Let’s start with a short editor’s note: On April 21 we reported about a new Sony CMOS sensor with amazing 120 fps at a full 12 Bit resolution (source DC.watch.Impress). I still noticed that many readers didn’t understand that this is a sensor for videocameras (and not for Alpha cameras). Thom Hogan explains why this is so: “It’s 24.3×12.8mm in size, which makes it slightly wider than APS-C but substantially shorter. The 2160 vertical pixels are exactly double what’s needed for 1080P (the horizontal component is almost exactly 4x). It looks like a straight-out video sensor to me. That’s confirmed by the other details: 60 fps 14-bit output, 120 fps max speed, and the first target: NHK’s Super Hi-Vision format.”
But the real interesting part of Thom Hogans article is the following one:
“To what end? Well, DSLR-enabled video is pretty compromised and about to be eclipsed by prosumer and pro video equipment, much as I predicted some time ago. Both Sony and Panasonic now have large sensor video products, with more coming. Those products blow away the capabilities within DSLRs for one simple reason: the sensor is optimized for video use. In order to get 1920×1080 (1080P) or 1280×720 (720P) out of our DSLR sensors, the cameras do two things: they line skip and they sub-sample within a line. The results are rolling shutters, artifacts, and less than optimal color information. Those dedicated large-sensor video cameras fix all those things. The output from the new Sony NEX-FS100E is noticeably better than any video-enabled DSLR I’ve seen to date. If I’m serious about video production, why would I compromise by using a DSLR? Plus the FS100E is configured to be a video camera from the get-go.”
What do you think?
Reminder: The Sony NEX-FS100U will be in Stock in July (Source: B&H).
43 Responses
Let’s hope he is right. Maybe then we can see a ‘real’ a700 relacement one day, not compromised by the urge to include video.
For relatively serious use and above, I would certainly agree.
But in the eyes of the average consumer with a limited budget this will still be an optimal solution.
I think Thom meant the future of video is videocameras with large sensors. I will have to agree. Not because of any technical reasons. Sony, Panasonic, and Canon have vested interests in keeping the prices of their video cameras high. The 5DII and GH2 set troubling trends for these guys. It’s too late to put the genie back in the bottle, but they can push the minimum specs of these video cameras even higher, thereby differentiating the low-end video 1080p DSLRs from the 4k and 8k videocameras.
It’s just the progression of technology trickling down. 1080p at maybe even 60p may be all that we DSLR users will get from these guys for some time. Maybe a decade from now, we’ll get 4k video in our DSLRs too, but only after the video guys extracted maximum profits from their video customers.
I agree. For the pro-market, DLSR’s doesen’t offer anything special compared to dedicated video cameras, such as the FS-100. However, for the normal consumers, they offer supreme image quality and flexibility (Different lenses, expandebale), compared normal consumer video-cameras.
Oh gash…. can we stop quoting mr. nobody with a website? Please? Cause with each quote of his works the ego of guy grows a bit – if you won’t stop it’ll eventually explode vaporising entire Earth.
And I want to live up till 2013. Mayans need to be wrong!
So, are you saying the guy is WMD?
Using a DSLR for video is not a pleasant experience. The user experience and handling is all wrong, the sensors aren’t optimised and nor, for that matter, are the lenses. On a video camera, any zoom action should be powered.
Once large-sensor, relatively cheap dedicated system video cameras come along it will just leave DSLR video facilities as a convenience add-on.
ps. The funny thing is that words of mr. nobody with a website in a matter of large sensor camcoders versus video DSLRs appear almost like a direct quotes from one of longer FS-100 presentations we’ve seen before on the SAR. He usually is little bit more creative with his imaginations….
This seems to always strike people as “news”, but to anyone actually in the business of professional filming this is blindingly obvious. Why do you suppose 99% of big-budget film-making is still done using celluloid film cameras and subsequent frame-by-frame film scanning? It’s certainly not for convenience 🙂
Just the fact that DSLR video is stored in a compressed format renders it entirely unusable for professional applications. Everyone seems to think rolling shutter is the main obstacle to it being used professionally, but that really is just the tip of the iceberg.
High end dedicated video sensors such as this might start to tip the balance away from film over the coming years, but DSLRs are a hell of a long way off.
Not to say they aren’t great for amateur filming. They’re a world apart from the cheap ‘n’ nasty camcorders and VGA video capture of years gone by. It’s certainly done a lot to put semi-pro quality video production within the financial reach of individuals.
+1
Mr. No body piss off by Panasonic? lol….
For the serious videographer for sure. But for me as consumer I doubt. I would not carry with me a DSLR and a camcorder so I need to chose what I want most, best video with PS quality photos or great photos with good (but not the best video). At least now I tend to make more photos with ocasional short video clips 1-4 minutes in length. I would say that 80 % is photos and 20 % video. But the video should also be good quality this is why I want a DSLR with good HD video. If I woul shoot hours of video, I would chose a camcorder of course.
+1
also depending on the needs.
99% of the content I create is for the web and anything over 1080 simply needs to be compressed anyway. My guess is that over 95% of the market uses their cameras the same way.
So, yes for the elite movie maker they can shoot 4k til the cows come home but for most consumers pushing 4k video around to edit video is going to be a major hassle.
I used to think I needed to have full HD 1080 to make my videos great but you’re not going to be streaming that on most computers over the internet without compressing it.
I’d rather have a 720p 24/30/60fps camera without 1080 than a 1080p camera with only 30fps and without 720.
Everyones needs are different but the meat of the marketplace IMO isn’t looking for 120 fps at a full 12 Bit resolution (yet)
As much I lock horns with Thom on most issues … I have to agree with him in some sense. One of the main reasons why Sony is transitioning towards the SLT and EVF is due to video capabilities and phase detection AF to make it stand out from competitors. Being someone who loves the traditional OVF on my a900 and does not use video besides on my NEX-5 just on occasion, I am worried about how SLT will be incorporated into pro models like the future A77 and A99. I just would really like Sony to offer an OVF version of whatever next releases they shall make for people like me that hardly ever use video.
I also feel that Sony need to maintain A900 type camera that is focused on still capture.
With the shift to EVF, i wonder if Sony is abandoning the market they inherited from K-M? Perhaps Pentax will step in and be the brand that offer satisfying still cameras? Oh how i wish the Pentax menu/ui is more elegant like Sony’s.
Finely it sounds like someone is writing sense about this trendy DSRL video thing.
Granted for some, having a camera and optics designed for capturing still images that allowed them to dip a toe into video was cool and possibly cost effective. And maybe having an all-in-one solution like an a55 or one of the lesser video implementations like the Canon is helpful for the mom and dad and the grandparents to capture the grandkids and maybe event/news reporters and war correspondents, But looking at all of the aftermarket brackets and odds and ends needed to turn a still camera design into video platform I wonder why not just buy a device designed out of the boxed to do video from the beginning.
But speaking for myself, a still photographer I want the device/system that allows me to make the best still images possible without compromise, added weight or device complexity. And if I believed the same way about video as I do about my still photography I would want the very best video camera I could afford. I’m sure that videographers would agree.
Personally, I wouldn’t be a bit upset if Sony’s a850/a900 replacement that hopefully will come around in a few years didn’t have video at all, but delivered a spectacular high resolution still image with best in class low-light performance.
Steve, you forget that to apply your own personal non-requirement for video to the rest of us is a bit of a rubbish argument. I don’t need my camera to do astro-photography but I can understand why some do, therefore to say DSLRs shouldn’t have video because you personally have no use for it is like saying we shouldn’t have the colour red because you prefer blue.
Its all about having the broad range of tools in your box…
Also this thing about video cameras being more ready in terms of rigging out of the box is nonsense. Every video camera is basically a box that makes pictures. You still need rigs for handheld work – like rails, rods and shoulder supports, you still need monitors and audio recording devices, you still need dollies and sliders, or even tracks and cranes if you really want to go high end.
Same for cinema or pro video as for DSLR video.
The rigs are not a hassle by the way… they are well designed and add greatly to the usability of the camera.
User theseday are very smart. They want cool video with shallow of DOF and they also want to save $ possible. They wish to invest a semi pro dslr with 1-2 prime lenses to gain ultra nice still, and same moment cound enjoy great HD video (eg video made in dslr: philips bloom-Cooking). of course the user are aware they could not ask too much about to remove those video artifacts (moire,RS,etc). and likely they will not carry a semi pro dslr & bulky vg10 for during traveling. Major of the cheap camcorder just cannot produce great shallow DOF in video.
well… I guess Thom’s brain needs update.
I checked with the manufacturer and there is no firmware update available yet. They suggest rebooting.
lol… 🙂
That new Sony sensor should not be a great mystery, they unveiled it in the Sony F65 at NAB and released several pages of brochures explaining the new sensor technology.
I did notice that on here you said it was an APS-C sensor at one point. If you need to run any video related things by someone you can always email me and I’ll be glad to help out.
Thom Hogan is really annoying… he hasn’t done his research into that sensor properly. Plus the implementation of video in DSLRs doesn’t hurt stills at all! It’s a by-product of the CMOS sensor technology speeding up and data-rates increasing to the point where they can allow video. No stills quality was lost in doing that, in fact the greater computing power and sensor speeds now available allow for better image processing of stills, better noise reduction methods and better continuous shooting rates.
Time Mr Hogan started checking his facts.
> Thom Hogan…hasn’t done his research into that sensor properly.
Please point out a fact that I made a mistake on.
> Plus the implementation of video in DSLRs doesn’t hurt stills at all!
Probably not in the sense of image quality. Because you’d have to bring more bandwidth to the sensor you might even get some modest benefits (assuming your buffer and write mechanisms can keep up). But your belief that spending time working on video in DSLRs doesn’t hurt the “still side” of the camera is incorrect in one very important sense: the engineering investment is not generating advances in still capture. I’ve got a huge list of things we still haven’t seen done in still cameras. If the engineers are spending most of their effort on improving video features, the still stuff doesn’t get done. Period. Nikon is masking that stall by upscaling the cameras. The D7000 looks like it is more of a camera than the D90 it replaces because of all the “new features.” But those new features were just moved down from the next higher model, not something that needed new R&D. Thus, the primary differences between a D90 and D7000 are: (1) stealing features from the next higher model; (2) the next generation sensor (18mp instead of 12mp, moderate improvement in still capability); and (3) lots of attention to fixing video. What I and others have been arguing is that we don’t need video fixed: put those resources into features we have yet to see on the still side.
> It’s a by-product of the CMOS sensor technology speeding up and data-rates increasing to the point where they can allow video.
By “it’s” I assume that you meant to write “video is a by-product of speeding up data rates on still sensors.” Very true. But we’re currently adding bandwidth we don’t need and aren’t using for stills. A 6 fps still camera doesn’t shoot any faster because you move data off the sensor faster. The one place where we did see a benefit was in 14-bit transfers, but I’m not even sure that was due to more bandwidth–my understanding it’s due to a different way of measuring the data on sensor.
> No stills quality was lost in doing that, in fact the greater computing power and sensor speeds now available allow for better image processing of stills, better noise reduction methods and better continuous shooting rates.
Uh, not really. There’s not really any evidence that this has happened, and plenty of evidence to the contrary. “Computing power”: cameras have dedicated CPUs and imaging ASICs. Both have clock speeds. You don’t necessarily get “faster stills” just because the data is moved off the sensor more efficiently. If you do move data off the sensor faster, you may need faster CPUs and faster ASICs. Oh, wait, that’s exactly what we got: faster ASICs. But most of the ASIC improvements went into handling and compressing video, not dealing with stills.
Moreover, frame rates are arbitrarily capped in still cameras. Nikon these days thinks 4 fps is entry, 6 fps is mid-class, and 8 fps+ is high-end. Has nothing to do with the sensor. Has to do with plenty of other design decisions, like that ASIC and how much memory you want to put into the camera. Indeed, we risk getting counter-productive moves by camera companies: if you can speed up the processing chain more, you can remove memory from the camera and still have the same buffer size. Voila, higher product margin.
I don’t know why you have such a bee in your bonnet regarding video on DSLRs. Surely a more positive approach would be to try your skilled eye in another discipline. Surely this is a good moment in history to see if you enjoy cinematography as well as photography. Instead you seem to be taking a very negative view of video. These DSLRs are capable of absolutely stunning results and many working professional photographers are now offering video as part of their services.
Also there is just no proof yet that Canon and Nikon are putting a dime of effort into improving DSLR video. Panasonic are the only company to do that and they are the only ones with a true stills / video hybrid product. The Canon DSLRs have a bolt on video mode with the bare minimum of effort put into it. Line skipping, crippled firmware, poor codecs, very old ASIC architecture – none of it changed with the 7D other than an additional CPU core and better HDMI output, and none of it improved with the 60D or 600D or 550D. So the 5D Mark II, which was 3 years ago, was the last major development in DSLR video. Nikon have put even less effort into it lately.
I think you worry that they divert masses of time and money to video – they don’t! Video is basically a hack of the existing live view system with a few bells and whistles on top. They put 90% more effort into live view than they did the video mode.
They haven’t neglected to implement the stills features you are asking for because of video – that is simply the wrong argument. I feel you are making video a scape goat for your frustrations on stills functionality. Believe me Thom there are a number of features cinematographers want to have as well… and a video camera is not always the answer.
Video cameras do not have large 35mm photographic sensors like the 5D Mark II or Nikon D3S. The Super 35mm standard is considerably smaller. Did you research into HOUSE and Gale Tattersall? Also for indie filmmakers video cams with large sensors start at $6+ when a 550D is $700 or less on eBay and video cameras can’t take stills. There are war photographers and photo journalists who need to carry around minimal gear and to have stills and 1080p in one package is a godsend, and one of the major reasons why we have video modes in DSLRs aside from the marketing pull of HD video for consumers – because these pros requested it!
Many more reasons are in my response to your blog here
http://www.eoshd.com/content/636
No factual complains about your Sony sensor coverage Thom but if you’d seen the NAB launch of the Sony F65 you’d know exactly what camera it is for, it’s no mystery sensor. You make no mention of the F65 and there is tons more info about the sensor in the brochure. I’ve been watching your blog for a while Thom and generally it’s an enjoyable read but the reason I think you don’t do your research into the video side very much is because you seem to be uninterested in the art and craft of cinema and producing cinematic images. It’s fine that you have no interest in that but please at least don’t resort to negatively belittling the vast and growing community of DSLR video users, many of whom are your own readers.
Andrew Reid – now i learnt your name. you have very good points there in this debate 🙂 I gives 10 points +3extra! (just kidding) 😛
> I don’t know why you have such a bee in your bonnet regarding video on DSLRs.
The issue is simple, there are three basic possibilities:
1. Still cameras should be still cameras.
2. All cameras need to be still and video cameras.
3. Still cameras are completely dead, all future cameras will be video.
I’m of the belief that #1 is where a company like Nikon–the one that I follow and report on–should be. They are not a video company. Despite a few minor dalliances in motion over the years, nearly 70% of Nikon’s current sales are still cameras. When you have that much reliance on still cameras, you’d think that you’d put more R&D effort into the still side of the camera, not the video side.
Nikon seems to think that #2 is the correct answer AND that they’ve already got the still stuff down pat. I disagree, and I’ve enumerated my disagreement in detail for two years now.
> Instead you seem to be taking a very negative view of video.
Considering that my undergraduate degree is in filmmaking and my graduate degrees are in broadcasting, I actually have a very high regard for video. But I also believe that a video camera should be a video camera, just as I believe a still camera should be a still camera.
> Also there is just no proof yet that Canon and Nikon are putting a dime of effort into improving DSLR video.
I’ll just take Nikon: most of the changes from EXPEED to EXPEED2 are video-related. We now have a Nikon-branded microphone for video. The way Live View works was changed, mostly for video reasons. Contrast AF has been worked on extensively, mostly for video reasons. So let’s look at the STILL CAMERA differences between the D5000 and D5100: new sensor (with more video capability), a couple of new post processing gimmicks in the imaging ASIC, and everything else is merely stolen from the D5000 or other still cameras has produced. No, I think the evidence is that more work is being done on the video side of these cameras than the still.
> Believe me Thom there are a number of features cinematographers want to have as well
I’m well aware of that. But the way you’ve framed your argument here I have to read that as “so if DSLRs give us a feature that the video cameras don’t, great.” You’re guilty of the same thing I’m accusing the camera companies of. By all means, let’s make the best still cameras we can and the best video cameras we can. If adding SOME video features to still cameras is a slam dunk, then great. If adding SOME still features to video cameras is a slam dunk, then great. But what I see in the development side of the DSLR makers is that a lot, if not the majority for some products, of the development time, effort, and money is going into the video side. It was even said to me in a direct way by one executive: “we’ve caught the video makers asleep and we’re going to steal their customer.” As I predicted when the D90 came out, we now have large sensor video equipment in response to the “DSLR threat.”
The notion that the DSLRs will shove aside the pro video equipment is a faux goal.
> Video cameras do not have large 35mm photographic sensors like the 5D Mark II or Nikon D3S. The Super 35mm standard is considerably smaller.
Correct. But is there really a move afoot to get more DOF isolation than movies, or was the move about getting the same DOF isolation as movies? Virtually every videographer/moviemaker I talk to speaks about achieving “equivalence” to Hollywood, not exceeding it. The one thing that the larger sensors of the 5DII and D3s give you is better low light performance, but that advantage won’t last long.
> Did you research into HOUSE
Yes. Did you? One of the reasons why the 5DII was chosen for use on that production was size. Indeed, we have a number of Hollywood uses were size was the determinant, including the use of m4/3 cameras in a recent horse racing film. But as mainstream, shoot everything on it cameras? DSLRs don’t win that battle.
> No factual complains about your Sony sensor coverage Thom
Yet you wrote “Thom Hogan is really annoying… he hasn’t done his research into that sensor properly.” Be more careful with your accustations.
> you don’t do your research into the video side very much is because you seem to be uninterested in the art and craft of cinema and producing cinematic images
Funny. I was just asked by my alma mater to come teach a class in video production using DSLRs. You seem to have missed something that admittedly is somewhat subtle: my writing about video in DSLRs is mostly pointed at the market my Web site serves: sophisticated STILL shooters. This group doesn’t want (nor do I) to see the advancement of still photography come to a standstill while manufacturers pursue a false dream of winning over the video market with HYBRID equipment. The fact that I argue that on my site has nothing to do with whether I value or shoot video at all. I do shoot video and continue to explore it deeply, but I do not have a site that speaks to that side of things, nor have I shared any of my recent video work (I’ve intentionally kept my name off it).
Simply put, I want still camera makers to make the best possible still cameras they can, in every variety that we need. I want video camera makers to make the best possible video cameras they can, in every variety that we need. The so-called “convergence” product (video/still/maybe more) is a mass market product at best.
And let me take just one of your arguments you make on your blog and expand on it: Larger Sensors.
You make the argument that the larger sensor of some DSLRs is better than the larger sensor of some video cameras. Take a sensor like the Canon 5DII versus the one in the Sony F3, for instance. Bigger is not necessarily better when it comes to low light, as Phillip Bloom just proved. Why? Because the DSLR video implementations are subsampling. They skip lines of vertical photosites and they subsample on the horizontal axis to get to 1080P. The dedicated video sensors use larger photosites that aren’t subsampled.
I’d also disagree with your assertion that the GH2 is better than the GH1 in low light because of gapless microlenses. In my measurements, the two cameras are very close to equal, with the GH2 being slightly worse in dynamic range.
Your comment “lower pixel counts actually produce more visible noise at the same pixel sizes as higher resolution sensors, because scaling a high resolution image down to 1080p reduces the size of the noise grain and makes noise less noticeable” isn’t quite right. For STILL use, two sensors with different pixel counts but everything else the same should produce EQUAL amounts of noise when viewed normally, because the overall sampling area is the same (and thus shot noise should be the same overall). In VIDEO use, your comment is incorrect about the results in DSLR video. You don’t actually SCALE a full still sensor down to 1080P. Instead, you subsample. You’ve therefore changed the area being sampled and thus random photon distribution will become more of a problem.
The DOF sweetspot is occupied by the APSC sensors – which is another reason that these cameras will outstrip full frame DSLR video, (from Luminous Landscape) “The smaller (but still large in cine terms) APS-C and Four Thirds sensors are at the sweet spot; small enough to allow shallow DOF when desired, yet not so shallow that a moderate aperture presents lighting issues. – Thus we will have both Sony and Panasonic with their new video cameras having sensors aimed at the DOF sweet spot.”
http://www.luminous-landscape.com/essays/mirrorless_video_convergence.shtml
Videocams captures crappy stills (no RAW etc.) and DSLR records crappy video |(no RAW etc). That’s it.
Video cameras can capture raws, the new Sony F65, that has a sensor related to the one refereed in the article, captures a 16-bit RAW at 4K in fact. Then there is the Arri, SI2km and Red. Sony also has what is called a ‘digital negative” S-log that captures all sensor dynamic range for grading (which is in the F3, which shares the sensor with the FS100 pictured above).
And we may see RAW video even in dSLRs down the road, we are already seeing 1080/60p in cheaper cameras, JVC is going to make a 4K video compact point-and shoot and low-end video camera.
Sensor bandwidth is already insanely fast. I think this is the main reason why Sony has move to SLTs, that removes that slapping mirror. 10fps of the A55 is nothing compared to what a cameras will do in the future. Imagine taking 120 fps with a 17.7MP camera on an SLT. The technology is here right now.
I’m ready to compare any video cam stills to A900 DSLR. None of video cams is not featured for pro quality stills.
Crappy video… har har. You haven’t seen the RedTails shots which make up a third of the film that Philip Bloom shot on a 7D then I take it? If it’s good enough for Lucasfilm… it’s good enough for HOUSE, who shot an entire episode and most of the latest season with a 5D Mark II.
Yes the Canons line-skip…
GH2 doesn’t… it subsamples and intelligently so. The scaling on that is superb and the resolution easily matches that of a dedicated HD video camera. You can also mount PL mount cine glass on it because it’s mirrorless. Ditto with Leica stuff and Zeiss rangefinder lenses… beautiful glass on a video / stills hybrid that shoots excellent cinematic 24p with full manual controls… for $800. What’s to complain about guys?
Just to be clear, I didn’t say that video in DSLRs has no future, I said that professionals seeking out DSLRs for large-sensor video will diminish due to the emergence of dedicated large-sensor VIDEO cameras.
At this stage in digital, basically every digital camera can or will do video of some sort. The question is whether the DSLR makers should spend time and energy trying to dislodge dedicated video pro companies with DSLRs by upping the capabilities of DSLR video. My answer is that “it is not worth the time and energy to try.” All you have to do is look closely at the output of the large-sensor pro video equipment versus the artifact-laden stuff the DSLRs are spewing out and you get to that answer very quickly. To EQUAL what the large-sensor pro video gear is producing you’d need to add enormous bandwidth AND change the way the sensor works in a DSLR. The problem? Spending all your time doing that just makes you equal to the pro video equipment and it does nothing for your still customers. You risk losing a generation of still customers not upgrading.
hi admin..
you might want to post an article from a real video guy who writes about the advantages of dslr video.
http://www.eoshd.com/content/636-Advantages-of-DSLRs-over-video-cameras
The headline says “DSLR with video has no future”. I wonder where the admin got that idea from Thom! It’s true some pros will go for video cameras over DSLRs but it misses the much bigger picture of how DSLR video is growing and converging with photography.
But my biggest bugbear in all of this is…
What do you say to the tens of thousands of aspiring filmmakers and talented artists who can’t afford to drop $10k+ on a pro video camera? Go back to shooting DV on a handycam?
The professional argument has an elephant in the room and that is the democratisation of filmmaking and making it an accessible art again.
DSLRs are responsible for carrying that forward, and they will do…
PS
AF100 contains the inner hardware of the GH1, a consumer camera. Consumer hardware trickling UP to the professional model. The margins between consumer and professional hardware is diminishing… law of diminishing returns as the hardware becomes more optimised, more advanced and more of a commodity. If you look at the difference between something like the Sony F3 and a Panasonic GH2, the main difference is price and featureset not image quality – the end result in terms of image is actually quite close especially if you are viewing it on an LCD panel, or via Vimeo.
Maybe a minor point but the FS100 also has rolling shutter.
Source: http://philipbloom.net/2011/04/24/upcoming-f3-vs-fs100-vs-af100/
(Philip Bloom in one of the commentaries at the bottom)
Not much news here. Panasonic and Sony have already launched big-sensor camcorders and off course there will be more and more professional oriented big sensor products which will be adopted by most professionals instead of the VDSLRs. But consumers are a different story and will keep using VDSLRs. And also there will always be some ‘on-the-budget’ move makers who will exploit the high-quality and lower cost of VDSLRs.
So not much foresight is required to make such “predictions”. What is more difficult to foresee is whether we’ll soon see a convergence of camcorder and DSLRs in the lower end (specially in terms of form factor and interface).
It is always like that. If you want to mainly shoot video then buy camcorder, if you mainly shoot still then buy dSL(R/T). The company will not sell all-you-can-do gadget that shoot best video/still even if they can because they will get less money when they do that.
In a few years (2 or 3), we’ll see 1080p video on ALMOST ALL cell phones.
DSLR with video has no future because, DSLR has no future.
No mirror, no shutter. That’s the future of (not only) still cameras.
Still cameras will not be still anymore 🙂
720p is good enough. Most video is shown on the ‘net, i.e. Vimeo, Youtube, Facebook, etc. Video shot on Flip, Sony Bloggie, camera phones, iPodTouch4 look fine on a 27″ monitor or a portable device – iPhone, Android phone or iPod/iPad.
Have you seen Goldilocks – shot with an iPhone4 ? http://vimeo.com/15608787 Story is more important than gear.
I’ve been involved with motion picture/video production since the 1970s. I seldom have a SLR with me, and if I do it is a film SLR. But I always have a Flip v2, Zoom Q3HD or an iPod Touch 4 in my pocket. For the web all three produce good video.
It seems a bit premature to mark the death of video DSLRs without seeing the next generation.
The GH2 is somewhere in the middle of 1st and 2nd generation and those who have used it will know that it has addressed a good deal of issues. Digic V will surely complete the cycle.
I can imagine an expensive video add on will be order of the day for the next gen DSLRs…. then it is up to the end user to make the choice. Having used many cameras, HVX, EX1, 5DII, Red One etc…. I can happily say I now prefer the small format.
I guess the Sony NX100 is trying to breach a middle ground, but without 24p/25p swappable models… it is dead in the water to me.
Thom Hogan isn’t exactly known for getting predictions right.
Sadly, Thom has completely lost it in recent years. There is absolutely nothing stopping “video camera sensor technology” to be applied to “DSLR sensor technology” (of course, there is not even a distinction, just a matter of design priorities). You don’t have to believe me, just think for yourselves; are Nikon and Olympus (and Pentax) going to roll over and die (well Pentax actually might)? They have no video products and will continue to leverage their expertise in consumer-friendly camera packages. This will force the others to do the same. Sure there will be differences in the pro video segment, but that NEX is six thousand dollars without a lens! I have no idea who he is talking to.