Pro-L 2 by FabFilter

[FabienTDR]

Thank you for chiming in. Would you be willing to talk a little about true peak detection from your own experience?

Sure. Any specific question I can help with?

[dangayle]

Thank you for chiming in. Would you be willing to talk a little about true peak detection from your own experience?

Sure. Any specific question I can help with?

Uh... all of it? Lol. I guess there’s the spec and then there are other, perhaps more optimal solutions. Where do you as a developer decide to go beyond the spec and what criteria do you look at?

[MogwaiBoy]

I'm starting to wonder if the whole 'True Peak' thing is a little overblown. What happened back in the late 90s when some mastering guy was slamming a Waves L2 to -0.3db and calling it done. Surely they had "ISPs" above 0db which were undetectable at the time or little was known or considered about them - are those recordings now defective? Are there any documented cases of ISPs damaging speakers or equipment?

(note: I've been an ISP obsessed snob for awhile)

[xx JPRacer xx]

I'm starting to wonder if the whole 'True Peak' thing is a little overblown...

You can stop wondering... It is overblown!

[plexuss]

Are there any documented cases of ISPs damaging speakers or equipment?
(note: I've been an ISP obsessed snob for awhile)

The facts are that ISPs can cause clipping in the analogue domain and that clipping causes wide-band intermodulation distortion. These facts have been proven and measured. However, whether the distortion is audible is not is an opinion: some will say they hear it and some will say they do not. To concern yourself with ISPs and all this comes down to a personal decision based on a subjective opinion.

In terms of speaker damage, I am not 100% sure about this, but I think these clips occur so quickly that a physical speaker can't react to the signal fast enough and avoids damage. It would be interesting to get more info about the nature of clips caused by ISPs in terms of real world impact outside of the work Benchmark did around the characteristics of the resulting distoriton.

[GrabtharsHammer]

Bought it today and I have to say that LUFS metering and most importantly the new audition mode are really super useful.

[dangayle]

Bought it today and I have to say that LUFS metering and most importantly the new audition mode are really super useful.

Yeah, I'm really digging the new loudness metering. It would be cool if there were a way to go negative on the gain slider though. If you're sitting at a few LUFS above your target and your dynamic range is acceptable, it would be nice to be able to just pull it down a little from within the plugin itself.

[dangayle]

I made a quick video running through the various limiter algorithms.



Is it just me, or is the new safe mode incredible? Can someone explain to me what is happening? How come I can push it 30 freaking dB and it doesn't bottom out?

[plexuss]

Is it just me, or is the new safe mode incredible? Can someone explain to me what is happening? How come I can push it 30 freaking dB and it doesn't bottom out?

In my video I show how you to use the 1:1 and inverse audition to hear more clearly what L2 is doing. You will be able to hear what is "limited out" and better understand all the modes including safe.

[FabienTDR]

Sure. Any specific question I can help with?

Uh... all of it? Lol. I guess there’s the spec and then there are other, perhaps more optimal solutions. Where do you as a developer decide to go beyond the spec and what criteria do you look at?

For metering, I think that 4x resampling is sufficient to cover most worst case situations.

But, I personally find it most problematic when processors assume that PCM values directly represent the signal. More clearly, I wonder why not all processors handle "ISP" (the whole signal) by default.

All of this independently of the product discussed in this thread (Fabfilter know well what they are doing)

[DrReacto]

Thanks for the video dangayle, that was very helpful. I'm glad you really slammed the levels.

[bmanic]

Is it just me, or is the new safe mode incredible? Can someone explain to me what is happening? How come I can push it 30 freaking dB and it doesn't bottom out?

Because it is very slow in it's release and has built in program dependency that doesn't allow it to distort.

You are not actually getting 30dB of gain/loudness though. The real amount of loudness you are gaining can be measured between the highest limiting peak and where the release stage "bottoms out" without ever returning to zero.

Other than that, yes I think the Safe algorithm is pretty damn impressive considering how pleasantly it does the limiting and incredibly useful in a lot of various situations like single mixing channels and various mixbusses or even FX returns.

[billcarroll]

Is it just me, or is the new safe mode incredible? Can someone explain to me what is happening? How come I can push it 30 freaking dB and it doesn't bottom out?

Because it is very slow in it's release and has built in program dependency that doesn't allow it to distort.

You are not actually getting 30dB of gain/loudness though. The real amount of loudness you are gaining can be measured between the highest limiting peak and where the release stage "bottoms out" without ever returning to zero.

Other than that, yes I think the Safe algorithm is pretty damn impressive considering how pleasantly it does the limiting and incredibly useful in a lot of various situations like single mixing channels and various mixbusses or even FX returns.

It is impressive. I really have to hand it FabFilter.

ps ... I think you win the signature quote competition that doesn't exist. I love the Elon quote.

[camsr]

Did anyone figure out how to use the sidechain?

[pekbro]

Did anyone figure out how to use the sidechain?

You enable sc input with a button on the output level knob. Says its "useful for stem mastering,
where you need to deliver stems with the same limiting as the full master".