Noisy ISO performance on 80D?

smr

smr

Suspended / Banned
Messages
2,047
Name
Joel
Edit My Images
No
Hi all,

I'm just reviewing my photos from RIAT yesterday and I thought instead of using shutter or aperture priority I'd use manual exposure - so for fast jets I set my shutter speed to 1/1250th of a second, aperture at 7.1 and auto ISO. On quite a few photos the ISO was mostly around 200-400, which should be fine for this camera to handle. But when cropping in at a 1:1 magnification the noise seems really bad. This is ISO 320. Any ideas?

ISO.jpg
 
The photo looks soft to me and slightly out of focus, would be good to see the photo before you cropped it.
My 80D seems fine at lower ISO settings but in some cases where the image isn't sharp and the ISO is quite high, I can see the noise easily as well.
 
hashcake said:
The photo looks soft to me and slightly out of focus, would be good to see the photo before you cropped it.
My 80D seems fine at lower ISO settings but in some cases where the image isn't sharp and the ISO is quite high, I can see the noise easily as well.
Click to expand...

Thanks. I think the photo is oof but even so the iso looks too noisy at 320? Will post original later on.
 
Its not the ISO doing that. Test ISO 200 - 400 on objects that are still and on the ground with the camera on a tripod to see the actual ISO noise effects. Not a Jet going hundreds of MPH
 
Have you done any shadow recovery? A BIG mistake I used to make until recently was to underexpose to protect the whites, then pull out the shadows, not the best thing to do with the 7D I used at the time.
 
Canon sensors do have a reputation of being noisier compared to Sony Nikon ones . It could be a combination of more than factor, incorrect exposure, heat in the sensor and pixel peeping.
 
smr said:
Ok thanks so its my technique thats wrong?
Click to expand...

I believe the distance and speed of the object together with some haze and lens zoom make it impossible or incredibly hard to get a crystal clear image at 1:1.

Your not Harry Houdini with a magic camera so I would not worry too much about it.
 
If you are shooting things in the sky (I shoot birds but same applies) the camera will meter the sky, not the object. I use the custom mode (C1 on the mode dial) which is set for manual, auto ISO and one stop positive exposure compensation. I find that gives me a pretty well exposed bird.

Contrary to common knowledge, it is not the ISO that causes most noise, it is underexposure of the sensor which can be a problem at any ISO setting. Shoot in Raw, expose for the object and slightly overexpose (according to what the camera tells you). Winding the tones back in a bright image will not increase noise, brightening a dull image will show you the noise.
 
strumstrum said:
Haze is probably not helping either
Click to expand...


This I reckon. As other suggested, shoot something on the ground, still life, at the same ISO - you should see cleaner results if you are exposing correctly. Shooting planes in the sky on hot days you will often get haze or heat shimmer - the dehaze tool in LR can help a bit,
 
sk66 said:
Almost all sensors show more noise in blue sky areas... I think this is due to demosaicing and the lack of blue sensitive pixels.
Click to expand...

My new word for today - demosaicing!

Is this something the manufacturers should resolve or is it just not technically feasible?
 
Archie747 said:
My new word for today - demosaicing!

Is this something the manufacturers should resolve or is it just not technically feasible?
Click to expand...

https://en.wikipedia.org/wiki/Bayer_filter

"Bryce Bayer's patent (U.S. Patent No. 3,971,065[6]) in 1976 called the green photosensors luminance-sensitive elements and the red and blue ones chrominance-sensitive elements. He used twice as many green elements as red or blue to mimic the physiology of the human eye."
​
There are other technologies, but most cameras use the above.
 
Archie747 said:
My new word for today - demosaicing!

Is this something the manufacturers should resolve or is it just not technically feasible?
Click to expand...
I think it's just a characteristic of the RGB bayer array... there is a similar effect in really dark areas. Basically, all of the pixels get a similar amount of the light spectrum they are sensitive to (a lot in blue sky, or next to none in dark) so the demosaicing algorithm isn't entirely certain what color all of the pixels should be.
 
You must log in or register to reply here.
Back
Top