going Manual!!!!

djorijun wrote in post #12218588
I use prime on manual. Are you saying it's better to be in av? I ask cause I'm new to manual.

I prefer to use Av so I just control the aperture primarily, but when I did use Manual for low light it really made the experience better. I guess Manual is better when the camera wants to meter a certain scene different than how you want it.

For general use, Av mode is fine, if you want more specific results, use Manual

Tony_Stark wrote in post #12218595
I prefer to use Av so I just control the aperture primarily, but when I did use Manual for low light it really made the experience better. I guess Manual is better when the camera wants to meter a certain scene different than how you want it.

For general use, Av mode is fine, if you want more specific results, use Manual

Ah Thanks

Tony thanks for taking the time to explain this in newbie terms.

I really like to read your posts and the way you make relatively "complex" things simple for people like me to understand. Plus the "how to use it" parts are just great.

I use manual if time isn't an important factor such as landscape work, but leave it in aperture priority for all other times - just makes life easier and works for me.

Now that the testing methodology has been explained in easy terms.

Took 2 shots - all in-camera settings off, auto wb, one shot, metering off a mid-tone. Took both at f/8, 1/3200 +1/3EV.

ISO 1600 - washed out and noisy as expected, disturbing yes but just that disturbing
ISO 100- almost black and totally noisy, very disturbing, in fact it was very very bad.

In ACR, noise in the 1600 looked relatively light and sparsely spaced but very tight and dense in the 100 shot.

Did minimal PP - the 100 pushed 4 stops and the 1600 normally. No NR applied within ACR. Tweaked the basic sliders a bit to get the shots "balanced" for the scene.

Took both into PS for only NR and basic sharpening.

Result - Both images are good but the 1600 one feels "gooder" because it seemed to require much less NR and sharpened up pretty well. The 100 is not bad but required more aggressive NR thus a shade more blur. Sharpening was the same for both.

But for both images, sharpening did introduce a slight grain which are not normally seen in normal low ISO images.

So now what ? Perhaps the next step is test in real world use.

For example, go out in good light late afternoon, set ISO to 1600 and keep it there for the duration until evening? Lighting will keep changing quite rapidly so what would be the thing to do in those situations? Keep ISO at 1600 but open up aperture and/or reduce ss as appropriate ?

bobobird wrote in post #12226010
Now that the testing methodology has been explained in easy terms.

Took 2 shots - all in-camera settings off, auto wb, one shot, metering off a mid-tone. Took both at f/8, 1/3200 +1/3EV.

For example, go out in good light late afternoon, set ISO to 1600 and keep it there for the duration until evening? Lighting will keep changing quite rapidly so what would be the thing to do in those situations? Keep ISO at 1600 but open up aperture and/or reduce ss as appropriate ?

How could both be +1/3EV? Do you mean 1600 was +1/3EV and 100 was -3.67EV?

You can't use 1600 by day, usually, unless you want high shutter speeds all the time, or want to use a tiny aperture. The point is not that 1600 is a better ISO to shoot at; the point is that if your chosen manual exposure doesn't blow highlights at 1600, you will be better off set to 1600 than something lower.

Another experiment you might want to try is to compare the same way, but with 1600 as the lower ISO. IOW, use something like +1 EC at the camera's highest ISO, then change only the ISO to 1600. The results will be much closer, even indistinguishable, from each other, in contrast to your 1600 vs 100 test.

People often speak of higher ISOs adding more noise, but as you can see, this is not true. What is true, is that at 0EC (or any constant EC), the metering calls for less exposure on the sensor, which is the only reason that higher ISOs tend to yield noisier images, in most cameras (some compact cameras use really bad amps that really do "add" more noise at higher ISOs).

Thanks John, getting my head round this as much as possible.

As a 6 months newbie, I started out with the "easy" ways Av or Tv, auto ISO, etc depending on what I wanted and adjusting as necessary. Only recently started understanding the ISO part. But the results were neither consistent or needed way too much pp. Good shots now and then but more luck or just happened to be a bit more disciplined at the time. Not technique in my opinion.

I thought that starting easy and working up was the way to go. Apparently not for me and all I was getting was more and more frustrated. If I cannot even manage to control the camera and relate the scene to the setting what am I doing taking pics in the first place and investing all that time and money ?

2 very frustrating experiences with blown highlights and pics in heavy shade me me rethink my approach. So last week decided to end the frustration and start fresh - start with the "hard" part and work down. So went fully manual for everything. May just add to the frustration or may not, hopefully not. Either way it can only help me understand the technicalities better.

So this discussion here is helping me a lot with regard to understand the relationship between noise and ISO.

Thank you (and Tony) for hand holding me.

Will certainly try out the suggestions again in the morning and come back.

bobobird, I have been trying to follow this discussion between John Sheehy and yourself, and frankly I am having a hard time even trying to interpret the specifics of the testing that John has suggested to you. So let me offer this instead, so you can comprehend the concepts of Signal and Noise ratios.

Let us first start with the understanding that all electronic circuitry has inherent 'noise' within, and ordinarily the signal being measured or amplified is high enough make that noise effectively 'invisible'.
Next, let us also stipulate that the ISO value which the camera processor is set to amplifies the signal BUT also the inherent noise.

The result is that ordinarily as the light levels drop and we increase the ISO value of the camera, the signal:noise ratio declines.
This scenario is shown in the upper half of this chart. Note the declining signal:noise levels in row 9. The problem with the above scenario is that the number of photos striking the sensor are dropping (with with lowered light levels of the scene), so the ratio of photons to inherent noise is dropping as well...we are not truly comparing inherent noise of different ISO levels.

No other way to shoot!

Thanks Wilts. I know you from other threads and very glad you dropped in.

Signal vs noise is not new to me albeit in a computer network environment.

But the challenge for most of us is getting an image that is the cleanest, sharpest and best focused. For that to happen we need to have a correct exposure. In good light that is not an issue, we use 100 and the sunny 16 rule is fine. As soon as conditions get darker we have to compensate with aperture, speed and iso. Here is when things become tricky and we (I) start to fumble.

We think naturally of the scene and composition at all times first and the tool second. We dont think tool first then scene which would make thing easier I suppose. We wish to be "creative" but we also want it to be easy and definable. That is why many started of with Av or Tv or even P.

What I have learned here from John, Tony, yourself and others is that

- manual is the way to go for consistent results.
- on a shoot, set aperture/speed and iso for the prevailing lighting conditions
- if highlights are blowing use a higher iso and set exposure for that iso (or increase EC substantially).
- do not just bump up the iso in declining light but set the correct exposure for that iso.
- for fast declining light go as high on the iso as you are comfortable with and expose for that iso.

Am I making sense ?

Will experiment auto iso tomorrow and see how that works out. I have only ever used it in Av or Tv and wide apertures in city light environments and has not been too bad - but could be better as it is just like Auto in a sense.

Thanks.

bobobird wrote in post #12228579
What I have learned here from John, Tony, yourself and others is that

- manual is the way to go for consistent results.
- on a shoot, set aperture/speed and iso for the prevailing lighting conditions
- if highlights are blowing use a higher iso and set exposure for that iso (or increase EC substantially).
- do not just bump up the iso in declining light but set the correct exposure for that iso.
- for fast declining light go as high on the iso as you are comfortable with and expose for that iso.

Am I making sense ?

Will experiment auto iso tomorrow and see how that works out. I have only ever used it in Av or Tv and wide apertures in city light environments and has not been too bad - but could be better as it is just like Auto in a sense.

Thanks.

If highlights are blowing, what is achieved by use of higher ISO?!


Canon implementation of Auto ISO is too fraught with flaws. For example,
in panning around just now in Auto ISO mode at various light levels,

• in low light and f/4.5 in Av mode it suggests shutter of 0"4 and ISO value of 800...why the heck didn't it selected ISO 1600 or 3200 instead, permitting a more handholdable shutter speed?!
  
• in bright light and f/8 in Av mode it suggests shutter of 1/2000 and ISO value of 400...why the heck didn't it select ISO 100 along with shutter speed of 1/500?!

Thanks Wilt

You are right about the Auto ISO, scrub that. Was probably making an excuse to confuse myself more them I am already. haha

Highlights - had bad experiences in that area (flamingoes in bright light) and no amount of +EC would correct. -EC did but then everything else went dark. Did not make much difference metering off something neutral or the whites. Really need to get this down pat as birds shooting is one of my favourites and there are a heck a lot of white birds.

Edit : whatever tried and the histogram not showing anything hitting the right the highlights were still there in ACR. There are usually quite a few birds in the frame (they like to be close to each other) and if this one was not blown the the other one is etc. Single white ie a swan came out fine though. So it is either a setting thing or those sort of things are impossible to get right and as long as it is minimal the should leave it alone. But.... that niggles at me as being "unprofessional". The pics can be seen here

https://photography-on-the.net …/showthread.php​?t=1022143
https://photography-on-the.net …p?p=12148574#po​st12148574
https://photography-on-the.net …/showthread.php​?t=1023338


Everything else makes sense then ?

Wilt wrote in post #12228775
If highlights are blowing, what is achieved by use of higher ISO?!


Canon implementation of Auto ISO is too fraught with flaws. For example,
in panning around just now in Auto ISO mode at various light levels,

• in low light and f/4.5 in Av mode it suggests shutter of 0"4 and ISO value of 800...why the heck didn't it selected ISO 1600 or 3200 instead, permitting a more handholdable shutter speed?!
  
• in bright light and f/8 in Av mode it suggests shutter of 1/2000 and ISO value of 400...why the heck didn't it select ISO 100 along with shutter speed of 1/500?!

If I could pick shutter speed and f/stop, and the camera could choose the right ISO to permit me to use that combination even as the light changed up or down in intensity, then 'auto ISO' capability has value. But today it is a PHD setting that makes stupid choices.

bobobird, sometimes the range of brightnesses in a scene are simply too wide for the camera to be able to handle...the 'wider dynamic range' that most photographers would rather have, rather than more megapixels! For example, just now I found areas with shadow detail at -6.8EV below a mid tone area (grass), and in the same scene there is an area with highlight detail that I might want to keep, at +3.5EV above the same mid tone area. It would take a camera with about 11.3EV dynamic range to cope with black-no-detail thru white-no-detail!

True, but the but remains.
Maybe bracketing will help but always only remember after the fact.

Missed a fantastic shot of a bird sitting a branch with nothing but a blue sky, the half-moon behind the bird and that one spindly branch. Very nice and clean. Forgot to bracket!

Please excuse the haloing as I dont know how to deal with that (yet).

To add to this discussion a little, here is a good resource of information concerning digital photography. This page in particular covers the three things that control exposure.
http://www.cambridgein​colour.com/tutorials/c​amera-exposure.htm

And another pretty good read
http://www.cameraporn.​net …e-good-kind-of-threesome/