You are here

What do I want in a 5d Mark 3 (next generation digital SLR)


I shoot with the Canon 5d Mark II. While officially not a pro camera, the reality is that a large fraction of professional photographers use this camera rather than the Eos-1D cameras which are faster but much bulkier and in some ways even inferior to the 5D. But it's been out a long time now, and everybody is wondering when its successor will come and what features it will have.

Each increment in the DSLR world has been quite dramatic over the last decade. There's always been a big increase in resolution with the new generation, but now at 22 megapixels there's less call for that. While there are lenses that deliver more than 22 megapixels sharply, they are usually quite expensive, and while nobody would turn down 50mp for free, there just wouldn't be nearly as much benefit from it than the last doubling. Here's a look at features that might come, or at least be wished for.

Better Pixels

More pixels may not be important, but everybody wants better pixels.

  • Low noise / higher ISO: The 5D2 astounded us with ISO 3200 shots that aren't very noisy. Unlike megapixels, there is almost no limit to how high we would like ISO to go at low noise levels. Let's hope we see 12,500 or more at low noise, plus even 50,000 noisy. Due to physics, smaller pixels have higher noise, so this is another reason not to increase the megapixel count.
  • 3 colour: The value of full 3-colour samples at every pixel has been overstated in the past. The reason is that Bayer interpolation is actually quite good, and almost every photographer would rather have 18 million bayer pixels over 6 million full RGB pixels. It's not even a contest. As we start maxing out our megapixels to match our lenses, this is one way to get more out of a picture. But if it means smaller pixels, it causes noise. The Foveon approach which stacked the 3 pixels would be OK here -- finally. But I don't expect this to be very likely.
  • Higher dynamic range: How about 16 bits per pixel, or even 24? HDR photography is cool but difficult. But nobody doesn't want more range, if only for the ability to change exposure decisions after the fact and bring out those shadows or highlights. Automatic HDR in the camera would be nice but it's no substitute for try high-range pixels.

Video & Audio

Due to the high quality video in the 5D2, many professional videographers now use it. Last week Canon announced new high-end video cameras aimed at that market, so they may not focus on improvements in this area. If they do, people might like to see things like 60 frame video, ability to focus while shooting, higher ISO, and 4K video. I'm also interested in some audio improvements. For example, if I plug in an external mic, the camera should record both the external and the internal audio -- space is cheap. As well as other audio channels I will outline below.

I want the hotshoe to add pins for power in both directions, and I want the data protocol in it to be beefed up and made public and supported. This would provide many things, but one thing it would be great for would be hotshoe mounted microphones, which now would not need a battery and would not need a cable.

I also want bluetooth in the camera (see below) and that includes bluetooth high quality audio for video shooting. Multiple channels of it if desired. Using a trick I outlined in a previous post, you can have the remote bluetooth microphones emit a light pulse and a data pulse together, so that the video and audio can be in perfect sync even with the digital delay. Audio sync could also be done by having a protocol to synchronize clocks and then timestamping the audio packets. Remote bluetooth microphones are not common but as they become common they would be cheap and decent quality, and far superior to on camera microphones for recording people. And far cheaper and easier to use than today's wireless mics. Filming a group of people? Just pin bluetooth A2DP microphones on their lapels and go.

Ordinary cheap bluetooth headsets and headphones could also be used by camera owners to leave annotations on photos, and to silently hear audio feedback from the camera, such as autofocus beeps, audio playback etc.

Bluetooth or even WiFi

Wifi is available as an expensive and bulky add-on for some cameras, though you can put in an Eye-Fi card cheap. Still, it's time to get away from having bulky and expensive add-ons. I think a lot of useful stuff could be done by making bluetooth standard in cameras. In addition to the audio functions above, you could do a lot of useful stuff:

  • Camera control, same as over USB. This allows simple bluetooth remote shutter releases, but in fact it also means most phones and laptops could act as not just shutter releases but full remote controls, with access to all camera settings. Tethered shooting without the tether.
  • Low speed image downloading and low quality video preview -- but ideally one would use 802.11 for this or the Bluetooth HS protocols that do high speed adjucts to bluetooth over 802.11
  • Wireless flash. Infrared control of flash just isn't as good because you can't put the flash behind you when outdoors. You would need a protocol for clock sync to synchronize shutter and flash firing, but I think these could be done. It's just silly that you need to pay $200 for an ST-E2 to do remote flash anyway. This probably means a new line of flashes since they will surely overprice the adapter that lets you convert your old flashes to bluetooth.
  • Access to bluetooth based GPS receivers. These are very cheap these days. Assuming they don't just put the GPS in the camera.

Be a platform

Camera vendors just haven't gotten a message that has been a great boon to other devices like phones. They should open up and be a platform, let people write apps that run on the camera to customize it and enhance shooting. You can even have an app store. Yes, it means you have to define and support a platform, but trust me, whoever does this will soon have the most capable and popular cameras out there. Forget about tethered shooting and even the bluetooth control -- you've got a nice computer platform, tons of flash storage and a nice screen. Make the screen a touchscreen if you can and let people run apps on it. If anybody says "Can the camera do that?" the answer will become yes. Some of the suggestions I have could readily be done as apps rather than putting them in the main camera firmware.

Fix my common mistakes and other features

In spite of the bravado, pro photographers make mistakes in settings on their camera all the time. Yes, I want full manual control, but that doesn't mean that if I set for tungsten white balance shooting at night that I want it to keep that setting 10 hours later when I am working in daylight. Warn me if it's been a while since I made a manual setting and that setting is causing photos to look wrong in some clearly defined way -- wrong exposure, wrong white balance, pointlessly high ISO etc. This is the sort of thing that apps would also be good at doing and letting me customize.

Another minor feature would be an on-screen digital level, using the accelerometer in the camera. And while we're at it, it would be nice to have a crude form of image stabilization based on today's cheap gyros. This would work in low light by holding down the shutter and having the camera wait until it reaches a point of suitable stability and then shooting -- and warning you if you moved too much during the shot. There's also been useful work done in correcting blurry photos based on gyro information.

I would also like to see a mode where if I shoot, it shows me on the display the sharpest section of my photo, at 1:1 resolution or near that. If that sharpest section is blurry, I know I have a bad shot. If it is not where I thought it was, I may have focused on the wrong spot.

Go mirrorless (well, not yet)

Most people agree that eventually Canon and Nikon will go into mirrorless SL cameras (no longer SLR, the R refers to the reflex mirror.) They won't do this in the successor to a current camera, it will come in a new camera line, but the time is coming. The mirror giveth a direct viewfinder view, but the mirror taketh away so much, and the balance has really shifted of late. We get mirror slap vibration but most of all we get serious limitations on lens design, since the lens has to stay far enough away from the sensor to leave room for the mirror. People don't know it, but this is one of the big reasons that wide angle lenses are so expensive.

Yes, we would have to go to digital viewfinding. But that's getting better and higher resolution, and it can even surpass the human eye in the dark in some ways. I find that if I want to do truly accurate manual focus, using the zoomed-in on-screen focus is better anyways. I actually suspect the right choice in mirrorless will depend on the the state of the art at the time the format develops in terms of getting low noise at high ISO in a given sensor size. It should actually depend on predictions for that in the future, because I want the lighter lenses that come from that.

Once you remove the mirror you can also consider a circular or cross-shape sensor. With such a sensor, you are shooting both portrait and landscape at the same time. Or if you are choosing which one to write to the card, you can do it in software and not have to tilt the camera. Tilting the camera makes it harder to mount on a tripod and harder to shoot naturally without a big battery grip. This precludes certain styles of lens hoods on wide angle lenses but it has a lot of advantages.

Of course if you want to get really far-out, you can consider the idea of lenses with the sensors built in.

Image stabilization

Right now if you want IS it comes in the lens. That's a good place to do it, but it comes at a very high price, and you pay it per lens. I would like to start seeing sensor based IS which would give me basic IS on all lenses, and possibly help the IS when I have a lens which also features it.

At some point, we will want to see cameras take longer exposures as a series of super-fast short exposures, with gyro-readings and software stabilization. This allows effectively a virtual tripod, and HDR at the same time. It's been a common technique in Astronomy for a long time, where the stars are not moving but the atmosphere is.

Power in the hotshoe

Above, I noted a desire for power in the hotshoe, along with high speed data. I want power in the hotshoe for a few reasons. First, it would be nice if you could get a very small flash to put there, useful for fill flash and low power flash applications. Right now your only options are huge monsters. One could even have a small LED based flash there for low power use. If you needed a flash transmitter like the ST-E2 you could power it from the camera as well, though again I think this should be built in and radio based.

Even with larger flashes, you could allow charging from the camera. My camera uses custom lithium-ion batteries but my flashes use AA, meaning I have to carry two kinds of batteries and chargers. I would rather just carry more of the one type and have a less heavy flash. Or I would like to still be able to use my flash if the battery in it dies, even though it's going to eat up my main battery. And finally, I want the power to be two way, which means that the flash can possibly power the camera, or I can even get a hotshoe battery pack to give me extra power, instead of the battery grip. The data bus would also allow the microphones I describe above, and other peripherals.


I'm happy with the Pentax K10D. Sure, there are better cameras, but a) it probably has more bang for the buck than any other camera and b) only very few people need something better. Note that it has shake-reduction in the camera, not in the lens. This means that lenses can be less complicated, one can use any lens (it is compatible with essentially all Pentax(-style) lenses as well) and if one needs better shake-reduction, just replace the camera, which is cheaper than replacing all lenses (especially if the lenses have shake-reduction). Many people will upgrade their cameras anyway every few years, while there is little reason to stop using a good lens.

It is no longer produced, but there are similar successor models.

Anyone who doesn't yet have a digital SLR should consider Pentax, especially if you have Pentax(-compatible) lenses. You will find more for your money.

Because it happens at the end of the light path and requires larger swings of the sensor to compensate for shake, moving the sensor is not as powerful a tool for IS as moving a lens element inside the lens. However, there is no reason not to make use of both.

Almost all pros shoot Canon or Nikon because they have decided to spend bucks to get the best, even though other brands are also quite good. But the selection of quality glass with the major brands is much higher, and there is an active aftermarket for them on eBay as well as all sorts of accessories.

I agree with both statements above. For the pros, or those who take photography as seriously as the pros, Canon or Nikon is probably the way to go, for the reasons you mention. However, for someone like me, who used a conventional SLR and a few lenses for decades and waited until DSLRs were good enough before switching, and plans to use this camera and at most a few lenses in the next, say, 10 years, I think one should seriously consider Pentax.

Pentax also supports DNG as well as their own raw format (and of course JPEG). I save both raw and JPEG on the card.

What do you think of DNG, both in terms of technical advantages/disadvantages and in terms of a future-proof standard (assuming it is)? Rune stones are still readable, 10-inch diskettes probably aren't in many cases.

Yes, there are old formats that it's hard to find readers for. I think the raw and DNG formats will retain decoders though since there are open source ones. On the other hand, I find all the raw/dng formats to be a pain to use in photo manipulation programs, and it would be nice if it was as fast and simple as jpeg.

What formats do you use for your own personal archiving?

I shoot either jpeg or raw+jpeg when in a setting of unusual light or contrast that might need raw. The truth is that raw is a pain to work with, so I use raw+jpg an use the jpegs for browsing, organizing and experiments and go to the raw if I need it. I archive as shot, I never change (other than to rotate or delete complete losers) my original files.

I also shoot both and save both (disks are cheap). The JPEG the camera makes is almost always all I need, but it's nice to have the raw data as well. I also save it as shot (of course copies can be manipulated). But what raw format do you save? The native camera format or DNG (assuming your camera can save DNG)?

Provide only their raw formats, usually. Fortunately canon raw is common enough that there are many tools to decode it, including open source ones, so I don't have too much fear of it not being decodable in the future.

The discussion of file formats brings up another issue. It would be nice if camera companies would promise to put the camera firmware in the public domain after a few years. That way other people could update the camera firmware to new formats and features without buying a new camera. I seem to recall that Canon did something like that with some of their low end point and shoot models. I could see where that could be a major selling point on new cameras where you could have a camera app store to let the user load what ever features they wanted into the camera. A library of camera functions on the SD card would be kind of cool.

I was shooting Canon for a while, but I went back to Minolta/forward to Sony because my best Canon glass was actually Tamron, and Canon dinked with their lens protocol in some way. I'd started with Minolta, but moved to Canon when the IS lenses came out. So when Sony introduced IS in the body (duh!), the cheaper move was a new Alpha 100 body, not buying new 28-300 lenses for my Canon. Sony's stayed close enough to Canon that I still don't see any benefit to jumping. IS in the body is just too useful. It can be like gaining a bit of low-noise/high-iso, or buys me a little deeper field, or whatever.

How about gesture triggers? Some years ago, when I was using the Canon and had an infrared remote, I got appointed to get a group shot at a family reunion. There were so many of us, that even the IR remote couldn't trip the camera. I had to set it on timer, run forward, trip it with the remote, run back, and get in position before it went off. It would be so handy if I could tell it "If I cross and uncross my arms three times, then go to a 2-second delay, then fire." The WiFi/smart-phone remote idea might work (Bluetooth wouldn't have had the range), except I don't have a smart phone.

The feature from my Maxxum7 SLR which I most regret not being able to move to my Alpha 100 (and, in fact, the ONLY accessory which didn't bolt right into place on the new camera) was the vertical grip. Not only did it make the camera big enough to actually be comfortable to hold, but it meant I could power the camera with 4 AA batteries. As you noted, being able to carry a couple spare sets of AA's and a USB charger was *way* more convenient than having to have spare exotic non-rechargable lithium camera batteries, or having to carry a separate special charger for the camera. And, worst case scenario, I could (and once did) just walk into a corner grocery and buy some AAs to power the camera (and flash).

How about replaceable/field-swappable sensors? I would def. appreciate being able to swap out the standard full-color sensor for an ultra-sensitive monochrome one. High-speed night shots in black and white would be very useful at times.

Low power mode? I don't have (or didn't bring) enough batteries, or it's really really cold and battery life is compromised. Shut down face detection or motion tracking and image stabilizing, use slow writes to the memory card (if that would help), default to non-auto-focus, default to reduced power flash, or whatever else might allow me to keep shooting longer.

That's all that comes to mind at the moment.

Whatever combination of mechanics, fast memory, and fast image processing is required to make my camera shoot what's in front of the lens NOW, not a quarter or half of a second in the future.

The DSLR cameras have never really had this problem. This is a problem found on many P&S but in many cases it is due to a misunderstanding about the two things which cause this delay. The first is the autofocus and the 2nd is the delay to actually open the shutter once in focus. I don't think any DSLR ever made has ever had much of the #2 delay. All cameras have some autofocus delay, but expensive lenses have better focus.

However, pro shooters tend to just put this into their flow. We always, when framing shots, focus in advance with a half-press of the shutter. You can do this on most of the P&S as well. If you focus in advance, even a decent P&S will take the photo pretty much when you press the button, though the DSLR are better at that.

I only recently made the shift from chemical analog photography to electronic digital and started with a 7D. My next quest is the 5D and waiting for the 5D3. Mostly my wants align with yours.

I do have one want that hasn't been mentioned. I'd like a special camera (call it 5B or 5M) specifically designed for black and white photography. The sensor would eliminate the color pixel filtering altogether (leave the usual IR blocking filter), but with the same pixel geometry as the color version. This would boost sensitivity, and the firmware will obviously need to be tweaked. The RAW format would just need a flag to say it's monochrome.

I too would like to have the mirrorless DSL camera. But I'd like to see the lens line start showing up with leaf shutters for quieter, more reliable operation, and faster flash sync. Then the focal plane shutter is optional. Even less vibration.

As for the platform idea, that could still be done by adding an ARM CPU along side the existing one, without the need for a major redevelopment. Adding interfaces would be the heavy work part. And the camera could still operate with the ARM CPU disabled (and off to save battery). The platform CPU should boot from the CF card if an image is on there (Android would likely be the popular system, but BSD and Linux builds could work, too), to allow setting up a custom app mix differently for different shootings (set them up on the cards in advance). If you get a trojan in your apps, you can just wipe the CF to remove it for sure.

I'd also like to see the video time limit issue solved.

So they have made custom B&W cameras for astronomical users, who really need that extra light. But few others need it so much as to make that a product, since you can do pretty well with the Bayer filter still on, and only a few specialized users will notice that difference.

Any way to get a platform is good, but frankly I see no reason to not let folks have-at the digic CPU.

Caveat: I'm a theoretician (or is that a theorist?), not an observer. However, IIRC, most CCDs used in astronomy (where they were used long before they became cheap enough for professional photographers, let alone amateurs) have just one type of pixel, but are mainly used with a broad-band filter, say red, which lets through a well defined spectrum. Some of the newer astronomical CCDs have hundreds of megapixels.

Add new comment

Subscribe to Comments for "What do I want in a 5d Mark 3 (next generation digital SLR)"