Recherche avancée

Sur d’autres sites (5382)

• FFMPEG - Width/ Height Not Divisible by 2 (Scaling to Generate MBR Output)

  15 avril 2020, par Sanjeev Pandey

  I am trying to generate multilple variants of videos in my library (Mp4 formats) and have renditions planned ranging from 1080p to 240p and popular sizes in between. For that I am taking a video with a AxB resolution and then running through a code (on bash) which scales them to desired following sizes - 
426x240
640x360
842x480
1280x720
1920x1080, with different bitrates of course, and then saves as Mp4 again.

  



  Now, this works just fine if source video has height and width divisible by 2, but code breaks on the following line for the videos with odd width and height :
-vf scale=w=${width}:h=${height}:force_original_aspect_ratio=decrease"

  



  Where 'width' and 'height' are the desired (and hardcoded) for every iteration : E.g. "426x240", and "640x360"

  



  The Error :
[libx264 @ 00000187da2a1580] width not divisible by 2 (639x360)
Error initializing output stream 1:0 -- Error while opening encoder for output stream #1:0 - maybe incorrect parameters such as bit_rate, rate, width or height

  



  Now approaches those are explained in this one doesn't work for me since I am scaling - FFMPEG (libx264) "height not divisible by 2"

  



  And, I tried this one too but it seems all qualities are getting the same size -ffmpeg : width not divisible by 2 (when keep proportions)

  



  


  

  • This is how I tried to use this one : scale='bitand(oh*dar,65534)':'min(${height},ih)'

  


  


  



  Kindly suggest how to solve this, keeping in view that :
1. I have a very large library and I can't do manual change for every video
2. I need to scale the video and keep the aspect ratio

  



  Thanks !

  



  PS : [Edit] One way that I can see is padding all of the odd height/ weight videos using a second script in advance. This however doubles my work time and load. I would prefer to keep it in single script. This is the script I see that I can use for padding :
```ffmpeg -r 24 -i -vcodec libx264 -y -an -vf "pad=ceil(iw/2)*2:ceil(ih/2)*2"`` (from : FFMPEG (libx264) "height not divisible by 2")

  


• How to visualize matplotlib animation in Jupyter notebook

  23 avril 2020, par anonymous13

  I am trying to create a racing bar chart similar to the one in the link (https://towardsdatascience.com/bar-chart-race-in-python-with-matplotlib-8e687a5c8a41). 
However I am unable to see the animation in my Jupyter notebook

  



  code

  



  import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.ticker as ticker
import matplotlib.animation as animation
from IPython.display import HTML

df = pd.read_csv('https://gist.githubusercontent.com/johnburnmurdoch/4199dbe55095c3e13de8d5b2e5e5307a/raw/fa018b25c24b7b5f47fd0568937ff6c04e384786/city_populations', 
                 usecols=['name', 'group', 'year', 'value'])

current_year = 2018
dff = (df[df['year'].eq(current_year)]
       .sort_values(by='value', ascending=True)
       .head(10))

colors = dict(zip(
    ['India', 'Europe', 'Asia', 'Latin America',
     'Middle East', 'North America', 'Africa'],
    ['#adb0ff', '#ffb3ff', '#90d595', '#e48381',
     '#aafbff', '#f7bb5f', '#eafb50']
))
group_lk = df.set_index('name')['group'].to_dict()


fig, ax = plt.subplots(figsize=(15, 8))
def draw_barchart(year):
    dff = df[df['year'].eq(year)].sort_values(by='value', ascending=True).tail(10)
    ax.clear()
    ax.barh(dff['name'], dff['value'], color=[colors[group_lk[x]] for x in dff['name']])
    dx = dff['value'].max() / 200
    for i, (value, name) in enumerate(zip(dff['value'], dff['name'])):
        ax.text(value-dx, i,     name,           size=14, weight=600, ha='right', va='bottom')
        ax.text(value-dx, i-.25, group_lk[name], size=10, color='#444444', ha='right', va='baseline')
        ax.text(value+dx, i,     f'{value:,.0f}',  size=14, ha='left',  va='center')
    # ... polished styles
    ax.text(1, 0.4, year, transform=ax.transAxes, color='#777777', size=46, ha='right', weight=800)
    ax.text(0, 1.06, 'Population (thousands)', transform=ax.transAxes, size=12, color='#777777')
    ax.xaxis.set_major_formatter(ticker.StrMethodFormatter('{x:,.0f}'))
    ax.xaxis.set_ticks_position('top')
    ax.tick_params(axis='x', colors='#777777', labelsize=12)
    ax.set_yticks([])
    ax.margins(0, 0.01)
    ax.grid(which='major', axis='x', linestyle='-')
    ax.set_axisbelow(True)
    ax.text(0, 1.12, 'The most populous cities in the world from 1500 to 2018',
            transform=ax.transAxes, size=24, weight=600, ha='left')
    ax.text(1, 0, 'by @pratapvardhan; credit @jburnmurdoch', transform=ax.transAxes, ha='right',
            color='#777777', bbox=dict(facecolor='white', alpha=0.8, edgecolor='white'))
    plt.box(False)

draw_barchart(2018)

import matplotlib.animation as animation
from IPython.display import HTML
fig, ax = plt.subplots(figsize=(15, 8))
animator = animation.FuncAnimation(fig, draw_barchart, frames=range(1968, 2019))
HTML(animator.to_jshtml()) 



  



  Below is what I tried using and the errors

  



  HTML(animator.to_jshtml())  <-- Static output with buttons unable to visualize animation
plt.rcParams["animation.html"] = "jshtml"  <- no error and output
HTML(animator.to_html5_video())  <---Requested MovieWriter (ffmpeg) not available 



  



  Note I have FFmpeg installed in my system.
Can you help me with the issue

  


• Revision 33402 : Amélioration du pipeline post édition

  29 novembre 2009, par kent1@… — Log

  Amélioration du pipeline post édition