Using ISO and ISO invariance

[Ysarex]

I am not in agreement with that statement.. not since the Canon 5D was my primary camera, back in 2006. With "some", older d-slr models, yes, lost-ISO meant BEST quality, for sure, but since ISO-invariant sensors were developed, and since SONY developed the Exmor generation of sensors, the old admonition to "use the lowest ISO" has been rendered an old-fashioned idea.

Honestly I think that "use the lowest ISO" was advice given in good faith to beginners, but wasn't good advice as it led many to fear the ISO. Expose to the Right theory I think was pushed hard by many because it did start to push away the "always use the lowest ISO" mantra. OR at least "expose to the right theory!" was as fast to type and yet if looked into properly by anyone learning, would yield them a greater basic understanding of how to work their cameras to understand what "the lowest ISO" actually meant (since rarely did it mean ISO 100).


I also agree that even without Invariant camera sensors, the technology has continued to advance to the point where many lower ISO's are very very good. Heck at the rate we are going I figure that with enough years we will likely reach a point where the ISO setting almost becomes meaningless to the average photographer. When you can go from 100 to 6400 and get very similar performance - or heck even higher

Yes, even higher.

I have the benefit of getting a crop of new students each semester and I rarely get one who's a complete virgin. In other words they're all Youtube polluted in one way or another. They come with a set of pretty uniform misconceptions. They have their camera's set to manual because that's how you get control of the camera instead of letting the camera do it for you. As a result half of them "meter" the shot by chimping the LCD. I pound my head on the table. They believe ISO causes noise and they resit raising ISO -- I literally have to raise my voice screaming "get the bleep bleep photo for bleep's sake and raise the ISO." I tell them to put the ISO on auto and just pay attention to it. That can help but the "pay attention to it" part goes out the window pretty quick. And as for trying to explain to them the finer points of what's actually happening under the hood (as in this type of thread) -- I don't dare, sad to say. Is "use the lowest ISO to not get noise still out there?" It sure is and it does substantial harm. I get a regular stream of blurry photos at shutter speeds from 1/60th to 1/10th sec. shot indoors at low ISOs.

As for even higher:

That's ISO 12800 from my Fuji XT-2. It's a low DR scene and I did noise filter it with a light hand. Seriously, if you need it learn to use it and use it.

Joe

 

[Overread]

Do they meter with the histogram at least?
I'll admit that's how I do it - use the camera needle plus what general experience I have to get as close as I can in the lighting (admitting that if I've been shooting lots recently I'm more accurate than if I've had a dry spell and am out of practice). But the histogram I use as a final aid to double check - mostly I find its more to control and keep an eye on the highlights more than anything else. Making sure if I do have blown highlights that they are in a safe spot that doesn't matter at all.

And yep I think a lot of youtube videos the issue isn't actually the lesson they give, but that (like a lot of online learning) they only go part way. They are popular because they are the entry level tutorials so they bloat the rankings on simple searches whilst the intermediate ones get lost or missed out. So they learn some basics but not the more subtle concepts that build from them. Plus I find that intermediate learning and teaching is a major weak point online and also in many books and general references. There's a bloat of beginner stuff out there, but far less intermediate and what there is is often hard to find.

Even in books you can see the very same pattern across a lot of subjects, if your'e lucky you might find a university reading list online which can highlight some of the rarer intermediate to advanced books (and their uni-set prices which are somewhat high).

 

[Ysarex]

Do they meter with the histogram at least?

No. This isn't all of them but a shockingly large percentage really have little clue what they're doing and their methodology is to set manual SS and f/stop and start at base ISO. They then take a photo and look at the result on the LCD. If the image is too dark they know they can slow the shutter, open the f/stop or raise ISO to brighten it. So they start that process until they get what they consider an acceptable result on the LCD. They resist raising ISO. They know that too low a SS will give them a blurry photo but they're judging by the LCD and have been told they can hand-hold down to 1/30 sec. (they can't) so they frequently end up there or worse.

A typical indoor photo of say their friends hanging out is taken at ISO 400 1/25th sec f/5.6 (zoom lens wide open). It's in the range of two stops under-exposed if you examine the raw file and the JPEG is dark. They're "photographers" so they do save raw files and then they head to Photoshop to try and salvage that mess. I tell them they should have raised the ISO but they point at the noise in the photo and tell me it'll only get worse if they do that. I devote an inordinate amount of total classroom time each semester trying to undo the damage done by the nonsense they pick up before I get them and I don't always succeed.

I'll admit that's how I do it - use the camera needle plus what general experience I have to get as close as I can in the lighting (admitting that if I've been shooting lots recently I'm more accurate than if I've had a dry spell and am out of practice). But the histogram I use as a final aid to double check - mostly I find its more to control and keep an eye on the highlights more than anything else. Making sure if I do have blown highlights that they are in a safe spot that doesn't matter at all.

You're using the meter and a secondary metering aide to determine exposure. AND you're trying to push toward more exposure by checking the histogram. You're not judging visually from the LCD or now in mirrorless cameras the EVF. That ISO 12800 shot from my XT-2: I also spend some time on a Fuji specific forum and I'll show an image like that and they'll want to know, how did I do that. They don't get that kind of result. They love their Fuji camera's because they can WYSWYG the EVF with the camera on semi-auto. Well Fuji has designed their cameras to create well exposed JPEGs (WYSWYG EVF) with the sensor under-utilized by at least 1 stop. Given the low DR of that scene you can likely add another stop and that's what I did. My exposure for that image pushes the raw file right to the clipping threshold. I'm likely giving the sensor at least two stops more light than they would if they set ISO 12800 on their camera and WYSWYG the EVF image.

Joe

And yep I think a lot of youtube videos the issue isn't actually the lesson they give, but that (like a lot of online learning) they only go part way. They are popular because they are the entry level tutorials so they bloat the rankings on simple searches whilst the intermediate ones get lost or missed out. So they learn some basics but not the more subtle concepts that build from them. Plus I find that intermediate learning and teaching is a major weak point online and also in many books and general references. There's a bloat of beginner stuff out there, but far less intermediate and what there is is often hard to find.

Even in books you can see the very same pattern across a lot of subjects, if your'e lucky you might find a university reading list online which can highlight some of the rarer intermediate to advanced books (and their uni-set prices which are somewhat high).

 

[Derrel]

[SNIP>>I tell them they should have raised the ISO but they point at the noise in the photo and tell me it'll only get worse if they do that. I devote an inordinate amount of total classroom time each semester trying to undo the damage done by the nonsense they pick up before I get them and I don't always succeed.

It seems that the old idea that, "any ISO-higher-than-the-lowest-possible-ISO-leads-to-poor-photos" dies hard!