Another important point you missed in regards to the Z9 specifically is it’s dual gain sensor basically shooting at iso 200-400 will yield an inferior image as opposed to going right to ISO 500. So for me anything over ISO 200 I go straight to ISO 500
I just laid this out in more detail in my comment. There are several issues missed here - the most important of which is that ISO is not part of exposure at all, but gain/amplification of the S/N . Also, as you mention, the sensor of the Z9 is not truly ISO invariant - it's variant at ISO 500 in stills, and elsewhere in video.
Scott, we’ll done. You presented an excellent explanation of ISO invariance. I am in full agreement with you. However, there is one other aspect about underexposing your images by using lower ISOs that most don’t consider. Whenever you raise your ISO, you are also reducing the available dynamic range. In some cameras, when you reach ISO 1600-3200, you cut your dynamic range just about in half. Besides saving your highlights, a good reason to keep those ISO values lower and raise the image brightness in your editing software. Again, thank you for the video.
Thanks Dennis, I think I briefly mentioned it here, but maybe I edited that out. Because I am not "tech" guy, sometimes I worry that someone will nitpick info. I try to do my research before making any video and double check my understanding.
@@WildlifeInspired You did well. No complaints. You have an ‘easy’ way about you and a clear way you present information. I wish you even more success.
Every stop you raise the ISO you reduce the DR capability by one stop. Think of it this way, DR is the difference between the darkest and brightest values a sensor (pixel/photosite) can record. If the brightest value exists in the scene, then adding one stop of analog ISO amplification pushes it one stop into overexposure (ADC clipping)... I.e. the minimum is still the same, but you've pushed the maximum over the edge.
We have know about this for Nikon Sensors for a very long time -- Lou Coetzer used to lecture on the subject -- and we were all taught to do shoot at EV-1 (or more) to gain 1 stop for free. He argued we should shoot with a faster shutter speed to freeze motion and the correct aperture to gain the DOF needed AND sacrifice ISO. When Highlight Protection/Priority/Weighted protection became available we gained from not blowing out highlights, but the results varied from shoot to shoot. NOW we know with the Z9 sensor that we can shoot to minus 5 EV and we can recover the image easily. BUT, this takes no account of when AF works the best -- which is when the subject is correctly exposed. So shooting consistently with minus 1 EV can help: BUT, with tools like DxO Pure Raw 3 and Topaz AI (denoise or photo) the end results can be spectacularly great even at ISOs over 10,000. Canon sensors are known NOT to be ISO invariant. Whereas one expects Song and Nikon camera to behave similarly.
You raise good issues, but neglect to appreciate that as ISO increases, you're also loosing dynamic range. If you're shooting a flat scene, it may be much easier on the eye when post-processing (and not printing) images of ISO 12,800, but if you print - depending on size - this becomes an issue.
As a wildlife photographer I've been shooting my Z9 in manual with auto ISO for about 2 months now. This video makes so much sense that I'll be using it here on in. Thanks for the tip and keep them coming.
At last! Your explanation is the clearest, at least for me, as I have watched so many videos about it and even as they are technical and detailed, I couldn't really grasp it!... THANK YOU! Yes, does make sense, I would use it definitely in the field and yes my camera it is, Z9...
Awesome video again!… I used the iso invariant concept for night photography. When I first listen to your video with the use of this technique, I was worried about seeing the image so dark that I would have a hard time to judge focus. But you came right out and proposed 1 or 2 stops and it makes so much sense. Why did I not think of that… ;-) I will definitely used this technique. Thank for sharing this great advice!
Scott, thanks for the videos and for making me a better artistic photographer. Using the Z9 metering set to Highlight Priority in Manual with Auto ISO, is the way to go! I've enjoyed successfully shooting that way for some time. Much appreciate your thoughtful videos!
Agreed! If I do use auto iso, I use the highlight weighted as well. They go hand in hand for me... Otherwise I use spot metering. Never liked matrix or center weighted
A very good explanation of ISO invariance - I'm sure any newcomer to this concept will find this a very useful instruction. I've been taking advantage of ISO invariance for a year or so - especially when I'm photographing certain waterfowl against a dark background. I find it works particularly well if you also apply a custom-made linear camera profile to your image at the start of the edit. It'll look flat and lack saturation so you have to push contrast and colour a lot more to get where you want to be - but the linear profile puts you in control of the edit instead of Adobe's canned profiles, which have way too much contrast across the board in my opinion which makes highlights harder to control. It's a great way of controlling contrast because you have more control and can even work selectively.
I've been using the ISO variance method since Nikon brought out the D750 going on to use it with the D500 ...... shooting auto ISO in manual mode at around -2 stops or more .. bringing up the individual shades of lower light while keeping full control of the whites for feather detail.... IMHO here should be only one fully white exposed part of a bird in an image and then only just ..... using some sliders but mainly the adjustment brush ...sometimes there are a dozen AB pins or more in my bird images for both light and dark areas ..... good to see a video about this ......ISO variance rules!
Thanks Scott. I've known about this and frequently get excited in the field and forget about it. I know my camera is iso invariant and am going to experiment tomorrow morning in a Grouse Blind. We got rained out this morning and the forecast is not much better tomorrow. In the morning I am going to concentrate on this to set it firmly in my brain. I already have plenty of images of Grouse in poor light. This frees me up to break new ground.
Been meaning to try highlight metering. I forgot the z9 is invariant and haven’t really experimented too much with that. Will have to give it a try, especially this summer with puffins, razorbills and loons!
Great job kind Sir. I shoot portraits, events and travel. What you just shared is so on point. You just gave me another reason to love my Z6ii. Not that gear matters. It does if you’re looking to implement this technique. Lol. Thanks again my friend. Cheers
Good job explaining a complex subject. I used spot metering with auto-iso with my DSLR but have found that matrix works better with my z9 using exposure compensation. I will give the highlight priority metering a try. Thanks!
OK, I just heard your first sentence and already I agree. When I'm shooting small birds I almost always keep my histogram to the left. I've found that I get more accurate colors, better small detail, and I don't wash out the bright feathers with over exposure. I almost always shoot with manual exposure. I say almost always because sometimes in woodland settings the light changes too quickly from one spot to the next as the bird flits around. Then I use auto ISO sometimes. This morning was overcast and rainy. I shot a house finch two stops underexposed. When I put it in PS there was the beautiful raspberry red throat and head along with the rich browns and sharp details of fuzzy feathers. The good thing was the way the background dropped back and the bird stood out in all of its splendor. Light meters are easy to fool for best bird exposures.
Scott, well done and well explained. Here’s an idea for a much needed video. I have been a photographer for many years and my photos are very good. The issues I have are with printing. I use a high end Mac to process and my colors seem to print well, BUT my exposure prints way too dark and not consistent. I have my screen brightness down to 2 out of 10 and some photos are still printing too dark. The exposure in camera is usually pretty good and the histogram shows good exposure. I send files to the lab for test prints and they come back consistently dark. HELP!!
I almost got there myself by shooting a bunch of Great Egrets (white as can be) against middle tone backgrounds. By dialing in underexposure a finally got photos with detail in the birds. What I haven't done is routinely underexpose EVERYTHING. I'll give it a shot, well maybe a million shots, and see how it looks...
OK, I've watched your entire video now, Scott. Only thing I have to add to my original comment is that wildlife photographers need to watch this, take it to heart and start practicing it now. This is an excellent video, Scott, with great information and exceptionally well explained. Wildlife photographers who learn this will surprise themselves, as I did, with the improved quality of their images.
Helpful commentary! My cameras are not ISO invariant and I wasn't certain what this meant, although I had a general idea. I have had instances photographing light colored hawks in the sky in golden hour where the camera blows out the highlight details. Sometimes this happens even when the histogram does not show any blown out highlights (or at least not enough for me to notice it). Although my cameras aren't ISO invariant, I still underexpose in such situations. One minor comment: ETTR does not always involve overexposing. In fact, sometimes to preserve all sky details in high contrast landscape sunsets, I have to go to around -3 exposure stops when exposing to the right. The histogram is then skewed (most of the area is to the left but there is a long tail of brighter points, which reach right towards the right end). (I would still consider this exposing to the right because I am pushing the exposure as high as possible until highlights are clipped.)
Great video Scott. It makes perfect sense and I am going to start shooting about 1 stop underexposed to see how that might improve things. Canon doesn't have the highlight metering like Nikon unfortunately, but shooting underexposed should help regardless. It's always the whites on the birds that get me! By the way, that Hoodie image is spectacular -- I can't believe the detail!
ISO Invariance came into play for me once I started astrophotography. Although I wasn't aware of it at the time I bought my D750, it turned out to be a very ISO invariant camera, and great with low light, extended exposures. Since that time, I started feeling more comfortable in most nighttime shooting situations with exposing for the highlights and keeping a higher shutter speed, especially with hand held night street photography. With wide angle astrophotography, lower exposure times can avoid having to lug a star tracker along, and I am okay with higher ISOs to get as much light as possible and cleaning up noise with stacking in post. I would say that I have found the Z9 to exhibit noise in low light situations more than my Z6ii or Fujifilm GFX 100S, but once one gets to know the threshold of the camera, it will still extend your low light capabilities much more than one might have expected. I have found that the GFX's dynamic range to by surprisingly good in low light, and routinely underexpose my shots with it, and even though it is a beast to carry, I have used it for landscape night sky shooting with and without a tracker to good effect. Thanks for sharing!
I switched from Canon to Nikon about 3 years ago when it was obvious that Canon wasn’t going to release an ISO invariant sensor. I had kept my 5D Mk II for far longer than I should have
Great info there. I’ve been using this technic myself for wildlife. Only potential issue I come across from time to time is the autofocus isn’t optimal if the autofocus point is in an dark part of the image (say a black head of the bird )
Excalty.. well said. In fact, I dont even use auto ISO. I have my camera setup so I can change ISO, SS, and Av, each with a separate dial. Mostly because I don't trust the Auto iso. I feel (on my A7RIV) it ranges too much. For the subject I'm photographing, I know what shutter speed I want/need so all I need to do is roll the ISO to adjust for exposure keeping an eye on the blinkies. When that's too much I'll then change shutter speed. This is the BIGGEST advantage to mirrorless and the EVF.
I agree whole-heartedly with your suggestion to slightly underexpose. I don't shoot a lot of wildlife, but I do shoot thousand of photos at dog shows as an enthusiast (I recently shot over 40,000 frames in six days at a National Specialty Show where I had been invited to shoot for the club.) My wife shows Dalmatians, which are white with black spots or white with liver-brown spots. Often the events are outdoors under the midday sun. I've been shooting underexposed for a couple of years now, knowing I can boost the exposure in post (LIghtroom Classic is my go-to for post-processing) without losing the data detail in the white of the dog's coat. Sometimes the problem exists when events go inside until lousy florescent lighting backlit by huge windows with very bright light shining through them. I love those times when we have an early or late ring time when I get to shoot in the Golden Hour, but unfortunately, I have little control over the time of day that I'm shooting or the direction the judge is asking for the dog to be moved, with zero consideration for what that means for the photographer. 🙂
I think protecting the highlights just makes sense. Not saying that everyone should go out and shoot under 3 stops but up to 1 stop has never once hurt me.
Very interesting and informative. I am amazed my 8 year old Canon 6D can do so well at bringing the shadows, but this is mostly landscape with low ISO. Thinking back, if I do have exposure problems it is blowing out the highlights and overexposing ½ stop sure didn't help. When in Shenandoah NP, I will do, or be ready to do wildlife under a variety of conditions and manual with Auto ISO is my go-to setting. In once instance I was on a dark trail with lots of laurel and rhododendron and there was an Appalachian Cottontail on the trail in deep shade but filtered sunlight. Oddly, this rare species is not at all wary. I laid down and took many photos at 15 feet. My main concern was focus and they were extremely sharp. But the highlights were way too bright. Still good photos but had I underexposed by a stop (and I had time to change setting), the photos would have been better. Bringing up the shadows in this ancient camera was not a problem at all.
I totally agree thats not a big deal to recover shadows with modern cameras, but you always have some exceptions. I shoot R5 and recently was trying to get a shot of black raven in flight. The evening sky looked great and as far as bird was so black tuned +1 expo correction. I was so surprised that even that didn’t help. It was absolutely black with no chance to recover any details from shadows. But yeah, in most cases some underexposing can save crucial highlights.
I think that the problem is, in most cases, lack of light (brightness). If you have a contrasty scene you gill get a noisy image. Under / overexposing is just a tool that might help in some situtations, but black bird in flight during evening (=getting dark) on a much brighter sky (it also requires fast shutter speed) so you get very high iso (and drop of contrasts in post). I think advise regarding +compensation in post works best when you deal with a subject that is not black (or at least eye, or any crucial details are not situated at black parts or subject has light-dark ptterns or mentioned details are colored differently than black).
You can raise exposure with two thirds of a stop when using the highlight weighted metering mode, there are few to none scenes you will blow the highlights.
Interesting video and thanks for the great explanation. I was under the impression that for ISO invariance you need to be working with 'real' iso values not the extended ones and for the most benefit using the base ISO up to the dual gain point. For example, with a Sony a1 ISO 100 is the base so invariance is best from 100-500 when 500 is the dual gain point where the dynamic range jumps back up to something like the same as ISO 200. The Z9 uses the same sensor with the same dual gain point (at 500). I tend to use zebras (Manual + auto ISO) with ec to control exposure (and the highlights).
A couple minor points... they don't change the basics of the idea/application; but are more technical/accurate. Noise is primarily caused by the lack of light recorded; which is why it is more apparent in darker regions of an image. With a non-invariant camera raising the ISO *reduces* the recorded noise level (i.e. the noise penalty for underexposing). But with an invariant camera it does not affect the recorded noise level. In either scenario, an ISO underexposed image is not *less noisy*... it is either just as noisy (invariant), or more noisy (non-invariant); it is just harder to see the noise level because it is darker. In general, ISO is not exposure and it is not *sensitivity.* In terms of signal (SNR/noise) it is more like a volume control... if you have a crappy/noisy signal, making it louder just makes it more apparent. In terms of visible noise, it is more like the brightness control on your monitor. Most sensors only have one sensitivity to light; base ISO. However, specific to the Z9, and other cameras with dual gain/dual ISO design sensors, there are two levels of sensitivity. The Z9 switches to low gain mode at ISO 500 which makes the sensor actually ~ 1 stop more sensitive to light than it is at base ISO. It also has less capacity in low gain mode... I.e. needs less light and can accept less light (not a problem when there is less light available). So with a dual gain sensor you have to be careful about that sensitivity step... for example, with the Z9 recording at ISO 400 and recovering 3 stops is a little worse than recording at ISO 500 and recovering 2 stops. IMO that difference is not really worth trying to manage in highly variable scenarios. But if you're working in lower light situations with auto ISO (for example) it could be beneficial to set your minimum ISO to 500. photonstophotos.net/Charts/PDR_Shadow.htm#Nikon%20Z%209
I've never seen a sensor with triple ISO... I have seen some (Z9 included) that use digital manipulation in conjunction with the dual gain to use different "base ISO's" for recording movies.
@@WildlifeInspired Something mush have been wrong with youtube, I have had some more comments disappear elsewhere too... Sorry for the unfounded accusation, If you want I'll remove that part of my comment
Actually this does work - the last few days I've been playing around with some Great Egrets and a Snowy Egret - extremely White Birds - Shooting them by crushing the the Histrogram all the way to the left delivered my best shots - gave them super moody tones - in my OP
On my Sony A7iv Highlight priority works well for me in most situations. I use it quite often. It kinda eliminates the need to use exposure compensation. It automatically gives me a slightly underexposed image with all the highlights in tact. Plus the ISO is always lower than when using evaluative metering in the same conditions.
An excellent video Scott. I have a Sony A9 and have been using a similar technique but with shooting indoor volleyball where the light is generally very poor. By under exposing 2 stops I'm able to use a much faster shutter speed to stop the action. It really is a game changer. I'm pretty new to bird photography and plan to use exposure compensation and auto ISO as you have described. Again thanks for posting this!
No need to underexpose. Assuming ISO invariance, underexposing two stops and then boosting by two stops in post is exactly the same as just exposing correctly with an ISO two stops higher. And if you cross the dual-gain ISO threshold (so the ISO invariance assumption doesn't hold), then what you're doing is significantly worse than just using the correct ISO to begin with.
if you have any articles about dual gain, please let me know. I was reluctant to add more info to this (I think I mention it in this video but may have deleted that in editing to avoid too much info). The Z9 gets a dual gain "bump" at 500, but explaining why in a video isn't easy.
Interesting that Tin Man Lee who shoot Sony mostly advise the old notion of Expose to the right. Overexpose without burning the highlights and then pull back the shadows.. That use to be the Canon and M43 recommendations. Still today many of the M34 shooters work on the expose to the Right concept. I shoot Nikon and for the past number of year have used the concept of using exposure compensation to under expose and recover in post.
I have shot all of my images starting when we all had too expose properly with slide film. I shoot at -1/3 under stop always. Even on my top Nikon cameras and my newest canon R5. Never blow out my highlights and shadows are always great. My seminars and workshops everyone always loves this concept after using it. Exceptions would be white out winter images of course.
This is how it looks to a mathematician. When you set the f stop and exposure time, the amount of light reaching the sensor is completely determined by the scene. The ISO does not affect this, and each pixel "contains" numbers representing the intensity for each primary color. If you post-process the raw file, adjusting the "exposure" simply increases the values representing the pixels. Or, you can have the camera do the same operation, (based on the ISO value) before it writes the raw file. Since both processes start with the same information for each pixel (along with the same amount of noise for each pixel), they both have to increase the pixel values by the same amount. That's in theory. In practice there may be problems if at any point the computed values overflow the maximum number that can be represented by 14 bits -- this is when highlights get blown out, and at the low end, it's possible to get banding if low light values are magnified too much. One way of thinking about adjusting the "exposure" in post is that you are essentially setting the ISO for each exposure. This analysis may fail if the in-camera ISO setting is used as input to the algorithm the cameral uses. This might be the case with dual gain sensors. Highlight Priority in Manual with Auto ISO works for me quite well.
A brief comment. If you use manual mode with auto ISO, exposure bracketing doesn't really do anything. If I take 3 exposures, with manual mode, the amount of light hitting the sensor is the same in each case (because only aperture and shutter speed actually affect the amount of light). Varying the ISO for the three exposures changes only how much the densities are stretched out. This stretching can be done in the camera or in post, but the starting data -- the information in the sensor -- is the same in all three cases. This is equivalent to taking the same picture three times and sliding the "exposure" slider in post by three different amounts. Of course, bird photographers don't use bracketing. Also exposure bracketing *does* do something if you use the shutter speed as the setting you change, since the amount of light hitting the sensor will change in that case.
I agree with what you said Scott but I think there is one variable to consider and that’s if you have a dual gain sensor. I‘m no tech expert so maybe someone can go into detail. My understanding is that it can be advisable to increase the ISO when you right close to where the dual gain kicks in. On my Sony A1 it is at ISO 500. ISO 500 produces a cleaner image than ISO 400. Above and below you are fine with the ISO invariance. Obviously other cameras may have a different threshold where it kicks in.
I had written a fairly lengthy explanation, but since I learned most of this information from Steven Kersting, and he answered the post below, I'll leave it to him. I was going to reference the Photonstophotos website and graphs, but I see he has already done this. He is an excellent reference and the graphs show a bit of what is going on in the camera.
What I’m getting from this, is an arguement against the ETTR theory; that with modern sensors, it’s safer to under expose than it used to be…regardless of where your ISO is set. I find every Sony I’ve used tends to over expose (using matrix metering) and ‘under-exposing’ somewhat leads to easier post processing.
@@thebinpustey6497 I wouldn't say that the theory behind ETTR (if that's the exposure you want) doesn't make sense with current cameras. You still can saturate the photosites as much as possible without filling them to the top. With mirrorless cameras it has been said by some that it's just much easier (depending on subject you are shooting and environment) to shoot and get your exposure right in camera (or as close as possible) to what your end desired result would be (i.e. less time processing the image).
Great video! Very few people talk about this concept! The only point I would add is that you can clip shadows too. I shoot the z9, generally in matrix metering and on cloudy days it can often miss by about a stop or even a stop and a half, depending on the scene. If you shoot with your exposure comp at -2 stops on a cloudy day it’s very possible that you will clip your blacks on a bird like a hooded merg for example. Not as big of a problem as clipped highlights but definitely something to be aware of. Also, the z9 isnt actually iso invariant till you get to iso 500. So for any large exposure pulls you really should be shooting above iso 500.
@@WildlifeInspired I agree it would be nice to hear what a tech guy has to say about dual gains. It would be cool to hear what an actual electrical engineer, or whatever you call people who design camera tech, has to say. I’ve done the tests for myself. They are easy to do, just set your base exposure so you have leeway to under expose your iso by 5 stops. Then shoot a frame at iso 400 and one at 500. Pull the exposure 5 stops in both images, then reduce the highlights 100pts and lift the shadows 100 pts and you will have no trouble seeing the benefit of dual gain sensors!
Thank you for your video, Scott! Do you know if Nikon has recalled any 180-600 lenses? The reason I ask is because I purchased this lens five days ago and spent the last two days on the phone or in a chat room with Nikon and Adobe. The colors are too desaturated, the noise is bad, and I noticed that for whatever reason my EV jumped from zero on my EVF to 1.33 in the Metadata. I shot in manual and shutter priority. My ISO has ranged from 100 to 800. My shutter speed has ranged from 320 to 1200. I used a tripod and my MC30 with our Z9. Nikon recommended that I do a "power cycle". When that did not work, they recommended that I clean the contacts on my lens and camera. All our other lenses work perfectly well. In fact, I borrowed a Tamron Z 150-500 from my local camera shop before I purchased the Nikkor Z 180-600 and it performed well when shooting geese. The Histogram looks good as does the image at time of shot, but the curves slider must be moved a good distance to the right in NX Studio. After moving the slider, my photos improve vastly. I plan to shoot without the hood and lower my EV to a -1.33 to offset what has been a positive 1.33 in my metadata. If the results remain poor, I will have to swap this lens out. Of course, Nikon wants me to upload my files to their website and wait a couple days for them to get back to me. They acted like I blasphemed them after asking if the lens might be a dud. Do you have any other sage advice? Would you trade in this lens for either another or go to something entirely different? Thanks, John
I have no experience with that lens and its not a story I have heard Feel free to email me a raw that your camera read as 0 EV and Ill take a look. The EV sounds really odd. skeysimages@gmail.com
There's another important detail that you didn't touch on Scott. While attempting to maintain detail in those white cheek patches, you didn't mention your challenges keeping the noise under control in the black areas. Well done on your examples Scott!
Hi Scott,great video and very interesting topic something I haven’t really thought about but it does make sense will definitely try this process our,new to your channel and I am enjoying your videos great stuff regards John
Small correction, most of the cameras you listed as being ISO invariant are only invariant throughout portions of their ISO ranges, e.g. the Z9 is ISO invariant from ISO 64 to ISO 400, and then when its dual gain kicks in, from ISO 500 onwards, so it is not invariant between say ISO 200 and ISO 500. Also, the older sensor in the Canon EOS RP is not ISO invariant throughout most of its range (100-2500), although above that it is.
Thanks for clarifying. I always try to say I am not a technical expert, but try to talk about concepts and generalizations. I have a pretty good understanding of most aspects of HOW things work, but the details about technology can be overwhelming.
@@WildlifeInspired No problem. I have taught photography at college level and have faced the same difficulties as to how much technical detail to provide, so I fully understand the need to simplify things so as to not overwhelm learners. My philosophy was to always provide that extra detail in an addendum and reference list, at the end of a presentation or printout, for those who were keen on developing a full understanding. Admittedly, I never got the impression that many did look into it, but at least it gave me the self-satisfaction that I was giving fully accuracy information.
This reminds me of the argument about the benifits of getting white balance right in camera. Because of color noise, micro contrast and proper color balance throughout in post process... and how all that affects sharpness too. So many just do it in post. In my experience or to my eyes, getting it right in camera, my images have a better look to them, like going from 1080 to 4k or shootinga zoom vs prime... just that little bit better edge/look to them. Curious if you've tested this out before and what your findings are. I could just be losing it 😂
I have nearly always shot at -.7 but when shooting coots I always take it a tad more, not sure if it hampers the af , I need to do some tests with the Z9. No problems with the Z6 with exposure but I never thought about impact on Af until the Z9.
Cool! Makes a lot of sense actually. I'm definitely going to try this on my modest Canon RP. I prefer to go fully manual with Auto ISO in many situations anyway, keeping an eye on the histogram and bearing the Exposure Area in mind.. I didn't consider this possibility though it does make sense. This phenomenon reminds me of 'Reciprocity' when push-processing film with chemistry back in the day. Some film emulsions responded better than others when 'baked' in the developer. Does anyone remember Kodak T-Max?!
@@WildlifeInspired ~ Late 80's/ 90's; I was studying photography at an art collage. We used to buy 35mm film in 50 foot roll canisters and recycle used Ilford cartridges. We needed to be a lot more mindful of not only camera settings and limited film ISO, but developer dilutions, temperatures and processing times affecting contrast & exposure for the ENTIRE ROLL! There was no opportunity to check results in the field so bracketing exposures was your only safety net. The one thing that we all gained from learning to shoot on film, was that it forced us to be technically mindful in the field and in the lab. I did a LOT of spot metering with my old Lunasix-F hand held meter. You not only had one crack to get things right, but also had to cough up a total of about $16 bucks to process 24 shots!
Literally had a convo this morning with someone about how "easy" digital photography has made the technical work, but still up to the operator to actually take a good frame!
@@WildlifeInspired I studied Fine Art, Life Drawing and Painting, Photography at The Ontario College of Art & Design in Toronto back in the mid-80's. If you are serious about visual composition and learning how to 'see', drawing and painting FORCE you into appreciating elements of design; colour, contrast, space, texture, rhythm, etc.. in a very DIRECT and visceral way through the limitations of the medium. Take studio drawing and painting classes for a year, and watch your photography take on a completely new dimension as you begin to come to terms with how you see the world and the UNIQUE way you interpret the magic around you.
I feel that while there's no penalty on post, the af will pay a penalty having to try and focus while underexposed. I'll have to check on my z9 if exposure comp is accomplished with ISO shift or just software. The later would require you to shoot in manual ISO control to see the benefits you're talking about? I wonder if the easy view mode would allow large underexposure but give the af system a good view. Interesting. Thanks.
This is a good point; anything that affects the image in the viewfinder will also affect the AI autofocus capabilities (e.g. subject recognition) as they are using the same data stream... that includes viewfinder image brightness/contrast (doesn't actually have to be the recorded exposure), subject size (i.e. FF/DX mode); as well as the in-camera jpeg settings (sharpening/contrast/etc) because they are used to process the EVF stream and image previews even when recording raw files. On any Nikon exposure compensation is accomplished by changing whichever variable the camera has control over at the moment... it is never "software" unless you are into the extended (digital) ISO's.
@@WildlifeInspired Easy way to test... set up a dark exposure w/ manual settings and then change the setting for d8; "show effect of settings" (dark viewfinder) will negatively impact AF, and "adjust for ease of viewing" (bright viewfinder) will improve AF. The interesting thing here is that you *can* underexpose the recorded image w/o affecting the AF performance... but you cannot preview the result in the viewfinder at the same time.
Hi Scott, love the channel! I'm trying to optimize my setup for travel photography, and wondered if you have any thoughts on 2-in-1 tripod monopod combos, e.g. those offered by 3 legged thing? I'm unsure if the monopod elements are actually good or if they're more of a gimmick Sorry, unrelated to this video but thought you might be more likely to see this here than on your older vids!
Maybe you mentioned it and I just missed it. But I read/heard a few years back, that iso invariant cameras were only iso invariant up to a certian iso.. I can't remember if the cap varied with different bodies or not but I feel like it was iso 800 for the body I had at the time. Alao, as someonementioned in the comments already, that theres sweet spots where the noise is better at specific iso settings.. Do you know if either of those are true or not?
Hi Scott, thanks for great video. I am not too much in birds but will be doing more once I get my Z8. Considering that Z9 and Z8 have the same sensors, I presume Z8 is also ISO invariant? Please correct me if I’m wrong. Thanks for sharing.
I would absolutely think it has to be. There is a concept called dual gain that the z9 employs but I cant say the z8 will have that ( i imagine it will) bacially you get an ISO boost (improvement) at 500
Hi Scott, I have been trying out this iso invariance thing the last last few weeks. On the whole it has been good, particularly shooting inside forests. One question I have though is that if I am shooting at -0.7 EV and then look at the image in Lightroom Classic, is if I adjust the exposure to +0.7 it seems way over exposed. Does this mean I have brightness too high on my monitor? The image in my mind is probably at +0.3. I am wanting to do a 2024 calendar but I am afraid of images being too dark. Thanks for this video as it has helped the way I shoot.
Shooting Manual and Auto ISO, what advantage have you experienced using exposure compensation at -1.0 with matrix metering versus exposure compensation at 0.0 with highlighted-weighted metering? I liked the video. Thank you Scott!
I think its more personal preference. Not sure either is better or worse. If you stay low on ISO and use highlight, there isn't much downside. The key for me with this is just protecting highlights on species with stark contrast, whichever way works best.
Hi Scott, thanks for this information. I am a “ walk around” photographer and often get a real mix of lighting in bushy areas with lots of shadows or open bright areas. This certainly gives me something to try. Maybe a bird eye AF issue but will see. I shoot with a Sony a7iv and 100-400 gm. I have been pondering about getting the 70-200 as at times in wooded areas I can be quite close to birds, two to three metres sometimes, and ISO is often above 4000. Many shots are for ID or a life list as I am only a hobby photographer. Cheaper to try this than buy another lense. 😊Thanks again.
Except the Z9 (and all modern Sony sensors) has dual native ISO and if you'd go over to the 2nd ISO by exposing correctly in camera, you'd have less noise than raising exposure in post. For the Z9 the 2nd ISO is 500 for photo, so if you shoot under-exposed at ISO 200 and raise it 2 stops in post, you'll have more noise than directly shooting at ISO 800, because the image would clean up a bit in camera when switching to the 2nd ISO at 500. Maybe that's why you picked your 2 examples on either side of ISO 500, without stepping over that... but you didn't mention anything about it. Also don't know where you got the list, but Canon sensors are not ISO invariant.
I was going to do a follow up with dual gain, but I have also tested real life and the noise difference at ISO 100 raised 2 stops in post was about the same as 400 in camera. Now I did NOT test this at high ISO3200 and above since I rarely if ever shoot that high.
Usually I cap max at 3200 unless I know i have to go higher even with advanced software NR, seems like when you push 6400+ with this sensor (or the D500 D850 Z7) its struggles The D4/5/6 and the Z6/ii both do a little better at high ISPO
@@WildlifeInspired thanks Scott. I've tried the auto ISO at Conowingo but find myself setting it manually instead. Next time I go, I have to try your method. BTW, I'm in York, PA, and would love meet and pick your brain some time If you go to Conowingo, let me know.
If light is consistent and you are shooting a subject like eagles, you can certainly set the ISO, I usually expose for the lightest rock on the far shoreline and then dial it down to protect the BRIGHT WHITE of the eagles head.
14-bits sensor can capture dynamic range of 14 bits. JPEG only uses 8 bits (Human eyes can not distinguish more than 8 bits dynamic range). There is 6 bits extra. Nikon leave more bits on low light side than Sony, so Nikon naw can recover low light side better than Sony's ISO is purely a multiple number applying to base ISO data. For a base 100 camera, ISO 400 is just timing 4 to the data from the sensor. So if exposure is with in 6 bits(i.e. 6 stops), there is no difference. Typically, under/over 3 stops exposure is ok. Then it comes exposed to the right. Exposed to the right only makes sense if it's due to larger aperture/slow shutter speed. It does not make sense if it's due to ISO increased. I got the conclusion on ISO nature was through ISO dynamic range analysis done by some others.
Jpegs can record 12 stops of dynamic range using the standard 2.2. gamma curve encoding. And a 14bit ADC (the sensor is not 14bit) can record up to 14 stops of DR theoretically... if it is given that level/accuracy of data from the sensor (which it is not at higher ISO's).
@@stevenkersting3494 , Yeah, ADC bits match the stops of sensor base ISO dynamic range. Eventually, the data is in RAW file. I don't know much about 2.2 JPEG. My understanding is that any thing more than 8 bit is for editing purpose. 2.2 JPEG could be regarded as a kind of universal RAW. But no matter what's bits inside, only 8 bits are presented to viewers.
@@WildlifeInspired Then something strange is going on, I was first to post because you came to the wrong conclusion because you chose ISO 500 for your test as base ISO which happens to be the very second base ISO your camera has. If you had chosen ISO 450 then your conclusions would have been different...
![ETTR and ISO Invariance - [Should You Care?]](http://i.ytimg.com/vi/jDLTybYbqr8/mqdefault.jpg)
Another important point you missed in regards to the Z9 specifically is it’s dual gain sensor basically shooting at iso 200-400 will yield an inferior image as opposed to going right to ISO 500. So for me anything over ISO 200 I go straight to ISO 500
I just laid this out in more detail in my comment. There are several issues missed here - the most important of which is that ISO is not part of exposure at all, but gain/amplification of the S/N . Also, as you mention, the sensor of the Z9 is not truly ISO invariant - it's variant at ISO 500 in stills, and elsewhere in video.
Scott, we’ll done. You presented an excellent explanation of ISO invariance. I am in full agreement with you. However, there is one other aspect about underexposing your images by using lower ISOs that most don’t consider. Whenever you raise your ISO, you are also reducing the available dynamic range. In some cameras, when you reach ISO 1600-3200, you cut your dynamic range just about in half. Besides saving your highlights, a good reason to keep those ISO values lower and raise the image brightness in your editing software. Again, thank you for the video.
👍🏼
Thanks Dennis, I think I briefly mentioned it here, but maybe I edited that out. Because I am not "tech" guy, sometimes I worry that someone will nitpick info. I try to do my research before making any video and double check my understanding.
@@WildlifeInspired You did well. No complaints. You have an ‘easy’ way about you and a clear way you present information. I wish you even more success.
Every stop you raise the ISO you reduce the DR capability by one stop. Think of it this way, DR is the difference between the darkest and brightest values a sensor (pixel/photosite) can record. If the brightest value exists in the scene, then adding one stop of analog ISO amplification pushes it one stop into overexposure (ADC clipping)... I.e. the minimum is still the same, but you've pushed the maximum over the edge.
Great informative video Scott. I thought the D500 is ISO invariant?
We have know about this for Nikon Sensors for a very long time -- Lou Coetzer used to lecture on the subject -- and we were all taught to do shoot at EV-1 (or more) to gain 1 stop for free.
He argued we should shoot with a faster shutter speed to freeze motion and the correct aperture to gain the DOF needed AND sacrifice ISO.
When Highlight Protection/Priority/Weighted protection became available we gained from not blowing out highlights, but the results varied from shoot to shoot. NOW we know with the Z9 sensor that we can shoot to minus 5 EV and we can recover the image easily. BUT, this takes no account of when AF works the best -- which is when the subject is correctly exposed. So shooting consistently with minus 1 EV can help: BUT, with tools like DxO Pure Raw 3 and Topaz AI (denoise or photo) the end results can be spectacularly great even at ISOs over 10,000.
Canon sensors are known NOT to be ISO invariant. Whereas one expects Song and Nikon camera to behave similarly.
You raise good issues, but neglect to appreciate that as ISO increases, you're also loosing dynamic range. If you're shooting a flat scene, it may be much easier on the eye when post-processing (and not printing) images of ISO 12,800, but if you print - depending on size - this becomes an issue.
As a wildlife photographer I've been shooting my Z9 in manual with auto ISO for about 2 months now. This video makes so much sense that I'll be using it here on in. Thanks for the tip and keep them coming.
At last! Your explanation is the clearest, at least for me, as I have watched so many videos about it and even as they are technical and detailed, I couldn't really grasp it!... THANK YOU!
Yes, does make sense, I would use it definitely in the field and yes my camera it is, Z9...
Glad it helped!
Awesome video again!… I used the iso invariant concept for night photography. When I first listen to your video with the use of this technique, I was worried about seeing the image so dark that I would have a hard time to judge focus. But you came right out and proposed 1 or 2 stops and it makes so much sense. Why did I not think of that… ;-) I will definitely used this technique. Thank for sharing this great advice!
Scott, thanks for the videos and for making me a better artistic photographer.
Using the Z9 metering set to Highlight Priority in Manual with Auto ISO, is the way to go!
I've enjoyed successfully shooting that way for some time. Much appreciate your thoughtful videos!
Agreed! If I do use auto iso, I use the highlight weighted as well. They go hand in hand for me... Otherwise I use spot metering. Never liked matrix or center weighted
Super easy to understand. Well done job! Thank you for sharing! Super helpful!
A very good explanation of ISO invariance - I'm sure any newcomer to this concept will find this a very useful instruction. I've been taking advantage of ISO invariance for a year or so - especially when I'm photographing certain waterfowl against a dark background. I find it works particularly well if you also apply a custom-made linear camera profile to your image at the start of the edit. It'll look flat and lack saturation so you have to push contrast and colour a lot more to get where you want to be - but the linear profile puts you in control of the edit instead of Adobe's canned profiles, which have way too much contrast across the board in my opinion which makes highlights harder to control. It's a great way of controlling contrast because you have more control and can even work selectively.
I've been using the ISO variance method since Nikon brought out the D750 going on to use it with the D500 ...... shooting auto ISO in manual mode at around -2 stops or more .. bringing up the individual shades of lower light while keeping full control of the whites for feather detail.... IMHO here should be only one fully white exposed part of a bird in an image and then only just ..... using some sliders but mainly the adjustment brush ...sometimes there are a dozen AB pins or more in my bird images for both light and dark areas ..... good to see a video about this ......ISO variance rules!
Probably the easiest to understand explanation ISO invariance i've seen yet. Well done
Thanks Vince. Im sure someone will correct my lol
@@WildlifeInspired Gives them a chance to be wrong and learn
This makes total sense to me. I'm going to try it out. Thank you. I'm having a blast discovering new things about my Z9!
Thanks Scott. I've known about this and frequently get excited in the field and forget about it. I know my camera is iso invariant and am going to experiment tomorrow morning in a Grouse Blind. We got rained out this morning and the forecast is not much better tomorrow. In the morning I am going to concentrate on this to set it firmly in my brain. I already have plenty of images of Grouse in poor light. This frees me up to break new ground.
Good stuff!
Been meaning to try highlight metering. I forgot the z9 is invariant and haven’t really experimented too much with that.
Will have to give it a try, especially this summer with puffins, razorbills and loons!
Great job kind Sir. I shoot portraits, events and travel. What you just shared is so on point. You just gave me another reason to love my Z6ii. Not that gear matters. It does if you’re looking to implement this technique. Lol. Thanks again my friend. Cheers
That's Z6ii is really great for lower light. Super sensor. Not my fave for wildlife but love a lot about it.
Interesting, I'm going to have to experiment with this. Thank you Scott.
Good job explaining a complex subject. I used spot metering with auto-iso with my DSLR but have found that matrix works better with my z9 using exposure compensation. I will give the highlight priority metering a try. Thanks!
OK, I just heard your first sentence and already I agree. When I'm shooting small birds I almost always keep my histogram to the left. I've found that I get more accurate colors, better small detail, and I don't wash out the bright feathers with over exposure. I almost always shoot with manual exposure. I say almost always because sometimes in woodland settings the light changes too quickly from one spot to the next as the bird flits around. Then I use auto ISO sometimes. This morning was overcast and rainy. I shot a house finch two stops underexposed. When I put it in PS there was the beautiful raspberry red throat and head along with the rich browns and sharp details of fuzzy feathers. The good thing was the way the background dropped back and the bird stood out in all of its splendor. Light meters are easy to fool for best bird exposures.
very informative. thanks Scott
Glad it was helpful!
I think you have convinced me to finally give Auto ISO a real effort. The flexibility has always been intriguing but this really makes the case.
Scott, well done and well explained. Here’s an idea for a much needed video. I have been a photographer for many years and my photos are very good. The issues I have are with printing. I use a high end Mac to process and my colors seem to print well, BUT my exposure prints way too dark and not consistent. I have my screen brightness down to 2 out of 10 and some photos are still printing too dark. The exposure in camera is usually pretty good and the histogram shows good exposure. I send files to the lab for test prints and they come back consistently dark. HELP!!
I've had a "printing" video on the tondo list for a long time. This sounds to me like a monitor issue still.
I almost got there myself by shooting a bunch of Great Egrets (white as can be) against middle tone backgrounds. By dialing in underexposure a finally got photos with detail in the birds. What I haven't done is routinely underexpose EVERYTHING. I'll give it a shot, well maybe a million shots, and see how it looks...
OK, I've watched your entire video now, Scott. Only thing I have to add to my original comment is that wildlife photographers need to watch this, take it to heart and start practicing it now. This is an excellent video, Scott, with great information and exceptionally well explained. Wildlife photographers who learn this will surprise themselves, as I did, with the improved quality of their images.
Helpful commentary! My cameras are not ISO invariant and I wasn't certain what this meant, although I had a general idea. I have had instances photographing light colored hawks in the sky in golden hour where the camera blows out the highlight details. Sometimes this happens even when the histogram does not show any blown out highlights (or at least not enough for me to notice it). Although my cameras aren't ISO invariant, I still underexpose in such situations.
One minor comment: ETTR does not always involve overexposing. In fact, sometimes to preserve all sky details in high contrast landscape sunsets, I have to go to around -3 exposure stops when exposing to the right. The histogram is then skewed (most of the area is to the left but there is a long tail of brighter points, which reach right towards the right end). (I would still consider this exposing to the right because I am pushing the exposure as high as possible until highlights are clipped.)
Great video Scott. It makes perfect sense and I am going to start shooting about 1 stop underexposed to see how that might improve things. Canon doesn't have the highlight metering like Nikon unfortunately, but shooting underexposed should help regardless. It's always the whites on the birds that get me! By the way, that Hoodie image is spectacular -- I can't believe the detail!
Thank you !
I’ve been doing this since I got my Sony a1. Now I know why it works. Very insightful!
ISO Invariance came into play for me once I started astrophotography. Although I wasn't aware of it at the time I bought my D750, it turned out to be a very ISO invariant camera, and great with low light, extended exposures. Since that time, I started feeling more comfortable in most nighttime shooting situations with exposing for the highlights and keeping a higher shutter speed, especially with hand held night street photography. With wide angle astrophotography, lower exposure times can avoid having to lug a star tracker along, and I am okay with higher ISOs to get as much light as possible and cleaning up noise with stacking in post. I would say that I have found the Z9 to exhibit noise in low light situations more than my Z6ii or Fujifilm GFX 100S, but once one gets to know the threshold of the camera, it will still extend your low light capabilities much more than one might have expected. I have found that the GFX's dynamic range to by surprisingly good in low light, and routinely underexpose my shots with it, and even though it is a beast to carry, I have used it for landscape night sky shooting with and without a tracker to good effect. Thanks for sharing!
By the way. I love the sensor on the Z6ii. The camera overall can't compare to the z9 for wildlife but that sensor is super.
I switched from Canon to Nikon about 3 years ago when it was obvious that Canon wasn’t going to release an ISO invariant sensor. I had kept my 5D Mk II for far longer than I should have
A very interesting and comprehensive video ! Thank you very much Scott ! Will keep underexposing my pictures ! 🤓
Glad it was helpful!
Definitely learned something important here. Thanks
Glad it was helpful!
Very very good advice! This is indeed great option specifically for us who shoot in the wild❤
Great info there. I’ve been using this technic myself for wildlife. Only potential issue I come across from time to time is the autofocus isn’t optimal if the autofocus point is in an dark part of the image (say a black head of the bird )
Agree
Excalty.. well said. In fact, I dont even use auto ISO. I have my camera setup so I can change ISO, SS, and Av, each with a separate dial. Mostly because I don't trust the Auto iso. I feel (on my A7RIV) it ranges too much. For the subject I'm photographing, I know what shutter speed I want/need so all I need to do is roll the ISO to adjust for exposure keeping an eye on the blinkies. When that's too much I'll then change shutter speed. This is the BIGGEST advantage to mirrorless and the EVF.
I agree whole-heartedly with your suggestion to slightly underexpose. I don't shoot a lot of wildlife, but I do shoot thousand of photos at dog shows as an enthusiast (I recently shot over 40,000 frames in six days at a National Specialty Show where I had been invited to shoot for the club.) My wife shows Dalmatians, which are white with black spots or white with liver-brown spots. Often the events are outdoors under the midday sun. I've been shooting underexposed for a couple of years now, knowing I can boost the exposure in post (LIghtroom Classic is my go-to for post-processing) without losing the data detail in the white of the dog's coat. Sometimes the problem exists when events go inside until lousy florescent lighting backlit by huge windows with very bright light shining through them. I love those times when we have an early or late ring time when I get to shoot in the Golden Hour, but unfortunately, I have little control over the time of day that I'm shooting or the direction the judge is asking for the dog to be moved, with zero consideration for what that means for the photographer. 🙂
I think protecting the highlights just makes sense. Not saying that everyone should go out and shoot under 3 stops but up to 1 stop has never once hurt me.
Very interesting and informative. I am amazed my 8 year old Canon 6D can do so well at bringing the shadows, but this is mostly landscape with low ISO. Thinking back, if I do have exposure problems it is blowing out the highlights and overexposing ½ stop sure didn't help. When in Shenandoah NP, I will do, or be ready to do wildlife under a variety of conditions and manual with Auto ISO is my go-to setting. In once instance I was on a dark trail with lots of laurel and rhododendron and there was an Appalachian Cottontail on the trail in deep shade but filtered sunlight. Oddly, this rare species is not at all wary. I laid down and took many photos at 15 feet. My main concern was focus and they were extremely sharp. But the highlights were way too bright. Still good photos but had I underexposed by a stop (and I had time to change setting), the photos would have been better. Bringing up the shadows in this ancient camera was not a problem at all.
I totally agree thats not a big deal to recover shadows with modern cameras, but you always have some exceptions. I shoot R5 and recently was trying to get a shot of black raven in flight. The evening sky looked great and as far as bird was so black tuned +1 expo correction. I was so surprised that even that didn’t help. It was absolutely black with no chance to recover any details from shadows. But yeah, in most cases some underexposing can save crucial highlights.
I think that the problem is, in most cases, lack of light (brightness). If you have a contrasty scene you gill get a noisy image. Under / overexposing is just a tool that might help in some situtations, but black bird in flight during evening (=getting dark) on a much brighter sky (it also requires fast shutter speed) so you get very high iso (and drop of contrasts in post). I think advise regarding +compensation in post works best when you deal with a subject that is not black (or at least eye, or any crucial details are not situated at black parts or subject has light-dark ptterns or mentioned details are colored differently than black).
You can raise exposure with two thirds of a stop when using the highlight weighted metering mode, there are few to none scenes you will blow the highlights.
Yes that would work as well
Interesting video and thanks for the great explanation. I was under the impression that for ISO invariance you need to be working with 'real' iso values not the extended ones and for the most benefit using the base ISO up to the dual gain point. For example, with a Sony a1 ISO 100 is the base so invariance is best from 100-500 when 500 is the dual gain point where the dynamic range jumps back up to something like the same as ISO 200. The Z9 uses the same sensor with the same dual gain point (at 500). I tend to use zebras (Manual + auto ISO) with ec to control exposure (and the highlights).
A couple minor points... they don't change the basics of the idea/application; but are more technical/accurate.
Noise is primarily caused by the lack of light recorded; which is why it is more apparent in darker regions of an image. With a non-invariant camera raising the ISO *reduces* the recorded noise level (i.e. the noise penalty for underexposing). But with an invariant camera it does not affect the recorded noise level.
In either scenario, an ISO underexposed image is not *less noisy*... it is either just as noisy (invariant), or more noisy (non-invariant); it is just harder to see the noise level because it is darker.
In general, ISO is not exposure and it is not *sensitivity.* In terms of signal (SNR/noise) it is more like a volume control... if you have a crappy/noisy signal, making it louder just makes it more apparent. In terms of visible noise, it is more like the brightness control on your monitor. Most sensors only have one sensitivity to light; base ISO.
However, specific to the Z9, and other cameras with dual gain/dual ISO design sensors, there are two levels of sensitivity. The Z9 switches to low gain mode at ISO 500 which makes the sensor actually ~ 1 stop more sensitive to light than it is at base ISO. It also has less capacity in low gain mode... I.e. needs less light and can accept less light (not a problem when there is less light available).
So with a dual gain sensor you have to be careful about that sensitivity step... for example, with the Z9 recording at ISO 400 and recovering 3 stops is a little worse than recording at ISO 500 and recovering 2 stops.
IMO that difference is not really worth trying to manage in highly variable scenarios. But if you're working in lower light situations with auto ISO (for example) it could be beneficial to set your minimum ISO to 500.
photonstophotos.net/Charts/PDR_Shadow.htm#Nikon%20Z%209
I wrote the same and pointed out that almost all of the cameras he listed have dual or even triple base ISO steps - and he deleted my post!
I did not delete any post ??? I'm just reading the latest batch of comments now. I don't delete unless someone is disrespectful or vulgar.
I've never seen a sensor with triple ISO... I have seen some (Z9 included) that use digital manipulation in conjunction with the dual gain to use different "base ISO's" for recording movies.
@@WildlifeInspired Something mush have been wrong with youtube, I have had some more comments disappear elsewhere too... Sorry for the unfounded accusation, If you want I'll remove that part of my comment
Great info Scott! I didn’t realize my D850 was ISO invariant, so I will try underexposing on my next outing.
Actually this does work - the last few days I've been playing around with some Great Egrets and a Snowy Egret - extremely White Birds - Shooting them by crushing the the Histrogram all the way to the left delivered my best shots - gave them super moody tones - in my OP
On my Sony A7iv Highlight priority works well for me in most situations. I use it quite often. It kinda eliminates the need to use exposure compensation. It automatically gives me a slightly underexposed image with all the highlights in tact. Plus the ISO is always lower than when using evaluative metering in the same conditions.
An excellent video Scott. I have a Sony A9 and have been using a similar technique but with shooting indoor volleyball where the light is generally very poor. By under exposing 2 stops I'm able to use a much faster shutter speed to stop the action. It really is a game changer. I'm pretty new to bird photography and plan to use exposure compensation and auto ISO as you have described. Again thanks for posting this!
No need to underexpose. Assuming ISO invariance, underexposing two stops and then boosting by two stops in post is exactly the same as just exposing correctly with an ISO two stops higher.
And if you cross the dual-gain ISO threshold (so the ISO invariance assumption doesn't hold), then what you're doing is significantly worse than just using the correct ISO to begin with.
if you have any articles about dual gain, please let me know. I was reluctant to add more info to this (I think I mention it in this video but may have deleted that in editing to avoid too much info). The Z9 gets a dual gain "bump" at 500, but explaining why in a video isn't easy.
The good old Adams under expose over develop ;) For those as old as me... Good job!
Interesting that Tin Man Lee who shoot Sony mostly advise the old notion of Expose to the right. Overexpose without burning the highlights and then pull back the shadows.. That use to be the Canon and M43 recommendations. Still today many of the M34 shooters work on the expose to the Right concept. I shoot Nikon and for the past number of year have used the concept of using exposure compensation to under expose and recover in post.
Great video Scott, really learned something new here! Thanks Rick
I have shot all of my images starting when we all had too expose properly with slide film. I shoot at -1/3 under stop always. Even on my top Nikon cameras and my newest canon R5. Never blow out my highlights and shadows are always great. My seminars and workshops everyone always loves this concept after using it. Exceptions would be white out winter images of course.
Great video Scott!
Glad you enjoyed it
This explains a lot! Thanks!!!
Glad it was helpful!
This is how it looks to a mathematician. When you set the f stop and exposure time, the amount of light reaching the sensor is completely determined by the scene. The ISO does not affect this, and each pixel "contains" numbers representing the intensity for each primary color. If you post-process the raw file, adjusting the "exposure" simply increases the values representing the pixels. Or, you can have the camera do the same operation, (based on the ISO value) before it writes the raw file. Since both processes start with the same information for each pixel (along with the same amount of noise for each pixel), they both have to increase the pixel values by the same amount.
That's in theory. In practice there may be problems if at any point the computed values overflow the maximum number that can be represented by 14 bits -- this is when highlights get blown out, and at the low end, it's possible to get banding if low light values are magnified too much.
One way of thinking about adjusting the "exposure" in post is that you are essentially setting the ISO for each exposure.
This analysis may fail if the in-camera ISO setting is used as input to the algorithm the cameral uses. This might be the case with dual gain sensors.
Highlight Priority in Manual with Auto ISO works for me quite well.
A brief comment. If you use manual mode with auto ISO, exposure bracketing doesn't really do anything. If I take 3 exposures, with manual mode, the amount of light hitting the sensor is the same in each case (because only aperture and shutter speed actually affect the amount of light). Varying the ISO for the three exposures changes only how much the densities are stretched out. This stretching can be done in the camera or in post, but the starting data -- the information in the sensor -- is the same in all three cases. This is equivalent to taking the same picture three times and sliding the "exposure" slider in post by three different amounts.
Of course, bird photographers don't use bracketing. Also exposure bracketing *does* do something if you use the shutter speed as the setting you change, since the amount of light hitting the sensor will change in that case.
I agree with what you said Scott but I think there is one variable to consider and that’s if you have a dual gain sensor. I‘m no tech expert so maybe someone can go into detail. My understanding is that it can be advisable to increase the ISO when you right close to where the dual gain kicks in. On my Sony A1 it is at ISO 500. ISO 500 produces a cleaner image than ISO 400. Above and below you are fine with the ISO invariance. Obviously other cameras may have a different threshold where it kicks in.
With the Z9 the dual
Gain is at 500. I'll try to get someone super techy to explain it and what the benefit is.
I had written a fairly lengthy explanation, but since I learned most of this information from Steven Kersting, and he answered the post below, I'll leave it to him. I was going to reference the Photonstophotos website and graphs, but I see he has already done this. He is an excellent reference and the graphs show a bit of what is going on in the camera.
What I’m getting from this, is an arguement against the ETTR theory; that with modern sensors, it’s safer to under expose than it used to be…regardless of where your ISO is set.
I find every Sony I’ve used tends to over expose (using matrix metering) and ‘under-exposing’ somewhat leads to easier post processing.
@@thebinpustey6497 I wouldn't say that the theory behind ETTR (if that's the exposure you want) doesn't make sense with current cameras. You still can saturate the photosites as much as possible without filling them to the top. With mirrorless cameras it has been said by some that it's just much easier (depending on subject you are shooting and environment) to shoot and get your exposure right in camera (or as close as possible) to what your end desired result would be (i.e. less time processing the image).
Great video! Very few people talk about this concept! The only point I would add is that you can clip shadows too. I shoot the z9, generally in matrix metering and on cloudy days it can often miss by about a stop or even a stop and a half, depending on the scene. If you shoot with your exposure comp at -2 stops on a cloudy day it’s very possible that you will clip your blacks on a bird like a hooded merg for example. Not as big of a problem as clipped highlights but definitely something to be aware of. Also, the z9 isnt actually iso invariant till you get to iso 500. So for any large exposure pulls you really should be shooting above iso 500.
im going to do a follow up on dual gain (your point at 500 ISO) Ill try to bring on a "tech guy" for that one as well.
@@WildlifeInspired I agree it would be nice to hear what a tech guy has to say about dual gains. It would be cool to hear what an actual electrical engineer, or whatever you call people who design camera tech, has to say. I’ve done the tests for myself. They are easy to do, just set your base exposure so you have leeway to under expose your iso by 5 stops. Then shoot a frame at iso 400 and one at 500. Pull the exposure 5 stops in both images, then reduce the highlights 100pts and lift the shadows 100 pts and you will have no trouble seeing the benefit of dual gain sensors!
Thank you for your video, Scott!
Do you know if Nikon has recalled any 180-600 lenses? The reason I ask is because I purchased this lens five days ago and spent the last two days on the phone or in a chat room with Nikon and Adobe. The colors are too desaturated, the noise is bad, and I noticed that for whatever reason my EV jumped from zero on my EVF to 1.33 in the Metadata. I shot in manual and shutter priority. My ISO has ranged from 100 to 800. My shutter speed has ranged from 320 to 1200. I used a tripod and my MC30 with our Z9.
Nikon recommended that I do a "power cycle". When that did not work, they recommended that I clean the contacts on my lens and camera. All our other lenses work perfectly well. In fact, I borrowed a Tamron Z 150-500 from my local camera shop before I purchased the Nikkor Z 180-600 and it performed well when shooting geese.
The Histogram looks good as does the image at time of shot, but the curves slider must be moved a good distance to the right in NX Studio. After moving the slider, my photos improve vastly.
I plan to shoot without the hood and lower my EV to a -1.33 to offset what has been a positive 1.33 in my metadata. If the results remain poor, I will have to swap this lens out. Of course, Nikon wants me to upload my files to their website and wait a couple days for them to get back to me. They acted like I blasphemed them after asking if the lens might be a dud.
Do you have any other sage advice? Would you trade in this lens for either another or go to something entirely different?
Thanks,
John
I have no experience with that lens and its not a story I have heard Feel free to email me a raw that your camera read as 0 EV and Ill take a look. The EV sounds really odd. skeysimages@gmail.com
There's another important detail that you didn't touch on Scott. While attempting to maintain detail in those white cheek patches, you didn't mention your challenges keeping the noise under control in the black areas. Well done on your examples Scott!
Very very interesting, thanks!
Hi Scott,great video and very interesting topic something I haven’t really thought about but it does make sense will definitely try this process our,new to your channel and I am enjoying your videos great stuff regards John
Glad it was helpful!
Small correction, most of the cameras you listed as being ISO invariant are only invariant throughout portions of their ISO ranges, e.g. the Z9 is ISO invariant from ISO 64 to ISO 400, and then when its dual gain kicks in, from ISO 500 onwards, so it is not invariant between say ISO 200 and ISO 500.
Also, the older sensor in the Canon EOS RP is not ISO invariant throughout most of its range (100-2500), although above that it is.
Thanks for clarifying. I always try to say I am not a technical expert, but try to talk about concepts and generalizations. I have a pretty good understanding of most aspects of HOW things work, but the details about technology can be overwhelming.
@@WildlifeInspired No problem. I have taught photography at college level and have faced the same difficulties as to how much technical detail to provide, so I fully understand the need to simplify things so as to not overwhelm learners.
My philosophy was to always provide that extra detail in an addendum and reference list, at the end of a presentation or printout, for those who were keen on developing a full understanding. Admittedly, I never got the impression that many did look into it, but at least it gave me the self-satisfaction that I was giving fully accuracy information.
Sehr schön erklärt!
This reminds me of the argument about the benifits of getting white balance right in camera. Because of color noise, micro contrast and proper color balance throughout in post process... and how all that affects sharpness too.
So many just do it in post. In my experience or to my eyes, getting it right in camera, my images have a better look to them, like going from 1080 to 4k or shootinga zoom vs prime... just that little bit better edge/look to them.
Curious if you've tested this out before and what your findings are. I could just be losing it 😂
I have nearly always shot at -.7 but when shooting coots I always take it a tad more, not sure if it hampers the af , I need to do some tests with the Z9. No problems with the Z6 with exposure but I never thought about impact on Af until the Z9.
Great explanation on ISO invariance, Scott. 👍
Thanks! 👍
Cool! Makes a lot of sense actually. I'm definitely going to try this on my modest Canon RP. I prefer to go fully manual with Auto ISO in many situations anyway, keeping an eye on the histogram and bearing the Exposure Area in mind.. I didn't consider this possibility though it does make sense.
This phenomenon reminds me of 'Reciprocity' when push-processing film with chemistry back in the day. Some film emulsions responded better than others when 'baked' in the developer. Does anyone remember Kodak T-Max?!
Not me. Lol.
@@WildlifeInspired ~ Late 80's/ 90's; I was studying photography at an art collage. We used to buy 35mm film in 50 foot roll canisters and recycle used Ilford cartridges. We needed to be a lot more mindful of not only camera settings and limited film ISO, but developer dilutions, temperatures and processing times affecting contrast & exposure for the ENTIRE ROLL! There was no opportunity to check results in the field so bracketing exposures was your only safety net.
The one thing that we all gained from learning to shoot on film, was that it forced us to be technically mindful in the field and in the lab. I did a LOT of spot metering with my old Lunasix-F hand held meter. You not only had one crack to get things right, but also had to cough up a total of about $16 bucks to process 24 shots!
Literally had a convo this morning with someone about how "easy" digital photography has made the technical work, but still up to the operator to actually take a good frame!
@@WildlifeInspired I studied Fine Art, Life Drawing and Painting, Photography at The Ontario College of Art & Design in Toronto back in the mid-80's.
If you are serious about visual composition and learning how to 'see', drawing and painting FORCE you into appreciating elements of design; colour, contrast, space, texture, rhythm, etc.. in a very DIRECT and visceral way through the limitations of the medium. Take studio drawing and painting classes for a year, and watch your photography take on a completely new dimension as you begin to come to terms with how you see the world and the UNIQUE way you interpret the magic around you.
I feel that while there's no penalty on post, the af will pay a penalty having to try and focus while underexposed. I'll have to check on my z9 if exposure comp is accomplished with ISO shift or just software. The later would require you to shoot in manual ISO control to see the benefits you're talking about? I wonder if the easy view mode would allow large underexposure but give the af system a good view. Interesting. Thanks.
This is a good point; anything that affects the image in the viewfinder will also affect the AI autofocus capabilities (e.g. subject recognition) as they are using the same data stream... that includes viewfinder image brightness/contrast (doesn't actually have to be the recorded exposure), subject size (i.e. FF/DX mode); as well as the in-camera jpeg settings (sharpening/contrast/etc) because they are used to process the EVF stream and image previews even when recording raw files.
On any Nikon exposure compensation is accomplished by changing whichever variable the camera has control over at the moment... it is never "software" unless you are into the extended (digital) ISO's.
Very good point. I will do a real life autofocus test at -3 exp comp vs 0 and record the results. I'll just have to do some controls.
@@WildlifeInspired Easy way to test... set up a dark exposure w/ manual settings and then change the setting for d8; "show effect of settings" (dark viewfinder) will negatively impact AF, and "adjust for ease of viewing" (bright viewfinder) will improve AF. The interesting thing here is that you *can* underexpose the recorded image w/o affecting the AF performance... but you cannot preview the result in the viewfinder at the same time.
Hi Scott, love the channel! I'm trying to optimize my setup for travel photography, and wondered if you have any thoughts on 2-in-1 tripod monopod combos, e.g. those offered by 3 legged thing? I'm unsure if the monopod elements are actually good or if they're more of a gimmick
Sorry, unrelated to this video but thought you might be more likely to see this here than on your older vids!
To be honest, I feel its a little gimmicky. Maybe I will try to review one soon!
@@WildlifeInspired Thanks Scott!
Maybe you mentioned it and I just missed it. But I read/heard a few years back, that iso invariant cameras were only iso invariant up to a certian iso.. I can't remember if the cap varied with different bodies or not but I feel like it was iso 800 for the body I had at the time.
Alao, as someonementioned in the comments already, that theres sweet spots where the noise is better at specific iso settings.. Do you know if either of those are true or not?
Yes, all that is true but he missed that in his crude test because he started above the ISO threshold for his cameras sensor...
Hi Scott, thanks for great video. I am not too much in birds but will be doing more once I get my Z8. Considering that Z9 and Z8 have the same sensors, I presume Z8 is also ISO invariant? Please correct me if I’m wrong.
Thanks for sharing.
I would absolutely think it has to be. There is a concept called dual gain that the z9 employs but I cant say the z8 will have that ( i imagine it will) bacially you get an ISO boost (improvement) at 500
Thank you for the information, Scott!!Very informative!!.
Glad it was helpful!
Thanks Scott very well explained.
Very welcome
Thanks
Thank you!!!!
Hi Scott, I have been trying out this iso invariance thing the last last few weeks. On the whole it has been good, particularly shooting inside forests. One question I have though is that if I am shooting at -0.7 EV and then look at the image in Lightroom Classic, is if I adjust the exposure to +0.7 it seems way over exposed. Does this mean I have brightness too high on my monitor? The image in my mind is probably at +0.3. I am wanting to do a 2024 calendar but I am afraid of images being too dark. Thanks for this video as it has helped the way I shoot.
Hard to say, but calibrating your monitor might be a good idea
Shooting Manual and Auto ISO, what advantage have you experienced using exposure compensation at -1.0 with matrix metering versus exposure compensation at 0.0 with highlighted-weighted metering?
I liked the video. Thank you Scott!
I think its more personal preference. Not sure either is better or worse. If you stay low on ISO and use highlight, there isn't much downside. The key for me with this is just protecting highlights on species with stark contrast, whichever way works best.
Hi Scott, thanks for this information. I am a “ walk around” photographer and often get a real mix of lighting in bushy areas with lots of shadows or open bright areas. This certainly gives me something to try. Maybe a bird eye AF issue but will see. I shoot with a Sony a7iv and 100-400 gm. I have been pondering about getting the 70-200 as at times in wooded areas I can be quite close to birds, two to three metres sometimes, and ISO is often above 4000. Many shots are for ID or a life list as I am only a hobby photographer. Cheaper to try this than buy another lense. 😊Thanks again.
Underexposing will not only make the images noisier if you get below one of the base ISO thresholds but also influence the focus tracking adversely!
Except the Z9 (and all modern Sony sensors) has dual native ISO and if you'd go over to the 2nd ISO by exposing correctly in camera, you'd have less noise than raising exposure in post. For the Z9 the 2nd ISO is 500 for photo, so if you shoot under-exposed at ISO 200 and raise it 2 stops in post, you'll have more noise than directly shooting at ISO 800, because the image would clean up a bit in camera when switching to the 2nd ISO at 500.
Maybe that's why you picked your 2 examples on either side of ISO 500, without stepping over that... but you didn't mention anything about it.
Also don't know where you got the list, but Canon sensors are not ISO invariant.
I was going to do a follow up with dual gain, but I have also tested real life and the noise difference at ISO 100 raised 2 stops in post was about the same as 400 in camera. Now I did NOT test this at high ISO3200 and above since I rarely if ever shoot that high.
Well done, Scott.
Waiting to get educated by a "tech guy" lol
My Wife loves the image at 12:54, any chance you sell prints?
another question, if you are using auto ISO how do you handle dropping the ISO while shooting? Do you change the EV as you're shooting?
I've been under exposing my images on purpose for years.. I much prefer the results..
do you set a minimum and maximum exposure on your auto ISO?
Usually I cap max at 3200 unless I know i have to go higher even with advanced software NR, seems like when you push 6400+ with this sensor (or the D500 D850 Z7) its struggles The D4/5/6 and the Z6/ii both do a little better at high ISPO
@@WildlifeInspired thanks Scott. I've tried the auto ISO at Conowingo but find myself setting it manually instead. Next time I go, I have to try your method. BTW, I'm in York, PA, and would love meet and pick your brain some time If you go to Conowingo, let me know.
If light is consistent and you are shooting a subject like eagles, you can certainly set the ISO, I usually expose for the lightest rock on the far shoreline and then dial it down to protect the BRIGHT WHITE of the eagles head.
14-bits sensor can capture dynamic range of 14 bits. JPEG only uses 8 bits (Human eyes can not distinguish more than 8 bits dynamic range). There is 6 bits extra. Nikon leave more bits on low light side than Sony, so Nikon naw can recover low light side better than Sony's
ISO is purely a multiple number applying to base ISO data. For a base 100 camera, ISO 400 is just timing 4 to the data from the sensor. So if exposure is with in 6 bits(i.e. 6 stops), there is no difference. Typically, under/over 3 stops exposure is ok.
Then it comes exposed to the right. Exposed to the right only makes sense if it's due to larger aperture/slow shutter speed. It does not make sense if it's due to ISO increased.
I got the conclusion on ISO nature was through ISO dynamic range analysis done by some others.
Jpegs can record 12 stops of dynamic range using the standard 2.2. gamma curve encoding. And a 14bit ADC (the sensor is not 14bit) can record up to 14 stops of DR theoretically... if it is given that level/accuracy of data from the sensor (which it is not at higher ISO's).
@@stevenkersting3494 , Yeah, ADC bits match the stops of sensor base ISO dynamic range. Eventually, the data is in RAW file. I don't know much about 2.2 JPEG. My understanding is that any thing more than 8 bit is for editing purpose. 2.2 JPEG could be regarded as a kind of universal RAW. But no matter what's bits inside, only 8 bits are presented to viewers.
Well done for deleting my comment on your errors about ISO invariance, almost no sensor is fully ISO invariant!
Didn't delete anything. Make sure you hit the send button.
@@WildlifeInspired Then something strange is going on, I was first to post because you came to the wrong conclusion because you chose ISO 500 for your test as base ISO which happens to be the very second base ISO your camera has. If you had chosen ISO 450 then your conclusions would have been different...
@@karlgunterwunsch1950 And I am pretty sure that you sir are shooting with a Canon and please do correct me if I am wrong!
thanks
Great video Scott!!
Thanks!