Features that more cameras should have
photographylife.com
photographylife.com
But here's one important feature the list misses: the ability to shoot uncompressed, or losslessly compressed, RAW images. The last time I checked, the introductory Sony cameras (the A6500 and its cousins) do not permit this. The customer can only access lossy compressed data. The same is true for lower-end Nikon DSLRs.
Most of the time, this is not a problem – though in some situations with heavy postprocessing or harsh lighting, you will certainly see compression artifacts. I just find it amazing that they're asking customers to pay almost a thousand dollars for a camera+lens kit that doesn't support lossless compression, or let me access the true "raw" data.
It's 2021! This is well-known technical knowledge taught in undergraduate CS classes.
While I agree with you, I feel like on a mirrorless, the Sensor Dust Protector Curtain is essential if you want to be able to change lenses while out in the field.
Many years ago I used to work for a small local paper, they gave me a DSLR so I had the luxury of the sensor being at least partially protected by the mirror. (I wish they gave me two bodies so I didn't have to swap lenses, but it was a small newspaper and they didn't have a lot of money).
Now I own a mirrorless (that I don't use professionally) and I would never attempt to change the lens unless I'm in an indoor, non-dusty room, without fans or a/c operating. Just looking at the naked sensor staring back at me makes me nervous ;-)
I can't imagine swapping lenses on the mirrorless while on assignments, I think I would have to resort to buying a couple of bodies just to avoid lens swapping.
The curtain protection would solve that problem.
For the average shooter, even one who uses their camera like a tool, maintaining it reasonably should make this a nonissue.
Perhaps you can create a macro to apply the same edit automatically to all your imported pictures. In a fast paced industry like news, that's still an unwelcome overhead though.
Honest question - is this a tech or a biz decision? I.e., is it more expensive to include output to raw? More i/o requirements or totally different chipsets or the like? Or is "raw" just deemed a "pro" feature that they feel they can upcharge for? (e.g. paying to flip a bit to unlock more battery life in an EV)
Most cameras offer lossless compressed raw these days, some (e.g. the low-end Nikon Zs, and also the flagship Z9) don't even have uncompressed raw anymore, only lossless compressed or a choice of lossy options. Much to the chagrin of some people, a particular Youtuber especially, who make a huge deal out of using uncompressed raw. The person in question doesn't understand what the word "lossless" means and claims there's a difference between losslessy compressed raws and "true uncompressed" raws.
Why it's lossy instead of lossless, I do not know. My guess is that lossy compression allows them to get more compression, and hence improve their effective buffer size for advertising purposes. I think 90% of the time this is a good trade-off. You can view some example artifacts here: https://stephenbayphotography.com/blog/sony-raw-compression-....
[Note that the author mentions being able to access uncompressed raw on his Sony after 2015, but it's a pro-level model. For the amateur models I believe there's no such option.]
Obviously, if you're trying to make them happen, you can do it, but for pics of your kids/pets/plants/etc. in normal daylight situations it shouldn't be a problem.
Still, as a technically minded person, it's infuriating to me that I can't access the original data.
There was definitely an R&D cost associated with developing the libraries to read/write that format, and additional QA steps required when implementing them, but no additional hardware was needed (it actually required less on-hardware processing to write to RAW).
If a company has RAW support on one camera but not another - especially for still photos, it is 100% a market segmentation strategy.
Casual photographers are fine with just using OOC JPEGs and apps like Apple or Google Photos. They aren't paying for Lightroom or Camera One and worrying about the sensor's dynamic range or ISO performance.
Also using RAW especially on higher megapixel sensors requires your entire stack to be more expensive. Computer needs to be fast with lots of hard drive space. You need a UHS-2 card reader as well as 300Mbps cards etc.
It's not a hidden feature by any stretch. Different vendors allow for different types. Some only allow a smaller frame size while still RAW. Some allow a compression to be applied.
Shorter answer, read the specs or RTFM.
You hold the shutter halfway and it takes photos at whatever speed you choose (I usually have it at about 10fps), then when you press fully, it stores up to 40 shots from before you fully pressed the shutter (you choose beforehand how many shots to store from before the full shutter press).
I would love to have a full-frame mirrorless with that feature, but alas.
I've got a lot of nice photos of birds taking off, fish jumping out of the water, street photography to capture people in an exact position of my choosing, the exact moment a bat hits a ball, and many other things that would have been almost impossible to capture otherwise.
It amazes me how stuck in the past camera companies must be, there's so little innovation.
Even something as basic as wifi/phone support is almost always half-baked and barely functional.
There's so many opportunities to improve with good phone integration too. You buy a $2k SLR, yet it's still difficult and time consuming to simply snap a picture and send it using your phone. It shouldn't be this complicated.
Given how simple it is (a ring buffer for images captured using an electronic shutter mode basically) it's surprising that every camera with a decent e-shutter implementation doesn't have this.
The big camera makers do throw in a bunch of "modes" on their prosumer lines (which they toss out the window on their professional models) to pander to the consumer. Lame stuff, mostly. Give me some real stuff like auto-masking foreground subjects or out-of-camera photogrammetrly created 3D models.
I get that prosumers are a dwindling niche and pros are where the money come from, but you need a source for more pros in the future, right? How does the camera compete if--as an artifact--it's inspirational value is outclassed by the mundane cellphone?
The beauty would be that they already have all the APIs to support doing this, and you could run one of the existing 3rd-party camera apps too. (Hi, Halide!)
I drop $2k+ for one in a heartbeat.
Lightroom's autostacking works pretty well for this if the bracketed shots are close together in time as they are when the camera is automating the bracketing. The lightroom feature augments the typical camera feature in this case.
The biggest challenge is that the hobbyist photography community is relentlessly, angrily furiously hostile to improving cameras in ways that move beyond pretending we're still using darkrooms, but with pixels.
The hostility to anything that is more like the "fauxtographers" who take excellent photographs with phones is one rich source of that, but there are many more.
Automatic shooting of HDR exposures. Stacking them requires some human intervention and shouldn't be done on the camera, I'm simply thinking of the shooting. You put your camera on a tripod, select HDR and push the button. It shoots a frame and checks the histogram--are pixels falling off either end? If so, adjust only the shutter speed and shoot again. Keep doing so until the slowest shot has nothing falling off the dark end and the fastest shot has nothing falling off the bright end.
Alternatively Sony has "DRO Bracketing" which basically does what you mean with the histogram, but a bit better.
FWIW I can take pictures in a room with a window looking at the sky and change the exposure to clearly see both the sky and the room without even bracketing on a full frame camera.
> At best the camera could convert everything to linear or logarithmic light and save in OpenEXR or something like that.
Yes exactly. That's what HDR is.
> Currently Found On: Sony A9 II and A1; Nikon Z9; Canon full-frame mirrorless cameras other than the EOS RP
That's not really accurate - the Sony and Canon are just using one of the shutter curtains for this, while the Z9 doesn't have a shutter so gets a dedicated mechanism not nearly as fragile as a shutter.
> 6. Bulb Mode Preview
Not sure how this is supposed to work without having a non-destructive readout sensor, unless it's just an approximation based on a quicker exposure before starting the bulb exposure.
> 8. Native Image Averaging
idk all of those guys are shooting raw anyway, so just stack them which you were probably going to anyway?
> 14. Raw Histograms
Yeah.
> 15. 16-Bit Raw
Probably in the next generation of landscape cameras like the Z7III
> 18. Base ISO Below 100
Mostly a fight with physics.
16 bit raw is another matter. If you're not getting 16 bits of dynamic range from the pixels, all you'd be doing is getting a more accurate reading of the noise. That just adds a lot of expense for no perceivable gain.
Why would anyone want this? Let me show you how I photograph wild bats in flight at night. I have many other use cases. For now I have to mount a shutter on the front of my lens to gain back the shutter response that i use to utilize from these film cameras of old.
It's pretty fun what you can do with it. Here's an example from a backpacking trip last year:
without Live ND: https://i.imgur.com/kJjEDmt.jpg
with Live ND: https://i.imgur.com/IhHzo4T.jpg
Both taken (handheld at 100mm) on my Panasonic 100-300/F4.0-5.6. I tend to carry it and my Oly 12-40mm f2.8 when backpacking.
I hadn't thought of sensor shift star tracking. That'd be useful, assuming it would work with, say, the R6.
Canon EOS 6D: Standalone GPS receiver on the body. Takes a long time to get first fix. Weak signal on airplanes. Heavy battery drain. Does not work underground or in big buildings.
Canon EOS M6: Post-shooting tagging with GPS when camera connects to smartphone, using timestamps for correlation. Smartphones are extremely fast (e.g. A-GPS) and accurate (e.g. GPS+GLONASS), can use Wi-Fi for indoor geolocation, and can tag basically 100% of photos without missing any. The cons are the need for accurate timestamps, the need to keep the phone on and logging (potential high battery drain), and the need to manually invoke the sync action after shooting.
Canon EOS RP: Bluetooth connection to phone. The phone's geolocation is fast and accurate. But the Bluetooth connection is terrible; takes 5~10 seconds to establish; cannot geotag missed photos after the fact (unlike the M6); the phone has a tendency to evict the Canon Camera Connect app which means no Bluetooth connection to the camera.
Besides people were doing long exposure since like forever with great success. If the sensor overheating was a problem, it would make sense to introduce a hard limit and some cameras have them, but from my experience the likely reasons for the prevalent 30s are: 1/ most people don't do several minute exposures 2/ they need to sell remote triggers to somebody.
30s limit was there even during nineties when you could do hours on film without overheating anything.
One thing that most people comparing a film camera and digtal camera by exposure lengths is the fact that a digital cameras image decreases the longer the chip is energized. Heat builds up, and the image gets more noisy the longer the exposure. One rule of thumb for time lapse is to have an interval in between shots as long as the exposure being taken to allow time for the sensor to cool back down. YMMV
I often need it for HDRs at night. I know I can do like wayloads of high iso noisy short exposures and average them in post, but it's clunky and even harder to set up automatically. I would prefer something user friendly. When I'm shooting at night, it's often cold and even not so safe. Fiddling with apps and cable releases is not something I'd like to do.
I know that the hardware is capable of it, but they just have to sell you that cable release or what not in order to unlock the capability, which is what I'm not happy about.
You don't have to do high ISO anything. If you're willing to shoot 1 hour exposure (which are you really doing that on DSLR?), seems like you'd be willing to shoot 30 2-minute exposures, or 60 1-minutes. What doesn't motion blur out in 1-minute shots other than maybe cars at a red light? Also, how many 1 hour exposure shots can you get out of a single battery charge (2 batts if using grip). I know doing 45s exposures with short intervals will chew up batteries for me.
It seems to me that Canon has decided that if you're going to do longer than 30s exposure, you're going to be using bulb mode and an external controller. I know MagicLantern allows setting custom bulb exposures, but I haven't messed with that option. Its built in intervelometer isn't reliable for short intervals (anything 5s or less is iffy regarding consistent shooting frequency). With my wired controller, I can do 1s gaps. Anything less, and I start getting issues with dumping to card reliably on the old body. Using my custom arduino controller for .5s gaps with 1s exposure with camera mounted to car for motion blurry goodness freaked the mkII out for some reason. newer bodies handle it much better.
Second, an hour was just an example. Most of my exposures are not longer than 4 or 8 minutes.
Third, I'm happy that you are fine with external controllers. But I'm not. I find them crude and unnecessary. I'm doing a lot of HDRs and I don't understand why I cannot simply preprogram my camera with sequence of shots and execute it without any gadgets and limitations. Say, there's this nonsensical limit for 1 minute or 30 seconds everyone seems to defend for some strange reason, (is there some sort of Stockholm syndrome at work?), at least the manufacturer could provide a better bracketing setting to compensate for it. It's not like it's rocket science. An intern could write the firmware in a few days. I cannot even patch the camera without voiding the warranty. I know they are always workarounds, but why make things complicated, when they could be simple.
The mkII got video basically because someone realized the chip could do it, and stuck his head in an office saying they could make it happen, and then did it. It pissed off the pro-video camera guys to no end. So, I'm guessing Canon is not going to let that happen again by letting the interns loose. It's goes against Sony's DNA to let features go out for free.
But seeing as this is HN, many a device has been launched of people tired of status quo and went and did their own thing to make what they wanted. You're welcome to post us a Show HN when you've fixed your problems that will make things easier for the rest of the world too. Or, better yet, sounds like it's ripe for disruption, so why not get a YC backed startup going to build a better camera. Try not to make the camera equivalent of Homer's car though.
Like when you shoot moving car or flying bird - use existing autofocus to detect the object and then track it over exposure time to keep it free from motion blur but get nice, blurred background. Pentax have simple version of that for tracking stars mentioned in the article but I hope some company figures out how to do it for arbitrary, fast moving objects in real-time.
Sony a9 series already got rid of blackout during shooting so it is technically possible to get image data in realtime, have continuous autofocus and save full size RAW files at the same time: https://www.youtube.com/watch?v=_ZXFI-eIXk8
But I think with much faster sensor readouts you will be able to compute sharp images sometime in the future.
For me, the only real competitor is the Leica M system :)
I also don't get this term "full frame". When I started photography, "full frame" was 120 format. 35mm was "compact camera". And I think you need to go to 120 (at least) to do noticeably better than Olympus micro four-thirds.
With respect to the Leica: if you can trust your self control, by all means try out a Leica, it is an experience one should have. But yes, deposit your credit card at a safe place, they are lovely. For me the downfall was that I bought a M-mount Voigtländer 40/1.4 as a short tele lens for mFT, back then when there where not many prime lenses to choose from. It works lovely on a digital M :)
Ok, thanks. That's a no for me, then! :-)
No auto-focus, no video, limited feature light metering, etc.
I drooled over getting a Leica M for 15 years and finally took the plunge about five years ago. Best choice I have made gear-wise in all this time. After getting the first, I went ahead and got the Monochrom as well (yes, B&W-only sensor).
To learn how to see, you have to do the seeing, not let the camera do it for you.
Others fall well into the amateur photography traps of "I must have this techno gadget in the camera to get better", whereas Pros have been getting that shot for decades without said gadget.
I have a pretty fancy system (but have been gradually losing interest the last 5 years or so) but I feel like it's all becoming irrelevant over time and the camera companies are going to be left with a very small pro market (that's much more pragmatic about upgrades) and the hard core amateur market that just buys everything to show off.
The biggest things that would keep camera systems relevant would be much better connectivity options to your phone or other device. Make it really easy and seamless, and wireless. Or put LTE right in the camera.
Sadly such an interface doesn't even exist on Android.
The idea would be to sync perfectly multiple cameras to capture an event in time exactly at the same moment, but from different angles. I know professionnal cameras can do this using a Genlock, but with the amount of sensors and radios on modern cameras, I'm pretty sure this could be made available to the masses.
Canon has https://magiclantern.fm/
I think they’re great - absolutely perfect for situations with young kids who are constantly moving, or just to capture little vignettes of movement that bring a photo to life for a moment.
But support for them is virtually non-existent outside the iOS Photos app.
You can extract them as a pair of HEIC and MOV files, but I know of no Free photo management software that is capable of associating them and displaying them together (and yes, there are internal IDs in both files that make it possible to link them, so they aren’t quite as fragile as most sidecar files, although they are still fragile because Apple for some insane reason decided not to make use of the HEIF container format’s native ability to store both in the same file).
I’m unsure if Android has an equivalent feature or how it’s implemented, but I’m pretty sure no DSLRs have anything like it.
It looks like the Samsung Camera has something similar, but I'm not sure if it's compatible...
I'm not sure what you mean - pretty much all of these can capture 4K video where each frame is about as good as a still from a phone camera.
At the levels below that you'll have to settle for full-res stills at somewhere between 14 and 20 fps or 4K30p/4K60p, for now.
Bulb Mode Preview
Vibration Detection Shutter Firing
Native Image Averaging "
Pretty much vote for those. Not that the other are unimportant. "Native Image Averaging" - with the live preview please and the ability to zoom into that preview.
Also would be nice to see more options for manual focusing. Pixii is interesting being a digital rangefinder but there isn't much innovation here.
That's a wrong assumption, it was certainly never "the whole point", just one of the factors.
Many of the "live viewfinder" benefits already existing in DSLRs that had live viewfinder modes, with the rest being excluded for cost or differentiation (focus pixels, live histograms, etc). I saw it as an more of an eventuality of the live viewfinder being made more featureful than the lens view, so removing the mirror just made sense. The only people I know that are still using DSLR are those that "need" the classic viewfinder, but there's no reason the features of a mirrorless camera couldn't be put behind a sometimes-flopping mirror. The only problem with that would be vibration and the increase in size of the camera and lens.
Yes, that's the primary driver for the size/cost of an equivalent lens. Smaller lens also means less distortion/non idealities, for the same cost.
Field curvature, astigmatism, coma, and chromatic aberration need to be handled optically. A larger sensor remains beneficial, even if the lens is physically larger, at least as long as you keep a wide aperture.
There is a hard limit at F/0.5 that you won't get closer to than ~ F/0.65, but the crop factor makes it so you have a smaller absolute aperture with a smaller sensor to get the same field-of-view.
MFT needs half the focal length and half the F-number to match full frame in FOV and DOF. E.g. an F/1.3 full-frame lens already matches what exotic optics can reach on MFT, and an F/1.0 one would match the theoretical limit of MFT.
Speaking about more realistic MFT lenses like F/1.4, that'd be a rather harmless F/2.8 in full-frame. Yes, the latter is more expensive, but only because F/1.4 is quite a bit away from F/0.7, the practical limit for "normal" photography lenses (anything faster is probably exotic).
Not sure what other benefits mirrorless brings.
It gets an accurate lock and moves the sensor about its 5 axes of magnetohydrodynamically-controlled freedom in order to emulate, to some degree, an equatorial mount. No image-domain tracking required -- purely physics. The things people get are insanely good (c.f. [1])
1. Find bright spots in last second of exposure (which is then added to a sum for the final picture)
2. Compare their locations to bright spots in the previous second
3. Move the sensor slightly so the bright spots remain "in the same place" on the sensor.
Note that cameras with integral anti-vibration do that by moving the sensor, so they already have the hardware for (3).
For star tracking there would be a max speed of stars based on Earth's rotation multiplied by focal length you could use to set a reasonable cap on movement. You could also expect similar movement for each second to the seconds before.
The camera could have a data structure per star being tracked, capping the number used to say 32 or 64 "stars" for accuracy. Each "Star" would "vote" how much movement has occurred. You'd gain faith in various stars as analysis frames go by if they keep giving valid results, a black-list of "not stars" would also be useful.
Some would be stationary, like lights on the ground. Some would be bogus, such as navigation lights on aircraft. Actual stars should expect similar movement and "agree" though if the zoom was low it might change across the image.
Pentax had had this for a while (as early as 2011).
1. https://blog.arduino.cc/2020/02/26/openastrotracker-is-a-tra...
An alternative are studio lights which are set up to trigger when the camera flash itself goes off.
OR wifi sync and phone sharing that doesn't suck. But no camera manufacturer has shown the ability to do that, so an android app mode is the next best thing so they can offload the software writing to much bigger tech companies instead.
Top 1 annoyance with non smartphone cameras is sharing the photos I just took with other people at the event. I'm not fucking around with the frankly awful wifi experience on most cameras and I'm not going to bring a cable or sd card dongle to do so either.
# 2 is the lack of built in GPS. It's 2021, adding GPS is extremely cheap, especially if you've already done wifi anyway.
#3 is qi charging. After a shoot put the camera on the wireless dock and everything is syncing well with the local desktop computer.
No, i just want org-mode.
My bet is Apple
Casual users who care about apps will just use their phones and the quality will be on par with Micro-4/3 and likely APS-C with computational photography advances.
Prosumers will stick with smaller full frame optimised for style and old school usability e.g. Leica. Professionals will either use full frame for speed or move towards medium format which Fuji is leading the charge towards.