The No.1 Website for Pro Audio
Logic Pro10.4 Release Just hit the App Store
Old 2nd January 2020
  #661
Gear Nut
 

Quote:
Originally Posted by Doc No View Post
Yepp, when you turn up the Volume all to the max you hear it it is brutal quite a light noise.
This is what they said:

Quote:
But louder genres like electronic dance music still need to be mastered fairly loud, which means these genres have a relatively small dynamic range. Ask yourself what happens when you reduce the dynamic range of a song by a large amount. That’s right… All the more quiet sounds in the mix become relatively much louder! This is why you typically have to re-level your mix after you squash it up to typical “loudness war” target levels.

And the problem is that truncation distortion can indeed become AUDIBLE in songs that are squashed to loudness war levels. But dithering noise added even to multiple successive 24-bit bounces is so incredibly small that the dithering noise itself will not become audible at loudness war levels!
Old 2nd January 2020
  #662
Gear Addict
 

There seems to be two parallel arguments going on here. One is about the value of dithering when exporting a 24bit file from a DAW which is running internally at 32bit and the other argument is whether Logic ought to import 32bit-float audio files without converting them to 24bit -- and if it does convert them, whether it should be dithering those files.

Regarding the import issue, I believe mu:zines is correct in stating that Logic doesn't damage a 32bit-float audio file when importing and converting it to 24bit. There are 8 bits of redundant information that Logic doesn't need, so it chooses to reformat the file without this unnecessary data.

I have tested it in the following way: Take a snippet of a Logic project and bounce it out as 32 bit float. Import this file into a new Logic project. A 24bit file gets created automatically. Bounce this new project at 32bits again. Bring both bounced files into Audacity and examine them closely at the level of individual samples. As far as I can see, they are sample-for-sample identical.

I haven't tried it but things might be different if the original project was peaking above zero db.
Old 2nd January 2020
  #663
Gear Nut
 

Quote:
Originally Posted by onerob View Post
There seems to be two parallel arguments going on here. One is about the value of dithering when exporting a 24bit file from a DAW which is running internally at 32bit and the other argument is whether Logic ought to import 32bit-float audio files without converting them to 24bit -- and if it does convert them, whether it should be dithering those files.

Regarding the import issue, I believe mu:zines is correct in stating that Logic doesn't damage a 32bit-float audio file when importing and converting it to 24bit. There are 8 bits of redundant information that Logic doesn't need, so it chooses to reformat the file without this unnecessary data.

I have tested it in the following way: Take a snippet of a Logic project and bounce it out as 32 bit float. Import this file into a new Logic project. A 24bit file gets created automatically. Bounce this new project at 32bits again. Bring both bounced files into Audacity and examine them closely at the level of individual samples. As far as I can see, they are sample-for-sample identical.

I haven't tried it but things might be different if the original project was peaking above zero db.
I know what you're all saying is logical/sound, just appears to be some confusing 'opinions' out there.

What about converting these "identical bounces" further to e.g. MP3s, lower bit depth, and multiple times. Copying exactly the same process to the realms where things become much more 'pronounced'?

I know it's unlikely if they are exactly the same or practically the same(?) waveform right now - but this might be conclusive (with some different genres of music, highly processed, etc.).

Is there anyway to analyze the actual code of the .wav files also? Are they 100% identical?

Some (& 'high profile') are definitely suggesting that there IS a difference when going from 32 bit float to 24 bit fixed without dithering...
Old 2nd January 2020
  #664
Gear Addict
 
mu:zines's Avatar
 

Quote:
Originally Posted by J3XS View Post
Some (& 'high profile') are definitely suggesting that there IS a difference when going from 32 bit float to 24 bit fixed without dithering...
There *can* be a difference at *extreme* levels, either above or below, as the levels drop above/below the "normal" range, and combined with the scaling factor, your wordlength can potentially increase - in these cases, you should dither. But that's extreme test cases, not regular normal mix levels. Nobody is suggesting bouncing out to +1000dBFS or -124dBFS!

(I was just discussing this with my homies over at: https://www.soundonsound.com/forum/v...p?f=16&t=69616 )

(And dithering from your 32float mix bus to 24-bit files, at typical levels will be extremely marginal - if you can hear on your material, and it sounds preferable with dithering, then by all means, go for it. You *should* definitely dither from your mix bus to 16 bit fixed files though.)

Like I say, once you grok the difference between 32-bit fixed and 32-bit floating (ie, 32-bit floating contains 24-bits of audio data, and 8 bits of scaling, which only really come into play above or below the typical normal levels used), you'll realise that 99.9% of the time, the audio data between a 32-float file and a 24-bit fixed file will be identical, and no dithering is required, because no wordlength reduction is occurring.

The misconception is thinking a 32-bit float file contains 32-bits of audio data. It does not.

Now, I'm not suggesting wavemantra isn't hearing *something* (I can't obviously say what he is hearing), but his thinking the cause is Logic's 32float to 24fixed conversion without dither is incorrect and seems to be based on a misconception of equating 32float files with 32fixed files (which could indeed cause the issues he describes under certain conditions).

In short, it simply doesn't work like he thinks it does. (Which, like I say, is surprisingly common... the maths of digital audio is often not intuitive.)

(I wonder if Logic dithers 32fixed files to 24-fixed on import, or whether it leaves them as 32fixed?)

Anyway, interesting thread. Always good to bust some digital myths from time to time!
Old 3rd January 2020
  #665
TNM
Lives for gear
Quote:
Originally Posted by wavemantra View Post
https://m.youtube.com/watch?v=uCiNSSa2oT8

@13:00

With all do respect, you are way off with some of the above statements. Most people are. I was until last week, sadly. I would rather not have to research all this technical stuff, I am a musician.
But I like audiophile sound. I noticed a slight degradation in my exports. That is what started it all.

All I wanted to know is if Logic 10.4.4 could import 32-bit f-p without auto-converting to 24-bit. That's it.

I am not interested in your challenges. I dont care.
This thread is whack. Typical for Logic boards.
No.. I exported a soft synth, one considered to be as close to real analog as possible, no samples involved, at 32 bit float. I reimported it and logic converted it to 24, and my null tests were done with the original 32 bit file playing against the 24 bit import.. Logic created the 24 bit import file right next to the original 32 bit, so all I had to do was drag and drop them in PT in a 32 bit float project.

It's, with all "due" respect.

And what am i OFF about? You just make a wild claim like that and do not give ONE example?

The reason you wouldn't spend 10 minutes of your time listening to simple audio files and trying to pick which are Logic vs Reaper, is because you know you'll fail and you'll never pick them more than say 2 or 3 out of 10, which is lucky guesses. If that test was done, *I* would have been the one that did all the work creating it.. but since it seems you missed that part, I will say it again.. I am not going to spend my time doing so *anyway*.. so i have no idea what your emphatic retort of "no way I am doing your challenges" is about.

I replicated *precisely* what you were complaining about in Logic, using an algorithmic synth. I used YOUR methods of export/import to trigger the "issue".

And the only difference is a bit of white noise at about -125DB, the way LOGIC does it. If my words are wrong regarding 32 vs 24 bit in general, that's not the point here.. I was showing the difference between 24 and 32 bit with the "logic limitation" you are specifically speaking of, and what it sounds like is a tiny bit of dither noise.

64 instances can not be heard, of that tiny bit of dither noise.., and the original files were loud, peaking at -0.1 DB, way hotter than most people mix. Yet the 24 and 32 bit of those files nulled to minus 125, and 64 tracks of that nulled result was not audible unless I put a limiter on the master bus and cranked it by 30DB, and even then, was gentle white noise.

I will admit.. over a large project, if one were to import a large bunch of 32 bit files into Logic, it could be ever so slight audible.. maybe... say, 128 tracks of imported 32 bit audio auto converted to 24 bit.

You argument still doesn't make sense though.. you just want to bash Logic and move to creeper and that's what it is about.. You are waiting for excuse after excuse. And you seem to have this desire to want to import large amounts of 32 bit tracks into Logic.. to what exactly? You can't freeze during the actual project if you run out of headroom, and at mixdown time, export to 32 bit? You HAVE to bounce stems to 32 bit as you go along and re import them? Really?

I'm going to create a video showing others exactly what logic does to a 32 bit file it imports, as people reading this might suddenly get scared and think "Logic isn't good enough". So it will all be in evidence showing your own actual methods of "exposing the problem".

I jumped in immediately to help you with *everything* you asked, and you have become obnoxious and basically called me a liar cause you didn't like the results..Classy!

To anyone else reading.. a DAW is a digital audio workstation, key word digital. If you are new to this and confused by any of it, but love using Logic.. please do not worry.. Null tests are 100% infallible. The result of a null test can only be the exact difference between 2 audio files and in Logic's 32 bit vs 24 bit situation, it's a tad of dither noise that is inaudible to any human being unless you duplicate tons of it and crank the volume. A tape plugin most of you use on your track has more noise. It's a flat noise that does not follow the melody of the sound at all so it should be consistent amongst different synths, but I will try it later with 16 different synths, all bounced to 32 and reimported to 24 bit, then null tested in Pro Tools, and see what sort of noise there is that could be "harming the mix" and I'll make a video of it.

If different plugins add different, audible noise, of course i will expose that. Note, bouncing samplers that are using 24 bit files, at a higher bit rate, is entirely meaningless.. I don't know of many if any samplers that use 32 bit audio files in their libraries.. The only things that I know of that are currently "affected" are virtual synths.

On another note, you really should put your money where you mouth is wavemantra.. just listen to 10, 30 second files.5 minutes of your time.. one each from different synths, one 32 bit, and one 24 bit Logic re import, and pick which ones have "lost quality". As I was writing this reply, I decided "to hell with it", i WILL waste my time after all and create the test, and no one will get it right within any margin other than standard guessing. You will see. And you won't either, even if you say you will not take the test and do in your private time.

The other thing I am going to do, is create a small midi file playing a synth in Logic, export it at 24 bit, then import the same midi file in Reaper and export at 32 bit.. in case Logic's "32 bit export is at fault".

No one will be able to tell the difference between Reaper and Logic. You'll see.

The key is 100% volume matching. I have nulled ALL daws against each other to infinity (at 24 bit a few years back) when the levels were done correctly, identical pan laws, etc. It could be the tiniest thing in Reaper by default that makes something sound better to your ears, for example. But, math wise, Reaper does not sound better than Logic, neither does PT, Cubase, or any other DAW currently existing.
Old 3rd January 2020
  #666
Gear Addict
 

Quote:
Originally Posted by J3XS View Post
What about converting these "identical bounces" further to e.g. MP3s, lower bit depth, and multiple times. Copying exactly the same process to the realms where things become much more 'pronounced'?
I guess if the starting position is sample-identical for both bounces then there are no small differences that might begin to accumulate further down the line. As far as I understand, the individual sample is the finest level of detail that audio data can be differentiated.
Old 3rd January 2020
  #667
Gear Head
 

TMN:

Okay, I'm in. Thing is I want to actually work on music tonight not address all your points.

For sure we can clear this up. I will read my words again and if its clear I owe you an apology I will give you one.

If nothing else we will at least understand each other. Aside from me being a grumpy freak, there is definitely misunderstanding and perhaps 3 different topics at issue.

I will be back tomorrow.
Thx

Last edited by wavemantra; 3rd January 2020 at 06:33 AM.. Reason: Olive branch
Old 3rd January 2020
  #668
Tui
Gear Guru
 
Tui's Avatar
Quote:
Originally Posted by J3XS View Post
Even tho you cannot hear the dither noise,
Yes, hold that thought. You are on to something.

Quote:
Originally Posted by J3XS View Post
... different types of dither change the perception of the material.
No.

Either you can hear it, or you cant.

At normal listening levels, nobody can hear dither - unless perhaps it is applied so may times - incorrectly - that it becomes audible as an artefact.

Consider this:

https://en.wikipedia.org/wiki/Confirmation_bias
Old 3rd January 2020
  #669
Gear Nut
 

Definitely sounds like a myth now (32 bit float to 24 bit fixed). Thanks for the discussion, I didn't know much if anything about any of this before!

Quote:
Originally Posted by Tui View Post
Yes, hold that thought. You are on to something.



No.

Either you can hear it, or you cant.

At normal listening levels, nobody can hear dither - unless perhaps it is applied so may times - incorrectly - that it becomes audible as an artefact.

Consider this:

https://en.wikipedia.org/wiki/Confirmation_bias
http://dload.sonnoxplugins.com/pub/p...ser_Guide.html

Quote:
Originally Posted by Sonnox/Paul Frindle
There is no technical (or philosophical) advantage to noise shaping above and beyond that which can be actually heard directly. Therefore the decision to use noise shaping (and to what extent) is basically determined by what might actually be heard in practice. If conventional TPDF dither provides sufficient audible dynamic range such that noise never intrudes within the programme material, it is safest to avoid noise-shaping altogether.
Again rather confusing, there's no advantage to noise shaping [above and beyond that which can be actually heard directly] BUT the decision to use it is determined by what might actually be heard in practice?

Is it that dither may be more 'perceivable' (audible?) in 16 bit than 24 bit?

But sounds like they're still saying it's inaudible?

What is the purpose of noise shaping? Why do they sometimes prefer it in mastering? How can it possibly change the material/song if it's basically completely undetectable?
Old 3rd January 2020
  #670
Gear Nut
 

Quote:
Originally Posted by mu:zines View Post
Like I say, once you grok the difference between 32-bit fixed and 32-bit floating (ie, 32-bit floating contains 24-bits of audio data, and 8 bits of scaling, which only really come into play above or below the typical normal levels used)
I came across the "32float to 24fixed DEFINITIVE THREAD" (lol)

32float to 24fixed DEFINITIVE THREAD

Some quotes:

Quote:
For a start the 32 bit float format has an effective mantissa size of 25 bits, not 24. There are 23 bits stored, and one sign bit, which would seem to make for 24 bits, but rules of the format mean that the highest bit in the mantissa is always a 1, so that isn't stored (it's known already, no need to store it), therefore there are 25 bits worth of precision.
Quote:
So dither will probably make little or no difference to the correlation of quantization error (it will already be very noise like) and that's before we consider the reality that the quantization steps in a 24 bit signal will be well below the threshold of audibility (assuming peak levels aren't set well above the pain threshold), and even further below the noise floor of the system.
Quote:
Great summary, Jon. I'd like to add that in practice dither is needed most at low signal levels — and this is where the “precision” of a 32-bit float format exceeds 25 bits and reaches 27–28 bits (in a fixed-point scale). This makes many of Nika's problems go away.
However:

Quote:
Originally Posted by FabienTDR
...One aspect is pretty clear however: Keep processing minimal. Truncation grows fast when it comes to processing.
Quote:
Originally Posted by Bob Olhsson
Indeed it does and we are almost never the last people to process the audio we handle these days. At some point it starts sounding crunchy. Best to not have that happen to the end listener!
Noise shaping:

Quote:
Bob Katz uses noise-shaped dither, while OpusOfTrolls uses flat dither. Go figure...
Quote:
Originally Posted by OpusOfTrolls
I consider noise-shaped dither to be a creative effect. Like those "hiss" buttons on plug-ins.
Old 3rd January 2020
  #671
Gear Addict
 
mu:zines's Avatar
 

Yep - point #1 and #3 (which relate to my side of the conversation) I already mentioned in my above post.

(Well, I sort of skipped the sign bit thing, it's not really important and it gets maths nerdy quite fast when you get deep into floats. I think it means there's effectively 25-bits, but that "25th" bit is an implicit 1 so not stored, so still 24-bits of audio. Not sure offhand without checking.)

Things often get funky at the extremes, but we can confidently assume relatively normal levels here - if someone's complaining that their audio sounds flat and one-dimensional when imported, and it's peaking at -100dBFS, I'd suggest an easier way to improve that than switching DAWs...
Old 4th January 2020
  #672
Gear Nut
 

Quote:
Originally Posted by mu:zines View Post
Yep - point #1 and #3 (which relate to my side of the conversation) I already mentioned in my above post.

(Well, I sort of skipped the sign bit thing, it's not really important and it gets maths nerdy quite fast when you get deep into floats. I think it means there's effectively 25-bits, but that "25th" bit is an implicit 1 so not stored, so still 24-bits of audio. Not sure offhand without checking.)

Things often get funky at the extremes, but we can confidently assume relatively normal levels here - if someone's complaining that their audio sounds flat and one-dimensional when imported, and it's peaking at -100dBFS, I'd suggest an easier way to improve that than switching DAWs...
Some interesting comments in the Good dither practices, what are yours?

Quote:
Originally Posted by chrisj View Post
Yeah, please don't start fires. It's been the same thing ever since people were arguing against dithering 16 bit audio, it's just a matter of degree. The conclusion you need to be drawing from this is, 'some people over there are not going to be able to help me work on word length issues in digital audio to my satisfaction'. In turn, you are not going to be able to fit into their 'good practices' box, as you'll be demanding stuff they consider bad.

Everything is relative. On this very thread, I worked out the details of what I'm currently using in all my plugins: dither of the fixed-point part of literally floating point signals, aka floating point dither. I didn't do it all by myself: Alexey Lukin of iZotope jumped in and even came up with code, though my final code didn't use his (so, we both have distinct implementations: don't know if he's using his, but he's free to as I feel that was sufficiently different from mine to not count as the same)

When I did that, people got really confused as to what they were supposed to do to take their existing plugins and dither THOSE to the floating point buss. meanwhile, people who don't think you should dither 16 bit audio were on fire

Compared to that, dithering 24 bit is simple, direct, obvious, and 100% technically correct. Dithering to ANY fixed point format is technically correct. Ever since that time I've been using floating point output dither on all my plugins, and though they have never been mainstream they're successfully doing what I wanted: pushing the limits of how un-digital you can get digital. Heck, the most recent one, 'Baxandall', had a poster say this:


Interestingly, the actual plugin EQ curves aren't what you'd think: there's a funny interaction up around the Nyquist frequency because the mechanic I'm using is interleaved biquad filters. But it does work, to get that result, and that result is absolutely what I intended. Might even be useful for someone like Bob O. so long as you're okay with the plugin being slightly quirky curve-wise

Digital is weird. I've got a gain scaling plugin that is absolutely, utterly lossless within its design parameters. So much so that it does NOT use the floating point dither as that would reduce accuracy. It's called BitShiftGain. What is the magic algorithm nobody else has, that lets me do this amazing lossless gain trim thing?

sample /= 2
sample *= 2
sample *= 16
and so on.

In other words, literally nothing. Nothing except gain changing, using simple math, on powers of 2. This bit shifts the audio file so that, if it was not distorting and there was room for it to be louder, it keeps exactly the same audio but X amount louder: it changes the exponent without touching the mantissa. You can put it back, losslessly. If you start with reducing gain you lose the least significant bits, but you've got a lot of room for error there too, and if you're boosting gain the most significant bits are going to be zeroes.

I will take a moment to observe that when your Logic-board poster said

Oh no no. That scaling is constantly in use. Every 6dB you're shifting to a new level of scaling. It's why we don't normally hear that 32 bit floating point goes as coarse as 24 bit at its loudest samples: all the quieter samples are continually being represented at finer word length, ad 'infinitum'. But the scaling is constantly in play, all the time, typically to your benefit. Any reverb tail is being represented at the equivalent of 32 bit fixed point or more, perhaps lots more. People have lots of reason to conclude that working with floating point is all they'll need.

For me, my 'best practice' and what goes into my plugins, is to dither the mantissa going into that floating point buss, and to run my internal calculations at greater than 32 bit. I'm pretty sure digital haters who like my plugins can thank choices like that. But it does tend to set forum threads on fire when I talk about it, so more often than not I just let the audio quality speak for itself. It does that, but people not attuned to peculiarly digital forms of distortion tend to not get what the fuss is about.

And that's fine. Takes all kinds.
Quote:
Originally Posted by Bob Olhsson/Motown
Nobody is arguing that 24-bit dither sounds worse. They just question its necessity. The folks at Bell Labs say that the math is wrong without it and that's good enough for me.
Quote:
Originally Posted by Bob Olhsson
That's the old myth.

The problem is that it takes the bit reduction completely out of context rather than pointing out that it happens multiple times before the final audio signal. Another problem is that if audio has ever been truncated, you'll not hear much difference because truncation distortion will mask the low-level information.
Also regarding null testing in Airwindows BitShiftGain: AU, Mac and PC VST

Quote:
Originally Posted by chris
I don't mean 'nulling out so you can't hear a difference', though, I mean literally every bit in the resulting files exactly the same. And this is difficult to do with floating point audio because so many things can affect the result.
I think they're suggesting that null testing are not accurate in this case. And a difference can definitely be heard according to them because they are not mathematically identical (it's not perfectly lossless with all of the processing stages) and there is more going on.
Old 4th January 2020
  #673
Gear Head
 

Hello all.
I haven't forgotten about this.
Got called to work yesterday, then home internet crapped out.

Good to see that last post. Still much confusion in the community about truncation.

We might not agree in the end but we will understand.

I have learned a lot about this. I think I know why I hear it so much and perhaps others do not.

I owe this board an apology for being a grump.
I do not owe Apple anything. My beef is valid.

Will be back asap...
Old 8th January 2020
  #674
Gear Head
 

Dither or no dither.
To each his own.
Would not even be an issue if Logic could simply import 32-bit files.
I would prefer no word-length reduction until the final mastering.

The fact is TD exists, whether you hear it or not.
You will not notice it much at all from loud signals, mixes with low dynamic range, from what I have been told and what I have verified for myself.

It is noticeable with music that has high dynamic range, lots low level signal components, verb tails, delays with high feedback.
It is noticeable on very accurate monitor systems.
At times my music can have 40db of dynamic range. Example: I feed some textures into delays and verbs and let it almost totally evaporate before the next cycle begins.
It is ambient music, the sense of space and width is important, sometimes I kick the sides up +0.5db - +1.0db.
In a mixing situation any reduction of sense of width is noticeable.

If I bounce the 2-bus without dither to 24-bit it is quite noticeable, the effects I have previously described.
If I bounce the 2-bus with dither to 24-bit the width and depth is retained, as is whatever sense of 'warmth and vibrancy' was there originally.
Bell Labs says all the math does not add up when reducing word-length, so dither masks the problem. I believe it.

Good news that Logic can bounce to 32-bit fp with 10.4.4. Not so in 10.2.2.
I can double bus, or export stems and render 32-bit fp, files, so there is a work-around.
Apple should allow 32-bit import. For those who would want that.
Screw Apple. Other DAWs perform better.
They all sound the same though, except maybe MixBus. It has a sound signature.

Quote:
The reason you wouldn't spend 10 minutes of your time listening to simple audio files and trying to pick which are Logic vs Reaper, is because you know you'll fail and you'll never pick them more than say 2 or 3 out of 10, which is lucky guesses.
I never said I could. It is all relative to was being heard before. I would not be able to tell the difference between Logic and Reaper.
I would be able to tell the difference between my mix bounced to 24b with dither and 24b without, regardless of the DAW. Logic, Reaper, S1, Cubase, PT, all of them.
At this point I do not think I need to go into more detail. It is elementary. It is not about blind testing.

Quote:
And the only difference is a bit of white noise at about -125DB, the way LOGIC does it. If my words are wrong regarding 32 vs 24 bit in general, that's not the point here.. I was showing the difference between 24 and 32 bit with the "logic limitation" you are specifically speaking of, and what it sounds like is a tiny bit of dither noise.
Bell Labs has research data to disprove that. If you don't notice anything, then for you it is not a problem.
I notice it, therefore it is, for me.

Quote:
You argument still doesn't make sense though.. you just want to bash Logic and move to creeper and that's what it is about.. You are waiting for excuse after excuse. And you seem to have this desire to want to import large amounts of 32 bit tracks into Logic.. to what exactly? You can't freeze during the actual project if you run out of headroom, and at mixdown time, export to 32 bit? You HAVE to bounce stems to 32 bit as you go along and re import them? Really?
Yes, due to limitations of my current system, I must. Does this bother you? If so, why?
DIVA will max out my machine with three notes, sometimes. I export, so I can import an audio file that can be heavily edited.
I want the imported file to be 32-bit. Does this bother you?

Quote:
I jumped in immediately to help you with *everything* you asked, and you have become obnoxious and basically called me a liar cause you didn't like the results..Classy!
I did not call you a liar. You are thin skinned. Obnoxious. On a forum. First time for everything.

Quote:
On another note, you really should put your money where you mouth is wavemantra.. just listen to 10, 30 second files.5 minutes of your time.. one each from different synths, one 32 bit, and one 24 bit Logic re import, and pick which ones have "lost quality". As I was writing this reply, I decided "to hell with it", i WILL waste my time after all and create the test, and no one will get it right within any margin other than standard guessing. You will see. And you won't either, even if you say you will not take the test and do in your private time.
I can tell which ones are which from my own audio files. I do not need yours.
I am not trying to convince you of anything.
You sound like whiner. You just don't like me. Good. I don't like you.

Quote:
No one will be able to tell the difference between Reaper and Logic. You'll see.
Yawn..... ^ refer to the comment I made above ^

Quote:
The key is 100% volume matching. I have nulled ALL daws against each other to infinity (at 24 bit a few years back) when the levels were done correctly, identical pan laws, etc. It could be the tiniest thing in Reaper by default that makes something sound better to your ears, for example. But, math wise, Reaper does not sound better than Logic, neither does PT, Cubase, or any other DAW currently existing.
Again I never said or implied DAWs sound different. The topic is TD, not DAW.
Math wise Bell say to dither. Good enough for me.

You do you. I do me.
Screw Apple and the fanbois....

No apology to you.
Ignore me. As I will you.

Oh, and bouncing to 24-bit without dither sounds like 5h1t compared to bounces using dither, regardless of what DAW.
Now go out and purchase that new Apple product!!
Godspeed people, Godspeed.

Last edited by wavemantra; 8th January 2020 at 08:44 PM.. Reason: lulz
Old 10th February 2020
  #675
Gear Nut
 

Dan Worrall posted a new video "WTF is Dither?"



I'm not all the way through yet but just got to the part where he mentions about applying third party dither plugins to the master fader.

If the plugin on that fader is not post fader, this can have an effect on rendering - if automating that fader!

So in Logic Pro there are no post fader plugins/inserts (except for being able to change a bus send to post fader - so all plugins in that send will be post fader) - however I believe the mix/stereo bus is PRE fader inserts.

Sometimes you want to automate the stereo bus fader, e.g. adjusting the overall mix to dip in volume in the pre-chorus. Now if using the pre-fader dither plugin it wont render correctly since processing of the fader automation is after the dither plugin!

---

Looks like it is the only option to use Logic's built in dither for mix bus renders!!!

Post fader inserts can also be useful for e.g. headphone calibration EQ / CanOpener - Logic seems to be the only platform that hasn't adopted this!?
Old 10th February 2020
  #676
Gear Nut
 

Quote:
Originally Posted by J3XS View Post
Dan Worrall posted a new video "WTF is Dither?"



I'm not all the way through yet but just got to the part where he mentions about applying third party dither plugins to the master fader.

If the plugin on that fader is not post fader, this can have an effect on rendering - if automating that fader!

So in Logic Pro there are no post fader plugins/inserts (except for being able to change a bus send to post fader - so all plugins in that send will be post fader) - however I believe the mix/stereo bus is PRE fader inserts.

Sometimes you want to automate the stereo bus fader, e.g. adjusting the overall mix to dip in volume in the pre-chorus. Now if using the pre-fader dither plugin it wont render correctly since processing of the fader automation is after the dither plugin!

---

Looks like it is the only option to use Logic's built in dither for mix bus renders!!!

Post fader inserts can also be useful for e.g. headphone calibration EQ / CanOpener - Logic seems to be the only platform that hasn't adopted this!?
But I guess, for anyone not aware of this, as a workaround you can add a gain plugin before the dither plugin, rather than automate the fader.

And you can use an app to be able to use the headphone EQ / spatial plugins for the whole OS sound system.

But really why not just allow post fader inserts, rather than all of these workarounds???
Old 10th February 2020
  #677
Gear Nut
 

If any of the Logic developers are listening:

Quote:
Originally Posted by J3XS View Post
Logic has a separate master fader also, as well as a final mix buss ("Output" channel).



If you put e.g. CanOpener/Morphit on the mix bus (Output) channel, it still affects the levels on the fader meter because all plugins are pre-fader! - in this case you have to use metering plugins before and the fader meters become practically useless.

On the Master channel there are no plugins allowed. However you've given me the idea that this is where the post fader [metering/monitoring/dither] plugins should be allowed, since as you can see in the pic it's just empty space anyway!

I currently use the SoundSource app to put my headphone EQ but there are other reasons to wanting post fader inserts in Logic.
https://rogueamoeba.com/soundsource/

I guess tho there are workarounds for all of these.
Old 27th March 2020
  #678
Gear Nut
90 days demo version available now

https://www.apple.com/logic-pro/

Old 29th March 2020
  #679
Lives for gear
 
jsblack's Avatar
 

Is the cursor bug gone in latest updates of Logic..? it's really hard to keep calm when doing automation editing particularly..
Would be so happy to keep my arrow cursor to work properly... (command+Click is set to Marquee tool for me)
I'm still working with 10.4.1 on Sierra FWIW.
i'm really tempted to upgrade both OS and Logic if this bug has been iron out..
Thanks
Old 30th March 2020
  #680
Lives for gear
 
Doc No's Avatar
Quote:
Originally Posted by jsblack View Post
Is the cursor bug gone in latest updates of Logic..? it's really hard to keep calm when doing automation editing particularly..
Would be so happy to keep my arrow cursor to work properly... (command+Click is set to Marquee tool for me)
I'm still working with 10.4.1 on Sierra FWIW.
i'm really tempted to upgrade both OS and Logic if this bug has been iron out..
Thanks

Never had this issue.
Old 30th March 2020
  #681
Lives for gear
 
jsblack's Avatar
 

Thanks for sharing, you're lucky i guess cuz it's a long time known bug
Old 30th March 2020
  #682
Gear Addict
 
mu:zines's Avatar
 

Quote:
Originally Posted by jsblack View Post
Thanks for sharing, you're lucky i guess cuz it's a long time known bug
What's the procedure to replicate this bug?
Old 31st March 2020
  #683
Lives for gear
 
jsblack's Avatar
 

Thanks for your interest.
Would you please have a look here.. When you've got some spare time.. https://www.logicprohelp.com/forum/v...c.php?t=130272
Old 31st March 2020
  #684
Lives for gear
 
Doc No's Avatar
Quote:
Originally Posted by jsblack View Post
Thanks for your interest.
Would you please have a look here.. When you've got some spare time.. https://www.logicprohelp.com/forum/v...c.php?t=130272
It seems to be fixed. I can't replicate bug.
Old 31st March 2020
  #685
Gear Head
 

Quote:
Originally Posted by jsblack View Post
Thanks for sharing, you're lucky i guess cuz it's a long time known bug
Does this happen with "Click zones" enabled? I can´t remember ever experience this.
Old 31st March 2020
  #686
Lives for gear
 
jsblack's Avatar
 

It happens on 80% of projects in my case.

Arrow shape being my main tool plus marquee tool as second one (Command key assigned)

the erratic behaviour occurs after some time on not so heavy projects just by overing audio or midi regions on the arrange window (automation view or not), the arrow shape cursor being replaced by other tools and stuck.

i need to double escape key stoke to get the arrow back and sometimes it doesn't fix the problem; i so need to go over the displayed track on the left (info - "I" key) and mimic some fake moves to recover my arrow.

@ Doc No - What version of Logic do you run and on what OS..?

@ marcello - Nice hint! no click zones enabled here.. i will to try to activate some to shake the bug

Thanks
Old 1st April 2020
  #687
Lives for gear
 
Doc No's Avatar
Quote:
Originally Posted by jsblack View Post

@ Doc No - What version of Logic do you run and on what OS..?

Thanks

Logic 10.4.8 and Catalina 10.15.3
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump