Hallo Tobias,
Tobias Ellinghaus schrieb am 25.08.2017 um 00:09:
Am Donnerstag, 24. August 2017, 22:34:10 CEST schrieb Bernhard:
Hallo everyone,
I want to benchmark my hardware and get a clou, which module uses most
time during export. For this I want darktable-cli to export each picture
given by .xmp-files, not by RAW-Input files, by using a bash script.
In order to get insight into runtimes you might also want to pass
--core -d perf
as the last argument to dt-cli.
That gives great information - and is human readable!
This works with RAW:
for file in *.nef
do
echo $file >> benchmark-dt_$HOSTNAME.txt;
start=$(date +%s)
darktable-cli "$file" "${file%.*}".jpg --width 3072 --height 3072
--hq true --verbose >> benchmark-dt_$HOSTNAME.txt
stop=$(date +%s)
echo "time used:" $(($stop - $start)) "sec." >>
benchmark-dt_$HOSTNAME.txt;
done
But this only exports version 0 of the RAW, not Version 1 ... n created
by duplicating in lighttable.
darktable-cli doesn't know anything about lighttable or duplicates.
How do I get darktable-cli to export any version available?
Iterate over the corresponding XMP files in an inner loop and pass the XMP
filename to dt-cli like this:
darktable-cli "$file" "$xmp" "${file%.*}_${i}.jpg" ...
$i being some counter to distinguish the output filenames.
I'll try this also even though I think it's not really necessary any
more thanks to the --core parameter.
What I originally wanted to do is process a given image differently and
then compare the results by measuring the total time. So thanks for your
tip about --core!
--
regards
Bernhard
http://www.bilddateien.de
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to [email protected]
