[Ratmagiclady]

Or disadvantage, depending on your point of view

I agree the fundamentals of composition are more or less the same for stills and video. But with video you have more options/factors to use/deal with because of the continuous nature of it. I haven't thought through it all the way, but it might be that video composition is a superset of still composition. That is, with video you have to deal with all of the factors of still composition PLUS additional factors due to motion, whether it be subject motion or camera motion. Thus I would think it would be much easier for a good videographer to do still photography than the other way around--the learning curve is much less from video to still. That's if we're talking about a real video artist and not just a video camera operator (basically a Director of Photography vs. a cameraman).

I generally agree. There's some overlap. ...and I'm sure video's a fun option to have, (though if you hand me a little video camera, I might make you seasick trying to shoot verticals.. )

[Yohan Pamudji]

I generally agree. There's some overlap. ...and I'm sure video's a fun option to have, (though if you hand me a little video camera, I might make you seasick trying to shoot verticals.. )

That would be me right there--flipping a video camera to shoot in vertical mode! I just know I'll do something goofy like that.

[falconeye]

don't you think the day will come when that section of sensor detecting the loss of photon hits could increase its dynamic range ?

There is no law of nature limiting the difference of photons counted by neighboring pixels. So, DR can be pushed to practically unlimited levels -- as long as enough light is captured, either by large enough sensor or by long enough exposure.
Personally, I am actually surprised that all consumer-grade sensors still have this problem of overflow. It would be so easy to detect, count and reset the overflows during exposure. Yes, it would make the sensor a processor. I.e., more expensive. But so what?

So yes, the DR will increase dramatically in the future. But only at very low ISO figures like ISO 10 and less. At normal ISO, you still hit the limit that you cannot count a photon which never hits the sensor.

Ctein over at TOP had a couple articles on the topic of future-tech for cameras, including radically different sensors.

I still have to see average-reader level articles on this topic which are worth a read. Most is reprinted hype originally targeted at venture capital investors. I can only say this: Don't believe it if you cannot verify it. That's a symptom left over from the dot.com bubble where venture capital investors believed in EVERY BS.

Besides this, I have no specific criticism with the two articles cited above. They seem to be rather correct. But don't promise a radical change of rules either. Unlike some other articles out there ...

[nanok]

electronic shooters: no, they don't (for dslrs, except some cheaper nikon models, in the case of fast shutter speeds), and yes they can (have almost arbitrarly high sync shutter speeds). i cannot explain right now why they don't, just that, as far as i know, there is a tradeoff which seems to be unacceptable for higher quality dslrs (take it with a grain of salt, i might be wrong, i have to read up on that). in any case, it would be interesting to see a camera which at least allows using electronic shooter as an option, for specialized flash work.

this is very arguable, but i disagree that video shooting is a superset of still shooting, they are simply different. i am not sure which is harder: to go from still to video or the other way around, but that proves nothing either way. in short, saying that is much like saying that it's easier to do offroad driving with a defender than to drive a rally car, because with the rally car you also have the speed to worry about: it's plain wrong, they are just two different things which happen to have a set of wheels in common. the fact that there's an overlap is a different story.

[monochrome]

It might turn out that the only way to have an exclusively still camera at prosumer prices will be film . . . .

And film will be reborn . . . .

And all those full frame lenses . . . .




Well, I can dream, can't I?

[nanok]

There is no law of nature limiting the difference of photons counted by neighboring pixels. So, DR can be pushed to practically unlimited levels -- as long as enough light is captured, either by large enough sensor or by long enough exposure.
Personally, I am actually surprised that all consumer-grade sensors still have this problem of overflow. It would be so easy to detect, count and reset the overflows during exposure.

hmmm. perhaps the problem is that, with the current implementations at least, you would need to read out that photosite, and store the reading in a buffer, during exposure. fore typical exposure times, perhaps the electronics are not yet fast enough to do that (and do it repeatedly), for long exposures, i honestly cannot see why not (but i reckon this would be astrophotographer land mostly, and that's a niche no dslr manufacturer seems to care about anymore ). astrophotographers do something similar: to get higher iso without the sideeffects, they take several exposures and simply add them (in the most literal sense you could imagine , as in, 1+1+0+..)

So yes, the DR will increase dramatically in the future. But only at very low ISO figures like ISO 10 and less. At normal ISO, you still hit the limit that you cannot count a photon which never hits the sensor.

that's the real world limit which physics imposes on us, obviously, but i doubt we are at that level (or even near it), maybe it would be more accurate to state that you hit the limit where you cannot read usefull signal from the sensor anymore (the sensitivity limit of the sensor, and i am not talking iso). there were some announcements about some breakthrough in that area (something about black silicon iirc?), while that might be bullshit, and/or take years to get a usable implementation, there is always a possibility something better (in sensitivity) than what we have now might develop.

on the other hand, there might be a chance to develop less "intrussive" ways of reading the sensors, which would effectively mean higher sensitivity for what we have now (less noise, thus better snr)


there is still no point, i think, in discussing if size matters or not. it does, because it's physics, for one any improvement in sensor technology will widen the gap rather than tighten it, and also, some things cannot be out-designed (for instance effective depth of field), at least not for now. for most uses, for instance, digital compacts, especially some of them, do an excellent job already, but i could never quite get the hang of them because of the huge dof, they are simply a different medium, so different for me that i have trouble using them. aps-c delivers a good compromise for me (mainly because i ahve a tendency to shoot longer lenses, and aps-c helps with that, while having still plenty shallow dof at small fov's), but that's just my personal choice, an i understand people wanting to have "full frame" (i myself miss medium format for some things i shoot, and i would probably have digital medium format if i could afford it). so, at the very least, unless "technology" can deliver arbitrary wide aperture lenses (or near it), or replace what we think of as lenses right now with something to that effect, dof will still haunt us, no matter how amazing the sensors will get, a different size of sensor will still mean a different medium of expression, and bigger will still mean better in some ways (it can be argued that at some point it will be good enough either way, but looking around at what's on the market, it seems we never have enough, and while we blame the megapixel wars on manufacturers, we do like more of them, as long as we don't trade too much for having them. i have yet to see a megapixel/dinamic range/sensitivity limit where photographers will say "okay, that is ENOUGH", take the phase one 60MP back for example -- and the ones before it; i think, as long as computing power and storage possibilities can keep up, we will never stop wanting more of _everything_, and for cheaper too)

[falconeye]

hmmm. perhaps the problem is that, with the current implementations at least, you would need to read out that photosite, and store the reading in a buffer, during exposure.
[...]
i doubt we are at that level (or even near it), maybe it would be more accurate to state that you hit the limit where you cannot read usefull signal from the sensor anymore

I agree to what you wrote.

In the first point, You're right, you need a counter and read-out electronics at each photosite, say 100 transistors, or 1bn transistors altogether. Not cheap but feasible. The best is that this way, you can start detecting single photon events ... Maybe even their energy, aka color (now, we're in future land)

In the second point, the important word is "quantum efficiency" (QE). You're right, there is room for improvement. But QE isn't that bad now. Even approaching 100% in dedicated B&W cooled astro equipment...

And again, we left the topic of this thread and I propose to continue this discussion some day in a more appropriate thread

[Slugger]

I honestly can't believe we need to have this debate about video functionality on a dSLR. The quality is just incredible on the 5D for a very affordable cost. If Pentax's video comes close to the 5D's quality i'll be laughing all the way to the camera store.

[Duplo]

I think people are forgetting about some of the downsides compared to dedicated video cams. It really is NOT a direct replacement for a dedicated video camera. Its something new.

Yes you can use fast lenses for shallow DOF, but you would also need huge telephoto lenses to cover the range of a normal video camera, and at long focal lengths shallow DOF is a real pain. Try keeping up with a racing car or BIF with a 600mm lens - an SLR's normal PD optical system can just about manage (though its not smooth) but a contrast detect system? Dream on.

SLR autofocus is nothing like as smooth and quite as video AF. It doesnt need to be - it needs to be fast and accurate. Perhaps the contrast AF system is being developed just for this, but its not "free". Do you want stills lenses to have smoother and quieter AF at the cost of speed? . Do you want to have buy new lenses just for video, especially as most of the old screw drive lenses will drown out any sound? Pretty soon it gets expensive.

You have to manually zoom, which means you really require a tripod much of the time.

You cant use the VF when shooting which means on really bright days you cant accurately verify exposure and focus. At least with micro 4/3 you dont have this problem. In fact without the mirror assembly, I see video as a far more natural extension of the micro 4/3 and micro APS concept. It just makes more sense to me.

Plus your still camera is carrying the additional cost of lots of technology licensing on the video side. Without this, the K7 would likely be a couple of hundred dollars less.

Now dont get me wrong, if I was a photojournalist, wedding or event photographer I would pretty soon need one to compete with all the other guys offering video. I could even see the point in it. But I just dont think most people other than indie movie makers are going to be able to exploit it that well. Its quite technically challenging to shoot movies on larger formats.

Personally I would like to have the OPTION of a camera without all this for a couple of hundred dollars less. Perhaps the K7 limited?

Very well said steve and I agree completely.

[creampuff]

Too many Pentax users are living in the past. Time to wake up to the current realities of technology and wean away from those memories when film was the be all and end all.

[OregonJim]

Too many Pentax users are living in the past. Time to wake up to the current realities of technology and wean away from those memories when film was the be all and end all.

Hogwash.

Was there any point buried in that statement, other than to provoke an argument?

[Jaboney]

I honestly can't believe we need to have this debate about video functionality on a dSLR. The quality is just incredible on the 5D for a very affordable cost. If Pentax's video comes close to the 5D's quality i'll be laughing all the way to the camera store.

Ditto. It's a fantastic addition.

[RawheaD]

I laugh at the "purists". You're the type that resisted computers (I'm doing fine on my IBM, thank you!), resisted Photoshop, resisted digital cameras, and in the end, you're shooting with a K20D, tweaking your RAWs with Lightroom/Aperture on your Mac, and, yeah, why not, use Photoshop now and again.

It will be the same for this "video thing", and 10 years from now, you will all be shooting a cam that does video, and importantly, you'll be loving it. You can quote me on that. 10 years from now. You really can, because I've been RawheaD on the 'net since 1995, so I'll be RawheaD on the 'net in 2020 :-)


The debate about whether or not video on DSLR will replace dedicated video cams (and/or vice versa) is besides the point. The point is, if you're not bound by technological limits, the possibility is there so that you can, e.g., depress the shutter, and it will record 4K HD video for the duration @ 120fps, and if you so choose, can select the best frame as your shot.


Some pompous commenter up there somewhere suggested this mode of photography as being blasphemous. I laugh. You've never used a motor drive with your film cams back in the day? You've never shot "continuous" or even "burst" mode on your Pentax? You can ridicule this idea if you answered no to all those questions, and will continue not to. If you have, then WTF? It's the same principle.


I'm not suggesting *all* shots should be taken like that, and I don't need lectures on the art of "capturing the moment." I also shoot 120 film, thank you very much. The question I posed was "Who wouldn't want such a capability in your cam". Now, if you only do landscape photography or highly controlled studio photography, maybe your answer is "Not I". But, if you're honest, and you do sports photography, bird or any kind of animal photography, anything with movement, really, then you are going to want it, and if the capability is there, you are going to use it. Hell, even in the studio, fashion photographers would definitely use it, like that Vogue cover (or whatever it was) that was shot with a RED cam.

Similarly, on the video camera side (RED is a great example), there have always been lame "still" shot capabilities, lame because the sensor size was always lame. But, when you get high quality video, like RED, you suddenly have the capability to extract a high quality still from video, and you can bet you'll see more and more such stills. E.g., movie posters where the still image is a direct extract from the motion picture that was shot using 4K HD. And a lot of those shots are gonna be better looking "stills" than what many of you could ever shoot (yeah, that includes me).


So, go ahead, be that last guy that switches over, like those last of the breed film photographers that are now, finally, moving over to digital (btw, I have a lot to thank those guys; I recently made several sweet, sweet MF lens acquisition from such folk). In the mean time, the world will move on, and unless you are among the cream of the crop photographers, it will have left you way, way behind.

[ajuett]

Too many Pentax users are living in the past. Time to wake up to the current realities of technology and wean away from those memories when film was the be all and end all.

I live in the present and have used present digital cameras in the past. Presently I use past Pentax in the present. The future might be different from the present, past history has presently shown that.

Now that we have that cleared up, I'm going to shoot Delta 100 in the future (tomorrow) with my camera from the past. I'll present the shots in the future.

Hogwash.
Was there any point buried in that statement, other than to provoke an argument?

I agree Jim, that was a pointless statement, a rare occurrence here. Maybe he/she was having a bad day.

[Venturi]

I live in the present and have used present digital cameras in the past. Presently I use past Pentax in the present. The future might be different from the present, past history has presently shown that.

Now that we have that cleared up, I'm going to shoot Delta 100 in the future (tomorrow) with my camera from the past. I'll present the shots in the future.

I'd just like to interject that there is a high degree of uncertainty that you just opened a rift in space-time with that post.