The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Ableton - sound quality DAW Software
Old 3rd September 2018
  #391
Lives for gear
Quote:
Originally Posted by lllubi View Post
You add this extra 64 samples buffers in Apollo´s control panel. Not sure why these extra buffers are needed in Reaper and not in Ableton according to UA.
OK, Indeterminate then.

Quote:
Originally Posted by lllubi View Post
I mean switching off multicore support in Ableton´s settings in Windows. Which seems to help even RMEs in some systems. .
As a stern RME user I'll take this "switching off multicore support" idea with a bowl of salt, please. Do I have problems with RME drivers? Nope. I think this is hogwash, but in another discussion focused on this issue (assuming there exists one) and solutions I'm open to hear the arguments for this action. I'm suggesting the RME forums arena. Post a link. I believe this is really about another problem getting solved indirectly by this very much "I'm giving up" action. And you can't do this in 10 anyway, so how is it relevant 2018? I'm just asking.

Last edited by Mikael B; 3rd September 2018 at 08:57 PM..
Old 3rd September 2018
  #392
Lives for gear
Quote:
Originally Posted by YourBestFriend View Post
ive always disliked the sound of the ableton eq. yesterday I could hear abletons filter pulling my track out of phase. The saturation is a sad affair. I wonder if itb effects are just pasted over the audio rather than having the audio ran through them like a traditional pedal. it sure feels like a pasted over sound effect.
You know, instead of "feeling" you could learn to identify problems and fix them. What is great saturation really depends on what kind you want and how much. I think the Saturation device is pretty good. This doesn't stop me from having 10 more plug-ins for that.
Old 4th September 2018
  #393
Lives for gear
 

Quote:
Originally Posted by Mikael B View Post
As a stern RME user I'll take this "switching off multicore support" idea with a bowl of salt, please. Do I have problems with RME drivers? Nope. I think this is hogwash, but in another discussion focused on this issue (assuming there exists one) and solutions I'm open to hear the arguments for this action. I'm suggesting the RME forums arena. Post a link. I believe this is really about another problem getting solved indirectly by this very much "I'm giving up" action. And you can't do this in 10 anyway, so how is it relevant 2018? I'm just asking.
It is not about RME. But about the behaviour of some Windows systems in Ableton, where it is not clear why an optimized soundcard for ASIO like an RME doesn´t work as intended. Would be interesting to know if people experience still Ableton´s sound engine as different when they switch off multi-core support.
Old 4th September 2018
  #394
Gear Guru
 

Quote:
Originally Posted by lllubi View Post
Would be interesting to know if people experience still Ableton´s sound engine as different when they switch off multi-core support.
This is the part I call "grasping at straws". It is clearly been demonstrated over and over that when audio is exported from DAWs that are set exactly the same, they sound exactly the same. They null.

Quote:
some weird behaviour in so-called real time versus a printed file.
So now it sounds different, but only to you in your studio? Like one of those movies where only the protagonist can see the ghost? Because if it actually screwed up the file, you would be able to post those files (and your methodology) and everybody could try it for themselves, and learn whether this claim was legit or BS.

But if you say: "it alters my "experience" - that allows you to cling to your placebos for a little bit longer. It sounds different in my studio because of buffers or whatever - but naturally I can't post any files, because it's only in the real time playback!

How convenient.

Now anyone who doubts the 'DAW sound mythology' will have to go over to your studio in person?
Old 4th September 2018
  #395
Lives for gear
 
Acid Mitch's Avatar
 

Quote:
Originally Posted by lllubi View Post
Would be interesting to know if people experience still Ableton´s sound engine as different when they switch off multi-core support.
Why would that change the sound instead of just giving less cpu power?
Old 4th September 2018
  #396
Lives for gear
 

Quote:
Originally Posted by joeq View Post
This is the part I call "grasping at straws". It is clearly been demonstrated over and over that when audio is exported from DAWs that are set exactly the same, they sound exactly the same. They null.



So now it sounds different, but only to you in your studio? Like one of those movies where only the protagonist can see the ghost? Because if it actually screwed up the file, you would be able to post those files (and your methodology) and everybody could try it for themselves, and learn whether this claim was legit or BS.

But if you say: "it alters my "experience" - that allows you to cling to your placebos for a little bit longer. It sounds different in my studio because of buffers or whatever - but naturally I can't post any files, because it's only in the real time playback!

How convenient.

Now anyone who doubts the 'DAW sound mythology' will have to go over to your studio in person?
Monitoring through DAWs never happens in real-time. I still prefer strictly analogue paths which mix in playbacks. Why people never complain about DAW´s sounding different in real-time or playback when they use UAD might be worth investigating I think.
Old 4th September 2018
  #397
Gear Maniac
 

Quote:
Originally Posted by Acid Mitch View Post
Over in the electronic music and production forum there is a current thread where people are stating that some daws and audio editors ( Soundforge and I think logic) do not display precise values. This means that you can change volume on a channel and then not get back to the same value you started with, due to the gui not giving the same values as what is happening internally.
So no, 1+1 is not always equal to 2 in all daws.
what has that got to do with summing and audio quality?
Old 4th September 2018
  #398
Lives for gear
 
Acid Mitch's Avatar
 

Quote:
Originally Posted by parricide View Post
what has that got to do with summing and audio quality?
If the display does not show true values and the controls don’t allow you to select values that reflect true audio levels then the sums will be off.
You won't be able to add 1 + 1 and get 2 as a result, as shown by some of the other posters with Ableton, Logic and Sound Forge.
Old 4th September 2018
  #399
Lives for gear
Quote:
Originally Posted by lllubi View Post
It is not about RME. But about the behaviour of some Windows systems in Ableton, where it is not clear why an optimized soundcard for ASIO like an RME doesn´t work as intended. Would be interesting to know if people experience still Ableton´s sound engine as different when they switch off multi-core support.
Wouldn't it be more interesting to measure the sound and compare what comes out with multicore enabled vs disabled? If one uses Live 9 that is.
Old 4th September 2018
  #400
Lives for gear
Quote:
Originally Posted by joeq View Post
It sounds different in my studio because of buffers or whatever - but naturally I can't post any files, because it's only in the real time playback!
which can be recorded…

We could even go as far as measuring in the ear canal. Where we can't measure, I believe, is in the brain.
1
Share
Old 4th September 2018
  #401
Lives for gear
 

Quote:
Originally Posted by Mikael B View Post
Wouldn't it be more interesting to measure the sound and compare what comes out with multicore enabled vs disabled? If one uses Live 9 that is.
Saved a friend a session after a Windows 10 update. While he never experienced any troubles on Windows 7 on the same computer.
Old 4th September 2018
  #402
Lives for gear
Quote:
Originally Posted by lllubi View Post
Saved a friend a session after a Windows 10 update. While he never experienced any troubles on Windows 7 on the same computer.
What does this have to do with anything? If disabling multicore improves things then this is only indicating there are other more serious issues that are now ignored. Disabling multicore, which can't even be done in 10 (because it's a stupid idea in 2018), doesn't really solve the root cause. It moves the problem elsewhere.

This is sound engineering, not black magic where you try a hodge-podge of actions and when something is different you declare that a miracle have occurred. I was hoping you had some actual meat to bring to this discussion, rather than some bland anecdotes.

Last edited by Mikael B; 5th September 2018 at 11:11 AM..
Old 5th September 2018
  #403
Lives for gear
 

Quote:
Originally Posted by Mikael B View Post
What does this have to do with anything? If disabling multicore improves things then this is only indicating there are other more serious issues that are now ignored. Disabling multicore. which can't even be done in 10 (because it's a stupid idea in 2018), doesn't really solve the root cause. It moves the problem elsewhere.

This is sound engineering, not black magic where you try a hodge-podge of actions and when something is different you declare that a miracle have occurred. I was hoping you had some actual meat to bring to this discussion, rather than some bland anecdotes.

Don´t understand why Ableton 10 doesn´t offer multicore to be disabled anymore. Can you change a software which was not optimized for hyperthreading to deliver the same latencies with hyperthreading on constantly
without a price in an update?

From my understanding not really. Since many users complain about Live 10 being unusable for them.
Old 5th September 2018
  #404
Gear Guru
 

Quote:
Originally Posted by Acid Mitch View Post
If the display does not show true values and the controls don’t allow you to select values that reflect true audio levels then the sums will be off.
You won't be able to add 1 + 1 and get 2 as a result, as shown by some of the other posters with Ableton, Logic and Sound Forge.
I could see this potentially causing a non-null in a null test. I make a mix in DAW 1 and then I want to duplicate that mix exactly in DAW 2, so I put all the levels "the same". To the same "numbers". If the readouts are off, then "the same" is not exactly exactly the same, and the null test will return a residue consisting of how much they are off by.

But of course that does not demonstrate that the two DAWs are actually summing differently. Only that there is an error in the GUI. Nobody in real life looks at the 'number' when he sets his lead vocal fader. He puts the fader where he thinks it should be by ear.

So a whacked-out display cannot in and of itself be sufficient to create the illusion that DAW 2 is "harsh" or "mellow" or "mid-forward" or whatever. Even if someone felt compelled to put his track at +1.5 because he thought +1.3 was "unlucky" - he would still have the freedom to adjust everything else. For a DAW to have a "sound" it must force that sound. A mis-calibrated display forces nothing.

In any case, this idea might be brought out to explain a non-null. But most of the non-nulls we see are errors in methodology. Lord knows there are a lot of people who do not understand how to do a real test. People who do understand HAVE nulled these DAWs. There is really nothing that requires this as an "explanation".

Except of course people's Placebos!

People can seize upon some minuscule residue and claim there's the 'quality' their golden ears are picking up, but they could not hear that residue if their lives depended on it. How many people have publicly picked out which DAW made which export?

Zero.
Old 5th September 2018
  #405
Lives for gear
 
IanBSC's Avatar
I've been using Ableton 9 for years. Realtime playback definitely sounds different from the exported files. Maybe they fixed this in 10.

My best guess is that they use CPU resources differently and realtime playback is competing with other processes/applications and is not bit perfect.

I honestly wish I didn't have to export my mixes to know how they really sound.
Old 5th September 2018
  #406
Lives for gear
Quote:
Originally Posted by IanBSC View Post
I've been using Ableton 9 for years. Realtime playback definitely sounds different from the exported files. Maybe they fixed this in 10.

My best guess is that they use CPU resources differently and realtime playback is competing with other processes/applications and is not bit perfect.

I honestly wish I didn't have to export my mixes to know how they really sound.

Obviously you thought about recording the playback, digitally before the DAC, and compared that to the export. What was the result? Please post a few snippets of these recordings, maybe 10 seconds will suffice. Make sure they start on the same sample. Thank you!
Old 5th September 2018
  #407
Lives for gear
Quote:
Originally Posted by lllubi View Post
Don´t understand why Ableton 10 doesn´t offer multicore to be disabled anymore. Can you change a software which was not optimized for hyperthreading to deliver the same latencies with hyperthreading on constantly without a price in an update?
What did i just say? Your assumptions here suggests you don't understand what "optimized for hyperthreading" even mean.

So what do you think happens when multi-core support is turned off in Live 9?

Quote:
Originally Posted by lllubi View Post
Since many users complain about Live 10 being unusable for them.
There are no established systemic performance differences between 9 and 10. When 9 was released people complained that it wasn't as efficient as 8. There certainly can be some plug-ins that run worse and need an update, but that's always a possibility in any major update. The same applies for hardware drivers.
Old 5th September 2018
  #408
Quote:
Originally Posted by IanBSC View Post
I've been using Ableton 9 for years. Realtime playback definitely sounds different from the exported files. Maybe they fixed this in 10.

My best guess is that they use CPU resources differently and realtime playback is competing with other processes/applications and is not bit perfect.

I honestly wish I didn't have to export my mixes to know how they really sound.
I've never heard that myself.
I just finished a commercial release only using Live.
The session in Live sounds the same as the exported stereo mixes.
I don't use a lot of plug-ins. I also record just about all my midi to audio.
I see people using many plug-ins on every channel. I might use no plug ins on most channels and two or three at most on other channels.
When I mix I put Softube Console on every channel.
Old 5th September 2018
  #409
Lives for gear
 

Quote:
Originally Posted by Mikael B View Post
What did i just say? Your assumptions here suggests you don't understand what "optimized for hyperthreading" even mean.

So what do you think happens when multi-core support is turned off in Live 9?



There are no established systemic performance differences between 9 and 10. When 9 was released people complained that it wasn't as efficient as 8. There certainly can be some plug-ins that run worse and need an update, but that's always a possibility in any major update. The same applies for hardware drivers.
I did hear more stable audio interfaces on Windows 10 with multi-core disabled in Live 9.

There are differences between Live 10 and 10.0.1 no matter how old a system is. The Ableton forum and the net is full of complaints.
Old 5th September 2018
  #410
Quote:
Originally Posted by lllubi View Post

There are differences between Live 10 and 10.0.1 no matter how old a system is. The Ableton forum and the net is full of complaints.
Differences between 10 and 10.0.1?

The net is full of complainers. If I saw artists of note complaining about 10.0.1 I would sit up and take note.
There's an old saying - a bad workman blames his tools.
If you look around, all you'll see is full time artists making great music using any recent iteration of Ableton Live (and Logic and Pro Tools), including 10 and 10.0.1.
Which brings me back to my original point from pages ago.
People who have to, just get on and make music with a positive attitude.
Other people get bogged down by testing and finding fault in their music tools.
Old 5th September 2018
  #411
Lives for gear
Quote:
Originally Posted by Acid Mitch View Post
If the display does not show true values and the controls don’t allow you to select values that reflect true audio levels then the sums will be off.
You won't be able to add 1 + 1 and get 2 as a result, as shown by some of the other posters with Ableton, Logic and Sound Forge.
This inspired me to do a test with unwarped pink noise. According to Melda MLoudnessAnalyzer it peaks at -11.3 dBFS before the fader, which is set to -6dB, and when the output runs into another track with the fader at 0dB this peaks as expected at -17.3dB.

The Live meters show a -17.3dB in the pink noise source and in the track taking this signal.

When I try with -3dB it's the expected difference in values.

If I enter "-1.65" on the fader this is rounded to -1.6. The results are that Live's meters peak at -12.9. MLoudnessAnalyzer displays -11.3 dBFS in the track as before and -12.9 in the track taking its signal, a difference of 1.6dB. Clearly the fader resolution is 0.1dB and it isn't lying about this.

Nothing of this indicates 1+1 isn't 2.
Old 5th September 2018
  #412
Lives for gear
Quote:
Originally Posted by lllubi View Post
I did hear more stable audio interfaces on Windows 10 with multi-core disabled in Live 9
So you think disabling multi-core is something without other side-effects, do you?

Quote:
Originally Posted by lllubi View Post
There are differences between Live 10 and 10.0.1 no matter how old a system is. The Ableton forum and the net is full of complaints.
You seem to assume these are all Live users or that the overwhelming majority simply are suffering in silence? 10.0.1? We're at 10.0.3 now.

How long have you been a Live user? How long have you been following audio software discussions on internet? You seem to make far, I mean very far, too much out of some anecdotes. Where is your analysis on what you read?
Old 5th September 2018
  #413
Lives for gear
 
stinkyfingers's Avatar
 

Quote:
Originally Posted by Mikael B View Post
Nothing of this indicates 1+1 isn't 2.
it has been discussed/shown, but Live is rounding/truncating fader values.
what you see is not always what you get.
and you can't make accurate measurements with a meter that is rounding.
Old 5th September 2018
  #414
Gear Maniac
Quote:
Originally Posted by lllubi View Post
Don´t understand why Ableton 10 doesn´t offer multicore to be disabled anymore. Can you change a software which was not optimized for hyperthreading to deliver the same latencies with hyperthreading on constantly
without a price in an update?

From my understanding not really. Since many users complain about Live 10 being unusable for them.
This doesn't have anything to do with multicore or single core. Live 10 has serious bugs. There are a lot of CPU/GPU related issues being reported in the beta on both Windows and Mac. Multicore isn't the issue, Ableton rushing out an immature version of 10 is. (This is also not limited to Ableton whatsoever. It happens with many software developers. As the saying goes, all software has bugs... Live 10's are unfortunately particularly bad.)
Old 5th September 2018
  #415
Gear Maniac
Quote:
Originally Posted by chrisso View Post
Differences between 10 and 10.0.1?

The net is full of complainers. If I saw artists of note complaining about 10.0.1 I would sit up and take note.
There's an old saying - a bad workman blames his tools.
If you look around, all you'll see is full time artists making great music using any recent iteration of Ableton Live (and Logic and Pro Tools), including 10 and 10.0.1.
Which brings me back to my original point from pages ago.
People who have to, just get on and make music with a positive attitude.
Other people get bogged down by testing and finding fault in their music tools.
Hold the phone. There actually could be some truth to them hearing something different in real time, as there is a major documented bug introduced in 10.0.1 that can cause random plugin parameter changes. The bug seems to happen on playback, but Ableton hasn't fully confirmed whether it impacts export or not. Many users including myself have experienced it and it's made me have to work in 9 until it's fixed...

The point is is that it is possible that what they're hearing in real time is, in fact, different than what is being exported.

That said, for people poised to pile on, this is a bug. That's how it goes with software... It's a brutal one, but it will be fixed...
Old 5th September 2018
  #416
Lives for gear
Quote:
Originally Posted by stinkyfingers View Post
it has been discussed/shown, but Live is rounding/truncating fader values.
what you see is not always what you get.
and you can't make accurate measurements with a meter that is rounding.
Yes, you can. By using another meter like for example Melda MLoudnessAnalyzer. Live is rounding fader values, not truncating and my little test clearly shows the Live meters are in agreement with third party meters. Also, the Live meters aren't rounding (the decimals are of course two) so is not lying! How can you conflate value entering with metering?

Don't try to stack this deck and mislead. The burden of proof is on you!

Last edited by Mikael B; 5th September 2018 at 11:28 AM..
Old 5th September 2018
  #417
Lives for gear
Quote:
Originally Posted by zedsdeadbaby View Post
Many users
There it is again. Define many. Compared to what? Zero?
Old 5th September 2018
  #418
Gear Maniac
Hadn't checked the beta in a few days, looks like they released 10.0.3 yesterday...
Anyone who's been experiencing issues should read the release notes, there are an absolute ton of bugfixes....

10.0.3 Release Notes
3
Share
Old 5th September 2018
  #419
Lives for gear
 
stinkyfingers's Avatar
 

Quote:
Originally Posted by Mikael B View Post
Don't try to stack this deck and mislead. The burden of proof is on you!
No idea what you’re talking about.
Old 5th September 2018
  #420
Lives for gear
Quote:
Originally Posted by stinkyfingers View Post
No idea what you’re talking about.
Your recent post. I embrace not accepting Live as it appears, but I expect statements illustrating a purported problem to be coherent. Your statements are incoherent.

Specifically that Live rounds fader values to 0.1 increments have nothing to do with metering rounding to 0.01 increments. The meters show the actual level in dBFS with two decimals. In other words the meters aren't lying.

**note I've changed my view on this. See later posts**

That you can't add for example 1.56dB or 3.78dB**., negative or positive, to the fader is a separate issue. 0.1 increments is an acceptable resolution I think even though I'd prefer meters and faders to have the same resolution.

If I'm missing something here I apologize and expect you to tell what you meant.

** You can use Utility for such two decimals level changes. Or the Push for three decimals.

Last edited by Mikael B; 5th September 2018 at 09:00 PM.. Reason: Update
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump