Check it with huge number of files, what is going on there?

The problem came from Margaux Dubuis.

It is interesting, whether it is useful to have Multi-line views like in Raven

https://ravensoundsoftware.com/video-tutorials/video-tutorials-english/04-multi-line-views/

Have a look at this sound annotation format.

:/

It actually means that the whole ideology of textgrid_to_df should be rewriten without counts of lines...

Would it be possible to create a report where some sounds have images, but others don't? (Or even one where it's just sounds and no pictures, or just pictures and no sounds?)

Thanks to Jonathan Keane

In textgRid author used the readr package in order to make encoding autodetection. It is better to use uchardet which has less dependencies.

thx to Sasha Shiryaev for this problem

It is possible to create another type of TextGrid files with the command Save as short text file in Praat. It would be nice to make phonfieldwork to read them. Here is an example:

File type = "ooTextFile"
Object class = "TextGrid"

0
0.6526757369614512
<exists>
3
"IntervalTier"
"intervals"
0
0.6526757369614512
5
0
0.012465825768947203
""
0.012465825768947203
0.247819135106803
"t"
0.247819135106803
0.39552362579469874
"e"
0.39552362579469874
0.5115771541923311
"s"
0.5115771541923311
0.6526757369614512
"t"
"IntervalTier"
"empty_intervals"
0
0.6526757369614512
5
0
0.012465825768947203
""
0.012465825768947203
0.247819135106803
""
0.247819135106803
0.39552362579469874
""
0.39552362579469874
0.5115771541923311
""
0.5115771541923311
0.6526757369614512
""
"TextTier"
"points"
0
0.6526757369614512
4
0.012465825768947203
"t"
0.247819135106803
"e"
0.39552362579469874
"s"
0.5115771541923311
"t"

Could you please allow user to center stimuli on the presentation slide in the create_presentation() function? I like this very nice and quick way of creating a linguistic survey, but words don't look very good in the corner.

👋 @agricolamz!

The pkgdown build for your package is currently failing because some of the documentation topics are missing from the pkgdown config file. Could you please fix?

> pkgdown::check_pkgdown()
Error in `check_missing_topics()`:
! All topics must be included in reference index
• Missing topics: add_leading_symbols, create_image_look_up, create_sound_play, create_viewer, df_to_eaf, df_to_exb, remove_textgrid_tier

👋 @agricolamz! I was looking at https://dev.ropensci.org/ and saw the docs building is failing for your package, it'd be nice if it were fixed before your blog post.

https://dev.ropensci.org/blue/organizations/jenkins/phonfieldwork/detail/phonfieldwork/26/pipeline

For context on the docs server https://ropensci.org/technotes/2019/06/07/ropensci-docs/

https://dopefishh.github.io/pympi/index.html

Correct multiblock problem.

e. g. this file can't be parsed because there is no any word-level annotations.

Thnks to Niko Partanen for spotting it.

while renaming files

As Eduard Klishinskiy proposed it is nice to have a function that will combine concatenate_sound and annotate_sound().

thx @Xenomirant

May be it is worth it to add ability to annotate Raven style rectangular annotations.

https://ravensoundsoftware.com/video-tutorials/video-tutorials-english/05-using-annotations/

"read_from_folder" cannot read files if the textgrid file that includes a tier with no points, and as a result, the whole textgrid file could not be read and is replaced with NA.

Zero point is actually important information in analysis; so I wonder if someone could show the solution to this problem.

For now it is possible only cut sound by tier in the textgrid. It would be nice to make possible to cut it using some part of textgrid's annotation.

whenever function has ..._from_folder = ... argument.

I'd like to make a viewer:

• searchable and sortable table with all contents
• select anything from a list and it loads
  • precomputed spectrogram (or any other picture)
  • sound player (play, stop, decrease rate)
• some common things for the website: title, citation info block, etc.
• password protection?
• it is important not to load everything on page call

R tools: htmlwidgets and crosstalk
JS tools: DT, ...?

It is important to have a possibility of non-verbal stimuli: pictures, gifs, may be even video...

transana

thanks to Oliver Ehmer and his Transformer tool.

Emu etc
https://clarin.phonetik.uni-muenchen.de/BASWebServices/interface/FormantAnalysis

For now draw_sound() is drawing the whole file that could be terrible, when you have a long file. It is could be nice to add a possibility to draw only fragment of the sound and TextGrid in order to see the result from R.

It is nice to think about creation of a document from any type of annotation.

thx Niko Partanen