cameramanben
Well-known member
Hi All,
this thread should really read 'The World's Most Expensive Lightmeter' - MLUT. The LUT's free, the camera to bung it in, the expensive bit! ;-)
I recently bought myself a DJI Osmo, and one of it's options is to shoot what is listed as 'D-Log'. The thing is, I can't find any tangible details about D-Log, so I thought I try to measure it.
To that end, created an MLUT for my Sony which takes S-Log3 in CineEI mode and maps each stop to 10% IRE, with 18% mid gray mapping to 50% IRE.
In other words, each 1% IRE on the waveform or zebras equates to 1/10 of a stop. Best accuracy is between stop -3 (20% IRE) and stop +5 (100% IRE). Rounding errors on interpolation start to drop things away below 20% IRE (though it is also hard to read zebras at that level!) and between 100% IRE and 107% IRE (the limit of zebras).
If you set zebra aperture to 1% and are consistent in how you read zebras it works really well. I set three daylight panels on a gray card, chose a zebra level, then dialled the lights so that zebras in the centre of frame were not quite appearing 1% above my target zebra. I then set my Osmo to manual, ISO 100 white balance 5500K (as with the Sony) and took clips at each shutter setting. I then repeated, taking my reference zebras down 1% each time between 100% and 90%. I then did the same yet again between 70% and 60%, to cover the whole potential dynamic range of the recording. Everything stayed nicely consistent; I reckon 1/10 stop is an achievable accuracy over the -3 to +5 range.
Not sure how generally useful it is - I could imagine checking the level of vignetting on lenses, or getting an idea of the realtive 'T'-ishness of lenses in a kit at a given 'f' stop. Anyway, since it seems to work, I thought I'd post it in case anyone found it handy. Instructions are in the zip file.
Dropbox link:
Sony S-Log3 LightMeter User 3D MLUT
Ben
this thread should really read 'The World's Most Expensive Lightmeter' - MLUT. The LUT's free, the camera to bung it in, the expensive bit! ;-)
I recently bought myself a DJI Osmo, and one of it's options is to shoot what is listed as 'D-Log'. The thing is, I can't find any tangible details about D-Log, so I thought I try to measure it.
To that end, created an MLUT for my Sony which takes S-Log3 in CineEI mode and maps each stop to 10% IRE, with 18% mid gray mapping to 50% IRE.
In other words, each 1% IRE on the waveform or zebras equates to 1/10 of a stop. Best accuracy is between stop -3 (20% IRE) and stop +5 (100% IRE). Rounding errors on interpolation start to drop things away below 20% IRE (though it is also hard to read zebras at that level!) and between 100% IRE and 107% IRE (the limit of zebras).
If you set zebra aperture to 1% and are consistent in how you read zebras it works really well. I set three daylight panels on a gray card, chose a zebra level, then dialled the lights so that zebras in the centre of frame were not quite appearing 1% above my target zebra. I then set my Osmo to manual, ISO 100 white balance 5500K (as with the Sony) and took clips at each shutter setting. I then repeated, taking my reference zebras down 1% each time between 100% and 90%. I then did the same yet again between 70% and 60%, to cover the whole potential dynamic range of the recording. Everything stayed nicely consistent; I reckon 1/10 stop is an achievable accuracy over the -3 to +5 range.
Not sure how generally useful it is - I could imagine checking the level of vignetting on lenses, or getting an idea of the realtive 'T'-ishness of lenses in a kit at a given 'f' stop. Anyway, since it seems to work, I thought I'd post it in case anyone found it handy. Instructions are in the zip file.
Dropbox link:
Sony S-Log3 LightMeter User 3D MLUT
Ben
Last edited:
