ISO facts for the Fujifilm user Locked

Started 4 days ago | Discussions thread
ForumParentFirstPrevious
Flat view
This thread is locked.
Doug Pardee
Doug Pardee Veteran Member • Posts: 9,495
ISO facts for the Fujifilm user

Aargh… the ISO wars were reignited by Tony Northrup’s popular (but incorrect technically) video and a number of response videos, most of them made by people who don’t even understand the ISO standards for digital photography.

What the ISO standard is and isn’t

ISO sensitivity ratings are not, and never were, a way to compare image quality between films or sensors. They were, and are, solely for estimating exposures.

The Japanese camera makers who are members of CIPA (Camera & Imaging Products Association) are required to rate their ISO sensitivity based on CIPA standard DC-004, published in 2004. You can read that standard for yourself, at the cipa.jp site (English PDF). CIPA DC-004 was incorporated into the ISO standard 12232 in 2006, so following DC-004 is the same as following the “official” ISO standard.

DC-004 specifies the relationship between exposure and straight-out-of-camera image files (JPEG or whatever). It doesn’t standardize raw files, which simply contain the raw sensor data that the camera’s processor used to create the JPEG (including the JPEG preview in the raw file). The ISO setting determines how your camera produces JPEGs; the raw data levels are just along for the ride.

The two options for ISO measurement

DC-004 provides two different options for desired JPEG tonality: “pleasing” and “accurate.” The various Japanese camera manufacturers have made different choices between the two.

Fujifilm, Olympus, Panasonic, and Ricoh/Pentax have chosen to use the “accurate” option, called Standard Output Sensitivity, or SOS. The functional intent of this standard is that if you use a calibrated incident light meter and set the camera (in manual exposure) to match its readings, an 18% gray card in the scene will come out 18% gray in the JPEG… provided the white balance setting is correct, of course.

Canon, Nikon, and Sony (and apparently Sigma) have chosen the “pleasing” option called Recommended Exposure Index, or REI. This allows the manufacturer to produce the JPEG however they think gives good-looking results. In practice, most people find 18% midtones to be about a stop too dark, so REI ISO numbers tend to be about one stop lower than SOS ISO numbers. Hence, for a given scene at a given exposure, ISO 100 on a Canon gives similar JPEGs to ISO 200 on a Fujifilm.

Nobody is “cheating” unless you count Sigma failing to disclose which option they’re using. The manufacturers made their choices and are doing what they say they’re doing. Four manufacturers went with SOS, four went with REI. But the Big 3 camera makers all went with REI, so obviously their choice is correct and the other guys must be wrong… after all, might makes right.

As far as I can tell, the only non-Japanese manufacturer to say which option they use is Hasselblad, who are using REI. Nobody else seems to say what they’re doing, but presumably they produce whatever JPEG tonality they like, so that’s effectively the same as REI.

Practical notes on SOS

The SOS option doesn’t say anything about how to handle the parts of the scene that are lighter or darker than 18%. This gives the camera plenty of freedom to play with the tonal curve depending on film simulation, highlight tone, shadow tone, DR setting, or the like. Only the 18% level is pegged.

The SOS option isn’t limited to manual exposure with light meters. On SOS cameras, the simpler metering modes (photometry) like Average and Spot are expected to normalize to 18% gray, so whatever you meter should come out at about 18% in your JPEG.

BUT! The SOS option isn’t applicable to multi-zone metering. When your Fuji’s photometry is set to Multi, the exposure will automatically be increased in order to achieve JPEGs that aren’t as dark as you’ll get with other photometry modes.

A quick Fujifilm test

I made some test shots with my X-T10. The subject was an 18% gray card set up in Sunny-16 lighting conditions, at f/16 and ISO 200. With the card filling the frame, Average metering gave me 1/200 second while Multi metering gave me 1/110 second. That’s just about one stop more exposure.

No surprise, that difference was reflected in the resulting sRGB JPEG. The average output level was 122 (19% gray) with Average metering, and 156 (33% gray) with Multi metering.

-- hide signature --

Co-founder of the open-source LightZone Project: http://lightzoneproject.org/

 Doug Pardee's gear list:Doug Pardee's gear list
Fujifilm X-T10 Fujifilm XF 18-55mm F2.8-4 R LM OIS Fujifilm XF 27mm F2.8 Olympus Stylus 1030 SW Canon EOS 350D +13 more
Fujifilm X-T10
If you believe there are incorrect tags, please send us this post using our feedback form.
ForumParentFirstPrevious
Flat view
Post (hide subjects) Posted by
rlx
rlx
ForumParentFirstPrevious
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow