The No.1 Website for Pro Audio
High Pass Filter at 20Hz
Old 13th May 2012
  #1
Gear Addict
 
🎧 10 years
High Pass Filter at 20Hz

I've heard so many opinions on this.

And I understand that there is no definitive practice, but I find that high passing at 20Hz cleans up my songs considerably. Not only that, but I can push a limiter harder with less obvious limiting. It frees up headroom and eliminates any DC offset problems just like that.

The cons... maybe some subfrequencies that are subliminally felt are gone.

It's not like my music will be played on a system that can produce below 20Hz, so....

Your thoughts, opinions, comments, techniques, etc?
Old 13th May 2012
  #2
Lives for gear
 
8 Reviews written
🎧 10 years
According to the Red Book the frequency response of an audio CD should be 20 Hz to 20 kHz.

I think that most encoders automatically cut at 20 Hz but sometimes it can help to use HPF.
Old 13th May 2012
  #3
Gear Head
 
dmsmastering's Avatar
 
🎧 10 years
Yes, if there are no song related productive frequencys this is a possibility. If you need to leave that subbass but having dc offset problems in the mix, the 16hz hp filter of the massive passive can do the trick well. I sometimes use to go with compressors that have a controllable side chain hp filter for leaving the bass/subbass untouched depending on the music. The limiters itself, input gain and settings can sound pretty different you may should try settings that do not affect the sound that much but only prevent for clipping. Thats the danger with limiters. If it affects the sound play around with the input gain, attack and release for limiting unwanted peaks but leaving life and dynamic. Auto-gain off. For processing only the low end try out multiband compressors and bypass what you do not need to process. Depending what you like to achieve this could be better instead firing the limiter so hard. But there are still a hand full mixes that have really problematic sub-bass frequencies for example rumble and pops that harm to the song and any dynamic processor. Problematic because the producer/mixer could hardly hear those frequencies on small near-fields. I like to get them listen to full range speakers first to talk about hp the trouble making tracks in the mix. This way you can keep some atmosphere in the low end instead cutting off the whole low end in the sum and dynamic processors can work more musical constantly to the song.
Old 13th May 2012
  #4
Lives for gear
 
DrAudioBot's Avatar
 
2 Reviews written
🎧 10 years
Quote:
Originally Posted by Rossd25 ➑️
I've heard so many opinions on this.

And I understand that there is no definitive practice, but I find that high passing at 20Hz cleans up my songs considerably. Not only that, but I can push a limiter harder with less obvious limiting. It frees up headroom and eliminates any DC offset problems just like that.

The cons... maybe some subfrequencies that are subliminally felt are gone.

It's not like my music will be played on a system that can produce below 20Hz, so....

Your thoughts, opinions, comments, techniques, etc?

Not always true about the "limiter/headroom" part. In some cases (quiet often) setting a 20Hz highpass filter may even INCREASE the peak with a minimum phase EQ.

Try it for yourself: Place a regular EQ on your Drumbuss and activate a HP filter at 20 or 30 Hz. Now watch the peak level on that channel!
(which doesn't happen with a Linear Phase EQ, that's why it's sometimes liked during mastering for exactly this purpose).

Cleaning up the low end is, of course, a cool thing and can be done with a regular EQ, especially on channels that absolutely don't need the low end information (below 100, 200 or whatever Hz).
IME just can be tricky, especially if the low cut is set very low. Highpassing channels shouldn't be a default thing IMO, though (only with obvious instruments, of course).
Old 13th May 2012
  #5
Lives for gear
 
MAzevedo's Avatar
 
🎧 15 years
I usually find the phase shift (if minimum phase) or pre-ringing (if linear phase) of a high pass filter too much of a tonal sacrifice to clean out low-end cruft with one. I'm much more likely to use a low shelf if there are subsonic problems that a low-cut.

Also, re: jetam's comment. There is no low cut at 20 Hz on a CD.
Old 13th May 2012 | Show parent
  #6
Gear Addict
 
🎧 10 years
Quote:
Originally Posted by jetam ➑️
According to the Red Book the frequency response of an audio CD should be 20 Hz to 20 kHz.

I think that most encoders automatically cut at 20 Hz but sometimes it can help to use HPF.
This is what I heard, too.

Yeah, and I understand that some EQ can add volume by cutting low, but I always use the Linear Phase EQ in Logic for mastering.

I'm still not sure about the whole DC offset thing, but after making this cut I have a pretty much centered wave form and I notice that on the Ozone DC meter it is evenly distributed on both sides. This is what I want, right?

I can't image anything below 20Hz being a big deal. Most of the low end rumble seems to be between 30Hz and a 150Hz, which I would never imagine rolling off. Maybe on some systems it can reproduce those details, but I was also under the impression that most system can't go lower than 20Hz anyway. And for mp3 distribution 20Hz is cut by the mp3 converter. Again, I don't know for certain, which is why I'm posting here. I have read this in a ton of places, though.
Old 13th May 2012 | Show parent
  #7
Lives for gear
 
dcollins's Avatar
 
Verified Member
🎧 15 years
Quote:
Originally Posted by jetam ➑️
According to the Red Book the frequency response of an audio CD should be 20 Hz to 20 kHz.

I think that most encoders automatically cut at 20 Hz but sometimes it can help to use HPF.

Incorrect. The Red Book doesn't specify any low cut and the frequency response of CD can extend to 0Hz. And to 22.050kHz for those keeping score at home.

Wrt 20Hz filters, in many cases it can range from inaudible to helpful, but you have to consider what rolloff rate you are using and the type of filter.

The general message is that there is never any default setting (just like there is never a " best" anything) and if the track needs it, feel free to attenuate the subs.

The DC offset question has been done to death here, and the answer is that it's extremely rare to have actual DC offset problems. What is common is the mistaken use of analysis software that shows offset, but there really isn't any issue that needs "fixing."


DC
Old 13th May 2012 | Show parent
  #8
Lives for gear
 
sat159p1's Avatar
 
3 Reviews written
🎧 10 years
Quote:
Originally Posted by dmsmastering ➑️
I like to get them listen to full range speakers first to talk about hp the trouble making tracks in the mix. This way you can keep some atmosphere in the low end instead cutting off the whole low end in the sum and dynamic processors can work more musical constantly to the song.
Agreed for the most part of your post, but I don't think you'll ruin atmosphere (or anything_ when put 20Hz (or even 30hz) HP filter on the sum.
Old 13th May 2012
  #9
Gear Addict
 
🎧 10 years
I have checked these both in Ozone and the DC meter is giving me two different things. First of all, on the unfiltered tracks the DC meter is showing my waveform moving up more than down, but on the tracks I filtered at 20Hz the track is evenly moving up and down. Is this a DC offset problem or something else? It's strange, because on most of the commercial CDs I am referencing the DC meter is evenly distributed between up and down. Or maybe I am reading this meter wrong?

I understand that DC offset is the actual waveform off of 0, but then why is my waveform acting so strange? Why isn't it more centered or even? Is this a DC problem or something else entirely?
Old 14th May 2012
  #10
Deleted 691ca21
Guest
How does it sound?
Old 14th May 2012 | Show parent
  #11
Lives for gear
 
Red Mastering's Avatar
 
Verified Member
🎧 10 years
Quote:
Originally Posted by Babaluma ➑️
How does it sound?
who cares here??))
this is 'talkers' forum, ppl like to talk about the SOUND,
instead of making it
but you have a point,
it's audio - listen&judge&use tools you think will benefit,
then listen again- sound better - good,
worst? - well then step back and try different thing
Old 15th May 2012
  #12
Lives for gear
 
Laurend's Avatar
 
🎧 10 years
Quote:
Originally Posted by Rossd25 ➑️
I've heard so many opinions on this.

And I understand that there is no definitive practice, but I find that high passing at 20Hz cleans up my songs considerably. Not only that, but I can push a limiter harder with less obvious limiting. It frees up headroom and eliminates any DC offset problems just like that.

The cons... maybe some subfrequencies that are subliminally felt are gone.

It's not like my music will be played on a system that can produce below 20Hz, so....

Your thoughts, opinions, comments, techniques, etc?
It just brings one more filter in the signal path, which introduces phase shift (IIR) or pre-ringing (FIR) issues in the remaining part of the signal.
In both cases, the limiter must work harder.
The only advantage of inserting a HPF is removing the DC offset if any.
Old 15th May 2012
  #13
Deleted 691ca21
Guest
I think another advantage is if you like the sound of it in/if you need to have some bass reduction/if cutting the subsonic mush makes the bass that is left sound better etc. I've found the last to be the case on numerous occasions.

I don't tend to use it to deal with DC offset, as just running the track through the outboard (with ten transformers in the chain), tends to do that.

Case by case...
Old 18th May 2012
  #14
Lives for gear
 
🎧 10 years
CD goes down to 5 Hz, not 0 Hz which is DC. I do not think there is a built in filtering in CD authoring software, though. System just can not handle DC and frequencies approaching it.

If the lowest end is left unfiltered (at 16 or 20 Hz) at least you need to make sure there is nothing there not intended to be there, either by having a monitoring system which goes low enough (who has a sub that goes to 10 Hz raise your hand) or running a frequency analysis. Subsonic signals are not nice to reflex speakers, and including unintended low frequency signals in the master is not polite to those few who happen to have systems which really go low.
Old 18th May 2012 | Show parent
  #15
Lives for gear
 
dcollins's Avatar
 
Verified Member
🎧 15 years
Quote:
Originally Posted by Petrus ➑️
CD goes down to 5 Hz, not 0 Hz which is DC. I do not think there is a built in filtering in CD authoring software, though. System just can not handle DC and frequencies approaching it.
The response does extend to DC. There is no built-in filter either in the red book standard or in software. This doesn't mean that there might be various filters introduced along the way, but they are not an inherent part of the system.

DC
Old 19th May 2012 | Show parent
  #16
Lives for gear
 
Adam Dempsey's Avatar
 
Verified Member
🎧 15 years
Quote:
Originally Posted by Petrus ➑️
(who has a sub that goes to 10 Hz raise your hand).
*raising two hands for stereo subs.*

Quote:
Originally Posted by dcollins ➑️
There is no built-in filter either in the red book standard or in software. This doesn't mean that there might be various filters introduced along the way, but they are not an inherent part of the system.
Although there used to be in Sonic classic, by default at 7 or 8Hz, from memory. Had to bypass it in prefs, which I'm sure you did too, Dave?

(also 10Hz filter by default in iTunes' mp3 encoder but that's barely worth mentioning...).
Old 19th May 2012 | Show parent
  #17
Lives for gear
 
🎧 10 years
Quote:
Originally Posted by dcollins ➑️
The response does extend to DC.

DC
It is my understanding that there can not be DC in the CD signal as that can not handled by the analog electronics of the system. Even the data stream actually put on the disk has one extra bit (out of 17) which is only used to flip the polarity of the digital signal when DC buildup starts to occur (the square wave is analog at this stage). Digital data would not mind having DC in it, but nothing in the analog side is capable of handling it and it does not serve any purpose either. It would only push the woofer cone out or pull it in, causing distortions. When filtering out DC there is no need of the filter to be sharp, maybe that is where the commonly referred 5 Hz comes in.
Old 19th May 2012
  #18
Lives for gear
 
Alexey Lukin's Avatar
 
Verified Member
🎧 10 years
Never heard of this 17-th bit. You may be confusing with CRC bits.
Old 19th May 2012 | Show parent
  #19
Lives for gear
 
dcollins's Avatar
 
Verified Member
🎧 15 years
Quote:
Originally Posted by Petrus ➑️
It is my understanding that there can not be DC in the CD signal as that can not handled by the analog electronics of the system. Even the data stream actually put on the disk has one extra bit (out of 17) which is only used to flip the polarity of the digital signal when DC buildup starts to occur (the square wave is analog at this stage). Digital data would not mind having DC in it, but nothing in the analog side is capable of handling it and it does not serve any purpose either. It would only push the woofer cone out or pull it in, causing distortions. When filtering out DC there is no need of the filter to be sharp, maybe that is where the commonly referred 5 Hz comes in.
The actual signal on the CD is not the 16 bit PCM word. It is encoded using a system called Eight to Fourteen modulation or EFM. While this code is known as "DC free," it does, in fact, allow the recording of DC through the system.

There is no requirement for the analog stages of the CD player to be AC coupled either, although it's good engineering practice to do so.

Users of early PCM systems will have stories of pops and clicks on the down- beat caused by DC offsets in converters of the day that traveled all the way to the D/A................

DC
Old 19th May 2012 | Show parent
  #20
Lives for gear
 
🎧 10 years
Quote:
Originally Posted by Alexey Lukin ➑️
Never heard of this 17-th bit. You may be confusing with CRC bits.
Each byte (8 bits) in the CD data stream is converted into 14 bits, then these 14 bit chunks are connected with 2 connection bits to make sure each 1 is followed at least with 2 zeros, then one more bit, the seventeenth, is added so that the cumulative number of zeros stays same on both upper and lower side of the resulting square wave = no DC. This has nothing to do with audio, but the actual data stream coding on the physical disk.
Old 19th May 2012 | Show parent
  #21
Lives for gear
 
Greg Reierson's Avatar
 
Verified Member
5 Reviews written
🎧 15 years
Quote:
Originally Posted by Adam Dempsey ➑️
Although there used to be in Sonic classic, by default at 7 or 8Hz, from memory. Had to bypass it in prefs,
I think it was 18Hz, and it caused a lot of grief for Sonic as many people noticed that Sonic had a 'sound' that wasn't quite right. Turning off the filter fixed the problem.

GR
Old 19th May 2012 | Show parent
  #22
Lives for gear
 
Greg Reierson's Avatar
 
Verified Member
5 Reviews written
🎧 15 years
Quote:
Originally Posted by Petrus ➑️
Each byte (8 bits) in the CD data stream is converted into 14 bits,

This has nothing to do with audio, but the actual data stream coding on the physical disk.
To say it another way, EFM ensures that the physical pit structure on the disc can be stamped and read reliably. Without it, discs would be harder to stamp and more susceptible to dust and scratches.

GR
Old 19th May 2012 | Show parent
  #23
Lives for gear
 
🎧 10 years
Quote:
Originally Posted by Greg Reierson ➑️
To say it another way, EFM ensures that the physical pit structure on the disc can be stamped and read reliably. Without it, discs would be harder to stamp and more susceptible to dust and scratches.

GR
Well, not quite. The start and end of each "pit" is one (1) and there has to be at least 2 and maximum of 10 zeros (0) between each 1. Those are the minimum maximum pit and flat lengths (pits and flats behave identically). 14 bits are needed to code all possible 8 bit combinations so that this requirement is fulfilled. This EFM translation lookup table is burned in CD iron. When putting these 14 bit chunks together 2 bits are used as "glue" to make sure ones do not get too close together, and the seventeenth bit is used to flip the phase as needed to keep DC at bay. You can visualize the signal as a square wave of sorts where the vertical points are ones (change of state in laser reflection) and flat sections are strings of zeroes, and the cumulative length of positive voltage zeros and negative voltage zeroes have to be kept the same, or there is DC present. 17:th bit is needed for this.

All above concerns the data coding, not audio signal filtering, though, so totally OT.
Old 19th May 2012 | Show parent
  #24
Lives for gear
 
Greg Reierson's Avatar
 
Verified Member
5 Reviews written
🎧 15 years
Quote:
Originally Posted by Petrus ➑️
Well, not quite. The start and end of each "pit" is one (1) and there has to be at least 2 and maximum of 10 zeros (0) between each 1. Those are the minimum maximum pit and flat lengths (pits and flats behave identically). 14 bits are needed to code all possible 8 bit combinations so that this requirement is fulfilled.
I thought I said that Those are the minimum and maximum pit and land lengths because shorter lengths would not produce reliably readable discs.

GR
Old 19th May 2012
  #25
Lives for gear
 
🎧 10 years
OK then.
Old 19th May 2012 | Show parent
  #26
Lives for gear
 
dcollins's Avatar
 
Verified Member
🎧 15 years
Quote:
Originally Posted by Greg Reierson ➑️
To say it another way, EFM ensures that the physical pit structure on the disc can be stamped and read reliably. Without it, discs would be harder to stamp and more susceptible to dust and scratches.
EFM tailors the frequency spectrum of the data to fit into the best range for the optical system. It also allows response to DC although the system Laser and detector is ultimately an AC coupled system.


DC
Old 19th May 2012
  #27
Lives for gear
 
Greg Reierson's Avatar
 
Verified Member
5 Reviews written
🎧 15 years
A much more elegant explanation than I could muster...
Old 19th May 2012
  #28
Here for the gear
 
🎧 10 years
When you decide to cut everything below 30 or 20 hz, what slope do you use ? does it depend of the style of music ? a standard 12db/oct is usually enough ?
Old 19th May 2012 | Show parent
  #29
Lives for gear
 
8 Reviews written
🎧 10 years
Quote:
Originally Posted by dcollins ➑️
Incorrect. The Red Book doesn't specify any low cut and the frequency response of CD can extend to 0Hz. And to 22.050kHz for those keeping score at home.

Wrt 20Hz filters, in many cases it can range from inaudible to helpful, but you have to consider what rolloff rate you are using and the type of filter.

The general message is that there is never any default setting (just like there is never a " best" anything) and if the track needs it, feel free to attenuate the subs.

The DC offset question has been done to death here, and the answer is that it's extremely rare to have actual DC offset problems. What is common is the mistaken use of analysis software that shows offset, but there really isn't any issue that needs "fixing."


DC
Well as far as I know the Red Book mentions 20 Hz, but it's not very clear on that topic. It is certainly possible to record well under 20 Hz on the CD.
Linear phase HPF at 20 Hz usualy makes more good than bad.
Old 19th May 2012 | Show parent
  #30
Lives for gear
 
MAzevedo's Avatar
 
🎧 15 years
Quote:
Originally Posted by jetam ➑️
Well as far as I know the Red Book mentions 20 Hz, but it's not very clear on that topic. It is certainly possible to record well under 20 Hz on the CD.
Linear phase HPF at 20 Hz usualy makes more good than bad.
Don't you think the damage the pre-ringing does to kick drum and bass instrument transients outweighs the benefits of getting rid of the subsonic energy, especially considering that there usually isn't much down that far anyway?
πŸ“ Reply

Similar Threads

Thread / Thread Starter Replies / Views Last Post
replies: 237 views: 81945
Avatar for fantomen
fantomen 2nd April 2010
replies: 3792 views: 816527
Avatar for andy-cytomic
andy-cytomic 7th April 2022
replies: 3 views: 3307
Avatar for calvinball
calvinball 18th October 2011
replies: 74 views: 19839
Avatar for JP__
JP__ 12th February 2019
Topic:
Post Reply

Welcome to the Gearspace 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…

Forum Jump
Forum Jump