Help with ETTR Please...

Pagman wrote in post #16642871
What do you see -more noise in the 100Iso shot ?

P.

I assume they are both 100%.

Tbh it's too small to tell, and based on what I can see not much difference I guess?

Pagman wrote in post #16642859
I do appreciate that and sometimes i aggree, but other times you do not have the time to flaff about, if you are sitting comfy taking a landscape almost like you are painting it then yes you would have the time to almost measure the scene for light and contrast and perhaps shoot several shots and keep adjusting till your happy, if your shooting something very fast and sudden like a fighter jet jumping at you from round a corner,or an animal suddenly coming into sight and moving away just as fast, it just One chance you get not a second chance.

P.

Again, each shot is unique to the scene. Each stop on the aperture is equal to a stop on the shutter speed or a stop on the ISO. If the fighter jet pops up, a lot of it is going to depend on what lens you have on the camera at the time and how close it is. If it is close and you have a long lens and it mostly fills the frame, an auto setting like Tv would probably be best. And if the sun is at your back you should be fine at a lower ISO. But as the light dims, it is gonna get trickier and trickier. You may only want to go down to 1/250 to avoid blurring the image and if you are already wide open on your lens, you are forced to push your ISO up. More noise is preferable to blur (in most cases). A faster lens would let you work in slightly dimmer light which would make a difference on all of this.

If that fighter jet only took up a small portion of your screen, either a spot meter or making an educated guess at the exposure compensation will be best because your light meter is going to read the sky and if that sky is bright, it will underexpose. So even pushing up +2 or +3 exposure compensation may not be enough to ETTR. If that jet is white, your meter is going to lie to you and tell you that you have more light than you do. If the jet is black, your meter is going to lie and tell you that you don't have enough light. If the jet is coming at you, you can probably push the shutter down a bit. If the fighter is backlit by the sun you have a completely different shooting scenario to deal with.

And in fast-moving situations like this, you have to be quick in making those judgments. You have to think and work fast. And that's why the pros spend a fortune on fast lenses, large sensors, and other fancy features. And that's why they get paid the big bucks.

I've seen some of your other posts and while I can appreciate what you are trying to accomplish, I really think you need to take a step back and go straight for the basics. Get a full understanding of exactly how the exposure triangle works and the specific benefits and tradeoffs of each setting adjustment. Then you will better understand how the meters work and you'll get to a point where you can do it all in your head and know almost what your settings need to be for a shot before you even pick up the camera.Really make it second nature before you try to complicate things by exposing to the right. Until you can really understand how a proper exposure is accomplished this will just confuse you.

AJSJones wrote in post #16642764
In M
Set aperture based on desired DoF
Set shutterspeed based on subject motion blur constraint
Set ISO so histogram just touches right edge of display.

(Use UniWhiBal if needed)

Pagman wrote in post #16642785
Thanks for the comment but not realy relavent, when out in the field with subject matter moving quickly, by the time i did the adjustments you mentioned, what i wanted to shoot would have been long gone, what i seek is a general rule of thumb, that i could use to help keep noise at bay.

P.

You should have much of that set anyway if you are going to take any pictures of fast moving subjects! Fast shutter speed would be required and it would already have been set; aperture would already be close, based on the subject matter and DoF range desired etc, even if not ETTR. If the lighting is changing rapidly too you need some auto mode that is capable of EC. Then the rule of thumb is "get experience in how much overexposure to use based on the scene's dynamic range". You can learn that by doing some tests before the fast moving subjects appear - do some shots with varying EC (of the surroundings under the field lighting conditions) and look how far the histogram is to the right - adjust as needed. If the fast moving subject's luminance falls within the range of the surroundings you're ready. Gotta be prepared, whether ETTR or not.

If I were doing BIF or AIF I would assume that the brightest element in the shot will be the sky background because whether it is a bird or an airplane you are seeing its belly which even if white colored is in shade. So before the B or A showed up I would do some test shots of the sky to find the combination of best speed, best f-stop and ISO that gave me a histogram where the right end of the graph touched the right side of the box and enter those settings in M and stop thinking about exposure, only about catching the B or A. If things happen too fast to adjust for them, it is better to be prepared in M than to depend on auto modes. Afterwards in LR you can use Highlights to make the sky a nice, natural blue and Exposure and Shadows for the B/A.

Hopefully Alan (aka Big Al) who specializes in A shots will be along to share his technique.

I have been through this with Pagman before but here goes again. When I'm deliberately shooting aircraft I will usually try to set the exposure up as following. I do not have an incident light meter, but I do have a fold up reflector that I got as a giveaway on the front of AP magazine a few years ago, this is supposed to be 18% grey, and it's close enough to not have to worry about. In manual mode I will meter off the reflector using partial metering (I have a 20D and thats as small an area as it will meter). The reflector is held so as it is directly in the light falling on the scene. So now I have a base exposure setting to use. For propellor driven aircraft I will set a shutter speed of no more than 1/160, helicopters usually benefit from being at or under 1/100th, for big ones (Sea King/Merlin for example) 1/60 is good and 1/30 would be better, but 1/60 is hard work at 300mm plus. For jets I like to get as close to 1/1000 as possible, but that will depend on the light and your max (non expanded) ISO and of course the max aperture of the lens, unlikely to be under f/5.6 though at the sorts of FL needed. I usually try for an ISO that will see me shooting at between max aperture and around f/11 depending on shutter speed. In really good British Summer weather with bright sunshine this will usually allow me to shoot at ISO 100 or 200.

So once I have made my choices of rough settings I will meter the scene as described and then use either +1 or +1.5 stops as my "zero" point, that usually take care of what is needed for ETTR to get a good exposure on the aircraft. Thats it I now have the camera set up ready to shoot. Even if the sun is going in and out behind cloud all it needs is for you to meter for both conditions and flick between them as the light changes. If the weather conditions are changing slowly, such as a layer of cloud building up over time, I usually check my metering intermittently anyway, it's not like it take long. An incident lightmeter would actually make this even easier.

Now as the aircraft are the subject of the images I will always bias the exposure to capturing the detail of the aircraft, so if I have to bump the exposure a bit to do that I will, I will worry about "sorting" the sky afterwards. Often the thin layer of overcast we get here in the UK in summer will make for a very plain background anyway.

Now if I'm not out specifically shooting aircraft and I hear one coming then I will usually switch to manual exposure grab a meter reading of some grass or other mid toned object (knowing that I want to use +1 to +1.5 as my "zero" setting) and go for it. Again I will know what shutter speed I want, which will determine my aperture/ISO combination, although if the camera is about I'm very likely to have a suitable ISO set for the general prevailing conditions.

If I have the camera in Av mode (which I tend to use quite a bit when just generally taking pictures) then again I pretty much know that I will need to add around a stop of +EC for shooting against the light and around a stop to a stop and a third of additional +EC to allow for the ETTR. That is usually for CW or Evaluative metering. If I'm using Partial it will depend on if the subject will decently fill the metering area, and of course on the colour of the aircraft, although it will probably mean using a bit less than the full amount of EC suggested above, which is good as the camera only offers +2 stops of EC anyway.

Of course sometimes I get it wrong and blow stuff I want to keep. I reckon that with everything added together, including hiring the 100-400L to gain reach and IS and on a good day I might have a 20% keeper rate overall. If I'm shooting with my Sigma 28-300 without IS then that probably goes down to 10%. Even with the 100-400 I'm still having to fix colour fringing in post at times, and with Sigma I probably spend most of my time fixing issues from the lens quality. The thing is that I accept that my kit is not the best for the job, and will struggle to produce results with the image quality of many of the shots some other members of POTN can produce. Yes I would like to upgrade my camera, and I saw a reasonable looking 1Ds MkIII recently for under £900, which I think would go nicely with the new Tamron 150-600 or even a 100-400 plus 1.4 extender for the way that I shoot. The burst speed and depth of the 1Ds MkIII is going to be fine for aviation, although possibly not so good for BiF.

I also accept that most of my images are only going to be posted to Flickr at 1024 on the long edge, or occasionally printed at A4 size if at all. The very best may get printed at 16"×12" so I accept that the quality of my gear is adequate for the use that will be made from them.

Alan

Which of your cameras are you using?

30D doesn't have Auto ISO.

Anyway, just stop using Auto ISO in combination with other auto exposure modes. EC cannot be applied to Auto ISO on any Canon camera, so using it will negate any EC you dial in. It's also quite hard to know what you are going to get when you mix Auto ISO with one of the other auto exposure modes.

So the first thing I'd recommend is stop using Auto ISO. Especially don't use it in conjunction with other auto exposure modes!

ETTR is really only approptiate to use with Av, Tv and P auto exposure modes. The reason for using it is to "err" on the side of overexposure. Try +1/3 stop.... that's usually plenty in lower contrast conditions, indoors or in shade. But in brighter conditions you might need +2/3.

You should still apply Exposure Compensation in the normal manner, just now including the ETTR setting. For example, if you have ETTR set to +1/3 stop and are shooting a lighter than average subject that's calling for +2/3 stop, set a total of +1 stop. Or, if the subject is darker than usual and calls for -2/3 stop, set -1/3.

Someone mentions shooting Birds and Aircraft in Flight and that's actually a different situation from ETTR. The sky is brighter than it looks and shooting any subject against it, you need to add + E.C. That's similar to what you need to do when shooting a snow scene. Neither of these are truly ETTR. These are Exposure Compensation corrections being made for unusually bright scenes that are "fooling" the refelctive metering of the camera into underexposing. You still might want to apply some ETTR on top of the E.C. corrections you are making. It's usually better to "err" slightly toward overesposure.

Big Al (above) mentions using an 18% gray target to determine exposures and, done right, that's certainly one good way to get more accurate readings in difficult-to-meter situations such as BIF or AIF. I'd only add that for the best results the target being used to take the meter reading needs to be angled similarly to your subject, so that the light falling upon the target is as similar to the subject as possible. For example, if you are shooting birds flying past you from the side, hold the target at approx. the same angle and meter it held out in front of you. If shooting up from directly below, hold the target over your head and take the reading up from below, too. Similar principles apply when using an incidence meter, it needs to be held in a manner that simulates the direction from which the subject is being illuminated as closely as possible.

Different camera models can require a little more or a little less ETTR. With my 30Ds, 50Ds, 5DII I pretty much kept them set to +1/3. With 7D I find I need +2/3 in brighter/contrastier conditions, +1/3 in other situations. These are when using Evaluative metering and mostly Av exposure mode.

You can't use E.C. with Manual (M) exposure mode. But you still can use an ETTR technique by slightly overexposing your images. You should not need to do more than +1/3 or +2/3 in most situations.

If you are looking at using ETTR with 30D for the purpose of minimizing image noise, you also might want to know that 30D is a "1/3, 2/3 ISO" model. What I mean by this is that the 1/3 or 2/3 stop ISOs are actually just as clean or cleaner than the next full ISO lower, until you get into the highest ISOs. When I was shooting with 30D I regularly used ISO 160 (instead of 100), 320 (instead of 200), and 640 (instead of 400). Those are what I call the "2/3 ISO stops". The "1/3 ISO stops" - 125, 250, 500 - on 30D are better than the full stops, too, though generally not as good as the 2/3 stops.

Up in the higher ISOs, 800 and 1250 are okay, but 1000 is not as good and the next full step up at 1600 sees a lot more noise but remains marginally usable (in my opinion) with extra post processing. I'd only use 1600 if necessary and would give my RAW files shot at that ISO special handling (at the time, processing through Canon DPP seemed best, now I'd use Noiseware on them). On 30D, H or 3200 was largely unusable (there are no fractional settings between 1600 and 3200, on 30D), except possibly as a black and white conversion where the image noise tends to look like film grain.

Take a look at this web page, especially the second chart: http://www.pages.drexe​l.edu …rawhistogram/30​DTest.html This chart illustrates the 30D's ISO performance pretty well and makes it fairly obvious why 160, 320, 640 and 1250 were my "go to" ISOs with that camera, as well as why I avoided the full ISOs: 100, 200, 400 and 800. (On this particular chart, the peaks prepresent "better" perfomance.)

Now, I don't know why this occured on 30D. I've heard various explanations or theories and I am certainly no technical whiz.

And, later camera models don't show as strong fluctuations in the ISO stairsteps as the 20D and 30D did. In other words, I've found later models to have more linear ISO performance. So I don't bother doing this with 7D, 5DII or 50D. (In fact I have my 7Ds and 5DII set to work only in full ISOs.) I also didn't do this with 10D, but that might have mostly been due to lack of knowledge on my part and ISO 800 was the highest I'd use with that camera anyway.

EDIT:

Pagman wrote in post #16642831
Here are two examples shot just, both handheld at 55mm available light no flash, both 1/6th second and f4.5 the first one is Iso 800 and with 1.5 stops of EC added in camera, the second is a base 100 Iso shot with no additional EC added but brought back up about 2 stops in post.

Both 100% crops

P.

Hey, these are shot with your Nikon!

I just looked at the EXIF and I am not sure any of the above applies to Nikon models. Certainly none of the 30D-specific advice applies.

I am not aware whether or not Nikon shooters regularly use ETTR. As far as I know, it's mostly a Canon-shooter technique and I've really only seen it discussed in relation to Canon cameras. But I don't go looking for discussion of Nikons very much, either.

I have minimal experience with Nikon but I thnk the D200 was one of their last CCD-sensored models... and not particularly good at higher ISOs.

A Nikon-shooting friend I used to do quite a few jobs with briefly used a pair of D200s. If I recall correctly, she had to keep to ISO 800 tops, and tried to keep below that as much as possible. ISO 800 was marginal and anything higher was either noisy or had to have such heavy noise reduction applied that her images lost a lot of fine detail, were only printable in relatively small sizes.

Meanwhile I was using my 30Ds (at that time) without much concern up to ISO 1250 and even at times at 1600.

As soon as the CMOS-sensored D300 came available, my friend traded in her D200s at first opportunity.

Nikon switching to CMOS with the D300 (following Canon's lead... all their DSLRs fitted with their own sensors have been CMOS) was sort of the start of an industry stampede away from CCDs. Pentax, Oly and Sony all made the same change about the same time or shortly thereafter. Only Leica kept using CCD, and even they finally made the switch to CMOS with their M that was intro'd in March 2013. Sony and Kodak used to make the CCD sensors that everyone was using. Sony shifted much of their production to CMOS, though they still make CCD for smaller formats and some vidcams. Kodak almost went bust, is still making CCD for small (sub-DSLR), video and bigger "medium format" digital cameras (Leaf, Phase One, etc.). AFAIK, Kodak doesn't make CMOS.

If I were you, I'd shoot higher ISO stuff with your 30D. Keep the D200 in the lower ISOs, where it's fine. But that's just based upon my own minimal experience with the Nikon, mostly through some friends use of Nikon models.

You really should be getting advice about getting the best out of the Nikon from other Nikon users.... preferably folks experienced with the older CCD sensor models.

Thank you Alan, mutch obliged, i no longer have my 30D just have my nikon D200, but again it was an impulse buy s/h off the net, price again was the deciding factor.


P.

Tried some outdoor shooting today no sun just overcast, i set manual Iso to 800 and set some EC to + 1.5, i must admit i got alot more flexibility with shutter speed and aperature than when i have just selected a low Iso, when i checked the raws in post all where showing more concentration near to the right of the histogram and although resting against the right end, none had over blown highlights and i could rescue them in Lr, i had to apply slightly more noise reduction, but as my shutter speeds were higher the pictures were slightly more sharply focused, so not a bad result i feel.

P.

amfoto1 wrote in post #16644101
I am not aware whether or not Nikon shooters regularly use ETTR. As far as I know, it's mostly a Canon-shooter technique and I've really only seen it discussed in relation to Canon cameras. But I don't go looking for discussion of Nikons very much, either.

I pro I know who shoots with Nikon (D4) uses ETTR

Yes, ETTR is useful on any digital camera, not just Canons.

Is there any suggestions over good ISO's as i know some are better than others if combined with Ec ettr etc.


P.

smythie wrote in post #16645244
Yes, ETTR is useful on any digital camera, not just Canons.

Yes, it is always good to maximize data capture at the sensor level. However, the HAMSTTR addition to the ETTR advice ( https://photography-on-the.net …t=744235&highli​ght=hamstr ) involves amplifying captured data that is less than maximal by raising ISO gain at the next stage in the pipeline and would be useful only if increasing ISO in-camera causes less noise than the subsequent tonal pull down in the converter reduces. Thus it is not valid for cameras in which the noise increase is linear to the ISO increase.

Pagman wrote in post #16645825
Is there any suggestions over good ISO's as i know some are better than others if combined with Ec ettr etc.


P.

160, 320, 640, etc. in the canon cameras. Don't know about others. But the differences are marginal and won't amount to a hill of beans if you continue to ignore the rest of my advice about taking a step back and actually learning how a proper exposure is made so you can really understand what is going on here.

Your image is too small for the differences in either ETTR or native ISO to make a difference. Your problem is basic exposure. In both of those shots, you have a bright white background. It is MUCH brighter than the light on the front of your subjects. Nothing you can do in your camera can change that. Even in overcast, the direction of the light matters. You may not be able to tell the difference with your eyes, but if the sun is right behind those clouds vs at your back, the camera can tell the difference.

You can increase your exposure to get the birds right but then you will likely completely blow out the background. You can do that and then crop down really close so that your frame isn't dominated by a big white blob of a blown out background, or you can get try to bring up the detail in the shadows in post but that's about all you can really do with this situation. In the end, there is just too much difference in the lighting between the front of the birds and the sky in the background. Your meter is taking that all in and trying to average it and is missing by a couple stops because of the huge amount of white space.

ETTR isn't really about noise or compensating for problems in exposure. ETTR is about maximizing the data to get the most out of an exposure that is already good. It is a technique far too advanced for someone who doesn't understand how to get a proper exposure in the first place. I don't say this to be mean or rude. I just think you need to start with the basics and get a solid grasp of the fundamentals.

kfreels wrote in post #16647043
160, 320, 640, etc. But the differences are marginal and won't amount to a hill of beans if you continue to ignore the rest of my advice about taking a step back and actually learning how a proper exposure is made so you can really understand what is going on here.

good advice.

Just for a quick clarification, I'm assuming, ideally you would use shutter speed or aperture to get the desired ettr histogram (stacked to the right without clipping). And only use ISO to accomplish it, if you had to. Or am I totally missing something, and changing the ISO is the "normal" way to get ettr? Thanks.