Importing a this midi gives loss a of subtilility of the song

If you want to import as recorded set the interpretation to static. You can do this for individual 'tracks', so you could keep the 'vocal' line as original then let synfire interpret other tracs. Or you can set everything to static, then if you atch the sounds and effects it should sound the same as the original.

I thought you wanted the imported midi to sound like the original? If that's so static is the way to go. It's also faster to import this way, but you can change that at any time after the import.

you are free to change the recognition of each instrument and try different things, keep as much of the original as you want, use the harmoniser to come up with a new chord progression which will affect any tracks that are no longer static, alter the strong/weak settings of those tracks. Break the song up into smaller containers, move them around, combine with original content played on your midi keyboard, add figures from a library or another imported song.. The possibilities are endless.

But by doing all the above it will no longer sound like the original midi source.

When we import tracks as 'static'. and then later process the " figure recognition" isn't that doing the same thing as doing figure recognition on import?  Being aware of course that if SFP figures it's a bass from the GM patch and range of notes it's processing will be different than a polyphonic chord track. ..

So if we imported a track static, and then individually processed each track with the proper recognition, wouldn't that be the same?  Or does SFP do more on input processing?  

I've been importing midi files from Logic as static, then dragging each instruement into a library (or could have imported MIDI file a 2nd time into a Library).

That way I have the original notes and new modified notes.. Sometimes however SFP makes some false assumption of what chord is being played.  Hence when it is interpreted, there can be clashing of notes..  I understand if we found the proper chord and dropped it in that 1/4 or 1/8th note slot, then Synfire could interpret the chord more correctly.

However even in the music world, there is some difference of opinion as to what exactly a chord was.  I was taught MANY MANY years ago.  That Diminished chords always had four notes.. (according to my music teacher)  Then there is something like a C2 or a C9th minus the 7th.  Putting the D note an octave above does somewhat change the character of the chord.  

So if we imported a track static, and then individually processed each track with the proper recognition, wouldn't that be the same?  Or does SFP do more on input processing? 

On the MIDI input SFP does more than just the figure recognition. For example it also calculates the harmonic progression and places the result at the Harmony parameter. At the MIDI import you can also choose which tracks of the MIDI file shall be included at these harmonic calculations, whether modulations shall be recognized or not, and so on. I surmise that the calculated harmonic progression also has an influence at the figure recognition process. I may be wrong, but I would say that importing tracks as static and then process each track individually does not necessarily give the same results as the figure recognition at the MIDI import. Maybe Andre can give us more insight.

When we import tracks as 'static'. and then later process the " figure recognition" isn't that doing the same thing as doing figure recognition on import?

Do not import as "Static" if you want do do figure recognition, because that would give you the "Drums & Effects" Interpretation and possibly other settings like instrument category would no longer match.

Instead simply turn off the "Create Figures" switch! Much recommended, if you want to save import time.

Otherwise, and in any case, after an import, you can redo the recogniton step for each instrument as many times you want.

I surmise that the calculated harmonic progression also has an influence at the figure recognition process.

Yes, a lot.

I would say that importing tracks as static and then process each track individually does not necessarily give the same results as the figure recognition at the MIDI import.

No, because the Harmony parameter is already there and it is being used for recognition (unless you check the "Independent" switch).

How to improve the quality of the segments in a measure?

Make sure small melodic snippets are grouped in a segment each. If there are blocked chords, use the chord abstraction switch. Try different settings and redo the recognition.

Sorry for the confusion, I only suggested using static as I thought janamdo was complaining that the result of his import did not sound the same as the source material. If I've misinterpreted what he said, please disregard the advice to use static imports.

Importing all tracks as static will give the same results as playing the song in Cubase, provided you're using the same patches..  if you want the stong to stay real faithful, - static import If you wanted to add tracks to modify some of the tracks, you could start with this, and add, drop modify individual tracks..  Or re-import just one instruement as a phrase and let SFP do it's recognition on it..

Also be aware of how SFP interprets the song chords.  In some instances they can be prettty off.  I printed out the chord in Logic.. Imported into SFP.  Most were fine, some pretty off, I corrected just those chords and it made a world of difference.  If you have 4 bars of G, you can play around with inversions, variations, descending bass etc.. You can make a plain arrangement a lot more slick by reharmonizing some of the chords. 

Also be aware of how SFP interprets the song chords.  In some instances they can be prettty off.  I printed out the chord in Logic.. Imported into SFP.  Most were fine, some pretty off, I corrected just those chords and it made a world of difference.

I'd like to know more about this, as a few times I've imported figures that just didn't sound right and I was unsure how to correct them.

Is the 'Confirmation' in the import dialog the tool? I assumed it only asked for the root key. Does it ask you to confirm the chord changes?

I also wondered what happens if there is a modulation in the SMF you import?

Mark, any 'tips/ tricks' about identifying and correcting inaccurate chord recognition? Or types of chord changes that are susceptible to these 'errors?'

I put 'errors' in parentheses, because more complex chords can be ambiguous and there could be more than one 'correct solution.'

I just looked at the chord progression SFP came up with.  I had created an interesting chord progression which couldn't be accomodated by any pallette, but the chord changes sounded good to me..

When I write a song, I'll record a track into Logic that is just the blocked out legato chord changes.. I than have Logic analyse it for display only, making sure to check 'do not transpose' on each track.  Otherwise Logic's 'Music for dummies' auto transpose takes over.  I leave it up on screen, or else just write the chord changes out by hand and compare to what SFP interpets.  Most of the time, it's right.  I'll use that to display chords in piano roll editor and score editor (which  I use a lot)

SFP had trouble with  some diminished, augmented, Am+2nd chords. Once I hand corrected them in  progression, song sounded fine. Other stuff like an F6 was intreperted as a Dm7.  I was taught the modalities when I was young, but I didn't quite grasp it, other than they sounded different.  I'm not aware of when I go from Ionian to aolian, etc.. Dorian I'm good at, cause that was my favorite. 

Interpretation is a hard subjecg, cause you can view it several possible ways,  One is that it's in one key, with a couple of exceptions, other interpretations, is song has temporarily shifted keys for one or two chods, or shifted modalities..

I was taught for instance diminished chords had 4 notes = a bass note and three notes each a minor third higher.. Yet most charts today have only three notes  in a diminished chord.  Same as with notation, there were a couple of systems, really just small things, that any musician could pick up one, once he played the song thru. = using just 'm' or 'min' or 'minor' how major 7ths were notated. etc. I think by the 70's music publishers unofficially formed a standard, which pretty much everyone follows now.

 I took some online music courses at Berklee in Boston, and have been following their conventions which are pretty standard now. 

No you can adopt SFP chord progression, just A/B it against the one from Logic, or from your printed noted,  when you play the piece you can immdediately tell if it suits your needs (not necessarily right or wrong).  Because of the latitude in chord progression analysis, SFP is ok.  Perhaps Andre will come up with some additional software ware patches, to bring in further accuracy down the road. 

I do wish I could directly import a chord progression into the chord progression, instead of going the round-a-bout way of importing as a phrase, cutting/pasting into arrangement.. 

Play with the subtleties of the chord progression can be very rewarding, when I become more comfortable with more pallettes and different modularities, I'll get closer to the music I want.. 

"I do wish I could directly import a chord progression into the chord progression, instead of going the round-a-bout way of importing as a phrase, cutting/pasting into arrangement.. "

By this you mean , importing a midifile ( with chords) and convert it to a harmony instead of a figure ?

By that I meant import a track of midi chords which I have played in Logic, and have them go right to the Progression section, rather thatn have  SFP 'calculate' the chords.  One import, SFP did away with my diminished and augmented chords.. It didn't sound right to me, so I manually put diminished and augmented chords  back into the chord progress ion window, and then it sounded all right 

BUT  at a later point, re-importing into SFP a 2nd time after working in Logic on the file.. I PREFERRED the chords that SFP chose over what I had originally written. I'm guessing  that with the intrepretation parameters I put in on both SFP takes, SFP 'smoothed' out the music, and thus changed the chord progression.  

I can only assume, that SFP looks at the whole picture of the MIDI file , all instruments, both forward and backwards in time, to come up with the chord interpretations.. So it is up to us ultimately what sounds 'right' and what sounds 'wrong'. 

Hey mark,
Haven't been able to test this but have you tried piping your chord track from logic via a midi drone or Virtual midi cable into synfire set as an input. Then go to the progression tab and turn on chord detection. Set synfire to record then play logic? Might need to play around with the sync if it gets in the way.

There might be an easier way to achieve what you want that super tonic or Andre will post....

Thanx for the idea.  I haven't had a lot of luck with drones.  mostly because I use MIDI hardware, but I do have virtual instruments..  Fiddled with IAC, can't remember what happened.. Well that will a task to do coming up this week.

I do it now, by exporting Logic chord track into instrument phrase static, then copy, paste into chord progression.  Even when you import a whole song static SFP will come up with a chord progression that can almost perfect to fairly off. Since SFP builds chord progression from what it sees in other tracks, By bypassing that and entering my own chords,, I wonder what I'm losing..

Being more familiar now,  I'm beginning to do tests that are a bit more revealing to me than in the past.