I am easing back into the photography world after taking a hiatus due to kids. I’ve found myself carrying dedicated camera a lot more frequently than I have in the past (even before kids), which has resulted in me spending some time thinking about what the right camera for me is. I’ve “short-term-rented”, via lensrentals.com, and “long-term-rented”, via buying and selling used equipment, a number of bodies and lenses over the past 1.5 years to try to find what clicks with me. So far, I’ve spent the longest amount of time with Sony’s A7III. I just started a long term rental on an OM-1.
The goal of this exercise is to help me understand how I’m using my gear empirically, to help guide my lens selection and body needs. The intent is to ignore feel and quality of life (for now).
To make this possible, I cobbled together a quick and dirty application that uses MetadataExtractor to aggregate EXIF data into a flat file that I fed into Power BI.
My only two lenses for the A7III are Sigma’s 35mm f/1.4 DG HSM and Sony’s 50mm FE f/1.8. Looking only at these two, the first insight is that I tend to grab the smaller and lighter lens. This is despite the fact that I like the way the Sigma 35mm f/1.4 DG HSM renders images. It’s also despite my enjoyment of the 35mm focal length. I know 35mm isn’t super popular, but I’ve found it works well with my kids. I’m usually “the right” distance away from them for the 35, and often have to back up for the 50. Backing up also isn’t always possible - especially when we’re visiting places like indoor museums.
That said, the Sigma 35mm is a fairly hefty lens at 755 grams. Sony’s 50mm FE f/1.8 is 186 grams in comparison. The results speak for themselves.
takeaway: maybe I should grab a more compact 35 (or maybe the 40mm G f/2.5) to use when we’re not going to be somewhere with very dim lighting
Next up is another obvious one, but one that’s worth saying anyway: I’ve shifted from shooting mostly outdoors in decent weather, before kids, to a more mixed environment in terms of lighting. This is clearly visible by calculating the exposure value from all my photographs and then looking pre-post 2022.
There are two stories to be told for the photographs taken in low light, but to get there let’s look at the aperture I used for low lighting levels (say under 6.0 EV, which is a brightly lit home interior)
Let’s then carve this into two groups: one that was shot (nearly) wide open (f/1.4, f/1.6, f/1.8) and everything else.
For the wide open group, it’s amazing what a fast lens can do for you in terms of keeping ISO down with low light. Of the 269 photos I’ve kept over the past year that were wide open in under 6 EV, ISO was at or below 3,200 on all but 35 of them.
takeaway: fast lenses are good for low light, but also maybe I don’t really need a FF camera for low light… at least with my current shutter speeds
A quick note on the scatter chart - the y axis is logarithmic.
For the second group, the majority were stepped down for increased depth of field. The photo below isn’t super interesting, but shows this. It was shot on the FE 50mm @ f/4.5 and 1/125 s shutter. The resulting ISO was 4000.
If I had shot this on say micro four thirds, I could have used a 25mm lens at f/2.2 for the same field of view and depth of field, knocking the ISO down to 1000. This sure sounds a lot better than ISO 4000, but when you consider that something like an OM-1 is 1 2/3 stops behind my A7III in terms of dynamic range the difference isn’t as large as it sounds. At ISO 1000, the OM-1 will return a dynamic range of 7.73 where as the A7III returns 7.15 at ISO 4000.
takeaway: darn you FF for your shallow depth of field, but maybe the advantages of a smaller sensor aren’t that big on overall image quality when you step the FF body down
As a final thought, I used to do a spend a decent amount of time taking photos of cars going by with my D5300 and the excellent VR Zoom Nikkor 70-300mm f/4.5-5.6G IF-ED.
takeaway: I could have probably used a longer lens for a number of these photos. A 100-400 with a teleconverter, or a 200-600 (or 200-800 in the case of the Oly 100-400) would cover 88% of these shots.
Interesting charts! Id love to see a jupyter notebook that presented all of this so it could easily be passed around and modified. Maybe I’ll make one.
Oooh, or this type of thing would be a great feature for immich or photoprism. Or the notebook could just use their database as a data source.
Tell me you work in data science at an enterprise level without telling me you work in data science at an enterprise level, lol.
That said, I completely agree with you, There really isn’t anything complicated here and it seems like a good coaching tool. Building it into existing software seems fairly obvious.
I’m also a little surprised that I haven’t run into this in the (many) debates between FF, APS-C, and M43. Yeah, FF performs better at high ISO but depending on what you’re shooting (eg not fast sports in crappy lighting) it’s pretty easy to avoid high ISOs. Conversely, if you compare fast M43 glass to slower FF glass, the size gap largely disappears until you start getting into the telephoto range.
Way too many debates on sensor size feel academic. My goal was to pull myself away what-ifs and say, “based on my current shooting patterns this is what makes sense for me”.
That’s a neat way to analyze what works for you without having to rely on memory or feel. The bit with the EV changes over the last couple years really shows how your range of photos and environments has changed.
Indeed! The thing that really surprised me is that ISO didn’t go through the roof. Since I was outdoors mostly before kids I never persued fast glass. As long as you don’t need a shutter above say 1/125, you can keep ISO pretty low on fast glass, even in very dim lighting.
Interesting project, and well done for making it reality. I’ve only had fanciful musings along these lines.
One question, do you rate your photos (e.g. 1 star, 2 star, etc.) and have you tried understanding the characteristics of what you tend to like? My musings imagine there is a level of ISO which caps my liking of photos to “That’s OK”, and another where it makes taking the photo almost pointless apart from as a record of what was seen.
I used to spend a decent amount of time filtering my photos into categories like “missed focus”, “sharp, but not interesting”, “sharp and interesting”, “the best of the sharp/interesting in this sequence”, etc in Photo Mechanic, but honestly I spent too much time categorizing photos into categories that didn’t really matter. I decided that I wanted to try to find a way to do this on my iPad for a vacation last summer, stumbled upon Photo Workflow, and have been using it ever sense. I find that it makes my culling flow/process quite a bit faster - photos go in the “yay” folder or the “meh” folder. It is a bit clunky to transfer files onto my iPad via USB and back off again to my NAS, but it works.
It seems like you’re asking if there’s a technical characteristic that I can point to and say “this is the reason why I don’t like such and such a photo”. I have three musing there:
- Even in very low light I’ve able to achieve low ISO values by slowing down shutter speed and shooting wide open on a fast lens. Keeping ISO low has not been a problem, but it has not always made for good photos
- In hindsight, I likely over-prioritized keeping ISO down. My photos over the past six months or so show that when I’m shooting in a low light environment, I have less preference for shooting wide open and/or I’m using a higher minimum shutter speed. I’m way less wary of high ISO
- It really comes back to basics. Low ISO + sharp focus don’t make for great images. Avoiding blowing highlights and framing are equally, if not more important
I’ve taken a number of photos that I really liked at ISO 6400+, including some at ISO 12800. One of the ISO 12800 photos was even taken on an X-H2s and not a FF body. ISO 12800 does feel like a bit of a cutoff though.
The photo below isn’t one that I necessarily really really like (I am trying my best to preserve my anonymity on Lemmy…), but it’s 23mm (35mm FF EQ), ISO 10000 (on the aforementioned X-H2s), 1/500, f/1.4. It shows that it’s certainly possible to retain a lot of detail at high ISO - especially given the right subject/scene. No fancy post processing done - this is a SOOTC JPEG.
Sorry, off topic, but I can’t help laughing. I’ve got a Sony A6000, and I get more noise at ISO 3200 than you’ve got at ISO 10000 >.<
It really is impressive what incremental progress has been made to reduce ISO noise over the years, even if there hasn’t been a single big splash. That said, I chose than photo fairly intentionally. ISO noise isn’t very obvious on the X-H2s if your subject fills the frame and your background isn’t that dark. If either of those isn’t the case, the noise becomes a lot more obvious.
FF is better again, somewhat obviously. I don’t hesitate to shoot at ISO 12,800 on my A7III.
Thank you for the content and effort put into that reply, you are generous.
My main interest is wildlife, and because I so rarely get more than a few seconds of shooting on any subject (and many that entirely get away before I notice, align lens onto subject and auto-focus does its thing on what I’m after) so I don’t have large numbers of photos from a walk so the culling process is not that bad.
I certainly agree that high ISO is not, in and of it self, a guarantee the photo is no good, though there is some loose correlation - and your cat photo makes the point well.
I’m just wondering whether your approach might be used to give pointers on how to balance settings “when every second counts”, and there’s a lot to be said for just taking the shot regardless. But, as ever, really I ought to just try it myself and see!
It was a good question and answering it afforded me the opportunity to do more reflection on what photos I like/don’t like.
For wildlife, I assume your biggest opportunities are going to be on the gear and technique side vs where you are in the exposure triangle. You need the appropriate amount of reach to fill your frame, or enough pixels on your sensor to make cropping not a big deal. You’ll need a fast enough shutter to freeze your target (or a slightly slower shutter and a willingness to have a lower keeper rate). Some of this is also equipment familiar and getting it set up ahead of time. If seconds count, how quickly can you turn on/wake up the camera, get it to focus where you want it to focus, and manage exposure settings? Maybe simply keep the camera in shutter priority if you’re newer to this? If you’re shooting in manual, a histogram is your friend but it will slow you down initially. As far as I’m aware, Olympus/OMD is the only company whose cameras will make blown shadows and highlights super obvious in the EVF.
Sorry - been away for a few days.
Your advice is good, but mostly already known to me. For birds, it’s rare for the bird to fill the frame on my 600mm lens, never mind a 200-800mm zoom - so it’s cropping that’s important. Shutter speed - as high as the light allows :) Generally around 1/1000s on the zoom, faster on brighter days. I walk around with the camera on: my walks are rarely more than an hour at a time, so no chance of flattening the battery (and a spare is carried) - the time is just noticing the bird, thinking “I want to photo that” (which takes remarkably long at my age), actually getting the bird in the frame (one major draw back of long heavy lens, and the real time killer), and then waiting for AF to kick in.
I seem not to be using manual focus these days, just single point auto-focus for birds in a cluttered shot. I think manual focus would be more relevant in an hide and/or with a tripod: the time to focus manually means the bird better be still or I need to know where it will appear.
I dodge most black/overexposed issues by shooting in RAW (so far more forgiving) - though I do use exposure compensation when shooting birds against the sky.
Are you generally happy with the photographs you’re taking? If so, keep doing what you’re doing. I could see wanting to do something like this to see if there’s a correlation between one of your exposure parameters, focal length, and how much you liked the resulting image though.
the time is just noticing the bird, thinking “I want to photo that” (which takes remarkably long at my age)
Have you tried skipping this step? Thanks to moving away from film, I tend to take a lot of photos these days. It does mean more culling, but I get a lot more interesting candid photographs because I didn’t miss the moment.
RE: autofocus, I wasn’t trying to advocate for manual focus, but rather knowing when to use each of the various focus options on your camera (wide area vs zonr vs spot). However, if you’re already using spot that’s probably the way to go for static birds. I’ve never tried moving birds before, but it seems like you might want a larger focus area to deal with erratic motion.
On time taken to think - it’s more the bittersweet admiring/wondering at wildlife vs. remembering what this weight in my hands is for!
On a related note, getting over the embarrassment of using burst mode freely (rather than working on having perfect timing) has definitely helped.
On swapping focus modes - that’s almost down to instinct now, my fingers know where to go and how many times to press.
Since you have kids, the lens of choice does not surprise me. Apple has trended wider with its 1x lens. That lens on my phone is 26mm. It is intended for use in the scenario you used as an example, inside taking pictures of people.
The thing that was somewhat surprising to me was that my emotional lense of choice didn’t match the lens that I use the most. The 40mm G f/2.5 looks like a pretty good alternative. Looking at focal lengths from when I had the Nikon 24-70 f/4 I very rarely shot at the far wide end.