Logic 10.4.3 Update

[Tribalman]

Let's say you want to use another software such as Ableton Live to play your drums, you can simply disable the MIDI Input of your drum controllers in Logic so that any MIDI those controllers produce will not reach Logic and potentially create havoc.

 

new-pref.png

David, you have a Nektar Panorama ? Just bought a P6 weeks ago. Seems nice

[Tribalman]

It is no longer necessary to hit play after loading a project that uses instances of ARA plug-ins to insure the data and edits are saved.

Really ? No longer ARA mess ?

[Glenn L.]

Addictive Keys has gone wonky on me. Not sure since when, as I updated the OS the other day, as well. But the pianos in both Logic 10.4.2 and 10.4.3 sound like they have a cheap chorus on them. Other Piano AUs, and AUs in general, seem to be normal.

[des99]

I noticed that many (Logic's native's/ custom user's / 3rd party plugin's) presets sound noticeably different from how they did in LPX 10.4.2. Somewhat like differently EQed. Some presets are also louder, which is confirmed by their channelstrip volume meters turning red while they weren't before...

Anybody else experiencing that?

 

Can you give some specific examples where something is not sounding right?

[kerochan]

I noticed that many (Logic's native's/ custom user's / 3rd party plugin's) presets sound noticeably different from how they did in LPX 10.4.2. Somewhat like differently EQed. Some presets are also louder, which is confirmed by their channelstrip volume meters turning red while they weren't before...

Anybody else experiencing that?

 

Wow, this could be a big problem. I already see one confirmation in this thread, anyone else? I think I'll hold off updating until this is more clear.

 

I agree, totally disappointing, everything seemed great with all the bug fixes, but this is crazy,

 

BUT, Is it 10.4.2 which was the problem, and 10.4.3 are how things are meant to sound?

[Eriksimon]

You all react like this (OOOOOO, it is such a YUUUUUUUGE bug again!!!!), but have you actually tested anything? You are aware of the phernomenon called perception bias? This effect is especially strong with audio.

OK, I'll do it. I'l test (4.2 vs 4.3) with audio tracks and a big stack of Logic plugins. And with that Beck demo.

 

See ya'all later.

Edited December 7, 2018 by Eriksimon

[Glendabed]

Heaven knows why but my projects are saved in the format (projectname).logicx and they won't open in 10.4.3 update.

 

Total pain in the ass, anyone know how I can reformat the project files without having access to them within Logic?

 

Hope that makes sense. Arg.

[stormy]

but have you actually tested anything? You are aware of the phernomenon called perception bias? This effect is especially strong with audio.

 

I haven't upgraded. I can see two posts (one with screen captures) where people say that channels that weren't in the red are now in the red. So, something might have changed.

 

EDIT: I appreciate you taking the time to test this!!!

[Eriksimon]

Yeah, people say. Sometimes people also cannot find their car keys, or remember their groceries (it is why they make lists). Peoples' memories are not exactly rock solid - that much is known by now.

In the ten years on this board and Apple's Logic Pro board I have seen innumerable examples of people being so damn extrasuper-dupersure of something sounding different - when it almost always turned out they had changed something, or that the difference they heard wasn't really actually there, so... do not underestimate perception bias paired with people thinking they are flawless and the best and whatnot else - ego's are powerful internal subconscious deception generators... and nobody is immune to them - that is why the scientific method has been developed: to rule out the conviction of infallible fabulousness that plagues mankind (nobody's perfect). A good null test IS rock solid. I am going to do some more. But my first result (see earlier this topic) is that the mixers' summing engine is unchanged between 4.2 and 4.3. Now I am going to try it with... well, I already outlined that in my previous post.

And if the result is that something is off or broken, well, congrats, and also: it can be reported because it can be repeated. It still doesn't mean that anybody is perfect.

[Blue Monkey]

nope I think you will need https://en.wikipedia.org/wiki/MacOS_Mojave to run: logic latest version….

 

You think? Do you have a source for your statement?

 

When I search Google all I find is this (see attachment) and it states Minimum System Requirements: macOS 10.12 or later.

 

[Tribalman]

- that is why the scientific method has been developed: to rule out the conviction of infallible fabulousness that plagues mankind (nobody's perfect)..

Lol, serioulsly I agree with your demonstration.

But ... keep in mind that even the scientist method is not an absolute rule since quantic research show that results depend of the observer.

Just to put a mess to the all the stuff.

Ok, I’m just kidding

[Tribalman]

nope I think you will need https://en.wikipedia.org/wiki/MacOS_Mojave to run: logic latest version….

 

You think? Do you have a source for your statement?

 

When I search Google all I find is this (see attachment) and it states Minimum System Requirements: macOS 10.12 or later.

 

Logic System Requirements.png

I’m on Sierra ( 10.12.6 ) and of course it’s ok with 10.4.3.

[Blue Monkey]

I’m on Sierra ( 10.12.6 ) and of course it’s ok with 10.4.3.

 

Thanks for confirming Tribalman!

[Eriksimon]

Thanks Tribalman. And yes, I know that the scientists' expectation can (and will) color the result, but a repeatable null test is pretty darn solid, and anyone (who has kept their 10.4.2 version of Logic) can do it for themselves.

Well, it is not easy... I ran two tests, in the first the two bounces of Becks Colors project "as is" did largely null, apart from the synth bass and some vocals and the cymbal swell track. So I bipped those tracks, but then I got different results (bass now indeed nulled, see waveforms), namely that now suddenly the chorus vocals (that I didn't touch) completely did not null at all .

 

 

But, it is a big project with lots of plugins, and probably also synths with random or free running LFO's - however, if there was something off with the EQ plugins, I would not expect any nulling whatsoever, and the fact that the drums and the synths (except the bass) completely disappear strongly indicates that nothing (or at best very little) is wrong. However, this conclusion is still falsifiable, so I challenge anyone to deliver a repeatable simple setup showing that two identically set up EQ (or other non-random) plugins do not null between 10.4.2 and 10.4.3, and I'll happily admit to it and also file a bug report. But anyone telling me that they don't need to test because they "can just hear it (because I am such a top projuicer/genius composer with golden ears)" I will dismiss. Because the real top producers and engineers will know that ears, even highly trained and experienced ones, are not trustworthy.

[danhirsch]

It's kinda crazy the you still can't manually punch-in record without the audio disappearing for a split second, isn't it?

[fisherking]

i can hear the difference between 10.4.2 and 10.4.3; am a top producer, so i have the ability to hear the nuances. there's definitely a little 'more' in 10.4.3. also, sounds have a more expensive quality, which doesn't surprise me; each new update to LPX has done this.

 

am doing new mixes on ALL my previous projects, to get that '.1' extra-ness. updates make our lives better...

Edited December 7, 2018 by fisherking

[schlomojorgensen]

Yeah you´re different alright

[Eriksimon]

Lol. Nulltest or it didn't happen. Even though I know you are indeed a top producer (or do I? No, I don't, I just know you are a good guy)

but, help me out: reveal which plugins are doing this? I will happily test it for you.

[danhirsch]

thanks for testing.

In your tests did you include a null test between 2 (or more) bounces from the same Logic version ?

If these don't cancel out perfectly, there's isn't much sense in doing a cross-version test. Right?

 

Thanks Tribalman. And yes, I know that the scientists' expectation can (and will) color the result, but a repeatable null test is pretty darn solid, and anyone (who has kept their 10.4.2 version of Logic) can do it for themselves.

Well, it is not easy... I ran two tests, in the first the two bounces of Becks Colors project "as is" did largely null, apart from the synth bass and some vocals and the cymbal swell track. So I bipped those tracks, but then I got different results (bass now indeed nulled, see waveforms), namely that now suddenly the chorus vocals (that I didn't touch) completely did not null at all .

 

pic 2018-12-07 at 18.48.12.png

 

But, it is a big project with lots of plugins, and probably also synths with random or free running LFO's - however, if there was something off with the EQ plugins, I would not expect any nulling whatsoever, and the fact that the drums and the synths (except the bass) completely disappear strongly indicates that nothing (or at best very little) is wrong. However, this conclusion is still falsifiable, so I challenge anyone to deliver a repeatable simple setup showing that two identically set up EQ (or other non-random) plugins do not null between 10.4.2 and 10.4.3, and I'll happily admit to it and also file a bug report. But anyone telling me that they don't need to test because they "can just hear it (because I am such a top projuicer/genius composer with golden ears)" I will dismiss. Because the real top producers and engineers will know that ears, even highly trained and experienced ones, are not trustworthy.

[Eriksimon]

In your tests did you include a null test between 2 (or more) bounces from the same Logic version ?

No. But a null test between two bounces would indeed reveal which track(s) have random or free running LFO's on them, and thus indicate which tracks need to be BIPped. Good thinking. I'll do that now.

[Eriksimon]

Hahahahaha... I'm inclined to give up, or blame the two glasses of port I've drunk, but the null test between two bounces in the same Logic version (Nulltest III 4.3/4.3 in the screenshot) does in fact null "less" than the other two nulltests, though it is pretty similar to the first "as is" - but LOUDER (the vocals)!

 

I'm going back to rocket science and brain surgery - much easier than audio engineering.

[triplets]

In the ten years on this board and Apple's Logic Pro board I have seen innumerable examples of people being so damn extrasuper-dupersure of something sounding different - when it almost always turned out they had changed something, or that the difference they heard wasn't really actually there, so... do not underestimate perception bias paired with people thinking they are flawless and the best and whatnot else - ego's are powerful internal subconscious deception generators... and nobody is immune to them

 

Well said.

The ego always says "I can hear a difference". So much of this is placebo, it's very entertaining.

[danhirsch]

hahah nice, Eriksimon !

That's why null test are solid in theory but not always as simple as one thinks to carry out.

As long as the two bounces don't cancel out, and they don't, as you have shown, no further testing cross-version makes sense.

On the other hand, if it's the summing engine you want to test why not bounce a project the consists of nothing but rendered files a.k.a. no plugins, no automation, faders at 0 etc etc.

?

[Ashermusic]

Our ears change not just daily, but hourly. Anecdotal evidence means nothing.

[Eriksimon]

I did that earlier (by freezing all tracks in a project and bouncing it in 4.2 and 4.3) the summing engine nulls completely.

viewtopic.php?f=1&t=140069&start=20#p721070

 

To test specific plugins one could devise a much simpler project (e.g. with just one audio track & file and one plugin). In these ones I used the Beck demo project, which is a project with many tracks and many plugins and therefore with many many many variables. I can't even capture them in one screenshot.

 

 

[danhirsch]

Perfect! so you can stick to making music after all

I did that earlier (by freezing all tracks in a project and bouncing it in 4.2 and 4.3) the summing engine nulls completely.

viewtopic.php?f=1&t=140069&start=20#p721070

 

To test specific plugins one could devise a much simpler project (e.g. with just one audio track & file and one plugin). In these ones I used the Beck demo project, which is a project with many tracks and many plugins and therefore with many many many variables. I can't even capture them in one screenshot.

 

Big mixer.gif

 

pic 2018-12-07 at 21.03.04.png

[des99]

Hahahahaha... I'm inclined to give up, or blame the two glasses of port I've drunk, but the null test between...

 

Friends don’t let friends null test while drunk.

 

Jus’ sayin’

Edited December 8, 2018 by des99

[danielsykes]

Nice, a Logic update! ... And of course there's a panic already!

This combined with the huge CPU hit from the meltdown and specter patches, not to mention the unthinkable mayhem from every OS upgrade and I don't even know how I make music anymore. It's hard to play an instrument from the fetal position.

[fisherking]

Nice, a Logic update! ... And of course there's a panic already!

This combined with the huge CPU hit from the meltdown and specter patches, not to mention the unthinkable mayhem from every OS upgrade and I don't even know how I make music anymore. It's hard to play an instrument from the fetal position.

 

wait... doesn't everyone play their instrument in the fetal position?..........

[Atlas007]

What is a fatal position?