It is proved that Cubase export has better quality than FL export ?

On 2009-09-21 06:50, Upavas wrote:
Nobody wants debate here.

The sample interpolation is not only important when rendering but also when playing back. I have found that with higher sample interpolation your mix in FL does significantly better, regardless of what the output sample rate and bit depth are. It does affect generators. I was merely wondering what interpolation you used when you rendered each version.

If you want to know what I mean, take a track mix with 25 tracks, make the sample interpolation rate in FL when rendering 6, then do the same thing again with 512 samples, I think you will notice quite a difference. Why that is so I am not sure, but I am sure it is so.

On 2009-09-21 03:37, 0hz wrote:

for the comparison to be accurate u need to mix at least 10 loops (with fx and reverb and stuff), but the more loops you mix the more difference you will hear, because the difference is in the SUMMING.

In the end what audio engine does is simply add and subtract number. lets say one daw when it get a number like 10.123456789 rounds the number to 10.123, the other daw rounds the number to 10.1234, might be really insignificant at first glance but the more rounded number you add together the farther you get from the real result.
so what im trying to say is that unless 2 audio engines use the same algorithms or methods to process information the result will be always different, not necessarily bad or significant different.

On 2009-09-21 03:37, 0hz wrote:
ok ppl im not here to debate but simply to help answer a question.
if u gonna take a kick in fl and export it, then take a kick in cubase and export it then the result will be identical. as i wrote before the difference is in the summing, again in big letters this time...SUMMING.
that means that you need the audio engine to actually work, move mixer faders, inserts fx, send to reverb etc... and then export the mix through the master bus (master fader).
for the comparison to be accurate u need to mix at least 10 loops (with fx and reverb and stuff), but the more loops you mix the more difference you will hear, because the difference is in the SUMMING.

In the end what audio engine does is simply add and subtract number. lets say one daw when it get a number like 10.123456789 rounds the number to 10.123, the other daw rounds the number to 10.1234, might be really insignificant at first glance but the more rounded number you add together the farther you get from the real result.
so what im trying to say is that unless 2 audio engines use the same algorithms or methods to process information the result will be always different, not necessarily bad or significant different.

in audio terms this can mean that unless you have a super quiet listening environment, heavily treated room, accurate quality monitors that are positioned/placed accurately and most important good trained ears, you might not hear the difference or wont be able to conclude what sounds significantly better or insignificantly better/worse.

heres a quote from an article sos june 2004 : "I think it is true to say that there were a number of cases in the early days of DAWs and digital consoles where flawed programming resulted in less than perfect results when mixing large numbers of tracks."
im to lazy to find the whole article so if anyone wants to read about digital summing then google it and check this link the article might be still there http://www.soundonsound.com/Contents.php?Month=6&Year=2004

also i would like to add that comparison that i did was FL7 vs cubase 4 vs ableton 6 (im not gonna say anything about ableton 6), so FL8 or 9 might output the same exact mix as cubase 4. also the comparison was done not for my benefit or peace of mind but as a research paper for my degree in computer science / digital audio and math related course. one of the questions that was asked there "do more accurate numbers provide a better listening experience?" , for those who are interested the short answer is yes but to a certain point after that its just a waste of resources. this was a 60 page paper with test subjects and shit...

in summery if u mix 50 tracks that play simultaneously in fl and in cubase, the cubase mix will be better not only on paper but to the ears too. if u mix 25 tracks the difference will be a little less noticeable. if u mix 10 tracks, the difference might be insignificant or unnoticeable in normal listening conditions. if u mix 5, tracks might be no difference. 1/2 tracks, identical.
conclusion : the more u make ur audio engine work the more important becomes the audio engine algorithms/ quality / accuracy.

If anyone wants to test this for himself, heres a basic idea how to do it.

you need, FL, cubase and a 3rd program that will act as a reference (something like digi check or any other program that shows audio information like volume, peak, rms, phase/stereo info).
u take 10 loops (the more loops u use the more u will hear the difference) put them each on its own track, solo track move fader, stick eq/comp/ whatever dial in some settings then look at the 3rd program while fl is playing live, write down all the numbers or (peak, rms....), go to cubase put same track same plugins with the same settings etc... check is all the numbers match in the 3rd program. do this procedure with all the tracks in solo. when all the numbers match in both daw, export in the same quality and then compare. also dont forget to mix in some reverb, because the difference is usually more noticeable with low level signals like reverb tails...

there is another way of matching tracks by exporting them individually in both daws then checking if they canceling each other by reversing the phase of one track, but in the end the whole mix has to be exported live for the comparison to be accurate.

as a reference program i used a self written one that shows more accurate numbers that consumer/pro programs have no need to do.
this shit is very time consuming and its easy to make a mistake.

for those who just want a simple answer and dont want to test shit : fl7 mix sounded more dense/less space then the cubase mix.
but the difference was not noticed by 6 out of 10 people that listened to the mix in normal listening condition (not super studio). also no one heard the difference on mid size PA.

hope this actually help someone to understand stuff...

p.s : upavas stop saying "sample interpolation rates" i think u just like how it sounds when u say it. if u mean at what sample rate i did the mix then it was 44.1.
interpolation is when you converting sample rates, i didnt do any converting..
here some info on interpolation for you knock your self out http://en.wikipedia.org/wiki/Interpolation

On 2009-09-23 13:57, Uedi wrote:
Yes. Almost every daw have the same export quality options. The issue is on the playback.. the summing, audio engine.

I used Cubase SX3 and really liked the sound.
The track elementws were well definded and clear.

Currently I'm use Ableton Live because I like the workflow.
But I get many differences.. especially on DrumMachine and Impulse, on percussion.
The velocity assigments are different by default.
Sometimes the sample preview sounds a lot better than the sample itself on the arranjement view!
Don't know why!!

FL Studio 8/9 has a good sound. But in all projects it starts with Fruity Limiter on the master.
There's some enhancement on playback because of this.
I'm going try to Reaper.
I really need clarity and punch.