Introducing T-RackS Stealth Limiter - Available now!

[MogwaiBoy]

As Peter @ IK quoted from the manual - Stealth can process the dithering noise, but does not actually bit-reduce the file.

Peter can you help clear this up for me... As I understand it...

- Working with 24-bit pre-mastered wav file
- Put Stealth last in mastering chain
- Set Stealth to 16-bit
- Export the song from DAW to 16-bit wav file
^^^^^^ Apply DAW dithering or not? No need to?

[simon.a.billington]

As Peter @ IK quoted from the manual - Stealth can process the dithering noise, but does not actually bit-reduce the file.

Peter can you help clear this up for me... As I understand it...

- Working with 24-bit pre-mastered wav file
- Put Stealth last in mastering chain
- Set Stealth to 16-bit
- Export the song from DAW to 16-bit wav file
^^^^^^ Apply DAW dithering or not? No need to?

You would think not to apply the DAW dither, but its a good question to ask so we can be sure.

[Peter - IK Multimedia]

From earlier in the thread:

The Stealth Limiter can add a dithering noise to the audio program to make the bit depth reduction that will eventually be applied at a later stage to sound best.

The controls in this section allow to set the level of the dithering noise depending on the bit depth the material will be reduced

Please note: the Stealth Limiter does not apply any bit reduction or truncation, which has to be applied elsewhere later in the processing chain by the T-RackS standalone application or by the DAW of choice

So you DO need to apply dither within your DAW (or T-RackS standalone, etc) when you export as Stealth Limiter does not apply any bit reduction or truncation.

[simon.a.billington]

Thanks man good to know.

Good piece of tech too.

[dmbaer]

So you DO need to apply dither within your DAW (or T-RackS standalone, etc) when you export as Stealth Limiter does not apply any bit reduction or truncation.

First of all, one should only apply dithering when going from a higher bit depth to a lesser one, most typically from 24 bits to 16 bits.

The Stealth documentation is unclear as to what level dither is applied, and this is very confusing. It makes no sense to apply dither to the lowest order bits if one is not reducing the bit depth, which Stealth does not do. Now, if Stealth applies dither assuming a subsequent truncation to 16 bits, it would apply the dither at the 16-bit lower boundary. This would be fine assuming the DAW is not also going to apply dither when doing the truncation.

Frankly, I'd just play it safe and have the DAW apply the dithering. That way there's no confusion. But whatever else, do remember that you only apply dither once and only when going from a higher bit depth to a lower one.

[Peter - IK Multimedia]

I believe the one part of my post you quoted is out of context - of course we are talking about applying dithering when going from a higher bit depth to a lower/lesser bit depth. That was assumed throughout the discussion.

Apologies, I sincerely thought the section of the manual I've quoted twice covers your concern clearly with "the controls in this section allow to set the level of the dithering noise depending on the bit depth the material will be reduced" but I see it does not. If you look at the controls to which that statement applies, you can choose 24 or 16, with should alleviate the confusion.

Sure, you can rely on your DAW and play it safe but the dither section of Stealth Limiter were made to augment that and get the best possible sound.

[MogwaiBoy]

I'm actually still confused - but that's not surprising given my average IQ

So if you're working with a 24-bit file and you want to master it to 16-bit. You set Stealth to 16-bit... but then you export as a wav from the DAW at 16-bit with DAW dithering applied.

This does seem like double dithering?

Or is Stealth not dithering at all? If not.. then what exactly is it doing with this function? Just preparing the file better for post-dithering?

[Christian_D]

As Peter @ IK quoted from the manual - Stealth can process the dithering noise, but does not actually bit-reduce the file.

Peter can you help clear this up for me... As I understand it...

- Working with 24-bit pre-mastered wav file
- Put Stealth last in mastering chain
- Set Stealth to 16-bit
- Export the song from DAW to 16-bit wav file
^^^^^^ Apply DAW dithering or not? No need to?

You've got it exactly right just don't add dither again with your DAW.

1) Stealth limiter adds the dither noise.

2) DAW creates the 16-bit file.

[MogwaiBoy]

Peter can you confirm that? Cheers

(I know... splitting hairs - OCD - no one will hear a difference anyway etc)

[dmbaer]

Peter can you confirm that? Cheers

(I know... splitting hairs - OCD - no one will hear a difference anyway etc)

I'm not Peter, but I will confirm it. That's 100% correct.

[sonicpowa]

@MogwaiBoy: It´s just adding noise. DAW or Stealth, it doesn´t matter (but not both at the same time). DAW will then truncate the bits.

[MogwaiBoy]

Right on - thanks for the clarification

[MogwaiBoy]

Hey Peter - I noticed that -0.1dB is sort of the default output level on Stealth. Is this perfectly safe to go with, or should we lower it further (-0.5 to -1) to compensate for possibly ISP and clipping during subsequent conversion to lossy formats like MP3?

[Lesha]

I'd go for -0.4 dB

[MogwaiBoy]

I'm working with -1.0 for mastering but I think it might be too drastic. Apparently -0.3dB is best practice for mastering to CD.

From the marketing material:

With traditional mastering processes, it’s normal to push the volume level towards a -0.1dB or even 0.0dB ceiling. Spikes above this ceiling are suppressed via the use of hard limiters and clippers. Yet these methods only ensure a stable digital level, they do not guarantee undistorted playback once the music is translated into an analog signal.

This is because the process of playback involves a number of conversion processes that, though the digital waveform may be below 0.0dB, can cause the analog waveform to exceed this threshold and introduce unwanted distortion and fuzz.

T-RackS Stealth Limiter features an inter-sample peak-limiting algorithm that gets around this issue by predicting the converted analog waveform of the digital source. Through this process it’s able to limit the sound in a way that ensures the final converted audio will very rarely exceed the set ceiling. This is useful in the modern era where music is often heard on different devices and via different services that all deliver their music via compressed formats, like MP3 or AAC.

Makes it sound like it's safe to bounce it to -0.1dB as long as you have ISP catcher on... but surely that's too close for higher lossy MP3 conversion later on?