Skip to main content

Synfire 2.1 Coming Soon

Comments

Fri, 2023-04-14 - 16:06 Permalink

You always listen to a take to decide if you want to keep it, I guess. 

Whether there are chords, that's not always obvious (they can be hard to spot between other notes sometimes). Bass also isn't obvious very often. I've seen thousands of phrases lately and their variety is staggering. Much depends on how you want to use a phrase (bass could also be a melody, chords could be counterpoint voices, etc).

Synfire doesn't change the take or its timing.

There are only the 4 symbol types you can select.
 

Fri, 2023-04-14 - 16:10 Permalink

@ Cliff, The original library new age midi library of PG Music  comes with  a own graphical interface for listening and choosing a particular mood /style.

Now it is a pool collection in Synfire without no idea as user what the takes are standing for if you don't have the original source.

That should take some time to prepare this for the users of  Synfire and limited in use i guess.
With the original program new age midi you could more easily listen to what kind of mood you are after for. 

 

Fri, 2023-04-14 - 16:26 Permalink

 I guess there should not be copy right problems.

If you use a few snippets for demonstration purposes, probably not. If you purchase a library and process all of it for your own music, fine. But if you share an entire collection online (imported or processed in any other way), you'll get into trouble. It's other people's work. If it takes away potential sales from somebody, then it's always an infringement.

Fri, 2023-04-14 - 17:14 Permalink

Demo Synfire 2.1 is available.  Fired Kim new algorithm on my  piano_86_cm.mid file

 

 

On the left Kim Synfire V2.1 on auto

on the right manual done by the master himself  Andre a few months ago

 

Still the work Andre did on this file a few months ago manual  (split the file in lower and higher part ) is still the best.  The Kim V2.1 is still  looking strange and sounding off .  

Fri, 2023-04-14 - 20:29 Permalink

I'm going to take some time to play around with the import process and let it all sink in.

However, I'm still having a problem with getting my imports to sound the desired instrument within a library. Until I resolve this, I'm just creating new problems for myself by importing a bunch of MIDI files.

I'm in a new library and the library rack has 2 modules: Omnisphere and an Integra-7 MIDI device. I used to have a FluidR Internal Soundcase VST (or some name like that) on the library rack, but whenever I tried to select a sound on it, I was told that sound was already in use and that I must accept a replacement sound, namely a drawbar organ, which I never selected don't want to use. 

So I select Omnisphere, which I also don't really want to use for a library rack because it makes too many demands on my CPU but hey, maybe it will at least work, and I import the file I previously uploaded to this forum. But what's this? After I press play, the rack (shown on the right, under Sound) has changed from the library rack to the global rack, with the sound being Soundcase FluidR3/DrawbarOrgan. There's no way for me to select a different sound, because all of the MIDI channels shown in the dropdown menu are greyed out. It's like that Twilight Zone episode where the guy is on trial for murder and then gets executed, but wakes up the next day and is on trial all over again. I can't escape the dreaded drawbar organ.

OK, let's try the Integra-7. I select it, and the channel shown is 01 Acoustic Piano. Great...press play, and once again it changes to the global rack and...wait for it...Soundcase FluidR3/DrawbarOrgan. At this point I run screaming into the night.

But in all seriousness, it's problems like this that caused me to set Synfire aside and hardly use it for about the first 3 years after I bought a license. (Even now I cringe when I have to do anything related to device descriptions.) If there's a simple solution that I, in my not-far-advanced-from-noob state am overlooking, please enlighten me. Surely Synfire can't be nearly so obtuse as I'm sometimes finding it to be.

BTW, whenever I attempt to select Edit Library Rack or Edit Global Rack from the dropdown menus under Sound (on the right side), I get a runtime error.

PS: I love Synfire and think it's a fantastic (and constantly improving) program, but it is more than a bit frustrating at times.

Fri, 2023-04-14 - 20:47 Permalink

<However, I'm still having a problem with getting my imports to sound the desired instrument within a library. Until I resolve this, I'm just creating new problems for myself by importing a bunch of MIDI files.>

Create a new global rack(s).
Omnisphere and Integra-Midi instruments are probably not multi timbral ?, so take (make)  a device description for one sound for each.
Then assign this rack to the library. 

Note: FluidR3/ seems te be using a multi-timbral DD ( device decsription) , that is handy for the imported assigned sounds, but not handy for changing a sound   

Fri, 2023-04-14 - 21:52 Permalink

When i made this afternoon a batch import into a new empty new age piano library, Synfire assigns a global rack with soundcase fluidR3 with a honky tonky sound and multitimbral device description.
Assigning a another piano sound in fluidR3 seems to be not possible.
Therefore i added a new empty global rack and assign a wanted piano vst here. 

But i did this for one sound as test and assigning a wanted piano sound doing this for all pool figures i must probably get rid of the fluidR3 global rack
But then it cannot be used anymore in a arrangement.

Probably if shou[d assign during import the right wanted piano sound then in a new made global rack, this sound assigment will be taken over for the pool figures? 
 

Fri, 2023-04-14 - 23:21 Permalink

Thanks for all your help with this, but I'm going to set this problem aside for a while and let everything sink in.

Sat, 2023-04-15 - 00:02 Permalink

Unfortunately for me it's also trial and error , because the setup you want may just now not be described by Cognitone as a video or manual, but you'll have to check that out
 

Sat, 2023-04-15 - 00:10 Permalink

Did a quick look at a "library rack" naming convention.
Synfire comes with an included library and then there is a library rack made for it ( must check for an arrangement setup if this is the case ) ?
Different story is if you create your own library and there are no library racks with it ( but global racks) as with my batch import of the new age midis as pool.
You can check this.

Sat, 2023-04-15 - 00:53 Permalink

There is no such thing as a "library rack" as a standalone rack. 
It is called that because the global rack is used in the standalone editor for libraries.
There are only two rack types: arrangement ( for 1 arrangement ) and global ( to be used for more than 1 arrangement.) and for the library editor.
 

Sat, 2023-04-15 - 12:18 Permalink

Note that Soundcase fluid 3D is a GM device description for a imported library.
You can replace this with a Omnisphere GM soundbank if is available : how to do this , don't know yet.
Probably you do want to replace the ( boring ?) GM sounds with custom sounds of Omnisphere, , don't know this either 

Sat, 2023-04-15 - 13:12 Permalink

Fired Kim new algorithm on my  piano_86_cm.mid file

The file has many issues. It is not in sync with time signature (off by 1/4). Harmony is 1/4 off too (Harmonize again). You also need to trim away the empty measures. The sound defaults to French Horns. It has chords in the bass range that should be chords rather than bass (disable bass segments).

If you fix this, you'll get better results.

Sat, 2023-04-15 - 13:14 Permalink

@janamdo

Libraries have their own rack now. This came new with 2.0.

Sat, 2023-04-15 - 13:25 Permalink

If there's a simple solution that I, in my not-far-advanced-from-noob state am overlooking, please enlighten me. Surely Synfire can't be nearly so obtuse as I'm sometimes finding it to be.

Hold on ;-) There's indeed something still weird with library racks and sound selection on the Import dialog.

Will fix it for the next build.

Sat, 2023-04-15 - 14:34 Permalink

Thanks
<Libraries have their own rack now. This came new with 2.0.> 

Someone was looking for "library racks" , but could not find them ?
I did find them as global racks.
Can't remember the library sound setup before the new 2.0.

A dedicated tab in the soundsetup : "library racks" could be a solution for misunderstanding were the library racks are to find ? 
 

Sat, 2023-04-15 - 14:37 Permalink

One thing:

There's no way for me to select a different sound, because all of the MIDI channels shown in the dropdown menu are greyed out.

You don't select a sound program/patch from the channel menu. That menu only works for fixed-channel sounds (same sound always on the same channel). Instead use the wrench button to open the wizard/browser and select or create the desired sound there.

Always use the wizard/browser when in doubt what the popup menus might or might not do. These menus are very context-sensitive and the browser is more intuitive to use.

Good news: I just found the bug with sound selection on the import dialog.

Sat, 2023-04-15 - 16:52 Permalink

I did, as part of the import process, use the wrench button and wizard to select the sound I wanted to use. However, that sound did not in fact get used. Instead the drawbar organ sound kept being selected. That's why, for lack of any alternative, I tried to use the channel menu to select the sound I wanted.

I hope this is the bug to which you're referring. 

Thank you for your help with this.

Sat, 2023-04-15 - 17:02 Permalink

"Probably you do want to replace the ( boring ?) GM sounds with custom sounds of Omnisphere, , "

Actually, I don't mind using "boring" sounds in a library. I just need to be able to hear if the import process has resulted in symbols that are usable. A VST that is light on the CPU is preferable, and I'm assuming the Soundcase VST meets this criterion better than Omnisphere. 

Which begs the question...if I later decide to switch to a different sound for an existing library or a specific phrase, can I do this? Andre says this cannot be done via a sound's channel menu, so how would it be done?

Sat, 2023-04-15 - 17:35 Permalink

I hope this is the bug to which you're referring. 

Yes, it is. All will be fine.

If I later decide to switch to a different sound for an existing library or a specific phrase, can I do this?

Of course. There's also a wrench button on the sidebar.

Sat, 2023-04-15 - 23:53 Permalink

Thanks...this is making more sense to me now. 

But can you see how having a dropdown menu with the choices grayed out might lead the user to think something was amiss? Like...what did I do wrong to cause the menu to be grayed out? Why can't I choose from those sounds? And what do I have to do to make the menu choices available to me so I can get the sound I want?

Suggestion: Add a text box under the rack that reads something like "Use the wrench icon shown above to select a sound for your [library, or phrase, or whatever is the proper term]. Do not use the dropdown menu."

This would avoid a lot of confusion.

Sun, 2023-04-16 - 00:09 Permalink

Also, what do you think of the idea of doing multiple instances of Figure Recognition on a take, and keeping all of them in a library rather than keeping only the "best" one?

One could hear something that works (or almost works) in one's arrangement, but then try out its "cousin" to see if it fits somewhat better. For example, one phrase might use C symbols for chords, while the other might use V symbols. One melody might rely more on the horizontal scale; the other melody on the vertical scale. These differences would be the product of different FR settings.

And the closely related phrases from different FR passes might provide variety (as in "variations on a theme") if both are used at different times in the arrangement. The listener would recognize the phrases as related but different.

Sun, 2023-04-16 - 00:29 Permalink

@Janamdo I finally got around to checking out those piano MIDI file links you posted. Interesting, and I might buy one or more of them. 

I agree with Andre that posting the processed files would violate copyright. Most sets like this have a clause in their license that states you cannot distribute derivative works. Not that you can't distribute music you make based on the MIDI the files contain, but that you can't sell the derivation as a product others would use to make music. This sounds reasonable and fair to me.

BTW, Big Fish Audio sells a bundle of piano MIDI files that I've used extensively. It used to be offered as 5 separate sets for $9.52 each (and maybe it still is), but now they have all 5 sets together for $9.52, which is ridiculously cheap for ~250 high quality MIDI files. The files are short but useful, especially in a program such as Synfire, where shorter is often better. (The bundle also includes .wav files.)

Big Fish Audio - Piano Melodies Bundle - So many doors are opened to your creative productions with this fat piano bundle

Sun, 2023-04-16 - 01:41 Permalink

@janamdo How would I access the MIDI files that comprise the various PG Music piano products? It looks like those MIDI files function only within PG Music's proprietary player.

I of course would want to access the MIDI files directly (rather than through the PG player) so I could import them into a Synfire library.

Sun, 2023-04-16 - 02:43 Permalink

@janamdo I took a chance and bought the New Age Pianist set. I then did a search for "new age pianist" and found a folder by that name. The MIDI files were in the folder. Easy.

Sun, 2023-04-16 - 10:27 Permalink

@THE ARTIST FORMERLY KNOWN AS HANGDOG CAT , With the player included, you get more of an idea of what the new age piano pieces stand for.
Thanks all for the links for the other midi's and is the problem of assignment of library sounds now solved to your satisfaction? 

Sun, 2023-04-16 - 11:19 Permalink

The dropdown menus can be confusing if you think they are for looking up and selecting a sound. They are not.

They are for assigning one of the already existing rack module's channels to the instrument. The upper menu selects a rack module. The lower menu selects a channel on that module.

I'll check if dynamic channels really need to be disabled here in all cases.

doing multiple instances of Figure Recognition on a take

Just duplicate the phrase. That's what duplication is meant for.

BTW, the new build is online.

Sun, 2023-04-16 - 15:52 Permalink

@janamdo Yes, I can see the usefulness of the PG Music player. But I needed to access the MIDI files directly for import (Didn't you?), and that turned out to be a simple matter of finding the folder into which they had been downloaded.

The problem of assignment of library sounds is at least preliminarily solved, although I still don't understand why one cannot assign sounds via the dropdown menu. But the wrench icon takes me where I want to go, so I'm not going to worry about those dropdown menus. I've made a mental note to ignore them.

Thanks.

Sun, 2023-04-16 - 15:58 Permalink

"Sorry Kim to interrupt your party with your AI chatbot friends..."

Really? 

@Andre Is there an ignore button on this forum?

Sun, 2023-04-16 - 17:01 Permalink

It was a reply on my "amazing good" synfire 2.1 in analyse midi file.  It is sounding good together with the original wav stem.  

But also in synfire 1.10 get the same good result.  So "amazing good" is for synfire 1 and 2. Sorry Kim

Sun, 2023-04-16 - 17:19 Permalink

@Nick I have to admit that most of the things you post make no sense to me, but going forward I'll atribute this to a language difference, and assume good intentions on your part.

But you should be aware that all over the internet, people are accusing each other of being AI chatbots, so when you start using that term it can be taken the wrong way.

EDIT: "...most of the things you post make no sense to me..."

I wish I hadn't written that...it comes across as harsh. It's not that the things you post make no sense, but that I usually have to read them several times to try to figure out what you mean. And often I don't bother, because other things are vying for my attention.

Sun, 2023-04-16 - 17:38 Permalink

@ The Artist..., 
<But I needed to access the MIDI files directly for import (Didn't you?), and that turned out to be a simple matter of finding the folder into which they had been downloaded.>
The player is for background information of the new age midis and fortunately they are usable in Synfire which was the intention.

Mon, 2023-04-17 - 12:29 Permalink

There is still a bug with batch import in "All Static" mode. Only the first file is imported static and subsequent files auto-estimate settings per track. Just in case you wonder why static batch imports takes so long.

Will be fixed in build #3

Tue, 2023-04-18 - 14:20 Permalink

looking at the blue figures from Synfire is for Kim top. For me the static pitch is visual better.  

And the Cubase chordtrack is telling us 3 bar same , last one a change.    

It is not clear what the left hand has played and the right hand on the piano?  I think the left hand played the bass (3 lower notes).  Synfire will not give us info. Should we split the midi in a lower part and melody part to get Bass and chords?

  

Tue, 2023-04-18 - 19:29 Permalink

Build #3 is online now. Here's what was fixed:

  • Library: Show estimated size on toolbar and a warning if library gets too big
  • Library: Fixed memory leaks
  • Library: Sound allocations have not been not reset after import
  • Import: Mark first phrase with a figure or take as default
  • Import: Static pitch mode still created figures (now fixed and much faster)

Tue, 2023-04-18 - 21:30 Permalink

@janamdo: I thought it might be pertinent to post my impressions of the PG Music New Age Pianist MIDI set you brought to the attention of this forum, and which I bought a few days ago.

The most relevant fact one should consider before buying this set is that PG Music intends it to be for listening, not for music production. Accordingly, the MIDI was performed live and not to a click track. So it was played "freely"...their term. The tempo appears to speed up and slow down as the MIDI plays. This can be a plus in the sense that the performance can therefore be very expressive. However, getting the MIDI to "play nice" with other instruments can be problematic. 

For example, I've started a project where I'm exclusively using a portion of a composition from the New Age set (no other phrases), and it's working well. I'm sending a phrase to two Omnisphere patches, and the lack of a steady tempo really adds to the emotion and human feel. However, I have yet to find phrases derived from the New Age set that work well in a project that incorporates a steady tempo (aka a click track).

I've tried quantizing, but so far I'm not happy with the results. Also, in a different project I have a busy bass line going, and I copied the bass Figure parameter and then pasted it into the Rhythm parameter of an Instrument using a phrase derived from the New Age set. This is more or less standard operating procedure for me, because it tends to make things really sync up. But so far I am also unhappy with the results in this instance.

This is all very preliminary. I haven't worked extensively with the New Age MIDI or its resulting phrases. Maybe there are other things I can do to make the phrases more "portable" in the sense that I can drop them into an Arrangement and have a reasonable expectation that they will sound good. For example, I want to try importing the MIDI into Cubase. I would then line the MIDI up (in small phrases) so the downbeats fall at the proper place on the timeline, and then quantize it. Then I'd import that MIDI into a Synfire library. Or maybe someone on this forum will have a suggestion.

But this all involves extra steps, and I'm not sure it will be a good use of my time even if it does work. The New Age set might turn out to be of limited use compared to MIDI sets played or programmed to a click.

Tue, 2023-04-18 - 22:49 Permalink

Thanks for effort,
<The tempo appears to speed up and slow down as the MIDI plays. This can be a plus in the sense that the performance can therefore be very expressive. However, getting the MIDI to "play nice" with other instruments can be problematic. >
Should be possible to extract a certain random rhythm from a piece of music in Synfire if I have it right, no problem.
This parameter can then be used on a new instrument rhythm.
Could be in one of @Juergen's tutorials describing this on the Cognitone website?
Paste Rhythm | Cognitone Users

Tue, 2023-04-18 - 23:40 Permalink

The first example from that article:

"1. Copy the figure of instrument A and execute Paste Rhythm at the figure of instrument B: The figure of the instrument B gets the rhythm of instrument A"

is very similar (but not identical) to what I was referring to when I wrote:

"Also, in a different project I have a busy bass line going, and I copied the bass Figure parameter and then pasted it into the Rhythm parameter of an Instrument using a phrase derived from the New Age set. This is more or less standard operating procedure for me, because it tends to make things really sync up. But so far I am also unhappy with the results in this instance."

 @Andre: Are these just two slightly different ways to arrive at the same result? Or is one preferable to the other?

Thanks to Juergen for writing the article.

Wed, 2023-04-19 - 08:14 Permalink

If a recording is constantly late or early, the take can be shifted manually to match the grid. If it's performed "freely", you're out of luck. You'd need to extract portions and move/stretch them to match the grid, then stich them together again. IMO that's too much effort, unless you have a very specific recording that you want to convert at all cost.

Rhythm transfer, as described by Juergen, is not a means to achieve that. It's merely a kind of soft quantization. It doesn't stretch a take or figure as a whole.

Some DAW support tempo detection and time streching to be applied after a recording. That may be a solution. Slow and expressive piano performances might not work so well though.

Workflow discussions about a topic like this are valuable. I'm always happy when they come up. That's what this forum was made for. It's a shame that  discussions get lost inside some obscure (and often unrelated) thread over time. Our search for a better community platform however has still not found anything. Still can't believe it's 2023. Feels more like 1998 in this regard.

Wed, 2023-04-19 - 09:18 Permalink

<An interesting concept. Did you use the Morph feature?>

I must confess that I made a few attempts to use the morphing feature, but it was rather laborious and I abandoned it, but who knows how it works for me now?
Slipping composing is actually quite simple: in an arrangement, gather a collection of music style containers. 

I had then used midi styles from yamaha and compiled some containers from that and then started combining.
In fact you are then one step higher than the figures alone, because you already have a complete piece of music.
It was all about the transitions between certain containers that you then shifted opposite each other and that I let repeat via a loop I thought?.

Listening if you were satisfied with that music passage and you can also start adding instruments with audio ( effect or with a voice now )
Go try it again the slip composing and it skips the detailled figure composing and it goes pretty fast.

But it is limited in use and whether you can call this composing.... 

Note: Could also be that I combined the new age piano midi containers with slip composing for the "promising day" composition and added other instruments to it 
I didn't keep any of it back then from the original composition. 
 

Wed, 2023-04-19 - 10:45 Permalink

Andre, would it be possible to develop an improvement when importing midi, maybe to have an option a bit like quantise, but as well as moving the notes to a grid also extracts a shift parameter? Appling the generated notes with the shift parameter would recreate the original timing.

 

Wed, 2023-04-19 - 12:12 Permalink

@blacksun

Automatically detecting rhythm and time-stretching a sequence accordingly is as hard an A.I. challenge as is figure recognition. Imagine a smooth jazz piano session with lots of syncopes, rests, runs, tuplets. Even most (untrained) humans have a hard time figuring out the beat. Sometimes its even deliberately obscured or fluid.

@janamdo

Thanks for the link. Of course we already checked all those forums. Forums are fine for short term, short form communication (help), but we need long term communication with blog-like posts that encourage users to show their work, talk about it and engage with others. There's so much to explore and share with Synfire, it deserves a place that can be browsed and searched more conveniently.

Wed, 2023-04-19 - 12:24 Permalink

Some DAW support tempo detection and time streching to be applied after a recording. That may be a solution. Slow and expressive piano performances might not work so well though.

For Cubase this would be the workflow for the automatic tempo detection. 

ttps://steinberg.help/cubase_pro_artist/v9/en/cubase_nuendo/topics/editing_tempo_and_signature/editing_tempo_tempo_changes_unknown_project_track_from_detection_setting_t.html

But as you say, the results depend a lot on the source material. In extreme cases, you have to edit every single bar line manually. 

Automatically detecting rhythm and time-stretching a sequence accordingly is as hard an A.I. challenge

An A.I.-based approach to this task is being taken by Scorecloud (formerly Scorecleaner).

https://scorecloud.com/

This claims to generate ready-made scores from freely recorded MIDI-data. But the results are also rather mixed here, of course. .

Wed, 2023-04-19 - 12:58 Permalink

<There's so much to explore and share with Synfire, it deserves a place that can be browsed and searched more conveniently.>

Yes, developing your own music platform requires expertise and a lot of time, but then you have something!
Has to be maintained.
Used to have some new functionality for my brother's webshop programmed in ukraine, cost very little and was well done.
Won't go there anymore to have software made ?  
Unfortunately, I haven't become a programming star myself and it was too much trouble for me to learn the entire php programming language for this one webshop.