Track sounds bad when limited.

Daniel Carroll

New member
Can someone tell me why my mix sounds good when I have -6db headroom then when I use a limiter to bring it all up it sounds harsh even when I'm not going over 0db?
 
A limiter is a rather extreme form of compression - usually a comp with a ratio of 10:1. So if you're limiting a lot, you are squashing your mix with a lot of compression.
 
Can someone tell me why my mix sounds good when I have -6db headroom then when I use a limiter to bring it all up it sounds harsh even when I'm not going over 0db?

A limiter typically has very short attack and very high ratio, this means it starts acting on the transient while it is still at a high amplitude, then it applies an extreme attenuation through the huge ratio and with the faster attack it also indirectly means a faster release, so when it releases it releases up to a higher amplitude (relative to having the attack slower). This very extreme signal processing degrades the signal, in combination with the threshold you are maximizing the harshness of software based processing. You can avoid it a little by basically putting the threshold near max, which will make the limiter act on the signal while the transient has not yet built up to a high initial amplitude. But that on the other hand does things musically that you might not like for other more artistic/creative reasons especially because the attack is so fast causing a very distinct "hammering" effect.

With a hardware based brickwall peak limiter you get a totally different impact because it applies the whole process with a much greater precision end to end. You can allow it to act on the signal at higher amplitudes, meaning you can set the threshold more mildly without having it degrade the signal due to it.

If you are using software, the best thing you can do is not to use any limiter, but instead a compressor with a very low ratio slow attack long release quite extreme threshold type of configuration that you also apply on as much signal as possible in parallel, so that each compressor can act on a more mild input.

Adding a software compressor on the mix signal is about the worst thing you can do to a mix, it can really suck the vitality out of it. Never do it.
 
Last edited:
As an all digital dude with midi controllers instead of legit analogue equipment I have no idea, Dark Red & Daniel.
I'm not sure what the differences between digital emulation & physical analogue are but that is interesting enough for some research purposes.

Alrighty, back. seems the difference between the two is that analogue has no samplerate.
Digital's maximum [As far as I know] is 192khz.
 
Last edited:
Alrighty, back. seems the difference between the two is that analogue has no samplerate.
Digital's maximum [As far as I know] is 192khz.

No that is not the difference (yes in a way it is but...), the difference is that those that create digital processes need to have deep understanding about nature, a level of understanding the developers have not yet reached, but it is fully possible. In the analog world the engineers do not have to have the same depth of insight about nature, the analog components encapsulate and take care of it by the way they naturally behave under various conditions. All in all hardware based processes output a more true end result.

Software is good at storing and passing the signal forward, either towards an analog process or towards consumption.

Have a listen to the music below, it is the result of hardware. Real and true all the way...

01:19:36 , 01:20:47 (focus on what you really hear, I hope you have good speakers/headphones, http://tidal.com/track/47068039)
 
Last edited:
Analogue has no samplerate, therefore the signal for those would literally be 100% accurate to whatever the analogue source is as analogue doesn't process sound the same way.
When the samplerate increases, the digital sound gets more and more smooth but would still not be 100% accurate.

If you are concerned with this, stick to analogue.
 
Analogue has no samplerate, therefore the signal for those would literally be 100% accurate to whatever the analogue source is as analogue doesn't process sound the same way.
When the samplerate increases, the digital sound gets more and more smooth but would still not be 100% accurate.

If you are concerned with this, stick to analogue.

Everybody should...

Although the sample rate is an important factor it is not the big one, if that was the big one pros would totally exclude software. No, the big one is the digital process of whatever that has been captured, at whatever sample rate and bit depth. That part of the equation is what everybody are missing out on and hence is why they are still trying to make their software work, although it is impossible at this point... With Big Data and Advanced Analytics using technologies like Azure Machine Learning and Cortana Intelligence Suite, some of the issues with digital processing will be partially resolved up to a level when more and more will move from hardware over to software with certain processes, but that is several years into the future from now. We are nowhere near a period of 100% software, we are going to be at 100% hardware at least 5 more years.

So if you are new and consider whether to put your money into hardware of software, the answer is hardware.
 
Last edited:
You prefer hardware. I prefer software.
Many people here have a bias for one or the other.
More power to you if you stick to hardware.

As a person who started on software, I'd recommend what I have experience with and what I know.
My analogue knowledge is zilch since I have not owned hardware outside of midi controllers.

Software will never be analogue because it is code and not circuitry.
 
You prefer hardware. I prefer software.
Many people here have a bias for one or the other.
More power to you if you stick to hardware.

As a person who started on software, I'd recommend what I have experience with and what I know.
My analogue knowledge is zilch since I have not owned hardware outside of midi controllers.

Software will never be analogue because it is code and not circuitry.

Yes, I understand, but I am still responding to your thread topic "Track sounds bad when limited.". The answer/resolution is hardware. Software means no resolution to the thread topic. The fact that you prefer software is the cause of the issue hence why I am pointing that out now, if you would prefer hardware you would not have the issue. Why do you stick to software limiters, can't you feel the sweetness in how hardware compressors handle dynamics? Change your false idea...
 
Last edited:
I'm not arguing with ya, and that other one's Daniel.
All I'm saying is, if you prefer one format over the other then stick to what you prefer. If you get good results with hardware then you stick to hardware.
There's many fellas who are loyalists like you. As there are many software biased people. I prefer digital.
 
The short answer is this will happen with over-limiting. Light, judicious use is the answer (or avoiding brick-walls altogether).

GJ
 
I'm not arguing with ya, and that other one's Daniel.
All I'm saying is, if you prefer one format over the other then stick to what you prefer. If you get good results with hardware then you stick to hardware.
There's many fellas who are loyalists like you. As there are many software biased people. I prefer digital.

Well, I don't prefer having the same issue as you do, so hence I prefer hardware and you are left with the software issue. So what becomes the point of your post (relative to the thread topic)? I was just trying to help you resolve the issue... To resolve your issue, you need to change your idea, that is how it is, not the other way around that I need to do anything about my preferences. So I think you are resisting change, the change of an idea that is false.
 
Last edited:
The short answer is this will happen with over-limiting. Light, judicious use is the answer (or avoiding brick-walls altogether).

GJ

I would say avoiding brick-walls altogether is better. (because that is one process of degradation less, a pro mastering engineer will handle it properly)

Time to bottle a beer guys... :cool:
 
Last edited:
It sounds also like a bitdepth thing the more I read these posts of the various fellas. [daniel, you, the others etc]
I'll do some research about the differences as it has gotten me curious but now I'm starting to think it's also bitdepth related as 24bitdepth and 32bit/64bit depth have immense headroom.

Buuuut this is probably just one other difference and there might be even more to it than that.
This what I find https://www.applied-acoustics.com/techtalk/sampleratebitdepth/
 
Last edited:
It sounds also like a bitdepth thing the more I read these posts of the various fellas. [daniel, you, the others etc]
I'll do some research about the differences as it has gotten me curious but now I'm starting to think it's also bitdepth related as 24bitdepth and 32bit/64bit depth have immense headroom.

Buuuut this is probably just one other difference and there might be even more to it than that.
This what I find https://www.applied-acoustics.com/techtalk/sampleratebitdepth/

No, the closest would be how many times of oversampling the limiter supports. You can for instance use KClip in rendering mode and achieve 256 times oversampling and yes it removes some of the aliasing artifacts, but the underlying issue about digital processing is of course still there. It is mostly the digital process itself that is the weak link because of current lack of understanding about nature (it is simply too advanced, we are not there yet) - the converter headroom, the sample rate and bit depth limitation adds to the issue. But that's it. You then need to change your idea about software relative to hardware and in doing so you are now working closer to the truth and in doing so you are achieving greater sound and greater emotion.

Engineers in general are too stubborn about mixing ITB. Although all pros tell them that no hardware is the backbone of a great sounding commercial mix and where the juice is, they are still stuck with the software idea. It becomes a false dream, an illusion. Then they mix, limit their mix and compare their mix to the commercial reference mix, every time comes the moment of truth, why go through that pain over and over when you can just change an idea...
 
Last edited:
DarkRed-- Nobody is more of a Luddite/analog appreciator than myself, but making this issue your one and only crusade (while it does have the benefit of at least codifying your ideas in a way that many readers can at least understand, but I digress) is ill-conceived on a site where 90% of the participants are a) dyed-in-the-wool digital producers, and b) mostly involved in Hip-Hop/Rap production. We have to help people with the basics, and help them make the most of what they have access to. It is a new world, a post digital revolution world, and you are not going to get 90% of laptop users to suddenly decide that they should sink $30,000 or more into analog gear that must then be tied to a physical location. Nah Gon Do. Let's try and help folks get the most out of whatever they're using (almost certainly an ITB DAW), rather than try to convert everyone to the joys of analog by saying "you'll never make a hit with software."

IT'S TOO LATE. Like it or not, people have been making software-based hits since the 90's. I'd wager that most of the Nashville-based CCM artists that you are fond of posting make their music in ITB DAW's, and they have it mastered by a mastering engineer using a computer-based mastering system. One of the most knowledgeable, old school mastering engineers I've ever met (he started at Motown in Detroit in the 60's) does a lot of his work on a laptop in his living room. I wish that everybody did everything to tape, but that's not how it's done. Let's move on from flogging this particular philosophically dead horse.

GJ
 
Last edited:
DarkRed-- Nobody is more of a Luddite/analog appreciator than myself, but making this issue your one and only crusade (while it does have the benefit of at least codifying your ideas in a way that many readers can at least understand, but I digress) is ill-conceived on a site where 90% of the participants are a) dyed-in-the-wool digital producers, and b) mostly involved in Hip-Hop/Rap production. We have to help people with the basics, and help them make the most of what they have access to. It is a new world, a post digital revolution world, and you are not going to get 90% of laptop users to suddenly decide that they should sink $30,000 or more into analog gear that must then be tied to a physical location. Nah Gon Do. Let's try and help folks get the most out of whatever they're using (almost certainly an ITB DAW), rather than try to convert everyone to the joys of analog by saying "you'll never make a hit with software."

IT'S TOO LATE. Like it or not, people have been making software-based hits since the 90's. I'd wager that most of the Nashville-based CCM artists that you are fond of posting make their music in ITB DAW's, and they have it mastered by a mastering engineer using a computer-based mastering system. One of the most knowledgeable, old school mastering engineers I've ever met (he started at Motown in Detroit in the 60's) does a lot of his work on a laptop in his living room. I wish that everybody did everything to tape, but that's not how it's done. Let's move on from flogging this particular philosophically dead horse.

GJ

Hmm... I know what you mean, it is a little tricky. I am going to try to find some more sweet spots for them, but it can risk becoming misleading, because they need or will need the hardware that's just a fact, but I hear you, they need a little push where they are currently also with what they have.

Here is what you can do to limit better with software... First reduce the total track count to max and remove all EQ and comp plugins on all channels, mix with the Brainworx bx_console channel strip on the individual channels, then use the KClip in Tape saturation mode and offline rendering at 256 times oversampling on only two master busses - 1) center, 2) side, just a little. On the final master bus you then run the
Brainworx bx_XL V2 mastering limiter just a little, mostly to set the ceiling and to slightly tune the stereo image. Any frequency issues anywhere are corrected using the Brainworx bx_console. Monitor the True Peak level with iZotope Insight.
 
Last edited:
I don't believe there is such a thing as analogue brick-wall limiting, is there? There are very fast compressors, saturators and clippers.

I suspect brick-wall limiters use all kinds of psychoacoustic tricks that can't be done in analogue in any practical way.

Anyway using the best analogue whatever in the world can still produce terrible results so assuming that as a factor is just that: an assumption.
 
try mulsistage compression with first a multiband and then a punchy compressor... after that you can limit your track because the limiter doesn't have to work as hard.
 
Back
Top