New TDR Special Filters Bundle

[El°HYM]

I did not say those are linear phase.

I didnt ask for your Opinion; nor did me suggest those Airwindows Filters instead of the new TDR ones.

Maybe try this: info@airwindows.com

[Dr.Gunjah]

I was about to look for my pitchfork when I saw there are actually demo versions this time (only limitation it has no recall). There's just no link on the bundle site, you have to check the individual plugins. Downloading now

[martinjuenke]

I'm still not convinced that I need this bundle...
Ideas?

[Digivolt]

In case you are wondering why Fabien and TDR havent posted in here for Decades now. There is a, lets say, more useful thread over at GS.

https://gearspace.com/board/new-product ... lyzer.html

The first page, what does the guy mean when he talks about 48k being an issue and Chris telling him to use 96k for the Airwindows plugs, I work in 48k so I'm curious now what this talk is about and how it applies to these new plugs, is it related to oversampling ?

[imrae]

Airwindows plugins are not oversampled, Chris generally uses and recommends working at 96k which gives some ultrasonic headroom and reduces the aliasing concerns. However, while running one nonlinear effect at 96k is roughly equivalent to 2x oversampling at 48k, there is a difference if running multiple plugins in series. The first plugin might create some HF content and aliasing at ultrasonic frequences, which is ok, but then second plugin might add further HF and more aliasing which could make it all the way back down to audible frequencies. Adding ultrasonic filters between the plugins helps avoid this cumulative buildup effect, and gives something more closely equivalent to a series of oversampled effects.

[Digivolt]

Airwindows plugins are not oversampled, Chris generally uses and recommends working at 96k which gives some ultrasonic headroom and reduces the aliasing concerns. However, while running one nonlinear effect at 96k is roughly equivalent to 2x oversampling at 48k, there is a difference if running multiple plugins in series. The first plugin might create some HF content and aliasing at ultrasonic frequences, which is ok, but then second plugin might add further HF and more aliasing which could make it all the way back down to audible frequencies. Adding ultrasonic filters between the plugins helps avoid this cumulative buildup effect, and gives something more closely equivalent to a series of oversampled effects.

That makes sense and perfectly explains why I might need those filters

[dionenoid]

So Arbiter seems closest to TEOTE, Sooth, Gullfoss, DSEQ etc, wonder how it compares, at €24 in the sale it beats them all on price

Arbiter is nothing like those. It's essentially a single band dynamic eq, but unique because of its relative nature ; it works purely on frequency so isn't depending on volume threshold like with traditional dynamic eq's.

This is very handy for cutting out specific frequencies while not having to care about volume of the signal : think of de-essing but without a threshold, so the esses will be cut out even when the vocals are at low volumes.

So, Arbiter is a very specific tool. And just like the other filters in this pack designed for very specialized tasks. They're not for everyone to be honest.

Ultrasonic is really only useful when you work on higher samplerates ; when you work at 44 or 48 then it does nothing for you.

Infrasonic and Elliptical are comparable to good filters in good clean eq's. Infrasonic is a very precise hi-pass filter, Elliptical cuts out stereo information from the low end. But both have some extra features, like adding harmonics or sub bass to compensate for loss of low-end energy.

But again, these are all very specialized tools for specific jobs. Not for everyone, and not for everything. But sure very handy to have when you need them, and in their functions pretty unique.

[ampetrosillo]

I did not say those are linear phase.

Just to inform other readers . In my opinion those Butterworths are a very bad choice. He even adds them everywhere or distributes the poles before and after non-linear processes. Even when using 44/48k they don't get disabled and have phase effects on the passband.

Note that they also had TDR ultrasonic long ago, which is a better choice than Airwindows.
https://vladgsound.wordpress.com/tag/tdr-ultrasonic/

In the analog realm those are everywhere. Preamps will feature lowpass filters in the ultrasonic band, processors likewise. With tape, I think you actually have a strong ultrasonic filtering so as not to disturb the bias signal. In other words, there is no problem with minimum phase filtering.

[rafa1981]

In the analog realm those are everywhere. Preamps will feature lowpass filters in the ultrasonic band, processors likewise. With tape, I think you actually have a strong ultrasonic filtering so as not to disturb the bias signal. In other words, there is no problem with minimum phase filtering.

The problem with minphase filtering for this application is:

1- For the same response than a FIR brickwall, it will use more CPU, because IIR filters don't parallelize. Each filter stage depends on the output of the previous stage.
2- Phases are affected on the passband. The more brickwall, aka aliasing rejection, the more pronounced the effect. If you want to use a chain with this in parallel and sum/subtract to the original signal the results will be different (for better or for worse).

These are digital plugins to combat digital aliasing by using a digital brickwall filter out of the passband. I don't see why a digital brickwall filter should introduce analog preamp or tape effects, isn't that the purpose of tape and preamp plugins?

[FabienTDR]

Airwindows plugins are not oversampled, Chris generally uses and recommends working at 96k which gives some ultrasonic headroom and reduces the aliasing concerns. However, while running one nonlinear effect at 96k is roughly equivalent to 2x oversampling at 48k, there is a difference if running multiple plugins in series. The first plugin might create some HF content and aliasing at ultrasonic frequences, which is ok, but then second plugin might add further HF and more aliasing which could make it all the way back down to audible frequencies. Adding ultrasonic filters between the plugins helps avoid this cumulative buildup effect, and gives something more closely equivalent to a series of oversampled effects.

TDR Ultrasonic really isn't about antialiasing, the old initial GS thread clearifies this aspect early on. Every AD will handle this fine, no need to DIY oversample AD/DAs that already operate in the MHz range anyway (usually 256 times beyond audible range, or more).

The idea is to prevent the (indirect) appearance of lower sidebands, a consequence of intermodulation distortion. In super hi res chains, ultrasonic content can freely accumulate. Especially in the digital domain, with an arguably perfectly flat magnitude response up to Nyquist, fully unrestricted. This ultrasonic content alone wouldn't really be much of an issue, it's once this content (in addition to the audible content we'd like to preserve) hits a non-linearity that problems can unfold, producing audible partials that have no reasonable relation to the original, partials that wouldn't appear if an economical environment would had been used (e.g. 44.1kHz or 48kHz) .

If you use super hi res chains (for whatever reason), it's worth placing an ultrasonic filter right before the final limiter, to make sure it produces the least amount of disagreeable distortion. Bypass often to find the sweet spot.

Aliasing is really just a special form of intermodulation, where the signal gets modulated with the Nyquist freq, producing sidebands mirroring back into the audible region. The audible and measurable effect of aliasing is very similar to intermodulation distortion (both produce sidebands extending to lower freqs).

[legendCNCD]

These are so slick, Infrasonic is godlike ahh and Prism - 60fps flow.

[ampetrosillo]

In the analog realm those are everywhere. Preamps will feature lowpass filters in the ultrasonic band, processors likewise. With tape, I think you actually have a strong ultrasonic filtering so as not to disturb the bias signal. In other words, there is no problem with minimum phase filtering.

The problem with minphase filtering for this application is:

1- For the same response than a FIR brickwall, it will use more CPU, because IIR filters don't parallelize. Each filter stage depends on the output of the previous stage.

A low pass filter is going to eat, like, 0.1% CPU. Big deal.

2- Phases are affected on the passband. The more brickwall, aka aliasing rejection, the more pronounced the effect. If you want to use a chain with this in parallel and sum/subtract to the original signal the results will be different (for better or for worse).

Yes, you can't mix dry signal in parallel (actually, you can; the response will essentially be a treble shelf, but OK).

These are digital plugins to combat digital aliasing by using a digital brickwall filter out of the passband. I don't see why a digital brickwall filter should introduce analog preamp or tape effects, isn't that the purpose of tape and preamp plugins?

It depends. Also, it can get academic very fast. We essentially can't hear phase, especially in the treble, and treble content is low enough in most audio that the waveshape is unlikely to change much (therefore not reducing headroom significantly).

[TBlake]

who has the best price on the bundle??

[0degree]

God damn it, missed the sale Any outlet still selling it at discounted price?

[Bartizanier]

They look really cool but I'm also too dumb to need them!
Love TDR