Output Dither

[Wolfgang Eller]

Originally Posted by NewHere

I wrote several times that a trip through my desk adds analogue noise that is well above dither noise (around -110dbFS). That is not the concern here.

This, exactly, is the point.
You are adding dither noise at -138dbfs which is masked by analog noise at -110dbfs.
It doesn't make any sense at all.

[Farhoof]

To answer your original question:

Originally Posted by NewHere

So whenever I route something out to the real world (including inserts and auxes) 24-bit dither is automatically applied to the output stream?

No, there is no dither automatically applied.
The only automatic dither is when using 'save copy as' and using a lower target bit depth than the original session or when using 'export audio' (or import audio) and using a lower target bit depth.

Bounce to disk does not apply dither nor does recording into another track within your own session. You need to add it yourself if needed.

The Protools guides recommends to _not_ use dither when going out the analogue output and into a 24bit recording system not because it dithers automatically (it does not), but because it is not needed. You are obviously still able to add it yourself if you want to or still need to.

[seanmccoy]

Originally Posted by JFreak

Last insert slot of master fader is the only place where dither has any use. Last insert slot of aux goes back to mixer and even though the fader was set to unity dithering there is futile.

I remember discussions from a few years back about adding a true peak limiter after the dither, because if flirting with 0Dbfs, the dither noise could actually create overs. Is there a reason doing this would negate the dither?

[JFreak]

Originally Posted by seanmccoy

I remember discussions from a few years back about adding a true peak limiter after the dither, because if flirting with 0Dbfs, the dither noise could actually create overs. Is there a reason doing this would negate the dither?

If it is purely a metering plugin, then it does nothing to the signal, just creates a graphical representation of it. Doesn't affect to dither.

Another solution would be using a limiter such as Sonnox Limiter that has reconstruction meter and dither, but obviously that is not the limiter of choice for most.

[seanmccoy]

Originally Posted by JFreak

If it is purely a metering plugin, then it does nothing to the signal, just creates a graphical representation of it. Doesn't affect to dither.

Another solution would be using a limiter such as Sonnox Limiter that has reconstruction meter and dither, but obviously that is not the limiter of choice for most.

When mixing hot I generally have Fabfilter Pro-L2 in TP mode at the end, but opt for GoodHertz dither a slot ahead when going to 16 bit. In these cases, the limiter is likely to be engaged on occasion. Am I still getting proper dithering?

[JFreak]

Originally Posted by seanmccoy

When mixing hot I generally have Fabfilter Pro-L2 in TP mode at the end, but opt for GoodHertz dither a slot ahead when going to 16 bit. In these cases, the limiter is likely to be engaged on occasion. Am I still getting proper dithering?

Don't know about Fabfilter plugin but you have an easy solution for your dilemma. Use it while mixing and disable it when bouncing. Assuming it is a metering plugin you need to be using while mixing. Dithering plugin really needs to be last.

[seanmccoy]

Originally Posted by JFreak

Don't know about Fabfilter plugin but you have an easy solution for your dilemma. Use it while mixing and disable it when bouncing. Assuming it is a metering plugin you need to be using while mixing. Dithering plugin really needs to be last.

Okay, you lost me. If a limiter at the end of a chain is engaging, preventing overs, how would disabling it during bounce not allow those overs into the bounce?

[JFreak]

Oh, it is a limiter. Well as said, dithering before limiter is futile, because limiter changes the signal.

[NewHere]

Originally Posted by Farhoof

To answer your original question:



No, there is no dither automatically applied.
The only automatic dither is when using 'save copy as' and using a lower target bit depth than the original session or when using 'export audio' (or import audio) and using a lower target bit depth.

Bounce to disk does not apply dither nor does recording into another track within your own session. You need to add it yourself if needed.

The Protools guides recommends to _not_ use dither when going out the analogue output and into a 24bit recording system not because it dithers automatically (it does not), but because it is not needed. You are obviously still able to add it yourself if you want to or still need to.

Although I have already figured that out by myself, I thank you for giving me EXACTLY the answer I was looking for in the first place and especially for providing it in a respectful manner and not like I've failed elementary school! I wish there was a thumbs up button.

Originally Posted by Wolfgang Eller

This, exactly, is the point.
You are adding dither noise at -138dbfs which is masked by analog noise at -110dbfs.
It doesn't make any sense at all.

No, that is your point, not the point of this thread, and I heard you the first time! Duly noted, thank you.

Originally Posted by JFreak

Point is to be able to hear this example you made you need to boost your listening environment 116dB to get comparable levels to -6dBFS sample. And by doing that you are boosting all noise that much whereas -6dBFS sample has the noise 116dB lower level. And when you listen to the -6dBFS sample you surely do not hear the noise at all, whether dithered or not.

I am aware of that, I gained the clips up myself. As mentioned several times I am talking about the stuff that happens at -140dbFS, way below any converters noise floor. The sole purpose of the clips is to illustrate the noise floor (that you normally wouldn't hear, it's at -140dbFS) because you said:

You can think of it (overly simplified) that when you create a fixed-point word from floating-point you get 23 solid bits and the last bit is calculated from the floating point mantissa. So in a sense it is dither in itself. Whatever "errors" there may be are about -140dBFS

The signal clearly doesn't (audibly) dither itself by any means (clip 1), there are lots of errors and dither preserves the signal integrity (yes, down at -140db below any noise floor). Further you said:

Last insert slot of master fader is the only place where dither has any use. Last insert slot of aux goes back to mixer and even though the fader was set to unity dithering there is futile.

This is simply not true (clip 2). If no processing whatsoever occurs behind the dithering plug, it works (listen to clip 2, dither on first slot on aux). Of course it makes most sense on a post-fader master, like you suggested, and I will implement it like that, yada yada yada, see earlier posts.


I didn't came up with this stuff by myself. I am following the suggestions of some great minds on GS, amongst them Bob Olhsson (who even chimed in here) and Paul Frindle (who knows a thing or two about this stuff) and want to check for myself if it makes any difference to me and my art (see? I said art, not work). So thank you all for not shoving your opinions down my throat instead of giving me a simple answer like Farhoof did...


Think I forgot to mention that I was talking about the noise floor at -140dbFS. Just so you know: it's below the noise floor of any converter or analog processor.

[Wolfgang Eller]

Originally Posted by NewHere

No, that is your point...

That is not my point that's a physically fact!

Originally Posted by NewHere

So thank you all for not shoving your opinions down my throat instead of giving me a simple answer like Farhoof did...

When Bob and Paul chimed in your same discussion at GS you knew the answer already.
I'm outta here I don't want to feed a troll.