Pro-L 2 by FabFilter

[do_androids_dream]

"We" being myself and other audio people who have deeper insight into this aspect of audio. Specifically, Bob Katz, Ian Shepard, et al. But if you don't care and don't feel it's important, good on 'ya.

Do you have extra big doorways in your house?

Hey well, I learned a lot from them. So what they know and shared I now know. I don't know everything they know and who knows I might know some stuff they don't know. But at least we all share information and experiences and come together towards making good sounding music.

Can't say the same for most of you wankers.

Get back to your distorted compressed -7 LUFS 4 PLR EDM.

Hmm.. so not worrying about ISP's (which no ME did up until very recently and only when plugins fully took over limiting duties) = 'distorted compressed -7 LUFS 4 PLR EDM. ' ... and a lovely insult to boot lol.

Okay..

[legendCNCD]

oh my

[skyshaver]

I have no horse in the ISP race but wanted to add I found reducing the output on the first iteration of the limiter by .1 db guaranteed the limiter would not go over O db. Also, I've found the metering in version 1 super useful for film/tv work.

[mutantdog]

Every time a new limiter comes out, its thread gets derailed into a debate about ISPs!

[samsam]

^ I blame it on net neutrality.

[do_androids_dream]

What did you use to measure the ISPs?

A good question actually. No meter is completely accurate so one can never be totally sure if there are no inter sample peaks going over 0dB. Thankfully nearly all playbacks systems have a good amount of headroom to ensure isp's don't cause problems. There are variances between all the meters I use.

[dreamvoid]

A System 6000 with mastering licenses can stop ISPs completely and I cannot remember other software/metering reporting one with such processed files. The MD 5.1 and MD4 in the System 6000 are still awesome. Most of the times I use software nowadays but some commercial mixes I send thru my System 6000s expecially for getting a totally save master.

[plexuss]

Scroll down about 1/2 to the Proof That it Works section. There is a graph showing the results of an ISP clip at 44.1k - the red trace was produced by an ISP clip using a more conventional DAC and the green showing their (Benchmark) "new" DAC with the headroom required to manage ISP clips. The point here is that unless a DAC is design for more headroom, it will clip on ISPs and that clipping is wide-band, audible and produces a mid-high brightness that masks PLR. It reads"

I'd love to see your research countering this!

"The green trace was produced by a high-headroom Benchmark DAC2. The DAC2 correctly reproduced an 11.025 kHz sine wave having an amplitude of +3.01 dBFS. The red trace shows the output of the Benchmark DAC1 under the same test conditions. Like most products on the market, the older DAC1 uses conventional interpolation and it will overload whenever intersample peaks exceed 0 dBFS. Note that the red plot shows many distortion products that are not harmonically related to the 11.025 kHz tone. These IMD (intermodulation distortion) products are produced by interactions between the test tone, the sampling frequency, and the oversampling frequency. This IMD distortion is not musical and does not occur in analog systems. It produces a sound that is unique to digital systems. Note that most of the distortion occurs between 5 kHz and 22 kHz. This can produce a false brightness or brittleness when overloads occur. This begs the question; how often do intersample overs occur?"


https://benchmarkmedia.com/blogs/applic ... tion+Notes

[do_androids_dream]

Every time a new limiter comes out, its thread gets derailed into a debate about ISPs!

Well our brains don't really work in 'threads'. I don't think many threads get truly 'derailed' - they just spread into wider discussion because that's the natural way we tend to talk.

[Armadillo]

New features in Pro-L2:

NEW - Extensive loudness metering with support for the EBU R128, ITU-R BS.1770-4 and ATSC A/85 standards
NEW - Surround support including Dolby Atmos 7.0.2 and 7.1.2 formats with flexible surround channel linking
NEW - Optional DC offset filter
NEW - External side chain triggering for stem mastering
NEW - Unity Gain option to easily listen to the effect of the current limiting at unity gain level
NEW - Audition Limiting option to listen to the difference between the input and output signal
IMPROVED - Eight different limiting algorithms, all with their own character and purpose
IMPROVED - Unique real-time level display with peak gain reduction labels and new innovative display modes
IMPROVED - Up to 32x linear-phase oversampling
IMPROVED - Highly accurate output and gain reduction metering including true peak metering
NEW - True peak limiting

Feel free to ignore the first 8 features.

[10bd01]

plexuss,

Thank you for spending the time and putting in the effort to test the plugin and post your findings. I referred to this thread over at FF support because I wanted their response. This is what I received:

Hi,

Pro-L 2 is absolutely a true peak limiter!

When comparing true peak level readings from different plug-ins, keep in mind that true peak metering is -not- an exact science; there is not one single method to implement true peak metering.

Therefore, different true peak meters may give slightly different readings. This is not a problem, and it does not mean one meter is wrong while the other is correct.

The EBU organization offers a set of test audio (tech.ebu.ch/docs/tech/tech3341.pdf), with which you can test metering for compliance with the EBU R128 metering standard. The true peak level tests actually have a tolerance of -0.4/+0.2 dBTP! This means that it's possible to have two meters that have a 0.6 difference in true peak reading, while they still both comply to the EBU R128 standard.

So, seeing differences of 0.1 or 0.2 dBTP is not unusual or wrong at all. But of course, if you see very large differences (e.g. > 0.6 dBTP), we'de surely like to know and like to investigate...

Feel free to re-post this on KVR of course.

Was wondering about your response to this. To me this is boiling down to a definition. That being said, I was already intending on going with Nugen for mastering and FF for mixing because I felt early on Nugen was putting more effort into market and standards compliance.

Thanks again.

[plexuss]

plexuss,

Thank you for spending the time and putting in the effort to test the plugin and post your findings. I referred to this thread over at FF support because I wanted their response. This is what I received:

Hi,

Pro-L 2 is absolutely a true peak limiter!

When comparing true peak level readings from different plug-ins, keep in mind that true peak metering is -not- an exact science; there is not one single method to implement true peak metering.

Therefore, different true peak meters may give slightly different readings. This is not a problem, and it does not mean one meter is wrong while the other is correct.

The EBU organization offers a set of test audio (tech.ebu.ch/docs/tech/tech3341.pdf), with which you can test metering for compliance with the EBU R128 metering standard. The true peak level tests actually have a tolerance of -0.4/+0.2 dBTP! This means that it's possible to have two meters that have a 0.6 difference in true peak reading, while they still both comply to the EBU R128 standard.

So, seeing differences of 0.1 or 0.2 dBTP is not unusual or wrong at all. But of course, if you see very large differences (e.g. > 0.6 dBTP), we'de surely like to know and like to investigate...

Feel free to re-post this on KVR of course.

Was wondering about your response to this. To me this is boiling down to a definition. That being said, I was already intending on going with Nugen for mastering and FF for mixing because I felt early on Nugen was putting more effort into market and standards compliance.

Thanks again.

I have a real hard time agreeing that Pro-L2 is a true peak limiter. The reason is because with 0dB ceiling it should not then throw ISP overs. But it does. However giving it a bit of ceiling, somewhere between 0.05dB and 0.1dB prevents all clipping as far as I can tell. So in the strictist sense, its not a true peak limiter. But for all practical purposes it's "good enough".

You mention Nugen. From my testing ISL2 is a true peak limiter. with 0dB ceiling it throws no clips.

Frankly, I don't think its worth pushing the point about Pro-L2 because as I mentioned for all practical purposes it does what it advertises just don't run it without a ceiling. It otherwise has a very versatile sound and good usability. I am experimenting with it as a first stage mastering limiter to add character and do a first pass softening of peaks before going to ISL2 which I hope I can run a little more lean and transparent.

Back to FF's comment about ISP detection not being an absolute - it's true. ISP clipping occurs in the DA process and it's impossible to predict every DA that will be used. Some DA's are designed with more analogue headroom to accomodate ISP's and prevent clipping where-as most DA's do not and hence there is a risk of a clip (and they usually do clip). So it depends on the approach used to detect these. FF may have a more strict detection algo in Pro-L2 which will give a more conservative ISP metering. From what I have read (but not tested) Apple's RoundTripAAC is less sensitive and will throw clip warnings less often. And of course use your years but accept that our audio preception is easily swayed by our psychology - this is why I prefer to use metering to give me a sense of any problems and try and use subject and objective evaluation together.

[DPhil]

I have a real hard time agreeing that Pro-L2 is a true peak limiter. The reason is because with 0dB ceiling it should not then throw ISP overs. But it does.

In metering plugins that have +/-0.2dB tolerance? And measured with algorithms which aren't strictly defined anywhere? I just don't get why you think your tools are the "right" ones. This is no front or so. It's good we talking about things like that

edit: great update btw! Thank you FabFilter for putting so much effort in existing plugins for your customers.

[plexuss]

I have a real hard time agreeing that Pro-L2 is a true peak limiter. The reason is because with 0dB ceiling it should not then throw ISP overs. But it does.

In metering plugins that have +/-0.2dB tolerance? And measured with algorithms which aren't strictly defined anywhere? I just don't get why you think your tools are the "right" ones.

edit: great update btw! Thank you FabFilter for putting so much effort in existing plugins for your customers.

There are no right tools when it comes to ISP clip detection. I use Apple RoundTripAAC because it's the only ISP clip meter that measure ISP clip "velocity" (# of clips over a period of time). A meter that tells me there was "a clip" is not useful - one predicted clip here or there is not going to break the bank. But a stream of clips is a sign of a problem. RoundTrip is the only meter that I know of that shows this "velocity". Also, RoundTrip apparently uses a more strict detection algo so it will trow a clip error less often than other meters, if this is correct. Therefore, if RoundTrip shows clipping there is more likely going to be clipping during DAC. This is why I prefer RoundTripAAC over other meters. But I use a number of meters and my ears together to make decisions about what is a problem and what is not.

[el-bo (formerly ebow)]

Every time a new limiter comes out, its thread gets derailed into a debate about ISPs!

Well our brains don't really work in 'threads'. I don't think many threads get truly 'derailed' - they just spread into wider discussion because that's the natural way we tend to talk.

This comment needs it's own 'sticky'