Mastering devices connected to mastering insert fx or master out ?

[Voyager]

Hello guys,

My mastering devices such as compressor and limiter should be connected to the mastering fx which is master bus pre-fader or to the mastering output which is master bus post-fader ? And globally when use one over the other one ?

[selig]

Hello guys,

My mastering devices such as compressor and limiter should be connected to the mastering fx which is master bus pre-fader or to the mastering output which is master bus post-fader ? And globally when use one over the other one ?

In one respect for me it has to do with workflow, such that you separate mastering from mixing as it would be done in hardware. So the "mix" involves ONLY the mixer, and "mastering" is everything AFTER the mix/mixer.

With that in mind, I've patched all mastering to the Master Outputs for some time now. It helps me separate the two processes conceptually and literally. Makes it easy to turn off all mastering, makes it easy to check "mix" levels "pre-mastering", makes it easy to use side-chain filtering on the Master Compressor (a process I've long associated with the "mix" coming from an SSL background for so many years), etc.

For me there are two types of mastering: the "single" and the "album". With the single, I just make the song sound good on it's own, following my basic loudness target of a crest factor of 12 dB max. With the album approach, it's important how each song sounds in context, which includes the previous and the following song. So I tend to make more adjustments in that mode to overall EQ or sometimes multi-band compression, whereas with the single approach I typically just use a brick wall limiter.

There is one subject which can trigger some debate, and that is where to do fade outs. Some say fade out the mix, then send it on to mastering and others say don't do the fade out until after mastering. The difference IMO isn't that huge, so it's really more subjective than objective, just like the debate over whether to compress first then EQ, or EQ then compress. In the end, you have to listen to both and choose the one you prefer, if you don't already have a working knowledge of the difference. What to listen for with regards to fade outs: If you fade "pre-mastering" some say the mix changes as you fade because you are below the threshold. But others (myself included) say we are used to hearing that, and it sounds natural for the song to get less squashed as it fades out - it actually adds to the effect of the fade out IMO! But you won't know for yourself unless you've done it both ways and chosen then one that works best for you.

One thing to consider: if you add dither, you must not fade the audio (or otherwise make ANY changes) AFTER the point dither was added or dither won't work correctly. This means that if you use a brick wall limiter that includes dither, you need to use two instances of the plugin; one for limiting, and then the fade out, and finally one instance of the plugin for adding dither.

[Voyager]

Great answer Selig, as usual lot of relevant points.

With that in mind, I've patched all mastering to the Master Outputs for some time now. It helps me separate the two processes conceptually and literally. Makes it easy to turn off all mastering, makes it easy to check "mix" levels "pre-mastering", makes it easy to use side-chain filtering on the Master Compressor (a process I've long associated with the "mix" coming from an SSL background for so many years), etc.

Yes i remember we've talked about your sidechain filtering on the master compressor method. Personally i prefer to have my mastering devices on master too, as you said easier for mix level check etc..

Now sounding wise do having the devices on insert fx is any different than having them on master out ? Didn't notice any difference but who knows maybe i've miss something here.

For me there are two types of mastering: the "single" and the "album". With the single, I just make the song sound good on it's own, following my basic loudness target of a crest factor of 12 dB max. With the album approach, it's important how each song sounds in context, which includes the previous and the following song. So I tend to make more adjustments in that mode to overall EQ or sometimes multi-band compression, whereas with the single approach I typically just use a brick wall limiter.

When you say "following my basic loudness target of a crest factor of 12 dB max" do you mean your gain staging level point ?

One thing to consider: if you add dither, you must not fade the audio (or otherwise make ANY changes) AFTER the point dither was added or dither won't work correctly. This means that if you use a brick wall limiter that includes dither, you need to use two instances of the plugin; one for limiting, and then the fade out, and finally one instance of the plugin for adding dither.

Interesting point, i use Kratos as limiter and it has a built-in dither. I actually never use fade-outs for the whole mix because it doesn't fit the genre of music i produce. But when you say using different instances of the plugin you mean that everytime you export to .wav then re-import the track and use the next instances ?

In my case since i mix and master only singles my though is that i don't need to export-import my project as .wav, just once i've finished my mastering stage i could apply the limiter and dither at same time and only then exporting my finished track as.wav. Is that a valid process ?

[antic604]

Very interesting discussion, but I'm not sure I understand the terminology (new to Reason, so maybe that's why)

Most demo songs have the mastering effects (typically the M-Class devices in some configuration) as an Insert on the master track. Does the above suggests to instead route them manually *between* the output of the Master Section and input of the Hardware Interface. If that's the case, then what is the difference? Is it only the matter of the actual master fader - that you usually don't touch anyway - that doesn't impact the mastering effects in the latter scenario?

[selig]

Now sounding wise do having the devices on insert fx is any different than having them on master out ? Didn't notice any difference but who knows maybe i've miss something here.

No difference, just a workflow preference and a way to keep dither as the very last process in the signal path.

When you say "following my basic loudness target of a crest factor of 12 dB max" do you mean your gain staging level point ?

No, I mean the crest factor of the final mix, which is a simple way to judge loudness and is included in Reason. Using the Big Meter, I set it to read VU + Peak. Since “crest factor” is the peak vs the average level, and the Reason meter uses RMS/average (not actually VU) and Peak levels together in this mode, it’s very easy to read the crest factor. Considering the brick wall limiter keeps peaks at 0 dBFS, the VU/RMS level automatically reads crest factor (peak minus average). So if the VU says - 12 dBFS, that’s your crest factor (removing the “minus” sign: 0 minus -12 = 12).

Now of course the VU level moves around, so I try to not let crest factor go above 12 dB or so on average; sometimes a little above, often a bit (or good bit) below).

Interesting point, i use Kratos as limiter and it has a built-in dither. I actually never use fade-outs for the whole mix because it doesn't fit the genre of music i produce. But when you say using different instances of the plugin you mean that everytime you export to .wav then re-import the track and use the next instances ?

In my case since i mix and master only singles my though is that i don't need to export-import my project as .wav, just once i've finished my mastering stage i could apply the limiter and dither at same time and only then exporting my finished track as.wav. Is that a valid process ?

I just mean use three devices: Kratos for limiting, something for the fade out, and another Kratos for adding dither. No reason to add extra steps for export/import just to add dither!

And yes, when adding mastering to a single song that doesn’t have a fade out, you can use a single device for limiting and adding dither - it’s what I do with Ozone for 90% of my work (which is delivered to my music library guy who can’t use fade outs!).



Sent from some crappy device using Tapatalk

[selig]

Very interesting discussion, but I'm not sure I understand the terminology (new to Reason, so maybe that's why)

Most demo songs have the mastering effects (typically the M-Class devices in some configuration) as an Insert on the master track. Does the above suggests to instead route them manually *between* the output of the Master Section and input of the Hardware Interface. If that's the case, then what is the difference? Is it only the matter of the actual master fader - that you usually don't touch anyway - that doesn't impact the mastering effects in the latter scenario?

I outlined my reasons, one of which has to do with dither/fade outs as you mention. The other has to do with workflow and organization, keeping the mastering section separate from the mixer section.

This is how I worked on the hardware SSL console, and how I work in Pro Tools. Old habits, I guess! But it makes it easy to quickly deliver a non-mastered version, and easy to see pre-post mastering levels as well.

And in the case of the master fader being used, I don’t have to make any other adjustments or worry about processing order.



Sent from some crappy device using Tapatalk

[antic604]

@Selig, thanks!

[Voyager]

And yes, when adding mastering to a single song that doesn’t have a fade out, you can use a single device for limiting and adding dither - it’s what I do with Ozone for 90% of my work (which is delivered to my music library guy who can’t use fade outs!).

So if i dither and limit with one device i guess that the device process dither after limiting, right ? seems obvious but i prefer to ask

Also when you brick wall what is your db reduction range ? any preference or advise in that regard ?

[Reasonable man]

Can someone answer me this please. You know the way the master outs of the master section in the rack always go into the main Audio I/O outs of the virtual hardware interface at the top of Reason? Well if i were to insert lets say a Kuassa Eve-Mp5 equaliser between the master outs and the I/O outs at the top of the rack how does reason interput that signal flow ?

Sorry im not being clear ... i mean if for eg im using one of Reasons many 'ready made' mastering suits ( like the default mastering suite for eg) in the insert section of the master section ..would the signal flow through those devices in the insert section first before it flows through the Eve-Mp5 which is connected to the master outs or does the signal go through the Eve first before it reaches the insert devices in the master section? I can't figure this out.

[househoppin09]

The signal flow is exactly the way it's represented in the rack--you can literally see the answer to your question by tracing the wires. So, yes, in your example the signal goes through the master insert section, then leaves the final outputs of the master section, then goes into the EVE-MP5, then leaves that, and finally gets sent to your audio hardware via the Hardware Interface.

[antic604]

i mean if for eg im using one of Reasons many 'ready made' mastering suits ( like the default mastering suite for eg) in the insert section of the master section ..would the signal flow through those devices in the insert section first before it flows through the Eve-Mp5 which is connected to the master outs

Yes.

[antic604]

BTW, since we're talking signal flow.

By default, signal that comes into Mixer, goes through Gain -> Compressor -> EQ -> Inserts, which corresponds with the layout of the channel strip top to bottom.

What about Sends, that are below ("after") Inserts? Is the signal that's being sent to Send FX a raw copy of what enters the Mixer or is it post Gain / Comp / EQ / Inserts? I'd expect it's the latter, since it's easy to patch the former version and the "PRE" toggle in FX Send section would suggest the copy is sent from the bottom of the channel strip, by default post-fader.

It's probably obvious, but I couldn't find the definitive answer in the 1300+ pages manual

[selig]

BTW, since we're talking signal flow.

By default, signal that comes into Mixer, goes through Gain -> Compressor -> EQ -> Inserts, which corresponds with the layout of the channel strip top to bottom.

What about Sends, that are below ("after") Inserts? Is the signal that's being sent to Send FX a raw copy of what enters the Mixer or is it post Gain / Comp / EQ / Inserts? I'd expect it's the latter, since it's easy to patch the former version and the "PRE" toggle in FX Send section would suggest the copy is sent from the bottom of the channel strip, by default post-fader.

It's probably obvious, but I couldn't find the definitive answer in the 1300+ pages manual

You are correct. Sends are in fact post everything by default: post fader, post panner, and even post mute (for some unknown reason).

And although sends lack pan capabilities themselves, you can take advantage of the post-pan routing when using stereo reverbs/delays by flipping the the send (or return for that matter) cables. Simply trade the left with right and vice versa) and then anything panned left will have it's reverb/delay appear on the right and vice versa - automagically!

[Voyager]

And yes, when adding mastering to a single song that doesn’t have a fade out, you can use a single device for limiting and adding dither - it’s what I do with Ozone for 90% of my work (which is delivered to my music library guy who can’t use fade outs!).

So if i dither and limit with one device i guess that the device process dither after limiting, right ? seems obvious but i prefer to ask

Also when you brick wall what is your db reduction range ? any preference or advise in that regard ?

Get lost in the discussion

[antic604]

You are correct. Sends are in fact post everything by default: post fader, post panner, and even post mute (for some unknown reason).

And although sends lack pan capabilities themselves, you can take advantage of the post-pan routing when using stereo reverbs/delays by flipping the the send (or return for that matter) cables. Simply trade the left with right and vice versa) and then anything panned left will have it's reverb/delay appear on the right and vice versa - automagically!

Thanks, and BTW that's a great tip

[selig]

And yes, when adding mastering to a single song that doesn’t have a fade out, you can use a single device for limiting and adding dither - it’s what I do with Ozone for 90% of my work (which is delivered to my music library guy who can’t use fade outs!).

So if i dither and limit with one device i guess that the device process dither after limiting, right ? seems obvious but i prefer to ask

Also when you brick wall what is your db reduction range ? any preference or advise in that regard ?

Yes, dither is always added last so it will work correctly, This has to do with it needing to sit down at the bottom bit to do it’s magic - and if you change levels it’s no longer going to be where it’s supposed to be.

My dB reduction range varies and depends on a few things. One is the quality of the limiter, based on how much GR can it provide transparently. That’s my upper “limit” of what I will allow. The second is the target crest factor of my mix, which I can achieve most of the time with only a few dB of GR on the limiter.

In general, 3-6 dB is the max most limiters can provide without crapping out to my ears, and also this depends on what you feed it. If the input is already a low crest factor signal, than further limiting will likely crap out sooner! If it’s a high crest factor mix (lots of spiky transients), then it will depend on what frequency range the transients reside in and the density of the transients (occasional spikes or constant repeating transients).

Transient density has to do a little with the time spent over the threshold for each crossing, as well as the total number of threshold crossings over the course of the song. If you only cross the threshold a few times in the song, and only for a few milliseconds, you’ll probably never hear it. Even if you cross the threshold every beat, but only for a few milliseconds or less, you’ll also likely not hear it (especially if it’s a consistent transient, like from a drum machine as opposed to a live drummer). But once you are above the threshold for more than a few milliseconds, our ears will tend to recognize it as distortion. That’s based on the fact our ears “average” things over time, so a short audio “clip” of a millisecond may go undetected and simply sound “louder”, while clipping/limiting lasting 3-5 ms or more will tend to be perceived more as distortion rather than loudness.

It’s actually a pretty simple and obvious rule I use: less threshold crossing both in length and frequency will sound less processed. And this in turn will relate to how much GR you’ll see - more time above the threshold for each crossing usually translates into a higher amount of GR.


Sent from some crappy device using Tapatalk

[33db]

In one respect for me it has to do with workflow, such that you separate mastering from mixing as it would be done in hardware. So the "mix" involves ONLY the mixer, and "mastering" is everything AFTER the mix/mixer.

With that in mind, I've patched all mastering to the Master Outputs for some time now. It helps me separate the two processes conceptually and literally. Makes it easy to turn off all mastering, makes it easy to check "mix" levels "pre-mastering", makes it easy to use side-chain filtering on the Master Compressor (a process I've long associated with the "mix" coming from an SSL background for so many years), etc.

For me there are two types of mastering: the "single" and the "album". With the single, I just make the song sound good on it's own, following my basic loudness target of a crest factor of 12 dB max. With the album approach, it's important how each song sounds in context, which includes the previous and the following song. So I tend to make more adjustments in that mode to overall EQ or sometimes multi-band compression, whereas with the single approach I typically just use a brick wall limiter.

There is one subject which can trigger some debate, and that is where to do fade outs. Some say fade out the mix, then send it on to mastering and others say don't do the fade out until after mastering. The difference IMO isn't that huge, so it's really more subjective than objective, just like the debate over whether to compress first then EQ, or EQ then compress. In the end, you have to listen to both and choose the one you prefer, if you don't already have a working knowledge of the difference. What to listen for with regards to fade outs: If you fade "pre-mastering" some say the mix changes as you fade because you are below the threshold. But others (myself included) say we are used to hearing that, and it sounds natural for the song to get less squashed as it fades out - it actually adds to the effect of the fade out IMO! But you won't know for yourself unless you've done it both ways and chosen then one that works best for you.

One thing to consider: if you add dither, you must not fade the audio (or otherwise make ANY changes) AFTER the point dither was added or dither won't work correctly. This means that if you use a brick wall limiter that includes dither, you need to use two instances of the plugin; one for limiting, and then the fade out, and finally one instance of the plugin for adding dither.

Be nice if you did a walk thru video of the things you mentioned.
I have trouble following the concepts that you guys are talking about because I have never really "mastered" or IMO "mixed".
I usually set volume in the mixer while recording and leave it there, I would like to try going a step further and work on the mix *after* the song is finished and then have a idea of what the heck to do when it comes to mastering.

[Voyager]

No, I mean the crest factor of the final mix, which is a simple way to judge loudness and is included in Reason. Using the Big Meter, I set it to read VU + Peak. Since “crest factor” is the peak vs the average level, and the Reason meter uses RMS/average (not actually VU) and Peak levels together in this mode, it’s very easy to read the crest factor. Considering the brick wall limiter keeps peaks at 0 dBFS, the VU/RMS level automatically reads crest factor (peak minus average). So if the VU says - 12 dBFS, that’s your crest factor (removing the “minus” sign: 0 minus -12 = 12).

Now of course the VU level moves around, so I try to not let crest factor go above 12 dB or so on average; sometimes a little above, often a bit (or good bit) below).

Just read this thread were you explain crest factor and it helped me to get a better understand of what it is, thanks for sharing such nice informations.

Here the link for anyone who is interested about this subject and who may have miss the thread :

viewtopic.php?f=5&t=7490715&p=213830&hi ... er#p213830

You said you do not let the crest factor go above 12db or so on average. Is this recomendable for all kind of music ? For Techno for example which is a very dynamic kind of music genre do you still recommend to not exceed a crest factor or ~12db ?

Edit : On the title thread you wrote "dynamic range aka crest factor". On this thread from another forum it seems that users agree that it's not exact the same thing :

https://www.gearslutz.com/board/masteri ... usion.html

Any thoughts ?

[selig]

No, I mean the crest factor of the final mix, which is a simple way to judge loudness and is included in Reason. Using the Big Meter, I set it to read VU + Peak. Since “crest factor” is the peak vs the average level, and the Reason meter uses RMS/average (not actually VU) and Peak levels together in this mode, it’s very easy to read the crest factor. Considering the brick wall limiter keeps peaks at 0 dBFS, the VU/RMS level automatically reads crest factor (peak minus average). So if the VU says - 12 dBFS, that’s your crest factor (removing the “minus” sign: 0 minus -12 = 12).

Now of course the VU level moves around, so I try to not let crest factor go above 12 dB or so on average; sometimes a little above, often a bit (or good bit) below).

Just read this thread were you explain crest factor and it helped me to get a better understand of what it is, thanks for sharing such nice informations.

Here the link for anyone who is interested about this subject and who may have miss the thread :

viewtopic.php?f=5&t=7490715&p=213830&hi ... er#p213830

You said you do not let the crest factor go above 12db or so on average. Is this recomendable for all kind of music ? For Techno for example which is a very dynamic kind of music genre do you still recommend to not exceed a crest factor or ~12db ?

Edit : On the title thread you wrote "dynamic range aka crest factor". On this thread from another forum it seems that users agree that it's not exact the same thing :

https://www.gearslutz.com/board/masteri ... usion.html

Any thoughts ?

Correct; it's a loose use of the term "dynamic range", which has a few definitions (see below). I will not use that comparison in the future to avoid confusion (sorry about that!).

I use a crest factor of 12 dB on average for my production music, as it's where I found most of my previous music to fall. So rather than choosing the number based on outside information, I choose it because it was what I already was doing for the most part. So it's really just a fail-safe for me to work a little quicker and double check my levels. Most of the time I don't adjust anything, as I'm already hitting around that range.

I could see some more aggressive dance music hitting a lower crest factor, understanding that a crest factor from 6 dB to 0 dB is pretty much approaching noise, so you're probably talking 9-10 dB if you still want to retain some punch.

Imagine getting punched in the chest from 1" vs from 12" - which one is likely to have more impact?

Acoustic music tends to have a lower crest factor, as low as 18 dB or lower.

You can also measure crest factor on individual instruments, which gives an indication of where your overall mix will sit. For example, a raw kick drum recording can have a crest factor as wide as 20-24 dB, and if you don't reduce it by a decent amount your mix will also tend to have that crest factor (assuming all instruments in a similar range, of course). Saturation is often more useful than compression to naturally reduce crest factor to more useable levels on a dense mix, since it's not heard as much as "compressed" but you still have to be careful not to sound "distorted" either!

As for the difference between crest factor and dynamic range:
Dynamic range is technically an audio specification, which is similar to signal to noise. Signal to noise is obviously the difference between the signal present and the noise floor, or the max signal allowed and the noise floor. Dynamic range is the difference between the highest and lowest level that can be represented by a system, the main difference begin the bottom range. Whereas S/N measures the noise floor as the lowest level, D/R can extend beyond the noise floor since it's possible to still hear a signal (and to record it) when it goes below the noise floor.

That's the technical definition of dynamic range. The "musical" definition is the difference between the loudest and softest note of a piece of music.

There is also a "micro" dynamic range (vs macro dynamic range, like over the course of a song), which is better described as crest factor. The difference is you can measure crest factor of a single musical note, but not dynamic range. So it would be more accurate to say dynamic range measures the distance between two extremes, and crest factor the peak to average ratio.

Why crest factor is useful
Crest factor is useful because it's the only way I'm aware of to measure a single note or an entire song. BUT, it IS an instantaneous measurement, rather than an overall average, so a song can actually have many crest factors depending on where you measure. That's why I say I go for a max crest factor of 12 dB, give or take a dB or two.

But even for a single note, describing the crest factor can go far towards knowing the dynamics of that instrument and thus your mix. In other words, if all your tracks have a low crest factor (say 10 dB, then your mix will also have a low crest factor. But if one of your tracks has a high crest factor, and that track is predominant in the mix, you will have a mix with a lower crest factor. This is because even one track with a high peak level will result in a mix with a high peak level, unless that track is lower in level than other tracks.

Or to reverse the process, if your track has a crest factor different from what you want, you first look at the individual tracks to see which one(s) is causing the problem. You often hear that if you want a "loud" mix, you start with "loud" tracks! Better in almost all cases to start with your goal in mind rather than to try to achieve it in the final process. So choose "loud" samples, and "loud" synth patches, and write "loud" parts. Chances are your track will practically mix itself if you build it "right" from the ground up.

Similarly if you want to produce a movie with saturated colors, best to film it that way, and best to dress the set that way, and to do makeup/wardrobe that supports that look - rather than to try to add a plugin on the final edit to give it the intended effect. Same goes for if you want a film to be in black and white - don't shoot it in color and change it later! That's all a part of the philosophy of starting as close to your goals as possible in any production, to try to avoid the "fix it in the mix" syndrome. Not only that but it's a more fluid workflow with less "surprises" and detours, which is always a more pleasant experience IMO.

[selig]

Be nice if you did a walk thru video of the things you mentioned.
I have trouble following the concepts that you guys are talking about because I have never really "mastered" or IMO "mixed".
I usually set volume in the mixer while recording and leave it there, I would like to try going a step further and work on the mix *after* the song is finished and then have a idea of what the heck to do when it comes to mastering.

What specifically are you interested in seeing? Maybe better to answer specific questions here?

One overall concept is to consider the general areas of a mix/master on a creative rather than purely technical level. This approach allows you to start with a useful description of your "sound" which will server your aspirations of achieving your sound both from a DIY approach OR if you work with an engineer/producer in collaboration. The better you can visualize and attempt to describe your sound, the easier and quicker you can achieve it IMO.

So rather than talking about EQ, compressors, reverbs, and delays, consider these descriptives:

Frequency response:
Is the mix dark or lo-fi, bright and crispy, or more mid-rangy?

Crest factor:
Is the mix dynamic, or limited in dynamics, (both at a micro and macro level)?

Color/cleanliness:
Is the mix a super-clean mix, or is there a little "color" - or lots of color (just plain dirty)?

Focus:
Is the mix clear and in focus for every element, or more like a soft focus, or even a blurry watercolor?
Or is the foreground in sharp focus and the background totally blurred out?

Density:
Is the mix sparse and open, or dense and full?

Width:
Is the mix more mono, or are there hard panned elements throughout?
Or somewhere in between?

POI (points of interest)
Is this a more hypnotic or meditative track, or are there POIs that get (and hold) your attention periodically?

Naturally, each of these can be applied to individual elements in the track as well as in different sections of the track. A mix can have some "clean" elements/sections and some "soft focus" element/sections and still work just fine - you don't need every element to fit into the overall direction.

[DoctoralHermit]

Wow, as a newcomer that guide Selig just posted is exactly what I've been looking for, just a simple rundown on different artistic decisions made during the mixing/mastering process. Thank you so much!

[33db]

What specifically are you interested in seeing? Maybe better to answer specific questions here?

I know literally nothing, so maybe just a basic start.
The stuff you posted I didn't even know about, I just mix so it sounds good (to me) on my system, of course if I play it in the car or some where else it doesn't sound the same.

[Voyager]

I know literally nothing, so maybe just a basic start.
The stuff you posted I didn't even know about, I just mix so it sounds good (to me) on my system, of course if I play it in the car or some where else it doesn't sound the same.

Start with those tutorials, there is 24 parts and he breakdown the different stages of a mix and it's very well explained in my opinion.

https://www.youtube.com/watch?v=kMYbTcUILvM

Saturation is often more useful than compression to naturally reduce crest factor to more useable levels on a dense mix, since it's not heard as much as "compressed" but you still have to be careful not to sound "distorted" either!

Talking about saturation to compensate crest factor can i use it on a parallel channel or it has a better result on the main channel ?

That's why I say I go for a max crest factor of 12 dB, give or take a dB or two.

You said you go for a crest factor of ~12db for the entire mix but do you apply the same (~12db) rule for each individual track as well ?

Or to reverse the process, if your track has a crest factor different from what you want, you first look at the individual tracks to see which one(s) is causing the problem. You often hear that if you want a "loud" mix, you start with "loud" tracks! Better in almost all cases to start with your goal in mind rather than to try to achieve it in the final process. So choose "loud" samples, and "loud" synth patches, and write "loud" parts. Chances are your track will practically mix itself if you build it "right" from the ground up.

I agree is best to start a project with quality materials to get an easier mixing and mastering route and overall result but couldn't this be compensated by limiting harder and thus having a lower crest factor ?

Similarly if you want to produce a movie with saturated colors, best to film it that way, and best to dress the set that way, and to do makeup/wardrobe that supports that look - rather than to try to add a plugin on the final edit to give it the intended effect. Same goes for if you want a film to be in black and white - don't shoot it in color and change it later! That's all a part of the philosophy of starting as close to your goals as possible in any production, to try to avoid the "fix it in the mix" syndrome. Not only that but it's a more fluid workflow with less "surprises" and detours, which is always a more pleasant experience IMO.

Agree and it's a good advise, as usual i'd say

[Voyager]

Selig, i bumped the thread in case you missed my later questions

[selig]

1-Talking about saturation to compensate crest factor can i use it on a parallel channel or it has a better result on the main channel ?


2-You said you go for a crest factor of ~12db for the entire mix but do you apply the same (~12db) rule for each individual track as well ?


3-I agree is best to start a project with quality materials to get an easier mixing and mastering route and overall result but couldn't this be compensated by limiting harder and thus having a lower crest factor ?

1-Saturation works well either way. For more extreme applications use it as an insert, for more subtle use it parallel.

2-I use -12 dBFS peaks for individual tracks, and -6 dBFS for a mix, and -1 dBFS for masters.

3-Not sure I understand the question 100%, but I feel I get better results by using a little dynamic reduction in multiple stages rather than all at once at the end.