both sides, max sample on both sides and DC offset are all identical. I have NEVER been able to hit a 16 bit file with 32767 in min/max samples yet. 32768 and you're clipping. At files over -2dB, I'm never even close to those numbers. I know, the numbers shouldn't matter...and they don't to me, but it makes me question why that baystid has them perfect on every one of his masters. So I decided to try and be as careful as possible to try and achieve them to see if it in fact made any type of sound difference.

Whether it's all in my head or whatnot, when I am in between -3dB and -2dB when I start mastering, my numbers are always nearly identical and quite close to Bob's by the end. I usually fall in at about 32760-32764 on all counts with 0 DC offsets and good averages throughout. This tells me I'm leveling everything to the numbers and have a good consistency...and the most important thing, I love how the stuff I'm doing is sounding. So at the end of the day, it's best to do what you feel is best brother Dug. If you like what you get at whatever level you choose, or decide to use a compressor on your 2-buss in moderation, that's what works for you and I totally respect that.

[venuestudio]

Originally Posted by Danny Danzi

OK, Now I'm gonna get all mathematical on yo azz. LOL!! Eventhough 24 bit recording is the norm for most people, some engineers like to push the meters as high as possible and even sometimes hit 0. It's totally unnecessary and here's why. A 16 bit recording fits in the bottom 91dB of the 24 bit. This means you would have to lower the peak of a 24 bit recording by 48dB to get an effective 16 bit recording. LOL!!!

There is a lot of room at the bottom so going to -3dBFS or even as low as -10 like we've talked about briefly, and you'll end up with a much cleaner recording. Since distortion accumulates when a hot recording comes in to be mastered, I may have to attenuate the level to prevent processing the DAC from overloading. A digital mix that peaks at -3dBFS or lower just makes it easier to eq and process without needing an extra stage of attenuation when mastering.

Danny, it is the normal practice in writing to credit the person you are quoting when you write something. This is lifted all but verbatim from Bob Katz book "Mastering Audio, The Art and Science", Chapter 5, page 72. Bob is a smart guy and deserves to be credited for the work he has done. That is why I put quotes and the name of the person in my post above. Please be mindful of others intellectual property.

[newkid]

Originally Posted by Danny Danzi

Please "belabor" all you want...and thanks for teaching me that word!! LOL!! I love talking about this stuff and learning from you as well man, so thank you for taking part in this. All that you say is completely credible, Dug. I'll try to explain this to the best of my knowledge and ability and why it works for me. That's really all I can offer bro.

OK, Now I'm gonna get all mathematical on yo azz. LOL!! Eventhough 24 bit recording is the norm for most people, some engineers like to push the meters as high as possible and even sometimes hit 0. It's totally unnecessary and here's why. A 16 bit recording fits in the bottom 91dB of the 24 bit. This means you would have to lower the peak of a 24 bit recording by 48dB to get an effective 16 bit recording. LOL!!!

There is a lot of room at the bottom so going to -3dBFS or even as low as -10 like we've talked about briefly, and you'll end up with a much cleaner recording. Since distortion accumulates when a hot recording comes in to be mastered, I may have to attenuate the level to prevent processing the DAC from overloading. A digital mix that peaks at -3dBFS or lower just makes it easier to eq and process without needing an extra stage of attenuation when mastering.

Could I master at near clip levels? Sure but I'd be willing to bet I'd have to lower the over-all mix because it would just not leave me any headroom to boost frequencies IF I need to. And, most people boost it up using compression too much on the 2-buss and it sometimes really makes it hard for me to compress MY work. Another example...

If you send me a metal tune that's kicking all through the tune and then it has a really quiet, intimate passage, I'm going to split that part of the clip and work on it as an entity. The transients in quiet parts are sometimes nasty to begin with. If added compression the wrong way is in there and the mix is automated to make that part close to the level of the heavy part of the tune, my judgement has to be extremely selective.

In a -3dB tune with no compression on the 2-buss, this quiet part might peak at -10dB. I can eq it, compress it and boost it up so it keeps dynamics and doesn't sound as loud as the heavy part. We want a bit of a break in volume there but we don't want it too low. As I'm working on this part, I'm raising in increments and watching everything I do. When it's already hot, IF it happens to be hot, my abilities are more limited really. I can't un-compress something to where you hear the compressor kicking in. There is just no way for me to stop that.

This is why I always ask people not to do that on their mixes. They can send me a reference mix of what they would hope it would sound like when I'm done showing me the type of compressor they want, but for the most part, it shouldn't be something you really hear. I compress for the peaks and valleys I create when eqing. Most times, a 2:1 ratio or a 4:1 with a -12 threshold works for rock or metal...but this depends on the eq curve I have to draw. To me, the -3dB just allows for a bit more control to where I have a little more flexibility if that makes any sense?

I understand a guy like you would probably just give me the mix without compression raising the gain....but as soon as I raise something, there's a clipping chance. -3dBFS doesn't let it get close to that because most times, I will not have a compressor plus an eq that exceeds more than a gain of 0.5 to 1dB. I try my best to make sure there is no volume change when I apply these effects. I don't watch a meter though....I use my ears. When I have used a meter, 05dB to 1dB is how off I was...which isn't too bad. So realistically, if the mix came in at -3dB I've raised the mix to, at the most, with compressors, eq's and maybe a multi comp and an imager, -2dB for the top peak.

Another thing I notice...and this is really senseless to mention, but it's been a pet peeve of mine...is guys like Bob Ludwig seem to get perfect analysis numbers. Meaning, if you run something they've done that isn't in mp3 format through an editor and bring up the numbers they've achieved, min sample on both sides, max sample on both sides and DC offset are all identical. I have NEVER been able to hit a 16 bit file with 32767 in min/max samples yet. 32768 and you're clipping. At files over -2dB, I'm never even close to those numbers. I know, the numbers shouldn't matter...and they don't to me, but it makes me question why that baystid has them perfect on every one of his masters. So I decided to try and be as careful as possible to try and achieve them to see if it in fact made any type of sound difference.

Whether it's all in my head or whatnot, when I am in between -3dB and -2dB when I start mastering, my numbers are always nearly identical and quite close to Bob's by the end. I usually fall in at about 32760-32764 on all counts with 0 DC offsets and good averages throughout. This tells me I'm leveling everything to the numbers and have a good consistency...and the most important thing, I love how the stuff I'm doing is sounding. So at the end of the day, it's best to do what you feel is best brother Dug. If you like what you get at whatever level you choose, or decide to use a compressor on your 2-buss in moderation, that's what works for you and I totally respect that.

+1

soem people wont like facts though

[newkid]

Originally Posted by Bigduggieface

I don't mean to belabor this point, but I am still unsure of any evidence of why a mastering engineer would require a digital file that peaks well below 0dBfs.

In these modern days of 32 and 64-bit floating point digital workstations, clipping a signal in the digital domain seems unlikely. Even for a mastering engineer works his magic entirely in the analog domain, there seems to me to be no harm in delivering them a hot file.

Also, an unmastered file that has been mixed to -12dBfs does not utilize the entire bit-depth nor dynamic range of a file that has been mixed to peak closely to 0dB. The mastering engineer is now forced to boost portions (or the entirety) of the spectrum +12 dB in order to get a hot signal that approaches 0dB without clipping. This 12 dB boost also increases noise and inherent low-level digital distortion.

It seems counter-intuitive to me to lower my mix by -12dB or even -3dB only for the mastering engineer to ultimately boost it back up to zero again. It seems more likely (especially in modern rock and pop) that the engineer will take my hot file, and compress it. Thus negating the need for any headroom.

For example, I can imagine submitting a stereo mix that peaks at -4dB. The mastering engineer ultimately compresses it another 3dB. He then boosts the signal +7dB to reach 0dB. All of the noisy low level crap has now been unnecessarily boosted by +7dB.

I'm trying to investigate the rationale of the mastering engineer for requiring that a mix need such a "window" to improve a mix.

---------------------------------

clipping in digital happens when you hit 0 dbFS
bit depth has nothing to do with that

in digital there is no harm in giving a cooler file to master.
you can always goose the final gain to make it as loud as you want.

the dynamic range accordign to chip makers is well over 100 db
there is no way you can use all of it.
not sure why that is counterintuitive.
sounds like you ahve decided that loudest is best and you ahve to do it at max level at every step in the chain.

you wont hear noise until you are down to a 55-60 db range
so you can give up 20-40 db for headroom and never miss it

as to dynamic range - there is way too much of it
you should compress so it is no more than 30 db total
else when it is loud it wakes the baby
and when it is soft you cant hear it for the traffic outside or the air blowing out of the hvac vents

with any reasonable dynamic range for listening
there will be no noise on the bottom
and with big headroom margin NO distortion or clipping at all on the peaks.

all that noisy crap is so low you cant hear it under the real noise.