SOUND ENGINE BUGS ???

[zee verkawound]

If you guys want Live to become better, there is a very simple thing you can do:
Find cases that can be reproduced and report those cases here or send a mail to bugs@ableton.com Then there is a chance that they can be fixed quickly.

This is a complete fallacy. It's pretty obvious that this IS the dept. that needs a MAJOR overhaul at Ableton. Sorry, just because it's the way things are supposed to work, doesn't mean it is working.

How about just simply allowing me to get my hard earned money's worth from Suite 8 by TRULY dedicating yourselves to FIXING Live 8 prior to releasing yet another public tested version? So, is Ableton going to release yet one more version of their program BEFORE the last version is even remotely close to being ready?

Ableton REALLY needs to get their poop in a group with respect to professional responsibility. One version after another WAY TOO quick prior to stability in previous versions means that Ableton owe their customers BIG TIME.

[agent314]

I found this funny in light of the thread at hand:

From here

Bangalter also turns to Ableton when he's looking for sonic warmth. "One of the things I prefer in Live is the proprietary Ableton effects plug-ins. They are amongst the warmest software effects. Sometimes I will import a Pro Tools track in Live, just to bounce it down with Ableton effects, to make it sound warmer and crisper, or dirty — I love dirt. I'm not a big ReWire fan, so I do it the old-school way, importing and exporting the sound files."

While Daft Punk record analog synths for thick melodies and lead lines, they're also fans of Ableton's built-in synthesizer. "Operator is one of the best soft synths out there, and I hate soft synths in general," says Bangalter. "It's great because it's got its own thing, it's not trying to emulate this or that." He adds: "Also, plug-in delay compensation in milliseconds is great. Live is getting better and better with each version."

[mr.ergonomics]

people seriously? burning witches? you wrap a big sheet of bullshit around a small core of truth (PDC could be better and it has been addressed). I really didn't want to drop names here before... but do you know a funny thing, cubase 5 DOESN'T compensate latency for it's automation too last time I checked it. I tested it some time ago when this problem in live was discovered.

and just this dumb "pro" arguments, being popular doesn't mean you have a clou. making good music is absolutely not automatically related to sound quality wisdom. and you know what rumours can do, right?

don't get me wrong, I would like to see improvements in the PDC department etc. too, but stay on the rational side...

ps: if you want to do something sensible try to reproduce this "ableton sound" and post it when you have something. really currious to test the scenario the OP suggested later.

[ollyb303]

Try climbing out from up your own arse mate.

Wow! That really brought a lot to the table. Please do come back a grace us once more with your infinite wisdome!!!

Needed to be said, in reply to a post suggesting that I'm a deaf noob for whom sound quality is not at all important, simply because I find Live's sound engine to be adequate for my needs. FWIW I have good hearing, I have been producing electronic music for over 15 years (with Live since version 3) and sound quality is very important to me.

Now if you'll excuse me, I'm going to leave you to your complaining and go and make some music with my perfectly acceptable DAW.

[catsandwich]

Sweet Jehu. Vague complaints won't get you anywhere. Isolate a case where the sound gets worse, save the project, and report a bug.

I can promise the Abes don't have time to follow red herrings. Give them a demonstrable case where the audio quality suffers due to an improper delay somewhere and they will likely fix it. If they don't fix it, there's either a fundamental architecture issue, or your bug isn't common enough to apply the engineering resources.

[dom]

Hi Guys!

And thanks, catsandwich. In a nutshell, this is more or less it.
But from time to time it may make sense to invest a view minutes and restate the usual...

Here we go!

Let's try to keep the thread on topic and also try not to get carried away too much emotionally.
The subject is "sound engine bugs" and therefore we should approach it from a pure technical point of view.
In his first post in this thread rcpunker quoted a mail 3phase posted on another forum with two technical observations/statements he made.
Those made him wonder if they are caused by bugs in the sound engine.
Let's get to it!

i was bouncing from one track to the other from within a group..solos on the group channels..

and this is the result of the phase cancelation test of an ableton internal bounce:

http://www.3phase.de/32BitProblem/non%20cancel.mp3

is this for others reproduceable?.. a bassdrum internally bounced should cancel perfectly... only a utility plug set to the left channel involved.

Bouncing and phase canceling a clip that is not altered otherwise should cancel the signal out completely.
If it does not, the files are simply not the same or something undesired happens during the mixing process.

I just tried bouncing and phase canceling again and the result was silence.
This means, the reason might be a wrong setting in the test set (fades, plugins, timing differences) or it might be a bug that only comes up under certain circumstanes.
Regardless of which reason it is, we need more information before we can discuss this further.

In order to give us (and others on the forum) the chance to comment on that technically, we need an exact description of the test and setup, or better, the project file. If there's some kind of error introduced by settings made by the user we will find them this way, and if it is not related to settings but really pointing to a bug, we will find it, too. No need to rant, as soon as we or anybody else can reproduce it at all, let's just find the reason and eliminate it.


The next one is...

Another bug i came along..the brown henke dither bug..

when you change the tempo around..best with the automation draw in the masterchannel... and than type in the original tempo again in the main tempo window.. because you never get back to the original temo by drawing an automaion line..

after going back to original tempo all tracks that was in the so called transparent setting in ableton live..where samplerate conversion is supposed to be off aslong the clip tempo nd projekt tempo match..

All this clips will have samplerate conversion engaged than..

you need to switch warping off and on again on each clip.. than it returns to the transparent setting..

is this an user exclusive bug as ableton wants to make me belive or are others experiancing the same?

its easy to test..when you switch warping on and of on original speed there shouldnt be a difference...

as there shouldnt be a sound difference between 32 bit float daws.. but theorie and praxis... not allways the same thing

In this description he is confusing a few terms and functions, but overall i guess we all get the point.
The general situation is, that besides when using the Complex and Complex pro warp modes, warping will be neutral as long as the clip is running at its original speed.
Regardless if you just never change the tempo, or if you change it and set it back to the original speed, just as explained in the manual:

"Unstretched Beats/Tones/Texture/Re-Pitch Warping

If the tempo of a Clip is the same as the tempo of the Set, that clip will play back unstretched.
In this case, if the Warp mode of the Clip is set to Beats, Tones, Texture or Re-Pitch (but
not Complex), playback will be neutral. Any Warping caused by changing the Set's tempo
is non-permanent, and audio that plays back unwarped at a given tempo will always play
back unwarped at that tempo, even if the tempo is changed and then changed back. For
example, if you've recorded some tracks at 120 BPM, but then decide you'd like to slow
the tempo down to record a particularly difcult solo passage, the original tracks will play
back neutrally again after returning the tempo to 120 BPM. Only the recording made at
the slower tempo will be stretched. Please note that the Global Groove functionality works
by modifying the positions of Warp markers. This means that playback of audio clips with
Global Groove applied will be non-neutral even at the original tempo.
The neutrality of unstretched clip playback is veried by performing cancellation tests on
rendered output."

I just tried it over here (not for the first time ;-)) and i can't reproduce a non-neutral behavior after setting it back to the original tempo.
I encourage everyone to do the same as in an ideal world such threads would be free from rants and just commented by people sharing their results of testing what the first post suggested instead of a heated up discussion.

So, let's get technical!
As always, if you find something in Live you don't understand, regard as a flaw, a bug or whatever - make it as reproducible as possible and send it over to support@ableton.com so that we can clear things up for you or fix bugs that might be found that way.
Also feel free to share such tests in the forum to get as many opinions and test scenarios as possible, while keeping it as technical and rant free as possible. This speeds up things a lot and makes the forum more enjoyable for all.

Cheers,
Dom

[Jesus Christ]

Give them a demonstrable case where the audio quality suffers due to an improper delay somewhere and they will likely tell you to workaround it. If they don't have a workaround, there's either a fundamental architecture issue, or your bug isn't common enough to apply the engineering resources.

fixed.

but one more thing, and correct me if I'm wrong but the bug either exists or it doesn't. It's extremely rare that, all things being the same, an operational bug would exist for one user but not another. That a bunch of novices and noodlers aren't bothered by it and thus don't report/complain about it doesn't mean the bug is non-existent. But I think I get the jist of what you're saying in that respect.

[merges]

Software is complicated. Some bugs truly are idiosyncratic and dependent upon the particular setup.

This is especially true when we get to a vague term like "sound engine bugs." An aural artifact may be a bug, or it may be the result of the confluence of specific factors in one's audio setup.

The above post about bugs not being widespread (or perhaps serious) enough to warrant the application of engineering resources to a fix, is a valid post. But it's also worth noting that the identification of a true "bug" is notoriously challenging in complex software, and particularly so in this world.

So, yes: Complaints and vague descriptions are actually unhelpful. Specific, carefully detailed, reproducible situations are helpful.

[Jesus Christ]

Specific, carefully detailed, reproducible situations are helpful.

in an ideal world. but there's SCORES of people, myself included, who have provided everything required to isolate bugs throughout the years and even when acknowledged by Ableton that they've spotted it - it can take years to fix, sometimes not fixed at all until the next commercial release.

So, sorry, but this "we're not getting detailed enough bug reports" line is kind of BS, although undoubtedly true in some cases I'm sure.

[merges]

in an ideal world. but there's SCORES of people, myself included, who have provided everything required to isolate bugs throughout the years and even when acknowledged by Ableton that they've spotted it - it can take years to fix, sometimes not fixed at all until the next commercial release.

Even if those scores of people have provided detailed, specific reports, and even if those situations are reproducible, the "bug" may be idiosyncratic; it may only be reproducible in a certain state, for example.

I'm not apologizing for any specific instance or bug. I am pointing out that finding and fixing bugs in software is more complicated than most people realize. From the outside, everything looks easier and sounds simpler than it really is on the inside.

After all, it's easy to make great music, right?

[mr.ergonomics]

i was bouncing from one track to the other from within a group..solos on the group channels..

and this is the result of the phase cancelation test of an ableton internal bounce:

http://www.3phase.de/32BitProblem/non%20cancel.mp3

is this for others reproduceable?.. a bassdrum internally bounced should cancel perfectly... only a utility plug set to the left channel involved.

however.
This shows somehow that strange things can happen within the live workstation.

For many people live sound in the the mix is inferior and muffled..other state that there are no problems..
Is it possible that this is to hardware platform related bugs? or do some people just have the better listening while others just dont get that there is something missing?


Another bug i came along..the brown henke dither bug..

when you change the tempo around..best with the automation draw in the masterchannel... and than type in the original tempo again in the main tempo window.. because you never get back to the original temo by drawing an automaion line..

after going back to original tempo all tracks that was in the so called transparent setting in ableton live..where samplerate conversion is supposed to be off aslong the clip tempo nd projekt tempo match..

All this clips will have samplerate conversion engaged than..

you need to switch warping off and on again on each clip.. than it returns to the transparent setting..

is this an user exclusive bug as ableton wants to make me belive or are others experiancing the same?

its easy to test..when you switch warping on and of on original speed there shouldnt be a difference...

as there shouldnt be a sound difference between 32 bit float daws.. but theorie and praxis... not allways the same thing

1.)
cloud not reproduce the first one, but maybe I missunderstand your setting? I used a project from me to get a "real" test setting.
I created a group with two tracks in it and one audio track outside with the group as input. then I changed the group to solo and recorded. I only tested it in the session view though...
I have one idea, was it a midi track in your group?

maybe you could upload a stripped down project to demonstrate?


2.) could reproduce it! but it's I think it's obvious why. the master tempo only has two(or tree?) decimal places. when you manually change the master tempo to, lets say 120,00, the real value is something like 120.00002. when I change the master tempo again to the wanted value by keyboard (typing 120) everything cancels out perfectly again.

but I totally agree, there should be an indicator that shows things like that. that said, I think it's relatively easy to prevent this, as long as you don't do master tempo automation. nevertheless, it would be lovely to type exact values via keyboard for automation points too!

edit: ups wrong post...

[Palmer Eldritch]

... 2 pages of ranting and not answering to the question.
Even a simple question from new users ...

[ikeaboy]

btw: if you are pro you learned how to work with the flaws of your equipment since all your tools have flaws. Muddy sound due to PDC??? You need to explain me this. What is "muddy" for you and how does it relate to which amount of delay from where exactly? Unless you can explain this to me, I have a hard time taking you serious. Digital audio is complex, but it follows quite straight concepts.

Call me a fanboy but Kaappoww!!

[ikeaboy]

Try climbing out from up your own arse mate.

Wow! That really brought a lot to the table. Please do come back a grace us once more with your infinite wisdome!!!

Needed to be said, in reply to a post suggesting that I'm a deaf noob for whom sound quality is not at all important, simply because I find Live's sound engine to be adequate for my needs. FWIW I have good hearing, I have been producing electronic music for over 15 years (with Live since version 3) and sound quality is very important to me.

Now if you'll excuse me, I'm going to leave you to your complaining and go and make some music with my perfectly acceptable DAW.

This is turning into my favorite thread ever

[ollyb303]