http://www.komponant.com/
I don't know if this post is in the right place, SING is not free software and even they do not plan to suport our OS at now, but...
..the demo is amazing, much more realistic than any VOCALOID/UTAU/whatever and I want this program on Linux. At this stage of development this is quite possible so if you want that too please join to the beta by filling the form on their site and for the question "What kind of features would you like to use in SING? " select other and enter "native Linux support"
Thank you in advance for any help, every vote can make that they will treat Linux serious as new growing platform for music production!
New realistic singing synthesis software, do you want Linux version?
Moderators: MattKingUSA, khz
-
- Established Member
- Posts: 2083
- Joined: Mon Sep 28, 2015 8:06 pm
- Location: Here, of course!
- Has thanked: 232 times
- Been thanked: 400 times
- Contact:
Re: New realistic singing synthesis software, do you want Linux version?
Impressive. I've heard some pretty good choir 'vocalisations' but this it quite different.
The Yoshimi guy {apparently now an 'elderly'}
-
- Established Member
- Posts: 81
- Joined: Mon May 30, 2011 6:32 pm
- Has thanked: 15 times
- Been thanked: 6 times
Re: New realistic singing synthesis software, do you want Linux version?
Hello everyone, some friends and I am already developing an opensource virtual singer platform "LINNE" (Means Linux's vocal).
Our platform is based on World synthesis algorithm,written by Dr.Morise. Its ability is between Vocaloid2 and Vocaloid3 and still move forward,but opensource + patent free. Its realtime synthesis is faster than Vocaloid2,even on older CPUs. It's possible to port to mobile devices.
In the platform,Now we have the synthesis engine,wavtool-pl(our opensource verison, now is in Debian),and we made a virtual singer vocal DB "ChiOnMeiLin" (Our vocal DB support Utau db spe. and extend some more)
Here are some demos:
Demo1 Honeyworks ママ ver.acoustic
Umbrella
Our platfrom lacks the final puzzle : an editor,so we Adopted Utau as our editor before. (Cadence has stopped development) But it have many unresolved problem. Especially unfriendly to non-Japanese language. For example, Our db file -oto.ini is UTF-8 for multi languages and can write comments for internet collaboration ,but Utau can't. and many other problem unresolved.
Now we starts to make a new editor by our own, based on QT. If someone is interested, please join us. I think It will speed up the editor development.
Another thing,actually speaking our voice talent is a Jazz singer.She also have a nice English accent. We have recorded Japanese and Mandarlin. We design a better recording way based on IPA, so different languages may share the same recordings. If someone help us to define English Syllables by IPA ,we also can record English in the studio.
Our platform is based on World synthesis algorithm,written by Dr.Morise. Its ability is between Vocaloid2 and Vocaloid3 and still move forward,but opensource + patent free. Its realtime synthesis is faster than Vocaloid2,even on older CPUs. It's possible to port to mobile devices.
In the platform,Now we have the synthesis engine,wavtool-pl(our opensource verison, now is in Debian),and we made a virtual singer vocal DB "ChiOnMeiLin" (Our vocal DB support Utau db spe. and extend some more)
Here are some demos:
Demo1 Honeyworks ママ ver.acoustic
Umbrella
Our platfrom lacks the final puzzle : an editor,so we Adopted Utau as our editor before. (Cadence has stopped development) But it have many unresolved problem. Especially unfriendly to non-Japanese language. For example, Our db file -oto.ini is UTF-8 for multi languages and can write comments for internet collaboration ,but Utau can't. and many other problem unresolved.
Now we starts to make a new editor by our own, based on QT. If someone is interested, please join us. I think It will speed up the editor development.
Another thing,actually speaking our voice talent is a Jazz singer.She also have a nice English accent. We have recorded Japanese and Mandarlin. We design a better recording way based on IPA, so different languages may share the same recordings. If someone help us to define English Syllables by IPA ,we also can record English in the studio.
Re: New realistic singing synthesis software, do you want Linux version?
Great! Please made your editor more vocaloid-like than UTAU (which is very crazy and unfriendly, at least for me) but with articulations based on easy editable bezier curves instead of steinberg-style events. And good English voicebank is a must to attract Western musicians.
Re: New realistic singing synthesis software, do you want Linux version?
How is this open source linne project progressing? Could not find anything.
Autotune/vocaloid type manipulator would be nice. Extensive manipulation of vocal recordings. Vocal analysis for re-creation with speech synth; not ipa, but pitch/formant curves etc
Autotune/vocaloid type manipulator would be nice. Extensive manipulation of vocal recordings. Vocal analysis for re-creation with speech synth; not ipa, but pitch/formant curves etc
-
- Established Member
- Posts: 81
- Joined: Mon May 30, 2011 6:32 pm
- Has thanked: 15 times
- Been thanked: 6 times
Re: New realistic singing synthesis software, do you want Linux version?
Sorry for late replying.
This is our official site. Meilin is our virtual singer. I am sorry that, English pages aren't translated yet. I hope we will prepare soonly.
"Linne" is introduced in "documents". Linne means a platform=
editor(old is Cadencii now change to make new one) and xvsqExec (a engine read utau song file to call the synthesizer and wavtool)
+synthesizer (World algorithm based , we now change to use tn_fnds_yc)
+wavtool-pl(already in Debian)
+tools(Linne analysis ,oto.ini diff tools,etc.)
+may include a arm-based robot doll hardware in the future.
We have a slack . If you are interested , I can add you in for discussion.
This is our official site. Meilin is our virtual singer. I am sorry that, English pages aren't translated yet. I hope we will prepare soonly.
"Linne" is introduced in "documents". Linne means a platform=
editor(old is Cadencii now change to make new one) and xvsqExec (a engine read utau song file to call the synthesizer and wavtool)
+synthesizer (World algorithm based , we now change to use tn_fnds_yc)
+wavtool-pl(already in Debian)
+tools(Linne analysis ,oto.ini diff tools,etc.)
+may include a arm-based robot doll hardware in the future.
We have a slack . If you are interested , I can add you in for discussion.
Re: New realistic singing synthesis software, do you want Linux version?
For those who are interested in Komponant, we have a new demo, or rather a short teaser soundcloud.com/komponant/teaser
Sorry if it wasn't clear, the previous demo "Take Me Away" was a proof of concept, not the actual output of our engine. We don't plan to release a linux version, but we would like to make our voices available as a web app.
Sorry if it wasn't clear, the previous demo "Take Me Away" was a proof of concept, not the actual output of our engine. We don't plan to release a linux version, but we would like to make our voices available as a web app.
-
- Established Member
- Posts: 2036
- Joined: Sat Jun 11, 2016 12:05 am
- Has thanked: 10 times
- Been thanked: 22 times
Re: New realistic singing synthesis software, do you want Linux version?
Quite nice. It seems that the release part of ADSR is a problem when synthesizing vocals, although this starts to go in the right direction.MGdesigner wrote:Hello everyone, some friends and I am already developing an opensource virtual singer platform "LINNE" (Means Linux's vocal).
Here are some demos:
Umbrella