Live's BIGGEST PROBLEM or why LIVE isn't a prof. audio soft

MM · Post by MM » Sun Sep 22, 2002 9:31 pm

Hello guys,

I really love LIVE !! It's the most innovative software long time ago !!

but there is just one thing that makes me really sad...

it's the sound quality ...

if you put Samples in Live they get a "live touch" that means
that they sound different...

ok, I think it's not easy to imagine (talking about sounds

...)
so i did 2 audio examples for you...

I used 3 sounds , all samples of my 808... bd, clap, cl hh..

I made a short loop in Ableton Live and the same in Fruityloops
using the same 3 samples and exported every loop to a wave file..
I just normalised the wave files and that's it.... (no compression,
eq, or something else...)

here you can download these 2 files:

http://people.freenet.de/fruitylive/808fruit.wav

http://people.freenet.de/fruitylive/808live.wav

when you use a headphone you can hear the differences much better...

I think the one did with live doesn't really sound as good as
the one did with Fruity..

the live one sounds not as "brilliant" and "present" as the fruity-loop...

just listen to the attack and the bass part of the 808 kick
and especially to the decay of the clap ...

the loop made with fruity sounds to nearly to 100% as the
single-samples do... there I can't hear any difference....

so what's the problem with ableton ???

is it thig, which can be fixed ??

is there a kind of in build data compression ????

It would be very fine, if one of the developer guys
find the time to answer...

I'm really interested....

so far...

cheers ---

Matthias

blaknboize · Post by **blaknboize** » Sun Sep 22, 2002 10:37 pm

The reason you're hearing such a difference is because the loop made with FruityLoops is louder. Normalizing simply brings the peaks to digital maximum, but by checking the average (RMS) level, you will see that the FruityLoops one is on average about 2dB louder than the Live one. If you used a VU meter (I use PSP's 'Vintage Meter') you would see the difference instantly. And louder sounds "better" to most people.

Post your files again WITHOUT normalizing so I can test and match the levels.

MM · Post by MM » Mon Sep 23, 2002 7:32 am

ok, I will post new files...

but the when the sounds are loaded into live they sound very different to the original samples... with fruityloops there's no problem..

M

Guest · Post by **Guest** » Mon Sep 23, 2002 10:06 am

Are you suggesting Fruity is pro software?

MM · Post by MM » Mon Sep 23, 2002 10:34 am

not really, but it sounds better than live... just listen !!!

blaknoize · Post by **blaknoize** » Mon Sep 23, 2002 11:41 am

Are you creating a loop in Live, or reprogramming the beat using individual samples?

MM · Post by MM » Mon Sep 23, 2002 11:53 am

hi,

i created a loop in live the same loop in fruity
using the same samples in each program...

MM

stereo · Post by **stereo** » Mon Sep 23, 2002 12:45 pm

i also recognized the same like matthias - for sure every instrument e.g. synth has its´ special sound ... but playback of samples should work 1:1.
using g4 / motu ... so i don´t think that this is a prob of soundcard.
re.mm: session ?

MM · Post by MM » Mon Sep 23, 2002 1:11 pm

it can't depend on the soundcard,
because the audiofile was rendered directly in live...

tonight I will make the files again without normalizing....

... so watch out...

MM · Post by MM » Mon Sep 23, 2002 4:50 pm

so I did new files without a hihat so you can better hear the difference...

here are the links:

http://people.freenet.de/fruitylive/808fruit.wav

http://people.freenet.de/fruitylive/808live.wav

nothing is normalized it's just the rendered loop of each program...
all levels in both programs were set to 100% or 0db ...

but just listen yourself....

SongCarver · Post by **SongCarver** » Mon Sep 23, 2002 5:09 pm

what you have there MM are two files which show the differences in the settings of the sample engines.

To me, it sounds like the fruityloops one has a longer decay, or a fade out, applied to the end of each audio slice, letting the kick drum boom longer (which makes the next kik clip, or maybe that is a much shorter 'open' gate) and leaves more of the decay, the reverb on the hand clap.

I don't know if your kik orginally has the small clipping sound at the begginning or if that is a byproduct of the fruity code.

The reason is sounds better is

a) The beggining of each fruity sound has a 'gate' which is triggered a little later than in live, giving a punchier sound. (The 'gate' is my way of explaining this type of course granular reconstruction)

B) The end of the gate has a longer decaying amplitude, giving a less dry sound than Live. (unless you have left some reverb on by mistake)

Is the loop playing back at exactly the tempo it was created in? try that to see how much difference there is between the two engines.

MM · Post by MM » Mon Sep 23, 2002 5:20 pm

hi,

first of all it's not a loop, I loaded into live... I arranged the loop using these two sounds... kick and clap...

here you can dl these two samples :

http://people.freenet.de/fruitylive/808Bt7d5.wav
http://people.freenet.de/fruitylive/808-Clap.wav

why does live manipulate the sounds much ??

in other programs like reason or logic, theres no problem..
it just sounds like the single sounds do...

regards,
M

kleine · Post by **kleine** » Mon Sep 23, 2002 9:44 pm

well...
Is Live a pro application ? as you can guess from it´s name it´s made for Live performances - and i heard amazing concerts from several people
using it - even with "good" and "fat" sound. If you compare it´s current
audio quality to other prg´s you´ll notify that you´ll a a better quality out
of protools/logic etc. why ? - because the files are "just" played straight
as they are. Live uses realtime-timestretching for the files with the
result that a) it´s incredible easy to play different source material in one speed and b) you have certain artefacts in the files resulting in this procession. Otherwise it wouldn´t be named Live...
However, If Live has the possibility to turn of it´s timestretch engine, you´ll
have a the same audio quality as the original file and/also if the timestrech
algorhythms will be improved, you´ll have a even better audio quality in this mode. so what ? i´d say: enjoy the possibility Live has right nowto "jam" with audio files and give the developers a bit time to implement
high level algorhythms....

...just my 2¢

dirtystudios · Post by **dirtystudios** » Mon Sep 23, 2002 10:21 pm

here's why the live loop doesn't sound as good. you say you imported the samples into live, oui? since live is meant to be a loop player, it assumes that anything you throw at it is gonna be a loop. so if you load up some individual hits that are not trimmed to be exactly an eighth note at 120 bpm for example, live will try to match it to the current tempo, so you get artifacts. if and when a version of live is released with the timestretch bypass option, then you can use it to program your drums. untill then, either use another app to make you loops, or, if you must use live to program drums, try adjusting the 'transients' setting in the 'warp' panel of the clip view to bar or 1/2.

of course, live wasn't built to sequence beats from scratch. it was built to sequence loopsthat you've already made. protools may be the finest audio recording program in the world (i don't want to start an app war here, i'm just using it as an example), but if i loaded it up on my tibook and try to play a gig with it, i'd think it was crap, because i would be using it for something it wasn't designed to do. i think a fairer (is that a word?) test of live's sound quality vs. fruityloops, would be to create that loop in a different program and then load it into fruityloops and live. though live might has a few minor oddities (every last one of which i'm sure will be fixed in the next release because gerhard is on the case ;) ), when you're talking loops and jamming on the fly (which is that live is supposed to do), i'm sure that you will find live to be a far superior program.

i think that was more like a nickel...

k

blaknoize · Post by **blaknoize** » Tue Sep 24, 2002 12:02 am

Okay - after running a whole bunch of tests and carefully matching levels, I can conclude that there is very little difference between the sound quality of Live and other programs. I tested Live on the Mac versus FruityLoops & Acid on the PC and Logic and ProTools on the Mac.

To start, you have to understand how the different programs implement their pan laws at their mixers. Carefully matching levels and checking with meters is the _only_ way you can be sure of what you are hearing. With Live, a full-scale (0dBFS) signal is attenuated by 6dB when panned center and returns to full-scale at extreme left or right. In FruityLoops, the same signal is only attenuated by 3dB (with channel faders full up). In practice, what this means is that unless you match your levels, the same signal will sound twice as loud in FruityLoops as in Live and thus "better". As a matter of interest, Logic does not attenuate at all, while ProTools attenuates by 2.5dB. Once I matched levels, there was very little difference between the Live and Fruity versions.

However, there _was_ a slight difference. This is entirely because Live does not offer a "One Shot" mode like Acid. Every sample is treated as a loop. If you're using single hits to program beats, you need to do the following:

1) Turn off the "fade" option in Sample Settings (this accounts for the differences in attack on the kick in your two examples). Turn on "Hi Q".

2) Set the transient resolution to 1 bar.

You need to save this setting on a per-sample basis or do this for every clip.

As I and many others create groves by dropping individual samples on a grid as well as using loops, Live really, really needs a "one shot" mode. Even doing the above I could still hear a slight difference in the sound and decay of the snare.

If you wish to try for yourself using the samples provided, put the kick and sanre channels at +3.0dB in Live in order to match the level of FruityLoops.

P.S. - I'm a "pro". I use FruityLoops. That makes it a "pro" app in my book. It can do quite a few things other "pro" apps can't. If they dropped the silly name and called it "SampleStudio Pro X 3000" or something and charged $500, everyone would want it.