8va/8vb not working- ledger lines galore

annawright18 commented 4 years ago

Example no. 744, both upper and lower ledger lines are way too much - score has 8va/8vb but this is not translating to KERN

craigsapp commented 4 years ago

The notation looks like this in VHV/Polyrhythm website:

https://polyrhythm.humdrum.org/example/?id=744

But target rendering in Sibelius PDF is:

The problem is that the translation of ottava marks in the Humdrum conversion of the MusicXML export from Sibelius is probably reversed (8va should be applied to top staff and 8ba should be applied to bottom staff).

There is also annoying incorrect tuplet number on the first beam group.

craigsapp commented 4 years ago

General comment about issue submissions: It would be useful to give Polyrhythm website links for the issue exemplars, and also screen shots of the rendering problems, along with the desired rendering from the Sibelius PDF and/or the original source images. (like the previous post I made in this thread).

Images can be dragged/dropped into these messages while in "write" mode, then you can click on "Preview" to see the actual image:

URLs to the Polyrhythm website can be copy/pasted into the text, and they will automatically be converted into hyperlinks.

craigsapp commented 4 years ago

Added issue on VHV repository to address the bug: https://github.com/humdrum-tools/verovio-humdrum-viewer/issues/263

craigsapp commented 4 years ago

Now fixed with commit https://github.com/craigsapp/humlib/commit/f998e31743b60b4765526b5bb093920373d7bdeb

craigsapp commented 4 years ago

There are 82 scores that have ottava lines:

I have reconverted them and in theory this issue should now be fixed. (Only the examples with multi-staff parts such as piano will have needed fixing).

craigsapp commented 4 years ago

Here is the updated exemplar with the corrected notation:

https://polyrhythm.humdrum.org/example/?id=744

Target rendering now matches to conversion:

craigsapp commented 4 years ago

The fix seems to have caused ottava lines to become backwards in some cases:

https://polyrhythm.humdrum.org/example/?id=716

So this will have to be fixed...

craigsapp commented 4 years ago

Need to fix position of ottava lines starting at beginning of music. Currently the marker for the start of the line is coming before the initial barline position. This should be move to after the initial barline:

Also, the ending ottava marker should not cause a contraction of the subspines is it is doing in this case (* markers should be given to fill in the subspines).

The ottava line seems to be ending one note two early as well. Here is the Sibelius rendering:

annawright18 commented 4 years ago

8va/vb also not working in example 655 - R655_Hin-w77p19m154-160

8 appears with bracket, but does not specify va or vb

annawright18 commented 4 years ago

Also in R678_Hin-w90p19m143-147

Last bar, bottom stave - 8va/vb missing

craigsapp commented 4 years ago

That will be OK. In the data there is an indication of 8va or 8vb. In verovio notation, the style is to just show "8" rather than add additional letters. The position of the ottava lines above or below the staff carry the information about 8va or 8ba (8va are always above the staff, and 8ba are always below the staff). So it is a style issue, and verovio only has one style for display of ottava lines.

The ottava line in R655 is poorly spaced, so it is easy to confuse which staff it is on, so I will report to verovio that it can be removed.

For R678, I will study why it was dropped at the end of the music. Probably this is due to the ottava line being cut off in the middle due to the example ending, in which case there may not be a closing for the lines (so it has to be implicitly added).

Boardbreak commented 4 years ago

ex. T439, Berg, Harp 8va needs to extend until the final bar. It stops one measure too short.

VHV:

Harp 8va - VHV

Desired:

Harp 8va - Desired

annawright18 commented 4 years ago

Same issue, example 492 - R492_Mau-w24p19m99-104

craigsapp commented 4 years ago

R493 is a style issue, so there will be nothing to fix. In both cases the last beat of the third measure is the last chord in the octave transposition, it is just that Sibelius is extending the line after the chord a little further than verovio.

rafaperlau commented 4 years ago

T693, Gershwin, Violin I, Violin II, mm7: missing octave line.

VHV: Sem título 1

PDF: Sem título 2

craigsapp commented 4 years ago

T693 is a special case that Sibelius MusicXML is not handling correctly.

               <measure number='7'>
                        <direction>
                                <direction-type>
                                        <octave-shift type='down' size='8' number='1' default-y='40' relative-x='-5'/>
                                </direction-type>
                        </direction>
                        <direction>
                                <direction-type>
                                        <octave-shift type='stop' size='8' number='1' relative-x='25'/>
                                </direction-type>
                        </direction>

Although it might be recoverable: when there is an ottava that is turned off immediately after being turned on, then that implies, in a Sibelius generated file, that the octave line goes to the end of the music.

polyrhythm-project / polyrhythm-website

8va/8vb not working- ledger lines galore #36