Ableton Forum

....so it's a bug with midi export function...
that means playback/rendering audio should still be fine... no?
so you still think logic sounds 'rounder' ?

3dot... wrote:so you still think logic sounds 'rounder' ?

I'm judging that based on sequences that were generated from an algorithmic composition environment in Ruby where I specifically set rhythms to Pi and Golden Ratio, etc..

Then I import the same file into both Logic and Live, and Logic sounds better.

I'll should print the audio from both and see where the difference is.

It may just be the way that Logic handles velocity compared to Live.

Tom, did you check to see if Live's hi res midi file import was ok?

3dot... wrote:....so it's a bug with midi export function...
?

if the OP's sequences were created in an external environment and then imported to Live it seems more likely to be an issue with the Live midi import. Or perhaps both.

My own tests were created, exported and imported in Live, so it's hard to tell whether the issue lies in the export or import.

my brain hurts.

kb420 wrote:Tom, did you check to see if Live's hi res midi file import was ok?

No, I'm working on that.. I hope so, since I'm generating a lot of MIDI algorithmically and bring it up in Live..
I would like to know just how Live is handling it, since sometimes my MIDI tracks sound better, sometimes they don't, and I would like to know
if a lower-resolution has anything to do with it, so that I can fix it.
I'd rather work fully in Live until the mix, but if my work can sound better by rendering the MIDI to audio first in Logic, then I'd take that route.
Of course I'd rather not do that unless I have to, and I don't have the time to make the same exact track in both systems to compare them.
There are too many other variables anyways.

It may just be the case that some stuff sounds better in certain configurations, etc..
for example, I used the same low-pitched drum sounds in both Battery and Live's Drum Racks, and Live sounds like it punching a lot more, and is overall much better sounding. Now that may be the fact in general, or it may be based on that I'm on a Mac with a Apogee duet interface.. I don't know...

I'm going to take some time off from this and run some more tests within the next few days..
I'll post any more findings.. and I hope that it's just the export function..

thanks for the help..

...surely enough , Ableton is gonna step in any minute now...
been trying to brain 'Rob' about this...but it's not working...the channels are blocked..

tomjordan wrote: I'm judging that based on sequences that were generated from an algorithmic composition environment in Ruby where I specifically set rhythms to Pi and Golden Ratio, etc..

neat..
how far are you going with the Pi thing... and how do you scale it?

I like making music with this soft. I don't care if it can play faster than Franck Gambale, joe satriani or their peers because I hate people who think speed is more musical. Sure midi resolution is important and I almost never quantize my stuff but by the way, who can really hear if some music was recorded at 480 or 960 ppqn ?

It looks like the import MIDI file works as expected, thankfully !
I imported both the original MIDI track at 960 ppqn, and the one that was exported out of LIve.
I set the tempo to 20 bpm and zoomed in on both tracks..
sure enough, there original track seems intact, and looks and sounds much different from the track that was exported out of Live.
So, the issue only seems to be with the MIDI file export.
Good news !

3dot... wrote:

tomjordan wrote: I'm judging that based on sequences that were generated from an algorithmic composition environment in Ruby where I specifically set rhythms to Pi and Golden Ratio, etc..

neat..
how far are you going with the Pi thing... and how do you scale it?

I like to use values of Pi and the Golden Ratio, etc, that are close to 16th-note divisions, but different enough to give it a nice groove.
Such as: instead of having a dotted-eighth note at 3 16th notes, I'll have it last 3.14 16th notes,
and combine it with another dotted-8th at 2.86 16th notes, so that when used together it equals 6 16th notes.
If you then use a regular 8th note with each pair of modified dotted-8ths,
you can do any permutation such as:

pi-3-2
pi-2-3
2-pi-3
3-pi-2
2-3-pi
3-2-pi

where the "3" is actually 2.86, to "make up" for the 3.14 pi value, since 3.14 + 2.86 = 6, and 3.14 + 2.86 + 2 = 8.
This may sound like a lot of unnecessary technicality, but it does sound awesome when using it for certain accents and parts..
it sounds really organic..

It's easy for me to generate these kind of timings now because I can just specify the values directly in Ruby and it will generate a MIDI file.
I used to figure out the tick values and input them by hand into Logic. The new way is much faster..

tomjordan wrote:So, the issue only seems to be with the MIDI file export.
Good news !

If that's truly the case now, then maybe you should change the title of this thread to, "Live's Midi export resolution is only 96 ppqn."

Regarding live sounding less tight then logic.. This is far fetched but just in case: did you check delay compensation?

Im switching it on and of all the time. If its on, midi keyboard latency is late but sequencer timing is good, and if its of lives midi timing is sloppy but controller latency is tight. So when recording midi from instruments I set it of, and when its done, I switch it back on.

this is a feature not a bug

d-track wrote:this is a feature not a bug

Yes of course, im just asking if he took it into account!