Sunny f16 rule doesn't work with 10D?

[jeremy_craig]

I was reading some posts (on another unnamed site) to the effect that

the sunny f16 rule doesn't seem to work with a 10D.

 

To recap, some posters were claiming that with ISO set at say 400, and

aperture at f16, that in a bright sunlight scene, the correct exposure

is achieved with about a 1/125 exposure time.... likewise with ISO at

100, correct exposure at f16 in same scene was found at about 1/30

second. This seems to imply that the rated ISO of the 10D is about 1-

1/2 stops faster than the actual ISO if comparing metering on a film

based camera with similarly set ISO settings. All shots taken using

the 1/400 (ISO 400) or 1/100 second (ISO 100) exposure (that you would

approximately expect following sunny16 rule) were indicated by these

posters to be very underexposed.

 

Can anybody with a 10D and a Canon film SLR either confirm or refute

this finding by metering bright sun scenes with both cameras? If true,

this IMO would very alarming news and probably turn me off from making

an investment in a 10D.

[beauh44]

Hi Jeremy, It's probably likely that the 10D might be a stop or so off from another camera or meter. I've heard that updating the firmware can help if you haven't done so already. And keep in mind that print film has *much* more latitude to accomodate something as generic as the Sunny 16 Rule - which is useful for when you're stuck without a meter using film, and remember - it was conceived long before digital photography was a reality and most cameras had internal light meters. In fairness to the 10D, Sunny 16 probably should not be used if you want accurate exposures, and unless the meter is broken, I'm not sure why anyone would want to. I just use the camera's meter and check my histogram. If it's a little off, I compensate. Best wishes . . .

[jeremy_craig]

I was hoping someone could do a side-by-side comparison to see if this is true. All the hoopla about the 10D having good noise control at higher ISOs would be meaningless if in fact you need to set ISO at 400 just in order to get camera to expose at a film-equivalent 125-160. ISOs, in my mind, should be somewhat universal, otherwise it's no longer a "standard".

[bobatkins]

Sounds like the typical BS often seen an some other unamed site!

 

I just did a test of my EOS-3 vs my 10D. At ISO 100 on green grass under somewhat overcast skies I get 1/30s at f16 on the EOS-3 and 1/20 at f16 on the 10D. My incident meter reads 1/30 at f16. Since I normally shoot with about -1/2 EC on the 10D, that gives me 1/30 at f16 with both cameras.

 

If anything the 10D is pessimistic about ISO, or at least it has a slight tendancy to overexpose highlights in high contrast scenes shot as jpegs (which is why I often shoot at -1/2, sometimes even -1 EC under really high contrast conditions).

 

The firmware update should have no effect on exposure. All it does is correct some typos in the Chinese language option and resolve some USB issues (though I haven't found any that need "resolving" myself).

[andrew_grant]

I used a grey card and a Minolta Spotmeter F to check my 10Ds meter against my Elan 7's meter. Using the handheld spotmter as the standard, the 10D was overexposing by up to 1/3 of a stop. The Elan 7 was under by up to 1/2 of a stop (the Elan is only adjustable in 1/2 stops).

[jeremy_craig]

So basically you're saying that the difference between an Canon Elan7 (film) and 10Ds meter could be about 1 full stop (ie. maybe if the Elan went to 1/3 increments it would read 2/3 "under" and it's rounding to nearest increment (being 1/2 under).

 

Anyway, I think you might be mis-using the term Under and Over exposed. From what I've read, if you actually TAKE the picture at the exposure recommended by each camera, my impression is that the scene will come out properly exposed. This is what I'm getting at... that the ISO ratings of the 10D are off of the equivalent film SLR ISO by significant margin. I think if you took the picture of the gray card with the 10D, at its metered "underexposed" reading, you'd get a properly exposed gray card, (although gray card photos are probably difficult to review objectively as they're by nature so flat).

 

Does nobody think that Andrew's approx. 1-stop difference finding is remarkable, in that the 10D's system is largely based on the Elan 7, isn't it?

[bobatkins]

I see 1/3 to 1/2 stop differences between film bodies all the time. One stop difference would be significant, but you'd have to take reading in a lot of different conditions using all the various metering modes to figure out if there is a consistant difference.

 

So far, after several thousand 10D exposures, I see no eveidence that the ISO ratings on the 10D are significantly in error.

 

Though the 10D does seem to be based on the 7e there's no reason to suspect it uses identical metering algorithms or even identical metering hardware. In fact I hope it doesn't use the same algorithm since optimal exposure for digital may be different from that of film since the two media have different dynamic ranges. If I was doing the programming, I'd take that into account.

[Ken Katz]

I am not surprised that the 10D was biased to slight overexposure, given the complaints of underexposure with the D60, and normal unit to unit differences. I don't think any of this should have an impact on a purchasing decision.

[andrew_grant]

I think you are missing the point of my post. BOTH the Elan 7 and the 10D were slightly miscalibrated compared to the Minolta Spotmeter F (I am assuming that my $380 handheld meter is likely to be more accurate than the meter in a $350 camera). Also, I used the words "up-to". The meter in the Elan 7 can only show differences of half a stop or more. The 10D's meter shows 1/3 stop differences. The metering segments of the 10D meter are also probably larger relative to teh size fo the frame. Clearly there are differences between the two meters.

 

Finally, I doubt that ANY two EOS bodies would have identically calibrated meteres unless they were specifically adjusted to acheive this.

[jeremy_craig]

ISO ratings are in fact meaningless? If its OK for 2 different same-brand SLRs to spot meter a grey card with 1-stop overall differences, then I don't see any other conclusion.

 

I understand that a Digital camera should be calibrated to expose a scene differently than a film camera -- mostly due to its more limiting latitude, BUT, this could certainly be programmed within the algorithms to STILL reflect apples-to-apples ISO equivalents to Film.

 

For example, imagine that if the most accurate $20000 meter available meters a subject at ISO 100 and gives proper exposure of subject as 1/60 second at f8. Also assume that Canon wants its consumer film bodies to match this, then the Canon film body likewise meters 1/60 at f8. Now assume that Canon's 10D returns 1/50 at f8 (set on ISO 100) as its metered reading BECAUSE Canon thinks the consumer will get higher percentage of useable returned pictures by making this adjustment.

 

WELL, my opinion is that Canon shouldn't do it this way, rather they should Adjust the ISO algorithms to be more Accurate to what their other film bodies meter. In this example, ISO 100 should really be renamed ISO 50, and ISO 200 should be renamed 100.

[johnlund]

Alarming? The sky isn't falling. Suppose you buy a camera and find that its meter doesn't agree with another camera you own. Make a mental note of the difference, tweak the ISO or meter readout (depending on what's supported by the camera), and you're back in business.

 

FWIW, my 10D and Elan disagree by 1/3 to 1/2 of a stop. Until I read several reports like Bob's I didn't think much of it. Now I think it's a curiosity, but it doesn�t get in the way of shooting. On my 10D I also tend to dial in an extra -1/2 stop.

 

It might save Canon some grief if they included a software setting to adjust the ISO or zero value of the exposure readout, but it's easy to make do without.

[jeremy_craig]

I understood your point. But I don't think the 10D necessarily mis-metered your gray card; I think entirely possible it correctly metered your gray card to provide you with a properly exposed shot.

 

My point is that the ISO-naming on the 10D is wrong, NOT that the Metering is returning unuseable shots.

 

John, my only worry on this is that to me it strikes of a little marketing deception by Canon... "ie. isn't our ISO 400 noise level great... mind you, you need to double your exposure time compared to a film camera"

[bobatkins]

The sun just came out here, so I checked my Kodak grey card with the 10D set to 1/3 stop increments. This is what I got:

<p>

<ul>

<li>ISO 100 - 1/100s @ f16

<li>ISO 400 - 1/400s @ f16

<li>ISO 1600 - 1/1600 @ f16

</ul>

Seems pretty much spot on to me!

[bruce_fuchs2]

Jeremy,

 

I agree with Bob. Until recently, I had 4 EOS bodies and there was variation between

them. Small, but it was present. If your photography demands better accuracy than

this invest in a handheld. My Sekonic is reproducible to within .1-.2 stops.

 

Also, agree with Bob's 10D experience. The meter is accurate. The experience is more

like shooting slide film than prints--be careful about blowing your highlights. The

amazing thing about the 10D is the histogram to immediately let you know if you are

holding your highlights or not.

[andrew_grant]

My 10D metered within a third of a stop of my Minolta Spotmeter. I would consider that to be a very minor difference. I use my Minolta Spot meter with my 10D and dial in the same ISO on the meter as the 10D is set to. My exposures doing this are accurate. It is therefore reasonable to conclude that the 10D's ISO numbers are correct.

 

The Elan 7's meter is within a half stop of the Minolta. It may even be within a third of a stop (the meter on the Elan only displays 1/1 stops). Since I use the Elan 7 for B&W negative film only, it doesn't really matter that much.

 

Finally the difference between my two bodies is LESS than a stop, probably not much more than a half stop. I don't consider this to be a big deal.

[beauh44]

It seems that I recall that the original Fuji Velvia is actually an ISO 40 film, rated at 50. I'm not sure that, as much as we'd like them to be, these sorts of things are absolutes. It wouldn't surprise me if a given brand of film varies 1/3 of a stop from batch to batch. That's the reason there's bracketing for film and histograms for digital cameras. Best wishes . . .

[david_kieltyka]

I'd noticed some differences in mid-day sunshine between what I expected the 10D's meter to read and what it actually did read. But this was in evaluative mode, and there's no way to know for sure what the camera is "thinking" when metering a complex scene in evaluative. So a couple days ago I did the grey card test, and my results were like Bob's. The meter was dead on, both with regard to the Sunny 16 rule and when compared with my Gossen Luna Pro.

 

In bright, contrasty conditions I usually set the 10D's exposure compensation to -2/3rds. I don't do this because the meter is "wrong" but rather because I'm smarter than it is and know better than it does what the camera will record when I press the shutter button. With digital I'd rather be safe with a bit of underexposure than risk blowing out the highlights.

 

BTW, there's a lot of bleepin' rubbish on that "other site." :-)

 

-Dave-

[jim_larson1]

Gosh. I wish I knew what you are all talking about. I am not sure I even care. . . .

 

Bottom line: **Does the camera overexpose or not?**

 

Sounds like those that know better set exposure -1/2 to -1. Sounds like good advice given the ability to dig details out of shadows, and the inability to do the some with blowouts. I will give that some thought!

 

Frankly, the 1.6 crop factor forces me to shoot my lenses completely differently because of the changed FOV and DOF. That is much more signficant of an adjustment than "gee, on my Elan at Av of 16 I shoot 125, and on the 10D I am shooting 90".

 

I find the whole concept of ISO is a bit bizarre on a digital body anyway. I recently made a Homer Simpson error and took a few flash shots at ISO 3200. Retook the same shots at ISO 100. I was impressed how nice the high ISO shots looked. After a few more advances in high ISO noise reduction, we will all be shooting everything at ISO 800 anyway.

[david_kieltyka1]

Well, ISO is still important. Not so much any more for determining image quality--I've got some real nice prints of 10D photos taken at ISO 1600 that don't look at all out of place next to prints of ISO 100 photos--as for determining the range of apertures and shutter speeds you can shoot at. If, for example, I want to take a panning shot of a train going past I need a slow enough shutter speed so I can pan with the train and blur everything else in the frame. With a high ISO I'm SOL...even f/22 won't give me an appropriate shutter speed. At ISO 100 I may still need to use a polarizer or neutral density filter to slow down the exposure.

 

In the future cameras may well have an auto ISO mode that will allow you to manually set both aperture and shutter speed. The camera would then choose the appropriate sensor sensitivity to yield a correct exposure. IMO this will be super cool if it comes to be. But it ain't here yet.

 

-Dave-

[adam_lang1]

>After a few more advances in high ISO noise reduction, we will all be shooting everything at ISO 800 anyway.

 

My wife already is, or ISO 1000 with her D60 and Neatimage Pro 2.5.