Is f/1.4 at ISO 200 really f/1.4 at ISO 200?

Luminous-landscape has a very interesting article on how camera manufacturers are silently increasing the ISO in order to compensate for light loss from the lens. The whole analysis is based on data from DxOMark:

click on graph for larger view

The above graph show the light loss at the sensor level for different Canon and Nikon cameras when using a f/1.4 lens wide open. For example, the D3s has -0.30 EV light loss, the D300s has -0.52, the D40 has -0.80, etc. In order to compensate for this light loss and produce correctly exposed images, camera manufacturers are compensating by "secretly" increasing the ISO. The author Mark Dubovoy goes even further by saying that you may get better results with smaller aperture lenses, since they have less light loss and that this phenomenon is also destroying the bokeh:

"One might be better off purchasing smaller aperture lenses and increasing the ISO.  Since these lenses have much less light loss at the sensor, one may well end up with virtually indistinguishable results. In fact, is not even clear that large aperture lenses will deliver a shallower depth of field as intended. The DxO measurements to date prove that the marginal light rays just don’t hit the sensor. The point regarding depth of field is that these rays are also responsible for a larger blur spot when out of focus. If they are lost, they not only don’t contribute to the light intensity at the sensor, but they also don’t blur the out of focus planes as much as you would expect at wide apertures."

Update - related links from the Web:

This entry was posted in Other Nikon stuff. Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • Just A Thought

    You’re a Sports Shooter. Indoors and outdoors and evening games.
    You use a DX camera for the reach and a newer model for better High Iso.
    Even so you still need more light – especially in dim gyms and evening football, soccer and baseball.

    The apologists are posting here that there is nothing to see here or that this issue is lauaghable abd that in thier mind it is a good thing. If you shoot pics of the family cat. then that maybe so. If you are the Sports Shooter (aswell as concerts, stage performances, shooting on movie sets, weddings in dim churches) mentioned above needing faster glass, it can be a big deal and even if shooting FX. So lets check prices at B&H:

    AF-S VR NIKKOR 200mm f/2.0 G AF-S IF-ED Lens
    # Price: $4,799.00

    Telephoto Nikkor 180mm f/2.8D AF ED-IF Autofocus Lens
    # Price: $899.95


    Telephoto EF 200mm f/2.8L II USM Autofocus Lens
    # Price: $739.00

    Telephoto EF 200mm f/2L IS USM Autofocus Lens
    # Price: $5,300.00

    Because of this apparent problem a f2.0 lens on the Sports Shooter’s camera performs as fast as a f2.8 lens. Tell the Sports Shooter who bought the f2.0 glass that this is a good thing or that the problem is laughable or that it is not an issue and that the photog should concentrate more on taking a good pictures.

    • There’s more to the price premiums than just aperture.

      Those charts also show that the higher-end your camera, the less this affects you (we’re talking film levels)… which basically excludes this being an issue for anyone who is/should be buying ultra-high end glass.

  • John

    Let’s ask Chase Jarvis or David Hobby or McNally if they care or even see the difference.

  • WOW! The third link here ( is on the money. This mythical phenomenon has NOTHING to do with bokeh, which describes blur circles across at least several pixels.

    And the angles that light hits a sensor in a camera with a mirror box are so close to 90 degrees that any arguments about light loss are stupid. With extreme angles like in an M9 there is a real reason to shift the microlenses. Not so with SLR type cams with significant flange distances.

    Large aperture lenses do suffer vignetting. That is real and has nothing to do with sensor design, layout or amplification.

    This is an old argument from like 10 years ago. Really.

    Stupidity of this level is kinda frustrating. But then again, it makes competition easy. : D

  • Really. People would be very surprised at what goes on with the image file processing to get the raw files we now get out of our cameras. Much of the improvement we have seen in imaging over the last few years is due to both sensor improvements and signal processing.
    Who cares what happens with signal processing as long as we get the images we want.

  • FakeKenRockwell

    This is idiotic. Repeat after me. ISO in digital is NOT a measurement. It is an index relative to exposure value. There is no speed as such in a digital sensor. You are always amplifying signal or attenuating it. There is no magic real bare iso value from which you are boosting. Film movie makers care about t stops because the NEED to know how different lens lose light. ALL lens lose light, period.

    Part of the problem is that in english we call lenses “fast” versus “slow.” it creates an imprecise expectation in our collective silly brains. Fstop is not a measurement of speed. It just tells you how big, physically, one’s aperture is independent of any light that reaches the sensor. If your cameras is automatically correcting for effective light loss, you should be cheering. This is a feature not a bug o in the old days you would wind up just bitching like film makers stressed over T stops because you’d have no idea why one lens behaves one way and another doesn’t.

    • i think… I totally agree with this.

    • taotejared


  • Not sure they do right conclusions from these data. Just thoughts from one czech photographer: there is light loss at sensor level probably given by sensor construction. Dense sensors (high megapixel count at small size senzors – small cells) seems more senzitive to this loss. So manufacturers needs to compensate this by raising GAIN (by the way – there is GAIN number in EXIF but i didn’t studied it – and not interested enough to do so now). There is maybe also some more loss when more diffused light comes at sensor from wide open fast lenses, so there is maybe some more gain used when you use these fast lenses and it means possibility of little more NOISE. But you still can mount some realy very old fast lenses and when used properly they give quite predictable results.
    By the way – ISO is just standart to compare senzitivity. But in digital era it’s not senzitivity of senzor, but of let’s say result. Cameras do many tricks on these senzor data before presenting them to users. Just try opening some RAW files in dcraw without anything and remember they are possibly already processed before saving!

    So what’s conclusion? Some DxO an LL marketing? Am i wrong?

  • Adam

    It may have been said but the article posted here is a little misleading. The cameras are changing ISO to compensate for light loss by the sensor. Different sensors work better or worse at gathering light. Thats one big reason why Id love to see backside illuminated SLR sensors.

  • QQ

    Frankly I don’t understand all the people trying to convince you that this doesn’t matter.

    This does matter, a lot.

    Not the fact that cameras try to compensate for light loss at lens/sensor. It’s a good thing they do that. What is very bad is that they hide it.

    Each camera should have a setting – compensate for t-loss using sensor gain? yes/no

    yes – they work like they do now
    no – they compensate using exposure (shutter speed)

    furthermore, if it’s “yes”, it should recording the actual “iso” speed used including compensation.

    What they do now is make us think that their lenses and sensor are perfect, losing no light, while in fact they do. Sometimes it would be preferable to have longer shutter speed instead of increased gain (noise, DR, etc come in mind).

  • Adorama had a test at 2009. february. There were interesting numbers too…

  • Back to top