Dslr ISO settings

[aGreyarea]

Hi all,

Just upgraded my 1000d to a 60d and am just returning to imaging after nearly a years break.

I used to set the 1000d to 800iso but don't know much about the newer sensor in the 60d and wondered what experience other users had and any tips for ISO settings on a 60d.

Thanks

[Auntystatic]

Basically you will have to suck it and see, the newer camera's are better but your first exposure will show you whether to increase the iso or time on your exposure, for me its how much grain or noise I can live, but 800 is a good starting point.

Dani

[Gina]

I've found 1600 good on my 1100D - going up to 6400 does tend to bring up the noise.

[aGreyarea]

Hi

Many thanks, I was kind of hoping there'd be an easy answer but I thought it might be test and see.

Thanks again!

[dph1nm]

Remember that changing ISO does not alter the sensitivity of the camera, so is does not affect the required exposure times.

NigelM

[paulk]

Yes it does What ISO denotes is how sensitive the image sensor is to the amount of light present. The higher the ISO, the more sensitive the image sensor.

[dph1nm]

What ISO denotes is how sensitive the image sensor is to the amount of light present. The higher the ISO, the more sensitive the image sensor.

Nope - absolutely not. 10min at ISO100 will give you the same s/n as 10min at ISO3200 (assuming read noise is not an issue and you are not saturated, and ignoring quantisation effects!). The CCD/CMOS in the camera has a fixed sensitivity to incoming photons. Changing ISO merely alters the number of photons each ADU represents (there are some intermediate steps involving electrons in there somewhere, but lets not worry about those), but you still detect the same number of photons - and it is the number of photons which determines the shot noise.

NigelM

[Gina]

Yes it does What ISO denotes is how sensitive the image sensor is to the amount of light present. The higher the ISO, the more sensitive the image sensor.

I thought so too in that the ISO is like a gain setting.

[Spikey]

I rightly or wongly think of it like this;

The ISO setting doesn't and cannot affect the sensitivity of the cameras sensor it makes the picture 'brighter' also increasing the visibility of the noise and grain. The amount of signal captured won't alter with ISO setting.

[Rob L]

So what is the benefit of shooting at a higher ISO setting?

[Gina]

I rightly or wongly think of it like this;

The ISO setting doesn't and cannot affect the sensitivity of the cameras sensor it makes the picture 'brighter' also increasing the visibility of the noise and grain. The amount of signal captured won't alter with ISO setting.

Right, I'm with you. The ISO setting changes the gain and therefore the sensitivity of the camera. It's something that happens in the amplifier part after the sensor. So yes, the sensor sensitivity itself remains the same. As with audio, the microphone sensitivity is the same but the gain control changes the overall effective sensitivity of the whole system. And as you say, the more you turn up the gain the more you amplify the noise as well as the signal.

[Spikey]

As I see it that's spot on Gina, the trick is to shoot at an ISO that gives you a decent signal to noise ratio for the exposure time.

For myself I tend to only use ISO 800 (or ISO 400 on something bright like M31) and adjust the exposure times to suit. I always tend to do a 5 min and 10 min test sub on the target to compare the difference in signal/noise.

[Gina]

So what is the benefit of shooting at a higher ISO setting?

It's like turning the gain up on an amplifier - you can see fainter objects. Conversely, when the object is bright, you can turn down the gain to stop overloading and whiting out.

[Gina]

As I see it that's spot on Gina, the trick is to shoot at an ISO that gives you a decent signal to noise ratio for the exposure time.

For myself I tend to only use ISO 800 (or ISO 400 on something bright like M31) and adjust the exposure times to suit. I always tend to do a 5 min and 10 min test sub on the target to compare the difference in signal/noise.

Yes, I do the same.

[themos]

So what is the benefit of shooting at a higher ISO setting?

The benefit comes from the reduced "read noise" at high ISO. If you're doing a quick shot to check framing, focus or merely to avoid trailing, your sky background photon count is low so your background "photon noise" is low and you might be limited by the read noise at a lower ISO.

People often think that because the high ISO JPEGS or screen previews look quite bright that they've bought some extra "sensitivity". They haven't. The lower ISO images can also be manipulated to look brighter and they would look pretty much the same (apart from the noise issue in the above para and issues of saturation)

[AstroGnome]

I might be being a bit thick here, but don't we shoot raw frames and avoid all the post sensor processing in the camera anyway? How does this sit with ISO settings?

(I miss 35mm film)

[themos]

The ISO setting is not a software setting. It's a hardware setting, controlling the analog to digital converter gain that measures the accumulated charge at each pixel, as I understand it. So we get different RAW files at different ISO settings. Dedicated CCDs have a more continuously variable gain setting.

[AstroGnome]

Gotcha, so it's pre-readout.

[iwatkins]

For the OP, I've used Canon DSLRs since day one. I've found for each different major version of the DIGIC system, you can bump the ISO up by a stop.

So for the 7D - DIGIC4 - I start at ISO1600

1D Mk3 - DIGIC3 - I started at ISO800

5D Mk1 - DIGIC2 - I started at ISO400

Etc.

New DIGIC 5 system is out but I think only is in point and shoot Canon cameras right now. If you can wait, wait for the first DSLR with it to come out. Early testing gave it a 75% reduction in noise over DIGIC 4 based systems, which is a big leap.

Cheers

Ian

[themos]

Early testing gave it a 75% reduction in noise over DIGIC 4 based systems, which is a big leap.

Is this written up somewhere?

[iwatkins]

Is this written up somewhere?

I first read about it in the Canon Pro magazine I get sent due to my CPS/CPN membership, but plenty of stuff on the 'net. Google "DIGIC 5 noise".

Planned to be in the replacement to the 5D Mk2 and to be announced any day now. That'll be around £2700. But I would expect the 7D Mk2 to follow quickly afterwards.

Cheers

Ian

[themos]

Are you sure that the DIGICs include the A/D converter? If the noise is reduced in the new DIGICs because of fancy new noise-reduction algorithms then no, thanks, not for astro!

[Gina]

I think "the proof of the pudding is in the eating" as they say - comparative tests have shown that cameras with the DIGIC4 processor have lower noise than earlier ones. I'm not concerned how they do it - I'm just interested in the result for AP. The DIGIC4 based cameras are the best so far and though it remains to be proved, it seems very likely that DIGIC5 based cameras will be even better.

[themos]

I'm not concerned how they do it

I am! We most certainly don't want "photoshopped" RAW files. Remember the problems the Nikon folks had with such techniques.

[Rob L]

Thanks for the explanations, I get it now (kinda )