did I apply the filter syntax "-sws_flags bicubic" correctly?

[sommers]

SET F1=force_original_aspect_ratio=decrease
SET F2=-q:v 2

aa=ffmpeg -i "%1" -vf scale=w=1680:h=1050:%F1% %F2% out1.jpg <------- no filter or default filter
bb=ffmpeg -i "%1" -vf scale=w=1680:h=1050:%F1% %F2% -sws_flags lanczos out2.jpg
cc=ffmpeg -i "%1" -vf scale=w=1680:h=1050:%F1% %F2% -sws_flags bicubic out3.jpg

did i apply the filter syntax as in "-sws_flags bicubic" right in bb= and cc=

the results were all the same in aa= bb= cc=, that is
the same file sizes
out1.jpg (251780)
out2.jpg (251780)
out3.jpg (251780)

even the best jpg quality setting is not good in FFMPEG, compared to other softwares.
i hope someone could explain this

[poisondeathray]

Include it in the -vf scale

e.g.

-vf scale=w=1000:h=1000:sws_flags=lanczos
-vf scale=w=1000:h=1000:sws_flags=bicubic

lanczos.png 971kb
bicubic.png 966kb

even the best jpg quality setting is not good in FFMPEG, compared to other softwares.

It might be partially due to chroma subsampling, for example are they yuv420 (4:2:0) ?
But I agree ffmpeg's jpeg encoder isn't the greatest. For free / open source, you might get better results with imagemagick

[sommers]

SET F1=force_original_aspect_ratio=decrease
SET F2=-q:v 2

AA=ffmpeg -i QQ1.JPG -vf scale=w=1680:h=1050:%F1% %F2% out1.jpg <------- no filter or default filter
BB=ffmpeg -i QQ1.JPG -vf scale=w=1680:h=1050:%F1% %F2% -sws_flags lanczos out2.jpg
CC=ffmpeg -i QQ1.JPG -vf scale=w=1680:h=1050:%F1% %F2% -sws_flags bicubic out3.jpg

out1.jpg (251780)
out2.jpg (251780)
out3.jpg (251780)

with your certainty that worked, i found that the chosen QQ1.JPG did not respond to the filters, QQ1.jpg decided not to change by themselves. so the file sizes were the same, so it made me think as a syntax error
----------------------------------------------------------------
AA=ffmpeg -i QQ2.JPG -vf scale=w=1680:h=1050:%F1% %F2% out1.jpg <------- no filter or default filter
BB=ffmpeg -i QQ2.JPG -vf scale=w=1680:h=1050:%F1% %F2% -sws_flags lanczos out2.jpg
CC=ffmpeg -i QQ2.JPG -vf scale=w=1680:h=1050:%F1% %F2% -sws_flags bicubic out3.jpg

out1.jpg (489056)
out2.jpg (503575)
out3.jpg (489056)

QQ2.jpg responded under same syntax, now they have the different file sizes.
----------------------------------------------------------------
now i have to be careful with the JPG i choose to test

DD=-vf scale=w=1000:h=1000:force_original_aspect_ratio=de crease: sws_flags=bicubic
EE=-vf scale=w=1000:h=1000:force_original_aspect_ratio=de crease -sws_flags bicubic

i learned the newway to put DD= and EE=, DD= and EE= were the same.

thanks

and lastly what other filter would you recommend or what is your favorite filter?

[poisondeathray]

i found that the chosen QQ1.JPG did not respond to the filters

Bicubic is default, unless otherwise specified. Thus you would expect 1 and 3 to be the same, which they are

and lastly what other filter would you recommend or what is your favorite filter?

Did you mean scaling algorithm ?

There are pros/cons to each of them. It depends on what your goals are and what the scenario is

For example , if source is sharp and clean, and you're downscaling, you might choose a neutral or softer resizer. But if source was "blurry" you might choose a sharper one. There are varying degrees of sharpness vs. blurring vs. haloing vs. artifacting. You can read up on scaling algorithm discussions in other threads and boards, some are 100's of pages long. Lots of academic research and papers in that area as well