Comments (9)
You should prepare: phoneme | pitch_midi | pitch_dur | is_slur ,then write to the data 'test“ just use IndexedDatasetBuilder like process_data() in base_binarizer.py.
You need to fix some data loading problems (getitem 、collater in fs2_utils.py). Just set it to None . They are not necessary in the synthesis stage.
from diffsinger.
This is my first exposure to singing synthesis. So I have some question about the terminology.
Does pitch_midi | pitch_dur
mean note & note duration
?
Should I set is_slur
through staffs ?
And I don't know how to set pitch_dur in a unseen song. Should I use Logic Pro to label it ? Or I can get this by some model or something like this.
from diffsinger.
This is my first exposure to singing synthesis. So I have some question about the terminology. Does
pitch_midi | pitch_dur
meannote & note duration
? Should I setis_slur
through staffs ? And I don't know how to set pitch_dur in a unseen song. Should I use Logic Pro to label it ? Or I can get this by some model or something like this.
Wait a minute. I'll find you a picture
from diffsinger.
We use the data marked by yellow box, phoneme | pitch_midi | pitch_dur
from diffsinger.
pitch_dur = 60 * NoteBeats / bmp
bmp : beats per minute --the speed
from diffsinger.
Thank you very much. I know how to do this. But I have another question.
There is silence in music. And it won't work if I simply turn text into pinyin?
Should I do singing - Lyrics alignment?
from diffsinger.
2001000005|面对浩瀚的星海我们微小得像尘埃|m ian d ui h ao h an an d e x ing h ai ai ai AP w o m en w ei x iao d e x iang ch en ai ai ai SP|C#4/Db4 C#4/Db4 D#4/Eb4 D#4/Eb4 C#4/Db4 C#4/Db4 D#4/Eb4 D#4/Eb4 E4 D#4/Eb4 D#4/Eb4 E4 E4 G#4/Ab4 G#4/Ab4 A4 G#4/Ab4 rest C#4/Db4 C#4/Db4 C#4/Db4 C#4/Db4 D#4/Eb4 D#4/Eb4 C#4/Db4 C#4/Db4 D#4/Eb4 D#4/Eb4 E4 E4 E4 E4 G#4/Ab4 A4 G#4/Ab4 rest|0.196990 0.196990 0.102120 0.102120 0.304680 0.304680 0.096780 0.096780 0.100220 0.150010 0.150010 0.361460 0.361460 0.221070 0.221070 0.183240 0.478670 0.384620 0.106510 0.106510 0.143020 0.143020 0.169480 0.169480 0.224180 0.224180 0.089360 0.089360 0.414460 0.414460 0.378050 0.378050 0.162790 0.207380 0.317260 0.297040|0.02765 0.16934 0.01874 0.08338 0.0821 0.22258 0.0693 0.02748 0.10022 0.07137 0.07864 0.12471 0.23675 0.12356 0.09751 0.18324 0.47867 0.38462 0.0405 0.06601 0.08303 0.05999 0.04687 0.12261 0.09778 0.1264 0.02321 0.06615 0.11958 0.29488 0.06723 0.31082 0.16279 0.20738 0.31726 0.29704|0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0
You should learn from transcriptions.txt
from diffsinger.
OK。 Thank you so much. I'll try.
from diffsinger.
@leon2milan did you succeed? can you share an example code?
from diffsinger.
Related Issues (20)
- I encountered an error while running DiffSinger on PopCS HOT 1
- 我发的数据集申请邮件没人回复 HOT 1
- Multi-GPU training & batchsize problem
- 什么叫声学模型? 什么叫唱法模型?这两个模型是什么关系? HOT 1
- 请问popCS里的歌曲,都是同一个歌手演唱的吗
- batchsize
- 云端训练时报错OSError: Unable to synchronously open file (truncated file: eof = 96, sblock->base_addr = 0, stored_eof = 2048) HOT 1
- English DiffSinger version
- 处理数据的时候报这个问题 HOT 1
- Training failed
- About Hyperparmeter "predictor_grad" HOT 2
- Binarize.py error, hparams not accessible from multiprocessing? HOT 4
- Any place to send donations? (aslo a message for gratitude) HOT 2
- AttributeError: 'LatestModelCheckpoint' object has no attribute '_save_model' HOT 1
- hugging spaces are not working HOT 2
- Cannot load testset for infer HOT 2
- 数据集标注 HOT 2
- the test infer using opencpop dataset isnot working HOT 1
- 为啥我发的数据集申请邮件没人回复 有大佬能给我一个数据集下载链接吗 HOT 1
- KeyError: 'hop_size' HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from diffsinger.