TDR Limiter 6 GE - Released!

VST, AU, AAX, CLAP, etc. Plugin Virtual Effects Discussion
Post Reply New Topic
RELATED
PRODUCTS
TDR Limiter 6 GE

Post

Fabien, again a great plugin! I'm a long time SlickEQ-GE user and do a lot of broadcast mixing lately with nova.
One question about TDR Limiter 6: the EBU meter is a great feature. Something even better would be the ability to integrate long term average EBU measurements, e.g. for a 20 minute audio mix. A typical specification for a tv show here is an overall ebu r128 value of -23db lufs with a maximum tp at -1dbfs.
The greatest thing would be to have Limiter 6 as the last plugin in the mix and once you bounced everything offline the EBU meter shows the overall value and the TP meter shows the highest over all peak in dB.
Such a feature would be extremely helpful and put Limiter 6 in competition with way more expensive plugins such as grimm audio and nugen...
this one goes to 11!

Post

Mackezwo, I suggest you have a look at Toneboosters' EBU Loudness as it does exactly what you're requesting.
Seasoned IT vet, Mac user, and lover of music. Always learning.

Post

wesleyt, thanks for the hint!
Toneboosters website gets me a 404 when I follow the googled links. But from what I see the Toneboosters plugin is an actual metering plugin. Which is great, but it would take another gain- and peaklimiter plugin before to achieve what TDR Limiter can almost do in one plugin. I have no clou about coding, but I could imagine it's not a big effort to add a longterm ebu measurement.
But anyway, TDR Limiter is of course great just the way it is :clap:
Last edited by mackezwo on Wed Aug 30, 2017 6:19 pm, edited 1 time in total.
this one goes to 11!

Post

They recently got a new website.

Post

mackezwo wrote:The greatest thing would be to have Limiter 6 as the last plugin in the mix and once you bounced everything offline the EBU meter shows the overall value and the TP meter shows the highest over all peak in dB.
Sounds like a good idea. We're probably resetting the meter too quickly. With EBU Integrated mode and max peak mode, you essentially get what you want, but it's probably reset after rendering. We'll try to fix it!
Fabien from Tokyo Dawn Records

Check out my audio processors over at the Tokyo Dawn Labs!

Post

Is there any reason why the True Peak Limiter is disabled in low latency mode?
I couldn't find any information in the manual on this.

Also +1 for mackezwo's suggestions :tu:

Post

FabienTDR wrote:Sounds like a good idea. We're probably resetting the meter too quickly. With EBU Integrated mode and max peak mode, you essentially get what you want, but it's probably reset after rendering. We'll try to fix it!
How about resetting only when playback (or recording) starts?
DarkStar, ... Interesting, if true
Inspired by ...

Post

krabbencutter wrote:Is there any reason why the True Peak Limiter is disabled in low latency mode?
I couldn't find any information in the manual on this.

Also +1 for mackezwo's suggestions :tu:
Probably because it requires looking into the future. Since that is still a challenge for our technological capacity, plugin developers opt to steal the necessary milliseconds from your life instead.

Post

FabienTDR wrote:We'll try to fix it!
Woa!
Just bought Limiter 6 GE :-)
Looking forward for the fix- but will use it anyway. Thanks for this plugin again, Fabien!

DarkStar: Good idea, I think. Some other plugins (like Toneboosters) use a start/stop button or even a sync start button. Your idea would work without any extra button- pretty smart.
When there's need for more ideas: such a plug may also include a threshold preference for these measurements for pauses in the material. Like in dialogue tracks or when measuring vinyl masters (without the gaps between the songs etc). Or an automated level aligning for a preset value after bouncing, or... nah, I'll stop it here ;-)
this one goes to 11!

Post

^^^
In my understanding, EBU-128 Integrated (and Range) measurements take quiet passages into account (by gating them out!, hence ignoring them). But I could be wrong.
Last edited by DarkStar on Wed Aug 30, 2017 5:18 pm, edited 2 times in total.
DarkStar, ... Interesting, if true
Inspired by ...

Post

DarkStar wrote:How about resetting only when playback (or recording) starts?
Yes, something like this. We simply oversaw the case.
Fabien from Tokyo Dawn Records

Check out my audio processors over at the Tokyo Dawn Labs!

Post

@mackezwo we could still open the gating thresholds to the user. Right now, thresholds follow the main EBU R128 recommendation, but certain branches use slightly different gate thresholds. We still have plenty UI space in the UI config view, so it would be realistic.

@DarkStar: When building time machines, there's no such thing as dead-lines :)
Fabien from Tokyo Dawn Records

Check out my audio processors over at the Tokyo Dawn Labs!

Post

.jon wrote:
krabbencutter wrote:Is there any reason why the True Peak Limiter is disabled in low latency mode?
I couldn't find any information in the manual on this.

Also +1 for mackezwo's suggestions :tu:
Probably because it requires looking into the future. Since that is still a challenge for our technological capacity, plugin developers opt to steal the necessary milliseconds from your life instead.
I'm pretty sure it's a niche usecase, but I was thinking about using Limiter 6 in live media production as a final true peak limiter. I know limiters need at least some amount of latency and Limiter 6 can do a helluva lot more, than just true peak limiting. But In my case 10ms latency would simply be too much. But as I said, I'm aware that's probably a niche :wink:

Post

Not specific to Limiter No6 but with lookahead in general, I wonder why a limiter would have any overshoots above the ceiling at all - isn't that the whole point of lookahead?

Post

MogwaiBoy wrote:Not specific to Limiter No6 but with lookahead in general, I wonder why a limiter would have any overshoots above the ceiling at all - isn't that the whole point of lookahead?
Well, it depends how you define overshoot!

If you're talking about overshoots of the PCM data (probes of the original waveform taken at the sample rate), sure, you can prevent any overshoot. You don't even need lookahead to do so.

The problem is that the PCM data is not the "music". PCM is the result of the super clever compression scheme called the "sampling theorem". In order to reconstruct the full original wave, one needs an analogue filter (a filter running at an infinite samplerate. Exactly what DA's do).

Now two issues appear:

A. The max values of the PCM data only weakly correlate with the max values of the underlying signal.

This means that in order to get an idea of the envelope and maxima of the underlying signal, have to use an analogue filter. We don't have this around, so developers approximate the true signal by resampling the audio to higher rates. With samplerates around 8x the audible bandwidth, we can keep the error below 1dB over the full audible range. So far, so good.

B. When we manipulate this PCM data, we are bound to the restrictions set by the sampling theorem. In particular, one cannot represent content having a frequency beyond half samplerate. If we do, this content will be mirrored at Nyquist, reappearing in form of an irreversible distortion called "aliasing".

More practically, you can of course manipulate the PCM values to taste, clip them or limit them to whatever values. But as soon you break any rule of the theorem, the true waveform will do something rather unexpected. While the PCM points will still represent totally correct points on the signal, everything in between these probes will be bent and squeezed around like crazy, in an effort to fit the probes.

Fast clipping, limiting and compression can easily multiply the bandwidth 10 times or more! Start with a 20kHz signal, clip it and you quickly end up with a 200+ kHz signal. Limiting the peak level of a signal generally extends the bandwidth.

If the samplerate can't handle the new bandwidth, aliasing will appear in the form of harmonically unrelated partials and missed overshoots. The trick is -again- resampling. We upsample, process the "pseudo analogue" data. All while making sure that the bandwidth is sufficiently wide.

Great, we now have limited the underlying waveform within close limits, and kept aliasing within control. It starts to sound good! :)

But we aren't done yet! The last step for us downsampling the data back to the original samplerate. This involves filtering away any context above the target Nyquist frequency, and thrashing most samples.

And here's the point. Limiting the bandwidth of a signal extends its peak level. These are the remaining overshoots you are wondering about!



tl;dr To summarize:
Limiting the peak level of a signal extends the bandwidth
and
Limiting the bandwidth of a signal extends its peak level

You can't have both at once! (DSP can be bitchy at times ;) )
Fabien from Tokyo Dawn Records

Check out my audio processors over at the Tokyo Dawn Labs!

Post Reply

Return to “Effects”