Cadencii users UNITE

[KitWistful]

--UPDATED POST--:

So this is a thread for Cadencii, it seems

Cadencii is a vocal synthesizer that one may use to host Vocaloid, UTAU, and Aquestone.

It's notoriously un-user-friendly, unstable, and unefficient. However, it's unique features include the beautiful following:

- A vocaloid-like interface (very much like V1)
- A unique properties window
- The ability to convert .vsqx to .ust
- The uniquely realistic resampler, vConnect-STAND
- The ability to use a variety of voicebanks together
- An engrish GUI out-of-box

It's not a program for the faint of heart. You have to be ready to save often, you have to be willing to live with some features mysteriously not working, and you'll have to be able to make compromises for the CPU load. However, it has it's good points, and is probably worth a try.

Here are some tips:

- You can decide what parameters you want to use from setting>preference>appearance. (Env especially is relevant when working with UTAU!)
- if you're tired of wait times, you can switch to a UTAU resampler.
- when working with a UTAU resampler, set mod to 0 on all notes
- when working with a UTAU resampler, remember that it likes resampler.exe best
- when working with a UTAU resampler, remember that it won't take some of Cadencii's parameter changes.


Cadencii can be found here

Ritsu has an HQ voicebank made specially for cadencii: demo ("Toss Up") (yt nico(" sm15561654 "). download/installation guide here.

Cadencii may be used to arrange a song for Sinsy.

It seems to update every 1-2 years.

I wish to learn many things about Cadencii, so posting your mis/adventures and suggestions is always welcome :3

--ORIG POST--:

So I picked up Cadencii the other day

and even though it's wanting in so many areas, I still love the idea of it. When it works right, it's pretty cool. Do any of you guys use it? Let's all list our complaints, as well as our reasons for loving it:

: ' O
1. Sweet Ann works fine, but Leon won't speak a word : ' O
2. Aquestone will produce sound, but I'm not able to make them pronounce things in the piano roll (I think I went looking for a way to get to the .dll so that I could set the text there, but if that's the only way Cadencii will work with Aquestone I don't care enough to use it).
3. It won't take either of my MIDI controllers, and there was a time when starting it up would actually knock my little LPK25 out. It's a bit discouraging to have to open up UTAU to use a piano.
4. I've actually never been able to use any exported .vsq/.ust. It just doesn't take for some reason.

<3
1. I love the hands on envelope editor (I still haven't found the crossfade function. Is it even there??).
2. And that "Devide Note" tool is pretty convenient for some rough engrish -- I wish it was cut at the end, but it's easy enough to rearrange.
3. And it's much easier to arrange and test a duet/chorus.
4. And I don't work with Sinsy, but the idea that I could make a song with her is just so tantalizing (I'm in love with Circus-P's cover of Tsukema Tsukeru that used her ;o

so what do you think?

 

[Yue Nagareboshi]

The key word here is "When it works right".... to to make it shorter "When it works".

The problem with Cadencii with UTAU VB (I can't speak for VOCALOID since I dont use it), is the fact it's extremely glitchy and tends to freeze/die at some point of working... in some cases to many times.

Also, even if it's rendering engine is nice in terms of the exported wav, it takes ages to do it. Even to check a part while tunning it will take you a while to listen how it sounds. Also dont forget the fact it DEVOURS resources of your PC like crazy.

VOCALOID paramtetrics in UTAU VB are nice... but also, the enveloping work goes 100% manual to sound ok; and 80% of the UTAU community hardly edit/tune their stuff in UTAU, so I find it hard that other UTAU users will spend their time into doing this. As example, some complain with CVVC for being "a troublesome method to use" because of otos and converting ust's manually. People expect to plug-in-Render-PRESTO. (Pretty sad I need to say).

Dont take me wrong... I dont mean that Cadencii is an awful program... It's just that it may not be suitable by the most of UTAU users due the whole manual work they need to do, and the constant glitching the program may give to the users.

On the other hand, users like HAL-P make awesome works with Ritsu-V.Connect (Ritsu VB recorded to an equivalent of VOCALOID-like) but as I say... so little people use it, and there isn't a lot/there's-not banks like that out-there.

[sub](NO.... Speedy Torane V.Connect does not count... he is just a VCV imported to Cadencii; not a 9 set of VCV's on different semitones and breathiness levels)[/sub]

I feel if Cadencii was a more stable program, it will viable to a lot of UTAU and VOCALOID users.

==

Out of topic: If you want to work with a bank that is PRETTY SIMILAR to Sinsy inside UTAU, check Kion Yura. She is available as CV, VCV, VCV whisper, and "English")

 

[KitWistful]

"English"? "English" is my favourite : OO I will have to check it out.

I'm willing to live with crashes and such because many of the programs I use to make music are pretty buggy. I'm in the habit of saving every five seconds (pretty much literally!)

....it's inefficiency and wait times are pretty brutal, though, to be sure. I'd like to give up on it, but something about it keeps drawing me back (it must be that strange cartoon orange it has for an icon. Why? Why is there an orange????).  A recent experiment I made of an engrish duet makes me think that it might be conceivable to make a UTAU song in it...weeelll, with a UTAU resampler, of course. V. Stand Connect might sound nice, but it's habit of dropping random notes, and it's ridiculous wait time -well, the dropping random notes is the real dealbreaker for me. I could try working around the wait time with a faster UTAU resampler, but if it can't do it's job then it's not worth it. That's probably the thing that bothers me the most about Cadencii e__e But it's not the only useful aspect of the program, thankfully.

The upside to trying to work with Cadencii is that the worst things about it (the bugginess, the inefficency, fancy functions not functioning, etc.) are sure to improve with each release. I haven't been with the program for long, but my experience with open source-y things is that it becomes better to work with over time.

 

[Umbreon126]

[retracted]

 

[Cdra]

Cadencii, I wish I could love you, but alas it is not meant to be... v_v

Cadencii is great for converting VSQx or VSQ to UST without errors (and the PIT stays in the UST as mode1 pitchbends!); however, it notoriously imports usts with funky pitching, and I don't just mean the mod.  There's no way to auto-VCV files for it, then, which is really irritating.  I do have a script written by Edgy that does the crossfade envelopes for you, but it blows up on negative overlaps (which I don't blame it for)...

Also it takes FOREVER to render.  Freaking hell.  Honestly, with that and the VCV issue (and the fact that it has vocaloid pitchbends ew), I'll just stick to UTAU+tn_fnds.  I do wish I could have the multi-track and the parameters though v_v  brightness baby please come to meeeee

And I have to go through some major pains to get Al to sing in it.  Oh well.

IF ONLY CADENCII WAS VOCALOID ENGINE + MODE2 FOR ALL SYNTHS, THAT'S ALL I EVER WANTED FROM IT

I feel like instead of taking the best from each program, it kind of took the worst... sigh.

 

[KitWistful]

Cadencii.... meh. Haven't experienced any bugs yet but one that's not too bad i guess. In Windows 7, Cadencii insists on staying on top of other windows. With the long rendering time.... bluh
I don't know anything about the Vocaloid part, I don't own one. Also it seems Cadencii has a X-SAMPA dictionary in it.
I haven't tried exporting vsq and ust. But when I usedCadencii a while back to make an xml for Sinsy, the xml was ALWAYS lacking the last part of the ust I imported into Cadencii.
And yeah what's with the icon what is that an orange with feet, an orange turnip what
And Aquestone works for me kind of i guess sometimes it gives me junk syllables and nothing else
Not saying Cadencii's bad, just okay.
Also who would control Cadencii using a game controller o.o

ONLY...THE HARDCORE O___O (I should try that out sometime XD)

I have two english vocaloids, and interestingly it seems to have imported my english dictionary. Unfortunately there's some instances where JPN voicebank takes precedence, such as default a [a], and I can't upload a custom dictionary...at least, I can't make one in Cadencii. I wonder if I could make a custom dictionary in VOCALOID and import it? That would be interesting.

If Cadencii successfully got Aquestone to pronounce then I think that would be a major success on it's part. I like Aquestone (especially the male voice!) but it's hard to tweak and fix the lyric output in it's VST form, which makes engrish nigh impossible. I mean, I could just use Defoko (and I think the male voice has a UTAU counterpart??? I think), but Aquestone has it's own sort of functionality (different parameters etc.) that I'd like to muck around with.

Cadencii, I wish I could love you, but alas it is not meant to be... v_v

Cadencii is great for converting VSQx or VSQ to UST without errors (and the PIT stays in the UST as mode1 pitchbends!); however, it notoriously imports usts with funky pitching, and I don't just mean the mod.  There's no way to auto-VCV files for it, then, which is really irritating.  I do have a script written by Edgy that does the crossfade envelopes for you, but it blows up on negative overlaps (which I don't blame it for)...

Also it takes FOREVER to render.  Freaking hell.  Honestly, with that and the VCV issue (and the fact that it has vocaloid pitchbends ew), I'll just stick to UTAU+tn_fnds.  I do wish I could have the multi-track and the parameters though v_v  brightness baby please come to meeeee

And I have to go through some major pains to get Al to sing in it.  Oh well.

IF ONLY CADENCII WAS VOCALOID ENGINE + MODE2 FOR ALL SYNTHS, THAT'S ALL I EVER WANTED FROM IT

I feel like instead of taking the best from each program, it kind of took the worst... sigh.

At least it's good for showing UTAUvocaloid initiates why they'd HATE HATE HATE vocaloid XD I remember how confused I was when I first got Sweet Ann, like, WHY IS THIS NOT SOMEHOW MIRACULOUSLY EASY ALL OF A SUDDEN??? And if it wasn't for the fact that it's Cadencii, it could be a great way for beginners to learn about tuning.

(off topic: Have you tried EFB-GT? It's very similar to tn_fnds, but I think it makes some of my 'loids sound better than tn_fnds. It might just be placebo-y-hype, but I like it ;o

I don't know if anyone's keeping track of Cadencii updates, but if I notice any then I think I'll post here with observations.

 

[Asteriski]

I think it's more trouble than it's worth right now.
-terrible with CV
-no auto-VCV
-can't use V3 banks
-can't actually import USTs(or maybe that one's just me)

The one upside is that it can be used to convert VSQx to UST. I want to love Cadencii, but it's not working out...

 

[Obakebaka]

I'm sure there's a plugin for utau where you can trace mode 1 pitching with mode 2.

I just have more issues with the resampler. I tried using Oto Juugatsu on a Cantarella UST with TIPS and he sounded like something out of Simon Cowell's nightmares...


Damn funny, though!

 

[KitWistful]

I'm sure there's a plugin for utau where you can trace mode 1 pitching with mode 2.

I just have more issues with the resampler. I tried using Oto Juugatsu on a Cantarella UST with TIPS and he sounded like something out of Simon Cowell's nightmares...


Damn funny, though!

That's one of the upsides of Cadencii, I think. As with all resamplers, the results are nigh unpredictable, but it can produce some really unique results... I plugged Macne Coco Kuro once, and she sounded like some kinda opera singer!

 

[Obakebaka]

That's not what I meant, lol:

[soundcloud]http://soundcloud.com/obakebaka/cantafaila[/soundcloud]

 

[KitWistful]

Quoi??? is that...auto-drunk singing????? I've never heard a drunk synthesizer before, What is this beautiful magic??? ))

(Did Cadencii put the song off-key, or was the .UST like that in the first place? XD)

 

[Obakebaka]

That's Cadencii and TIPS. The original one was fine.

There should be a parody version of Happy Synthesizer with this. Lol.

"I'm a Drunken Synthesizer and I will sing a song for you.
"Please don't be alarmed if I fall over or puke."
*Flailing dance*

 

[KitWistful]

That's Cadencii and TIPS. The original one was fine.

There should be a parody version of Happy Synthesizer with this. Lol.

\"I'm a Drunken Synthesizer and I will sing a song for you.
\"Please don't be alarmed if I fall over or puke.\"
*Flailing dance*

Come on

Make it happen

bring this beautiful cover to me now XDDDDDDDDDDD

 

[Obakebaka]

I could maybe get Billy to sing it when I'm done otoing him. I'll need to write the lyrics first, though. Lol.

 

[FastSpeedy]

I have a question.
I've seen Ritsu for Cadencii has about 9 banks, and that's interesting. I plan to do that for Speedy as well.

However, I did record a 4-pitch bank for Speedy recently, and it works perfectly on UTAU. My question is: how do I use a multi-pitch bank (that requires a prefix map) in Cadencii? I tried to load it, but Cadencii only reads one of the pitches. Can anyone tell me how to import a multi-pitch bank properly into Cadencii?

Thanks!

 

[KitWistful]

I have a question.
I've seen Ritsu for Cadencii has about 9 banks, and that's interesting. I plan to do that for Speedy as well.

However, I did record a 4-pitch bank for Speedy recently, and it works perfectly on UTAU. My question is: how do I use a multi-pitch bank (that requires a prefix map) in Cadencii? I tried to load it, but Cadencii only reads one of the pitches. Can anyone tell me how to import a multi-pitch bank properly into Cadencii?

Thanks!

Well, we'll have to get to the bottom of this! I fear there's not much to be done about it, but let's get a few things out of the way...

1) Have you successfully used the Ritsu bank before?
2) what resamplers did you try? Both V. Stand Connect and resampler.exe?

I will look more into Cadencii's options, and into this Ritsu bank. There may be something in the note property, settings, etc. that may help. If I dig anything up (or fail to do so) I will update the thread.

You are a brave sort to develop a VB with Cadencii support! It's very admirable, and I wish you good fortune!

EDIT: incidentally, on the subject of updates: the last update was last year, in May. The update before that 2 years before, the update before that yet another year before. I will see if I can find a development forum or something similar; if any of you guys have anything to say on this subject I would be very pleased to learn! ^w^

Topic: what do you think of the allegedly upcoming V3 support? I imagine it will be difficult, as far as I know V3 doesn't come with the vst instrument versions that the other versions do.

If they successfully manage to host V3 voicebanks, I wonder what that means? An update like that is sure to come with a slew of other features (and instabilities)

 

[FastSpeedy]

[quote author=FastSpeedy link=topic=2774.msg21318#msg21318 date=1343749044]
I have a question.
I've seen Ritsu for Cadencii has about 9 banks, and that's interesting. I plan to do that for Speedy as well.

However, I did record a 4-pitch bank for Speedy recently, and it works perfectly on UTAU. My question is: how do I use a multi-pitch bank (that requires a prefix map) in Cadencii? I tried to load it, but Cadencii only reads one of the pitches. Can anyone tell me how to import a multi-pitch bank properly into Cadencii?

Thanks!

Well, we'll have to get to the bottom of this! I fear there's not much to be done about it, but let's get a few things out of the way...

1) Have you successfully used the Ritsu bank before?
2) what resamplers did you try? Both V. Stand Connect and resampler.exe?

I will look more into Cadencii's options, and into this Ritsu bank. There may be something in the note property, settings, etc. that may help. If I dig anything up (or fail to do so) I will update the thread.

You are a brave sort to develop a VB with Cadencii support! It's very admirable, and I wish you good fortune!

EDIT: incidentally, on the subject of updates: the last update was last year, in May. The update before that 2 years before, the update before that yet another year before. I will see if I can find a development forum or something similar; if any of you guys have anything to say on this subject I would be very pleased to learn! ^w^

Topic: what do you think of the allegedly upcoming V3 support? I imagine it will be difficult, as far as I know V3 doesn't come with the vst instrument versions that the other versions do.

If they successfully manage to host V3 voicebanks, I wonder what that means? An update like that is sure to come with a slew of other features (and instabilities)
[/quote]

Thanks for the reply!

As for what you asked, I've not tested Ritsu's bank. For some unknown reason, I can't make my Cadencii read vvd files (which are files Cadencii should read, and the files Ritsu's bank is composed of). I've converted my first bank (standalone VCV bank, not the 4-pitch) into vvd format, but I can't render sound with it either. A friend of mine tested it, and it worked perfectly on his computer. Yes, we tried EVERYTHING we could think of (even having him send his whole Cadencii folder to me) to make it work, but I just can't get my Cadencii to render using vvd files. Using Ritsu's V.Connect crashes my Cadencii.

OBS: I don't have the best computer ever, but I don't think that's what is affecting my Cadencii's performance...

And thanks! I really like what Cadencii can do with UTAU voices, so I hope to get the best out of it ^^

As for the V3 support... I don't know. V3 seems to be quite complicated in terms of it's technology. I think that Cadencii's V3 support will have the most glitches...

 

[KitWistful]

^^

That's good that it works on your friend's computer ! It's a good sign. Uhh, so let's look at what might make the difference...Okay, so what is your friend running, and what are you running?

1) Operating System (Mac, Windows 7, Windows XP, etc.)
2) Japanese locale?

And if you could give some details on what happens when you load Ritsu or your bank, that could be helpful. Like, when you load your guy up, does it stop responding, or does it just not put out noise? If it loads a note, is it the neutral pitch or does it do something weird? Is your bank double-aliased with hiragana and romaji? (strikeout because thinking about it, these things would be affected on your friend's computer. If it works correctly on your friend's computer, it's something in your enviroment.)

It might not be the kind of thing that can be fixed, but we might find some kind of answer! )*___*)9

EDIT: and you probably referenced this already, but the guide for Ritsu's configuration might save the day....you probably looked at it already, though ;;;orz

EDIT: and I'm doubly unsure of what might save the day, since the folder your friend sent you didn't work. That would suggest either it's not routing properly, or your enviroment's not able to support it.

 

[FastSpeedy]

Oh, wow, I haven't seen that guide!
I will try it, and then I post here if it worked or not :3

As for my computer, I use a notebook with Windows 7. And the system is set for Japan in its locale.

Thanks!

EDIT: I followed that guide, and when it asked me to double-click the vConnect-STAND.exe, it didn't give me any errors. Still, once I try to render using Ritsu Connect's banks, it crashes the vConnect-STAND.exe. It doesn't crash Cadencii, but it doesn't render either.

Also, I don't know if it's normal, but, whenever I delete everything Cadencii related from my computer and unzip it again, it loads all the voices I had before, even though I deleted all Cadencii related files before. Is it normal, or is my computer doing something wrong?

 

[KitWistful]

Oh, wow, I haven't seen that guide!
I will try it, and then I post here if it worked or not :3

As for my computer, I use a notebook with Windows 7. And the system is set for Japan in its locale.

Thanks!

EDIT: I followed that guide, and when it asked me to double-click the vConnect-STAND.exe, it didn't give me any errors. Still, once I try to render using Ritsu Connect's banks, it crashes the vConnect-STAND.exe. It doesn't crash Cadencii, but it doesn't render either.

Also, I don't know if it's normal, but, whenever I delete everything Cadencii related from my computer and unzip it again, it loads all the voices I had before, even though I deleted all Cadencii related files before. Is it normal, or is my computer doing something wrong?

There must be something somewhere storing the voices, otherwise it wouldn't recognize it.

I'm running the same op on a netbook too, and it seems that the new V. Connect stand won't work for me either! I believe we've found the answer! ! I'm gonna see if running it in compatability mode helps anything BRB

EDIT: hmmm, it looks like my CV Ruko bank works fine with it. I wonder what that means? (I still have yet to run it in compatability mode)