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
Peak Headroom Dynamics Plugins
Old 2 weeks ago
  #31
Lives for gear
 
chrischoir's Avatar
 

Verified Member
Quote:
Originally Posted by Trakworx View Post
transparent limiter[/I]?

I have several limiters but I have no idea which ones would fit that definition, or maybe they all do...?
For analog, Crane Song has some transparent stuff.
for digital any plugin will be transparent if set-up properly
Old 2 weeks ago
  #32
Lives for gear
 
bcgood's Avatar
 

Verified Member
The title of this thread, "Peak Headroom" reminded me of this guy..


Old 2 weeks ago
  #33
Lives for gear
 
Apostolos Siopis's Avatar
 

Verified Member
Quote:
Originally Posted by Trakworx View Post
Can you elaborate on bit perfect/transparent limiter?

I have several limiters but I have no idea which ones would fit that definition, or maybe they all do...?

sure

what saxnscratch said:
Quote:
A bit perfect/transparent limiter is a limiter that doesn't change anything to the PCM stream as long as the signal doesn't reach the threshold.
If there is oversampling for example, it's not bit perfect anymore.
I would also add to this, the limiters that have saturation options and possibly multiband limiters.
I havent tested with any of the above though.

Voxengo's limiter is not affecting anything below the threshold, at least at 2 modes that I have tested .
I suppose, more or less, most brickwall limiters are the same.

I confirm transparency below threshold, with a simple null test.

Having said all that, lately I have been trying new ways of limiting and I end up with only very few overs, that need to be dealt with after SRC, so I usually try to find some time to automate a volume fader.
Old 2 weeks ago
  #34
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by bcgood View Post
Quality over loudness is what I say
Absolutely!

I'll give you my entire contact list if you can convince all my clients of that.
Old 2 weeks ago
  #35
Lives for gear
 
Trakworx's Avatar
I'm not really seeing why a limiter has to be bit perfect to be used at that final post-SRC stage vs. any other stage.

I generally just go with whatever limiter sounds best for the music at hand. That approach should still apply, I think.
Old 2 weeks ago
  #36
Lives for gear
 
Apostolos Siopis's Avatar
 

Verified Member
Quote:
Originally Posted by Trakworx View Post
I'm not really seeing why a limiter has to be bit perfect to be used at that final post-SRC stage vs. any other stage.

I generally just go with whatever limiter sounds best for the music at hand. That approach should still apply, I think.
I guess it doesnt have to, but I prefer it to be so.

I cant listen to this limiter before SRC (there are no overs) so I dont want it to change what I already did while at 96KHz.
I do ALL the processing at 96K, heavy limiting included. I then SRC and there are some peaks caused by the filter. These peaks are usually 3-4 per minute and they live at the loudest parts of the track.
When I produce the 44.1kHz file I could just normalise it to -0.1dBFS or gain maybe half a dB extra if I find these peaks and tack them in.
So I need a tool that will only work for a brief moment and only touch the peaks that were not there at 96Khz...best tool is gain automation and second best is a limiter .
The null test between the SRCed file against the SRCed file with the stray peaks tacked in, is basically -inf, except a few times that the limiter kicks in for a couple of msec and the residue is at -40 or lower. Moreover, this happens at the loud parts of the track so it is literally inaudible.


had to rewrite as I may not be making sense...feel free to ask for moredetails if I still dont make any sense

Last edited by Apostolos Siopis; 2 weeks ago at 05:33 PM..
Old 2 weeks ago
  #37
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by Apostolos Siopis View Post
had to rewrite as I may not be making sense...feel free to ask for moredetails if I still dont make any sense
It made sense to me both times you wrote it!

Thanks for the insight. I agree that any limiter with saturation or multiband would probably be too unpredictable for that application. Oversampling, however, may be alright with me if it helps the GR to be more transparent without changing the sound below the threshold. I can check on that easily enough with a null test.
Old 2 weeks ago
  #38
Lives for gear
 
Apostolos Siopis's Avatar
 

Verified Member
Quote:
Originally Posted by Trakworx View Post
It made sense to me both times you wrote it!

Thanks for the insight. I agree that any limiter with saturation or multiband would probably be too unpredictable for that application. Oversampling, however, may be alright with me if it helps the GR to be more transparent without changing the sound below the threshold. I can check on that easily enough with a null test.
i cant see the point of oversampling when gain reduction is so low and for such a brief moment...I think elefant is oversampling the whole signal so I use it at 1x.

A limiter which oversamples only the sidechain may be a little bit more appropriate.
It is very rare that I can really hear this instance of "limiting" though.

Last edited by Apostolos Siopis; 1 week ago at 08:54 AM..
Old 1 week ago
  #39
Lives for gear
 
Trakworx's Avatar
I just did a null test with Pro-L 2 and it appears to be bit perfect below the threshold with or without oversampling turned on.

I used a mastered limited track that had been SRCd from 96 to 44.1 so there were ISPs to be tamed.

If I set it to do just 0.2dB of GR then I hear a tiny difference in the null. Just a little crackling where it's skimming the ISPs.

Interestingly, that sound changes with different oversampling settings. The more oversampling the more little peaks it seems to be catching.
Old 1 week ago
  #40
Quote:
Originally Posted by Trakworx View Post
I just did a null test with Pro-L 2 and it appears to be bit perfect below the threshold with or without oversampling turned on.

I used a mastered limited track that had been SRCd from 96 to 44.1 so there were ISPs to be tamed.

If I set it to do just 0.2dB of GR then I hear a tiny difference in the null. Just a little crackling where it's skimming the ISPs.

Interestingly, that sound changes with different oversampling settings. The more oversampling the more little peaks it seems to be catching.
I'm not sure exactly but FF ProL2 may just oversample the key of the limiter not the signal itself (I think elephant does the same), I'll test that tomorrow.
Old 1 week ago
  #41
Lives for gear
 
Trakworx's Avatar
That would explain my results!
Old 1 week ago
  #42
Quote:
Originally Posted by Trakworx View Post
That would explain my results!
But instead of a null test, you can feed the limiter with a 16bits signal and check the output with a bit counter like Bitter << Stillwell Audio
And if it stay 16bits, it's bit perfect !! Otherwise it would go 32f
Old 3 days ago
  #43
Lives for gear
 
SmoothTone's Avatar
 

I've done some detailed listening tests here this week and my results - although far from conclusive - have been interesting.

I prepared 3 files: jazz, indie rock and EDM, each at -1, -0.3 and 0dBfs in both 320 and 128kbps mp3 for 'real world' consumer formats. I listened to them on my mastering mains, studio headphones, car stereo, bluetooth speaker, 10 year old PC motherboard DAC and cell phone with earbuds. All files were quite loud, ranging from -8.4LUFS (short-term) for the jazz (I know!) to -4.9LUFS (s-t) for the EDM. The jazz track had a loud brass section, the indie rock had big drum transients that had to be tightly controlled and the EDM was fairly laid back but had a thick, subby low end and low mid resonances. All files had originally been mastered to -1dBfs using true peak limiting, so they were optimized for that setup and I simply backed-off the limiting to reach the alternate ceilings of -0.3 and 0dBfs. The EDM and indie tracks had used a combination of clipping and limiting, in the latter I had leaned quite heavily on the clipping. I measured the true peak level and number of 'possibly clipped samples' of each of the 0dBfs 128kbps files in Izotope RX and the results were +0.8/379, +0.12/2 and +2.17/1850 for the jazz, indie and EDM files respectively.

I listened blind where possible (not possible for all scenarios). To my surprise, I found the results to be consistent across all playback systems and between 320 and 128kbps. There were a couple of slight variations between the different genres - largely, I think, due to how the masters were processed - and my results were generally as follows.

Blinded, I could consistently hear the greater amount of limiting on the -1 files and this was the most immediately noticeable difference. The -1 files generally sounded more dense and squashed while the -0.3 and 0 files had more movement and depth. (The EDM file was an exception).

I consistently picked the 0dBfs files out as having a mild veneer of crunch that was not unmusical. That is, it sounded part of the music (not separate) and I could imagine that some listeners might actually find this more engaging. I consistently found these files more stressed/fatiguing in sound. (Interestingly, this was most pronounced on the file with the least TP overs but which had been clipped more heavily). However, these differences were much smaller than I expected.

The files sounded progressively cleaner at -0.3 and -1dBfs (less 'crust'). In a couple of instances, I could hear slightly increased distortion from the additional limiting at -1dBfs.

These results were much less pronounced than I expected. The differences were very small and I can't be sure that I would notice them if I wasn't listening for them in a comparison. My results seem to suggest that the difference between 0 and -0.3dBfs makes a small but significant difference in terms of lower distortion but this improvement doesn't increase relative to the reduced ceiling at -1 and is offset by the the impact on musicality by increased compression at -1dBfs for masters with higher loudness. In short, in most listening scenarios -0.3 was a welcome compromise between the more compressed sound at -1 and the more fatiguing sound at 0dBfs.

In hindsight, I should have also done some files at -0.1 and I'm curious whether I could hear a difference between -0.1 and 0dBfs.

So, while I haven't made my mind up yet and my practice will still be to send low to moderately loud files out at -1dBTP, I guess I'm kinda concluding that when things have to be pushed for maximum loudness, -0.3 might serve the music better in the grand scheme of things.

How are things going on your end Justin?
Old 3 days ago
  #44
Quote:
Originally Posted by SmoothTone View Post
I've done some detailed listening tests here this week and my results - although far from conclusive - have been interesting.

I prepared 3 files: jazz, indie rock and EDM, each at -1, -0.3 and 0dBfs in both 320 and 128kbps mp3 for 'real world' consumer formats. I listened to them on my mastering mains, studio headphones, car stereo, bluetooth speaker, 10 year old PC motherboard DAC and cell phone with earbuds. All files were quite loud, ranging from -8.4LUFS (short-term) for the jazz (I know!) to -4.9LUFS (s-t) for the EDM. The jazz track had a loud brass section, the indie rock had big drum transients that had to be tightly controlled and the EDM was fairly laid back but had a thick, subby low end and low mid resonances. All files had originally been mastered to -1dBfs using true peak limiting, so they were optimized for that setup and I simply backed-off the limiting to reach the alternate ceilings of -0.3 and 0dBfs. The EDM and indie tracks had used a combination of clipping and limiting, in the latter I had leaned quite heavily on the clipping. I measured the true peak level and number of 'possibly clipped samples' of each of the 0dBfs 128kbps files in Izotope RX and the results were +0.8/379, +0.12/2 and +2.17/1850 for the jazz, indie and EDM files respectively.

I listened blind where possible (not possible for all scenarios). To my surprise, I found the results to be consistent across all playback systems and between 320 and 128kbps. There were a couple of slight variations between the different genres - largely, I think, due to how the masters were processed - and my results were generally as follows.

Blinded, I could consistently hear the greater amount of limiting on the -1 files and this was the most immediately noticeable difference. The -1 files generally sounded more dense and squashed while the -0.3 and 0 files had more movement and depth. (The EDM file was an exception).

I consistently picked the 0dBfs files out as having a mild veneer of crunch that was not unmusical. That is, it sounded part of the music (not separate) and I could imagine that some listeners might actually find this more engaging. I consistently found these files more stressed/fatiguing in sound. (Interestingly, this was most pronounced on the file with the least TP overs but which had been clipped more heavily). However, these differences were much smaller than I expected.

The files sounded progressively cleaner at -0.3 and -1dBfs (less 'crust'). In a couple of instances, I could hear slightly increased distortion from the additional limiting at -1dBfs.

These results were much less pronounced than I expected. The differences were very small and I can't be sure that I would notice them if I wasn't listening for them in a comparison. My results seem to suggest that the difference between 0 and -0.3dBfs makes a small but significant difference in terms of lower distortion but this improvement doesn't increase relative to the reduced ceiling at -1 and is offset by the the impact on musicality by increased compression at -1dBfs for masters with higher loudness. In short, in most listening scenarios -0.3 was a welcome compromise between the more compressed sound at -1 and the more fatiguing sound at 0dBfs.

In hindsight, I should have also done some files at -0.1 and I'm curious whether I could hear a difference between -0.1 and 0dBfs.

So, while I haven't made my mind up yet and my practice will still be to send low to moderately loud files out at -1dBTP, I guess I'm kinda concluding that when things have to be pushed for maximum loudness, -0.3 might serve the music better in the grand scheme of things.

How are things going on your end Justin?
Really interesting test !!

When aiming to this kind of loudness, 0.7dB of GR on the limiter is a quite significant amount.
Did you tried to get the LUF inegrated value of the files ?
Because the 0dBfs and -0.3dbfs files might be slightly louder (talking about 0.1dB here)

I found that when backing up a limiter threshold by 1dB, I don't loose 1dB but somewhat 0.9dB (depending on the material and in the situation when the limiter is really working) which make sense since the GR will more often going back to 0 than when pushed louder.
So with -0.3dbfs you get a little bit more loudness while being slightly more alive, which is kinda cool !

And you know that I'm a -0.3dBfs guy
Old 3 days ago
  #45
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by SmoothTone View Post
How are things going on your end Justin?
Nicely done. Thanks for sharing your results!

On my end - I've settled on -.3 for now, but for a different reason than you. Workflow. I wanted to go with -.1 but that requires another stage of limiting after SRC in a second session at 44.1, otherwise the SRC causes overs. -.3 allows room for the ISPs from SRCing after the limiter as is my practice. I'm undecided as to whether that extra .2dB of dynamics is worth the extra work. It might turn out to be worth it but I haven't had the spare time to do the listening tests yet (There's construction next door so my studio hours are restricted lately). Your results point to -.3 being a sweet spot, which is appealing because it would mean less work for me!

But all this thinking about it has lead to some other discoveries made during sessions. I'm getting better perceived punch/attack by tweaking my limiter settings - turning off oversampling and using long attack times/short release times. It's helping about as much as tweaking the peak headroom does. I know there's a greater risk of distortion with these settings but so far so good. This is with Pro-L 2. What limiter did you use for your test?
Old 3 days ago
  #46
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by Saxnscratch View Post
Did you tried to get the LUF inegrated value of the files ?
Because the 0dBfs and -0.3dbfs files might be slightly louder (talking about 0.1dB here)

I found that when backing up a limiter threshold by 1dB, I don't loose 1dB but somewhat 0.9dB (depending on the material and in the situation when the limiter is really working) which make sense since the GR will more often going back to 0 than when pushed louder.
So with -0.3dbfs you get a little bit more loudness while being slightly more alive, which is kinda cool !
That's good information - thanks!

I'll have to check for that loudness difference...
Old 3 days ago
  #47
Lives for gear
 
SmoothTone's Avatar
 

Quote:
Originally Posted by Saxnscratch View Post
Did you tried to get the LUF inegrated value of the files ? Because the 0dBfs and -0.3dbfs files might be slightly louder (talking about 0.1dB here)
The integrated LUFS difference between the -1, -0.3 and 0dBfs files was 0.2dB max. But I only used 45sec of each song.

Quote:
And you know that I'm a -0.3dBfs guy
Old 2 days ago
  #48
Lives for gear
 
SmoothTone's Avatar
 

Quote:
Originally Posted by Trakworx View Post
Your results point to -.3 being a sweet spot, which is appealing because it would mean less work for me!
lol

Quote:
What limiter did you use for your test?
Limitless for the jazz and EDM tracks. For the indie rock track I had clipped the ADC and saturated a tube stage just before the ADC because it suited the vibe. This was followed by the safety limiter in Limiter6GE.

I agree that at these levels, the way we set up our loudenators has a massive impact on the sound. I use Limitless almost exclusively now and always spend a couple of minutes tweaking it for each mix.
Old 2 days ago
  #49
Lives for gear
 
JP__'s Avatar
 

Verified Member
Limitless here to in 98,56% too, still. I own/tried a lot of other Li iters too, but they are mostly good for tweaking Limitless to perfection over the years.
Im a 0.3-1dB guy, depends on oversampling too.
Old 2 days ago
  #50
Lives for gear
 
Trakworx's Avatar
I really like the tone and stereo imaging I get with limitless but Pro-L 2 preserves more punch and attack, so I keep going back to it. But then I miss the sound of Limitless. It's been a constant source of inner conflict for me ever since Pro-L 2 came out LOL. Maybe I need to spend more time tweaking Limitless to see if I can get it to give me the best of both worlds... Any tips?
Old 2 days ago
  #51
Lives for gear
 
Kimotei's Avatar
Quote:
Originally Posted by Trakworx View Post
I'm getting better perceived punch/attack by .. turning off oversampling
Nice.

SmoothTone: Did you find out if src only altered the sound at the isp’s, or if it was constant with Pro L?

Youtuber ME Streaky (previously working from Metropolis in London) said he hears a loss of hf using src, and recomends not using it with Pro L. However so far I cant hear much difference. Or im not sure, as I havent done a proper test like you. He also use gr up to -4db or so on edm which on the other hand can sound horrible to me on Pro L, but can be ok depending on the mix I guess..

Edit: ah, I see you use Limitless..
Old 2 days ago
  #52
Lives for gear
 
SmoothTone's Avatar
 

Quote:
Originally Posted by Trakworx View Post
I really like the tone and stereo imaging I get with limitless but Pro-L 2 preserves more punch and attack, so I keep going back to it. But then I miss the sound of Limitless. It's been a constant source of inner conflict for me ever since Pro-L 2 came out LOL. Maybe I need to spend more time tweaking Limitless to see if I can get it to give me the best of both worlds... Any tips?
These are essentially the same differences I hear between ProL2 and Limitless. It's funny how we're sensitive to different things. I like how solid ProL2 sounds but I can't tolerate the hit to depth and image. I feel like I can set up Limitless to retain punch. It really just takes a bit of time to learn what the different parameters do to the sound and how they interact. It's worth it!

I recently did another round of trying my other limiters and it just confirmed that Limitless is most transparent for my purposes.

I wrote some detailed stuff about it in a couple of PMs a while back so I'll try to dig it out and forward it on. In the meantime, Macc wrote some detailed posts about how he sets up Limitless in the thread about it here which are well worth a look.
Old 2 days ago
  #53
Lives for gear
 
bcgood's Avatar
 

Verified Member
Have you guy experimented with the different modes in Fab Filter? I liked it a lot better in transparent mode as I was using it last night. I hadn't messed with that drop down setting before. I liked it better than the modern setting.
Old 2 days ago
  #54
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by SmoothTone View Post
I like how solid ProL2 sounds but I can't tolerate the hit to depth and image. I feel like I can set up Limitless to retain punch.
Yes, I prefer the imaging of Limitless, but "Modern" mode in Pro-L 2 is just so damned punchy! I've been doing various things to compensate for the imaging problem. But I'd love to figure out how to get that punch with limitless! I now have a new mission...

Last edited by Trakworx; 2 days ago at 12:25 AM..
Old 2 days ago
  #55
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by bcgood View Post
Have you guy experimented with the different modes in Fab Filter? I liked it a lot better in transparent mode as I was using it last night. I hadn't messed with that drop down setting before. I liked it better than the modern setting.
I like "Transparent" on material that doesn't have loud drums, but in cases like that I'll just use Limitless.
Old 2 days ago
  #56
And again, no one is talking about the look-ahead
Old 2 days ago
  #57
Lives for gear
 
Trakworx's Avatar
Bringing the topic back around to peak headroom - has anyone noticed that Limitless' true peak setting seems to let more ISPs through compared to Pro-L 2?
Old 2 days ago
  #58
Quote:
Originally Posted by Trakworx View Post
Bringing the topic back around to peak headroom - has anyone noticed that Limitless' true peak setting seems to let more ISPs through compared to Pro-L 2?
yes it does and it might be the reason why it sounds better.
True peak limiting option on the fabfilter is awfull but the oversampling can sound great while avoidibg ISP.
Old 2 days ago
  #59
Lives for gear
 
SmoothTone's Avatar
 

The safety limiter on the TDR is a great transparent failsafe to stick at the end of the chain just to catch TP overs. You might want to experiment with using it after SRC Justin.
Old 2 days ago
  #60
Lives for gear
 
JP__'s Avatar
 

Verified Member
Quote:
Originally Posted by Trakworx View Post
Bringing the topic back around to peak headroom - has anyone noticed that Limitless' true peak setting seems to let more ISPs through compared to Pro-L 2?
I just used Limitless OS feature on a master yesterday and it didnt let any appreciable peaks come through. I used Digicheck and RX own analyses.
The big prob with all this "True Peak" stuff is, it isnt really normed nor do all meters show the same. Its not a real value but more a raw approximation. I would not waste too much time looking at those small numbers, really.
Personally I choose OS/True Peak features by sound, not because of their ability to reduce those theoretical peaks.

Last edited by JP__; 2 days ago at 11:06 AM..
Topic:
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
Forum Jump
Forum Jump