The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Digital Limiter Quality Test Dynamics Plugins
Old 13th April 2019
  #61
Lives for gear
 
zmix's Avatar
 

Quote:
Originally Posted by johannburkard View Post
The FFT makes it easier to see what distortion products are being generated by the limiting. Many of the limiters have time constants that make it hard to show a time domain screenshot.

Harmonics generated under gain reduction are IM byproducts of the sidechain's interaction with the signal. In my tests I have observed that hardware compressors typically generate both even and odd order harmonics under gain reduction and that all plugin compressors only generate odd order harmonics, and though there are a few that will add even order harmonics, they are "faking" this response by adding asymmetry to the audio path, so these faked even harmonics are present even when no gain reduction is being applied. In my hardware compressors the signal path is typically very clean and generate no harmonics when not compressing.


As for your comment that is is "hard to show a time domain screenshot", you should know that there are standardized tests using level shifted tone bursts specifically designed to show the attack and release behavior, and please note that the behavior of the sidechain is the sole factor in determining the "sound" of a limiter.

See the posted video above to see how this has been incorporated into a series of plugins to test compressor behavior.
Old 13th April 2019
  #62
Gear Maniac
 
dione's Avatar
@ johannburkard :

Will you please be so kind to delete all my test results from your test ? It's pretty clear that i didn't understand what your goal was with this test.

I was simply using my ears to get a clean result. Starting with some basic presets, trying to get a louder output with no audible distortion. That's what clean transparent limiting is for me in the end.
I also left out some options, like putting Pro-L2 in Safe Mode or trying different settings in Limitless. I posted the ones that gave me the best audible result quickly while still raising the levels by a big amount, with obvious limiting going on. Because i was curious to see if those settings would show weird behaviour in the frequency plots you posted. Not to see whether they meet your requirements that you didn't post.

It was also not clear from your first post that you also did other tests, which you base your findings on. The posted graphs show different results and there's no way in finding out whether the conclusions you draw are true or false or whether they are based on simple personal preference.

I also seriously doubt your recommendations, as stated a few times before, simply because people that read this thread ONLY see the graphs that are posted.
You simply cannot draw conclusions of something and show us something completely else. That's like showing us pictures of a few cars and tell us that one drives faster than the other.

So again : please delete the results from all the limiters that i contributed. Maybe someone else can try. Or yourself because you can demo all these tools, so you can set them up to meet your requirements.
Old 13th April 2019
  #63
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by dione View Post
@ johannburkard :
Will you please be so kind to delete all my test results from your test ? It's pretty clear that i didn't understand what your goal was with this test.
Just deleted them.
Old 13th April 2019
  #64
Lives for gear
 
kuasalogam's Avatar
Quote:
Originally Posted by johannburkard View Post
@ kuasalogam I've added the samples you posted. Thank you!
That's interesting, because I think Maxim sounds 'bland'.
I believe Kratos and AOM might perform better with the right setting
Old 13th April 2019
  #65
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by kuasalogam View Post
That's interesting, because I think Maxim sounds 'bland'.
I believe Kratos and AOM might perform better with the right setting
I haven't used Maxim yet but I could imagine that what you're hearing with the other limiters is distortion.
Old 4 weeks ago
  #66
Lives for gear
 
DrAudioBot's Avatar
thanks for the effort.
Old 4 weeks ago
  #67
Lives for gear
 
kuasalogam's Avatar
Whhat app did you use to examine and generate the graphs?
Old 4 weeks ago
  #68
Quote:
Originally Posted by waltercruz View Post
Some limiters that I want to see: Limiter no 6 GE and Voxengo Elephant 4.6. Is someone have them, please add to the test!
I have Elephant 4.x (I don't know if it's 4.6, though), D16 Group's Frontier, Slate Digital's FG-X and Waves' L1 and L1 Ultramaximizer (which are probably pretty much the same thing as far as this test is concerned), and some Abbey Road limiters, if anyone wants to see those.

I'd be interested to see some of the UAD or Acustica Audio results.

Steve
Old 4 weeks ago
  #69
Gear Maniac
Quote:
Originally Posted by ShadowsOfLife View Post
I have Elephant 4.x
Please post Elephant 4!
Old 4 weeks ago
  #70
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by kuasalogam View Post
Whhat app did you use to examine and generate the graphs?
Just Audacity. Nothing fancy. It's a 2^16 FFT.
Old 4 weeks ago
  #71
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by ShadowsOfLife View Post
I have Elephant 4.x (I don't know if it's 4.6, though), D16 Group's Frontier, Slate Digital's FG-X and Waves' L1 and L1 Ultramaximizer (which are probably pretty much the same thing as far as this test is concerned), and some Abbey Road limiters, if anyone wants to see those.
Since there's been some confusion before: Try to get the cleanest result you can. Use Voxengo SPAN (or similar). Shortest release possible.
Old 4 weeks ago
  #72
Lives for gear
 

I decided to get the Acon Digital Mastering Suite and uninstall my Toneboosters Barricade. I’m impressed by all the software in this suite and now I feel confident that I’m not damaging my track by running it through the limiter to gain some dBs.
Thank you for this thread!
Old 4 weeks ago
  #73
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by johannburkard View Post
DMG Limitless has visible release. Pro-L2 in "Transparent" mode has almost none which is why I've included it. If you can get Limitless to output what Avid Maxim can do, I'll include it, too.
That wouldn't be good. Maxim is useless as a mastering limiter IME. Limitless is one of the best.

I think you are confusing the visual look of what is going on with how things actually sound. For instance you have mentioned several times that such and such a limiter lowers the signal before the transient as though it is a bad thing when in reality it is a GOOD thing. The loud transient psychoacoustically masks whatever soft stuff happens before it. This means it is much better sonically to lower the level before the loud transient, a change in volume that will be entirely masked to our hearing, than to do it during the loud transient which will be clearly audible.

So basically you are not accurately interpreting the test results and drawing the wrong conclusions. Sorry to be blunt but this entire thread is a waste of time and is misleading.

Alistair
Old 4 weeks ago
  #74
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by UnderTow View Post
For instance you have mentioned several times that such and such a limiter lowers the signal before the transient as though it is a bad thing when in reality it is a GOOD thing.
The reductions in volume that I have seen are so short to be probably inaudible. Many times, they are happening on the slope of the previous sine wave as it's going down.

Quote:
Originally Posted by UnderTow View Post
The loud transient psychoacoustically masks whatever soft stuff happens before it.
You said that lowering what happens before the transient is good and now you say that the transient is masking that? So why make what happens before the transient even more quiet?

Quote:
Originally Posted by UnderTow View Post
This means it is much better sonically to lower the level before the loud transient, a change in volume that will be entirely masked to our hearing, than to do it during the loud transient which will be clearly audible.
A limiter gets rid of transients (if it's limiting, that is) so we don't have too many options there, do we?

Quote:
Originally Posted by UnderTow View Post
So basically you are not accurately interpreting the test results and drawing the wrong conclusions. Sorry to be blunt but this entire thread is a waste of time and is misleading.
I suppose there is an "ignore thread" option somewhere.
Old 4 weeks ago
  #75
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by UnderTow View Post
That wouldn't be good. Maxim is useless as a mastering limiter IME. Limitless is one of the best.
And just to clarify that: The samples in this thread from DMG Limitless have a visible release. If you can bring that to zero, then DMG Limitless would also be an option.
Old 4 weeks ago
  #76
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by johannburkard View Post
The reductions in volume that I have seen are so short to be probably inaudible. Many times, they are happening on the slope of the previous sine wave as it's going down.
Yes that makes sense. Basically it depends on the look-ahead time set in the limiter (or some internal parameter that the user can not control).

Quote:
You said that lowering what happens before the transient is good and now you say that the transient is masking that? So why make what happens before the transient even more quiet?
If there is a loud peak that will pass the ceiling, a brick-wall limiter needs to lower the level of the signal to avoid it passing the ceiling (or it isn't a brick-wall limiter). To do that it needs to do a gain change. The gain change itself is a modulation of the original signal. This modulation causes distortion (IMD). If you do that modulation just before the loud peak, the loud peak psychoacoustically masks whatever distortion that gain change might be causing. This makes the whole process more transparent which is of course a good thing.

Quote:
A limiter gets rid of transients (if it's limiting, that is) so we don't have too many options there, do we?
Yes we do. We can do it just before the peak with the advantage explained above or you can do it as the peak hits (so basically the gain change tracks the signal curve in the opposite direction which makes it closer to a clipper). Doing it this way tends to be more audible, hence smarter algorithms doing it slightly before the peak.

(There are other choices too. You can clip loud transients and do slower gain changes to deal with slower fluctuations in signal level. You can separate out the peak and RMS behaviour of the signal (and thus how the limiter behaves on that signal) and many other options and nuances).

Quote:
I suppose there is an "ignore thread" option somewhere.
It isn't about me. It is about this thread being based on a lack of understanding of how limiters, sound and psychoacoustics work. You haven't even listened to the limiters! This thread is spreading misinformation.

I already see some people making purchase decisions based on this misinformation. This is not a good thing! That is why I am commenting.

Quote:
Originally Posted by johannburkard View Post
And just to clarify that: The samples in this thread from DMG Limitless have a visible release. If you can bring that to zero, then DMG Limitless would also be an option.
This is exactly the problem with this whole thread. You are looking at pictures and not understanding what they mean and you are not LISTENING to what the limiters sound like!

This thread is literally the deaf (because you are not listening) leading the blind. Please educate yourself instead of spreading misinformation.

Alistair
Old 4 weeks ago
  #77
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by UnderTow View Post
The gain change itself is a modulation of the original signal. This modulation causes distortion (IMD).
Changing level in the digital world is a multiplication operation which may introduce miniscule quantization errors.

Why you think that multiplication causes IMD is beyond me.

Quote:
Originally Posted by UnderTow View Post
This is exactly the problem with this whole thread. You are looking at pictures and not understanding what they mean and you are not LISTENING to what the limiters sound like!
How do you know what I'm doing?

Again, this thread is not about what limiters "sound" like. It documents with which limiters you can have (more or less at times) distortion-free limiting and with what limiters you don't. All I want is giving myself and other GS a choice.
Old 4 weeks ago
  #78
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by johannburkard View Post
Changing level in the digital world is a multiplication operation which may introduce miniscule quantization errors.

Why you think that multiplication causes IMD is beyond me.
It is about changing the gain. (The actual change itself, not a steady-state gain reduction which indeed only introduces tiny amounts of quantization distortion or dither). That is a modulation of the original signal. That causes IMD. All compressors/limiters cause IMD.

Quote:
How do you know what I'm doing?
Because 1) You have stated yourself you don't have some of the limiters yet you are making comparisons and pronouncements 2) Your description of what you are doing and how you are analysing and judging the limiters indicates that you are not listening. So please stop being disingenuous.

Quote:
Again, this thread is not about what limiters "sound" like. It documents with which limiters you can have (more or less at times) distortion-free limiting and with what limiters you don't. All I want is giving myself and other GS a choice.
You are looking at irrelevant stuff and ignoring the relevant stuff. Distortion that you can not hear (because for instance it is masked) is completely irrelevant. You do not understand the subject enough to draw any useful conclusions and you are sharing these erroneous conclusions with others. The deaf leading the blind.

This thread is worse than useless. It is misleading.

Alistair
Old 4 weeks ago
  #79
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by UnderTow View Post
This thread is worse than useless. It is misleading.
The door is open. --> [_|]
Old 4 weeks ago
  #80
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by johannburkard View Post
The door is open. --> [_|]
I will keep alerting people of the futility of your test methods and conclusions if that is OK with you.

Alistair
Old 4 weeks ago
  #81
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by UnderTow View Post
I will keep alerting people of the futility of your test methods and conclusions if that is OK with you.
Only in the corridor.
Old 3 weeks ago
  #82
Lives for gear
 
zmix's Avatar
 

Quote:
Originally Posted by johannburkard View Post
Changing level in the digital world is a multiplication operation which may introduce miniscule quantization errors.

Why you think that multiplication causes IMD is beyond me.

Clearly it is. Let me explain: When you make an instantaneous amplitude change, you introduce ampliitude modulation to the signal. Amplitude modulation produces sum and difference products of the carrier frequency and modulation envelope known as sidebands or Intermodulation distortion (aka IMD).
Old 3 weeks ago
  #83
Gear Addict
 
johannburkard's Avatar
Quote:
Originally Posted by zmix View Post
Clearly it is. Let me explain: When you make an instantaneous amplitude change, you introduce ampliitude modulation to the signal. Amplitude modulation produces sum and difference products of the carrier frequency and modulation envelope known as sidebands or Intermodulation distortion (aka IMD).
Splitting hairs... but okay. I was talking about static level change, you're talking about limiter action. Two different things.
Old 3 weeks ago
  #84
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by johannburkard View Post
Splitting hairs... but okay. I was talking about static level change, you're talking about limiter action. Two different things.
This has nothing to do with splitting hairs. Even though I had been very clear and explained exactly what I was talking about, you misunderstood what I wrote about IMD because you have no understanding of the subject. You then talked about static level changes out of sheer cluelessness.

You responded to me remember. Pretending that you were talking about something else just shows you can't follow a simple conversation let alone perform technical tests with useful conclusions.

You are not equipped to report on limiters because you don't understand the very basics of how these tools or audio in general works.

Alistair
Old 3 weeks ago
  #85
Lives for gear
 
zmix's Avatar
 

Quote:
Originally Posted by johannburkard View Post
Splitting hairs... but okay. I was talking about static level change, you're talking about limiter action. Two different things.
You stated you were trying to determine which plugins gave "distortion free limiting". I can see that you don't understand how limiting actually works, and it's unfortunate to see you persistently deflect any new information given here to help educate you.
Old 3 weeks ago
  #86
Gear Addict
 
johannburkard's Avatar
@ UnderTow and @ zmix It seems you don't like what I'm doing here. The question I'm starting to ask myself is why you are still posting in this thread?
Old 3 weeks ago
  #87
Gear Addict
 
saagedal's Avatar
 

Hi guys,
Quote:
Originally Posted by UnderTow View Post
This has nothing to do with splitting hairs. Even though I had been very clear and explained exactly what I was talking about, you misunderstood what I wrote about IMD because you have no understanding of the subject. You then talked about static level changes out of sheer cluelessness.

You responded to me remember. Pretending that you were talking about something else just shows you can't follow a simple conversation let alone perform technical tests with useful conclusions.

You are not equipped to report on limiters because you don't understand the very basics of how these tools or audio in general works.
Quote:
Originally Posted by zmix View Post
You stated you were trying to determine which plugins gave "distortion free limiting". I can see that you don't understand how limiting actually works, and it's unfortunate to see you persistently deflect any new information given here to help educate you.
The OP measures the distortion using the mean spectrum over the whole time period as far as I understand and the test is perfectly valid in my opinion. You see the typical increased distortion in many of the limiters on the low frequency input which occurs during the steady state and has nothing to do with the level changes. Of course there are other aspects to limiting, but that doesn't invalidate the test.

Best,
Stian
Old 3 weeks ago
  #88
Lives for gear
 
zmix's Avatar
 

Quote:
Originally Posted by saagedal View Post
You see the typical increased distortion in many of the limiters on the low frequency input which occurs during the steady state and has nothing to do with the level changes.
What you're seeing is specifically due to the action of the gain reduction mechanism. It's not distortion due to nonlinearities in the signal path.
You can easily test this by sending a 100Hz sinewave into the limiter, disconnecting the sidechain while observing the spectrum produced.
Old 3 weeks ago
  #89
Gear Addict
 
saagedal's Avatar
 

Quote:
Originally Posted by zmix View Post
What you're seeing is specifically due to the action of the gain reduction mechanism. It's not distortion due to nonlinearities in the signal path.
You can easily test this by sending a 100Hz sinewave into the limiter, disconnecting the sidechain while observing the spectrum produced.
I know that, but why does it matter where it occurs? Users would obviously like to see undistorted steady state signals and don't really care where the distortion occurs. Digital limiters with sufficient and properly designed look-ahead avoid the problem.

Best,
Stian
Old 3 weeks ago
  #90
Lives for gear
 
zmix's Avatar
 

Quote:
Originally Posted by saagedal View Post
I know that, but why does it matter where it occurs? Users would obviously like to see undistorted steady state signals and don't really care where the distortion occurs. Digital limiters with sufficient and properly designed look-ahead avoid the problem.

Best,
Stian
You claimed the "distortion" : "occurs during the steady state and has nothing to do with the level changes."

This is incorrect. In fact it has everything to do with "level changes".
If you disconnect the sidechain (thus removing any gain reduction or "level changes")it will go away.

I didn't expect that these fundamental concepts would be so challenging.
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…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Luke44 / Electronic Music Instruments and Electronic Music Production
44
Afroham / So Much Gear, So Little Time
2
kundevelop / Mastering Forum
2
space2012 / Gear Shoot-Outs / Sound File Comparisons / Audio Tests
2
PaRaNoId / So Much Gear, So Little Time
12

Forum Jump
Forum Jump