Canon 50D Iso settings

[FTO_gpx]

Hi,
Ive got a canon 50D which I upgraded from a 20D and I have noticed that I seem to have to use higher Iso settings to get a decent shutter speed to hand hold the camera. I am using the same lenses as on the 20D and I'm not talking about taking pictures in the dark, just in the afternoon on an overcast july day with the aperture fully open for my sigma 28-300mm at f3.5 I'm finding I need a minimum of 400 iso if not more. Where I used to use 100 or 200 on the 20D in the same conditions.

I know manufacturers have been pushing up the available ISO settings on cameras and the relative noise as a selling point but has there been some adjustment of the figures going on to enhance the claims. In other words is 400 on the 50D just the same as 200 on the 20D.
I'd be interested to hear if anyone else has noticed this or if its just me and I've got a possible fault on my camera.

[chris-p]

ISO 400 is ISO 400, regardless of which camera you use (in theory). Basically, the same general settings shout be applicable across cameras.
It's really difficult to tell if you're shooting under the same conditions though. Do you still have your 20D? If so try some back to back, tripod mounted shots with the same aperture and shutter speeds and see if you get a difference.

[FTO_gpx]

Hi chris,
no sorry the 20D was swapped for the 50D so I cant test it back to back. I know that ISO is supposed to be standard and I cant give empirical evidence of any difference. It just seems that I am always bumping up the ISO speed to stay above 1/60th at my widest aperture which I didnt do before. I have still got my film eos500 so if I put the same film speed in I should get the same meter readings shouldnt I?

Perhaps my 50D is Ok and the news stories about inreased pollution blocking sunlight (global dimming) is happening and this is the evidence- higher ISO! (LOL)

[chris-p]

The thing is, all of the ISO settings are "estimates". If you actually shoot test charts under highly controlled lighting conditions you can measure the actual ISO of a camera (what a dull life).

For example, for the EOS 50D and 20D the measured "actual" ISO values are (listed with the ISO setting on the camera):

Indicated - EOS 50D real ISO
100 - 157
200 - 158
400 - 306
800 - 618
1600 - 1172
3200 - 2342
6400 - 4433
12800 - 9210

If you look at the same data for the 20D, the real ISO values are actually slightly higher (177 real ISO when the camera is set to ISO 200) so that could certainly explain it.

[FTO_gpx]

Thanks again for the reply Chris, it certainly does go some way to explainining it. It also seems to confirm what I said at the start that the estimation is getting more liberal so that it appears the camera has a larger range, as your figures show at 800 a real value of around 600 and all the other higher values similarly 25% over estimated. Especially as can say with a new camera they have better noise at higher ISO and a larger ISO range they are always going to get that if they can fiddle the figures. I'd be interested to see how other manufacturers compare and if the estimations get more acurate the higher you get up the range. Is there anywhere that does a comparison or I could get the figures from ?

Thanks
Adam

[GeoffWessex]

There's some more 'explanation' in this DP Review Editorial. It's all very technical stuff but it's saying that, as a piece of electronics, each sensor would need to be 'tuned' to do certain things with different amounts of light (and also to 'tell the camera' that it's working at a particular sensitivity). It's probably fair to say that the sensor manufactured in their thousands and would be fairly reliably all at the same sensitivity. A different sensor, even from the same manufacturer, may not follow quite the same pattern and be 'off' by what we might translate to 'a few ISO points'. If you used film you may remember that they had different sensitivity characteristics and some would require perhaps half a stop more or less exposure. In the end, though, we're talking about tiny differences and - if you think your exposures are too bright or too dark, adjust accordingly, with Exposure Compensation..... using perhaps 1/3rd of a stop either way as a regular setting.

[chris-p]

Originally Posted by FTO_gpx

Is there anywhere that does a comparison or I could get the figures from ?

DxO Labs are the people. Their tests go into minute detail and they check every aspect of image quality in lenses and sensors. They completely ignore all aspects of the camera or lens which don't relate to the image quality so you get some very interesting results.

Originally Posted by GeoffWessex

In the end, though, we're talking about tiny differences

Thats the thing, we're really not. The actual sensitivity you get when you set the camera to ISO 6400 is actually closer to ISO 3200.

This is often the problem with the sort of work people like DxO labs do. The question you have to ask is "does this actually matter?". Almost everything in this field is an estimate. Focal lengths aren't exact, aperture values aren't exact (and this is compounded by the inaccuracies in focal lengths). Even shutter speeds are "best guesses".

[Oly Paul]

Originally Posted by chris-p

DxO Labs . The question you have to ask is "does this actually matter?". .

Not to me it does not, but i guess it does to manafacturers marketing techniques.

As most magazine reviews obsess about High ISO noise then over claiming your ISO values is a neat trick to give you a marketing advanatage

I remember when the 450D came out and reveiwers raved about its lack of noise at ISO1600 compared to others, the fact that it was actually a little over ISO1000 might have had something to do with how it faired in comparason. And from your figures for the 50D it looks like Canon is playing its favourite game again to negate the higher pixel density

Mind you other manafacturers are catching on to Canons marketing ways.

For example all Olympus ISO's used to be nearly bang on and always higher than claimed for years

E-520 Claimed ISO 1600 true ISO 1702

E-620 Claimed ISO 1600 true ISO 1043

[FTO_gpx]

Hi all, thanks for the replies. The DxO stuff makes interesting reading. As you say Chris the differences are not just small discrepancies that can be put down as part of a normal tolerance that you would get in any production process. In a way it doesnt matter what they are because you have a camera and it works by balancing aperture, shutter and ISO to get a good exposure and the actual values don't matter except to allow you reproduce the results given the same conditions with that camera. Everytime you get a new camera you workout by trial and error how you need to alter your settings to achieve the same results but isn't that why the ISO system was started, so you didnt have to go through all the trial and error everytime you tried a new film, its suposed to be a standard. So it does matter because as marketing men fiddle the figures to make their new model appear better then it really is, other manufacturers do the same but by ever larger amounts for each new model and very soon we end up with ISO being a meaningless figure. Which then makes it impossible for consumers to compare brands and models when buying.

[chris-p]

Originally Posted by FTO_gpx

As you say Chris the differences are not just small discrepancies that can be put down as part of a normal tolerance that you would get in any production process.

Part of the ISO system (not just sensitivities but the whole quality accreditation system) requires known tolerances and acceptable variance, effectively margins for error. I don't know what these are for sensor speeds but, in my field at least, these would be unacceptable!

Originally Posted by FTO_gpx

In a way it doesnt matter what they are because you have a camera and it works by balancing aperture, shutter and ISO to get a good exposure and the actual values don't matter except to allow you reproduce the results given the same conditions with that camera. Everytime you get a new camera you workout by trial and error how you need to alter your settings to achieve the same results but isn't that why the ISO system was started, so you didnt have to go through all the trial and error everytime you tried a new film, its suposed to be a standard. So it does matter because as marketing men fiddle the figures to make their new model appear better then it really is, other manufacturers do the same but by ever larger amounts for each new model and very soon we end up with ISO being a meaningless figure. Which then makes it impossible for consumers to compare brands and models when buying.

You're right. It makes no difference to you and me and anyone with a camera because, as long as the meter works, you can get a metered exposure.

From a sales and marketing point of view it's, in my opinion, of questionable honesty.