ISO settings on 5D

As a new 5D user I was interested to note the wide range of ISO settings available:
100, 125, 160, 200, 250, 320, 400, 500, 640, 800, 1000, 1250, 1600 as well as L (50) and H (3200).
My tendency is to stick with the settings I used on previous dslr, 100, 200, 400 and 800. I realise that noise increases with higher settings, but am wondering if 5D owners have noted any particular problems with the intermediate settings - 125, 160, 250, 320 etc., do they work the same way?

I shoot studio shots at 100 ISO, weddings at 400 ISO, although for bride getting ready at home I sometimes use anything from 1600 -3200 with no problems. Also same 1600-3200 inside the church during ceremony. Noise is never a problem as the grain effect often adds to the whole ambience of the shot. If grain is a problem it's sortable with various sotfware, I use the noise reduction in Bibble Pro.
These days noise is not a problem with high end slr's & yes treat it the same as film ASA/ISO. And I do use a 5D.

Do you have noise reduction on the 5D turned off or set to auto?

Is it better to have it turned off and do the NR in PP?

See this thread from a day or two ago:
http://75.126.234.18 …hread.php?t=354​967&page=2

To asnwer the OP's question - The intermediate settings are just increases in the sensor gain (just like the standard settings) and any increase in gain will cause an increase in noise, but it should be no more of an increase than a similar increase to get to a whole ISO setting.

The comment from the other thread about avoiding 1/3 ISO settings doesn't ring true to me (all ISO settings in a digital camera are the result of gain changes to the signal amplication and therefore meet the criteria of "fake" if you compare it to the natural ISO settings of film).

Mark

Mark_Cohran wrote in post #3636032
The comment from the other thread about avoiding 1/3 ISO settings doesn't ring true to me (all ISO settings in a digital camera are the result of gain changes to the signal amplication and therefore meet the criteria of "fake" if you compare it to the natural ISO settings of film).

Mark

Not when you get more noise at say iso640 then you do at iso800. The true full stop isos give better IQ than 1/3 stop. This is widely accepted. When the 5D and 30D first came out, I saw a lot of 1/3 stop iso's in exif of posted pictures. Now I mainly only see full stop iso pictures unless the user is new.

RedHot wrote in post #3636950
Not when you get more noise at say iso640 then you do at iso800. The true full stop isos give better IQ than 1/3 stop. This is widely accepted. When the 5D and 30D first came out, I saw a lot of 1/3 stop iso's in exif of posted pictures. Now I mainly only see full stop iso pictures unless the user is new.

Do you have any objective data to support that? My experience with the 5D does no support your contention. Not saying you're wrong, but your assertions seem purely subjective.

Mark

Strange as one of the first tests I did with my 5D was to try every ISO setting on a test subject, all I saw was a very gradual increase in noise, it was quite a smooth increase.

I never saw more noise at 640 compared to 800.

Nick

I would like to see samples of that to prove the point. If the image were truly affected one would think that Canon would avoid the option. And I have not seen much comment on it (which means little, actually). My inclination is to suspect that people revert to the ISO settings they are accustomed to using with film or previous digital cameras. However, I am open to the possibility that its true and am highly interested.

I think people generally find it simpler to choose between ISO200 400 800 etc.. especially when you need change it quickly, with more time I mess around with other settings, but I definately have not seen anything other than a gradual increase of noise over the entire range.

Nick

I don't have any reservations about using the "intermediate ISO points".
I tend to stick with the "traditional" ones...so if ISO 640 is spot on, then I might use ISO 800 and shoot to the right a little bit.

I ran a quick test to see for myself, got me wondering now about ISO640 & 800, well here are my results.

Reference Image:

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

RedHot wrote in post #3636950
Not when you get more noise at say iso640 then you do at iso800. The true full stop isos give better IQ than 1/3 stop. This is widely accepted. When the 5D and 30D first came out, I saw a lot of 1/3 stop iso's in exif of posted pictures. Now I mainly only see full stop iso pictures unless the user is new.

I've read that on the 30D, the 1/3 stop increments are software controlled, but the full stop increments are handled with transistors as they should be. I do not believe that the 5D and 1D series cameras handle the 1/3 stop increments the same way.

Either way, I use the 1/3 stop increments and have never had a noise issue with them... guess I'm new.

timnosenzo wrote in post #3637237
I've read that on the 30D, the 1/3 stop increments are software controlled, but the full stop increments are handled with transistors as they should be. I do not believe that the 5D and 1D series cameras handle the 1/3 stop increments the same way.

Either way, I use the 1/3 stop increments and have never had a noise issue with them... guess I'm new.

What differences there may or may not be are hardly likely to be seen or spoil an image, nobody is going to say "ahh damn, I used ISO640, if ONLY I had used ISO800".

Differences like this are only noticed when you have a direct comparison.

Nick

timnosenzo wrote in post #3637237
I've read that on the 30D, the 1/3 stop increments are software controlled, but the full stop increments are handled with transistors as they should be. I do not believe that the 5D and 1D series cameras handle the 1/3 stop increments the same way.

Either way, I use the 1/3 stop increments and have never had a noise issue with them... guess I'm new.

Reading something on the internet is not the same thing as having a white paper or a technical reference. Frankly, I don't know if 1/3 ISO stops are software controlled on the 30D or not, but it seems unlikely to me since it would be simpler to just incrementally increase the sensor amplifier gain across all ISO settings.

Like you, I use 1/3 ISO increments all the time and never noted an issue with noise. While this forum is a very useful place, there is a lot of inaccurate, or at best, not validated information that gets tossed around as "widely accepted." That the earth was flat was widely accepted many years ago - that didn't make it true. It's fine to have an opnion as long as it is presented as such, but we need to be careful when we state things as facts and should be ready to back up our facts with the appropriate references and technical data.

Mark