Video equal, will you sell your 7D for a T2i?

[shaocaholica]

Do you go with a still camera, and get all the aliasing/moire/lousy form factor that goes along with it, compromising your video, but getting a cheap price?

Form factor, yes. However, I wouldn't say that moire and aliasing are fundamentally attached to still cameras. The issue of moire and aliasing can be fixed, in theory, without compromising stills by over sampling. My prediction is that in the near future, say 3-5 years, we'll have DSLRs without any major visual artifacts in video. Maybe even RAW video. The only major gripes will be audio and form factor.

DSLRs have lots of room to improve, RED does not. RED is already pretty maxed out on quality and functionality. However, as DSLRs improve, their costs will say in place. The performance gap is closing but the price gap will remain.

 

[Michael Olsen]

It's interesting that the concept of modularity has been brought up here. I recently thought about that in regard to the DSLRs we are using right now. In many ways, they are nearly as modular as the RED will be. Of course, this doesn't stem from design, but from adaptation to cinema production.

Lenses are separate.
Lens adapters are modular.
Rails and baseplate mounts are separate.
Additional handles are separate.
External monitors are separate.
The EVF equivalents (Z-finders, LCDVFs) are separate.
Additional battery capacity? Separate.
External recording? Separate.
Audio...separate!

So if I got rid of my 7D and bought a t2i (or an "8D" - which looks like some goofy emoticon, btw) I would be able to keep virtually all of my accessories and use them there.

In many ways, as it is right now, Canon is selling us Brains with an attached CF card reader.

Similar in use, very different in purpose, design, and execution.

 

[Richard J. Johnson]

^interesting take. That thing on the left kind of looks like a unicorn.

 

[Michael Olsen]

However, I wouldn't say that moire and aliasing are fundamentally attached to still cameras.

They are fundamentally attached to cameras that emphasize stills. So until a majority of photographers also become videographers/cinematographers/filmmakers and demand good video from their HDSLR, we won't get it. To do otherwise would be a poor business decision!

Why spend a bunch of money making Product A have Feature B if only 5% of Consumer A uses it? Especially when you have Product B that already does Feature B? And both are selling well?

The issue of moire and aliasing can be fixed, in theory, without compromising stills by over sampling.

Indeed they can! And it can come cheap - $5000! But for Full Frame its going to cost you.

To do this well on a Full Frame sensor requires a LOT of processing power. And, even using a great code like REDCODE, a LOT of storage space.

My prediction is that in the near future, say 3-5 years, we'll have DSLRs without any major visual artifacts in video. Maybe even RAW video. The only major gripes will be audio and form factor

Sooner than that. RED will be shipping DSMC (Digital Still and Motion Camera) within the year. And add one handle essentially transforms the Scarlet or Epic body into something close to DSLR format. In fact, their prototype "Tatoo" cameras are heading out to beta testers soon. If other camera manufacturers don't follow suit with something similar they'd have to be crazy, in my opinion, or really counting on the "current state of the industry" to hold them up.

DSLRs have lots of room to improve, RED does not. RED is already pretty maxed out on quality and functionality. However, as DSLRs improve, their costs will say in place. The performance gap is closing but the price gap will remain.

I beg to differ and RED begs to differ. They are undoubtedly pushing a new direction (and so are pretty much on the edge of quality and functionality [imho]), but they have always said that there will be much more for them to do. Once 2/3" and S35 come out, there is FF35. Then 617. Then 645. And all along there will undoubtedly be progress made to sensor function (Mysterium -> Mysterium-X -> Monstro -> ??), hardware design (processors, ASICS, etc.), and software (REDCINE-X, RED4K).

 

[Zim]

I'm just glad there are good choices for video. Not everyone can afford a $5,000 or $20,000 camera.
I think the HD-DSLRs have a market. A big market too. I would like something like the camera's listed below but not for those prices. I can see if you had "real money" to make a film, it would be better to go rent a Red One or something.

Panny HPX170-$4300
Panny HMC 150-$3500
Canon XL1s -$9,000
Sony's HVRV1-$4200
Sony's HVRZ1U-$6,000
Red One (Body Only) $17,500

 

[hepabst]

In my last post I mentioned those 2 "Pros" using "Canon" not Nikon, Panny, or any of the others.

(IMO) VDSLR's will soon deliver a high quality moving "still" image that will be widely used by pros all the way down to a weekend hobby filmmaker. It looks like Canon could be the first in line to do that unless the other manufactures beat them to it. Red is already there.

Can't wait til the moving image is as good as the still. <-- That's what we're all waiting for, right?

 

[Barry_Green]

Form factor, yes. However, I wouldn't say that moire and aliasing are fundamentally attached to still cameras.

No, but those issues *are* fundamentally attached to the way a still camera has to read its chip out to deliver video.

The issue of moire and aliasing can be fixed, in theory, without compromising stills by over sampling.

Yes, but it doesn't happen at video frame rates, not at this price point. The whole reason we have this awful aliasing/moire problem is because the only way they could read out the chip at video frame rates is to skip 90% of the chip, reading out only 10%.

My prediction is that in the near future, say 3-5 years, we'll have DSLRs without any major visual artifacts in video. Maybe even RAW video. The only major gripes will be audio and form factor.

Five years? I'd buy that it'd be possible they could have this all sussed out properly within five years.

 

[Barry_Green]

Can't wait til the moving image is as good as the still. <-- That's what we're all waiting for, right?

Yep. If the SLR could deliver in motion, an equivalent shot to what it can do in stills (even a still downconverted to 1080p) it'd be phenomenal.

Of course, waiting is silly anyway. Jack and I have a film premiering at South By Southwest (SXSW), one of the top-10 film festivals on the globe, and it was shot on a GH1.

So ... what again are people waiting for?

 

[Michael Olsen]

Jack and I have a film premiering at South By Southwest (SXSW), one of the top-10 film festivals on the globe, and it was shot on a GH1.

Didn't know that! Congratulations, Barry...I wish I could be there.

 

[Richard J. Johnson]

So ... what again are people waiting for?

...For quality movies to come out and move them. Damn the camera it was shot on.:laugh:

 

[Zim]

Sounds good Barry. What is the name of the film?

Yep. If the SLR could deliver in motion, an equivalent shot to what it can do in stills (even a still downconverted to 1080p) it'd be phenomenal.

Of course, waiting is silly anyway. Jack and I have a film premiering at South By Southwest (SXSW), one of the top-10 film festivals on the globe, and it was shot on a GH1.

So ... what again are people waiting for?

 

[sidjastudios]

Jack and I have a film premiering at South By Southwest (SXSW), one of the top-10 film festivals on the globe, and it was shot on a GH1.

WOW Congrats! SXSW is HUGE! Here's hoping you get picked up!

 

[Kholi]

Form factor, yes. However, I wouldn't say that moire and aliasing are fundamentally attached to still cameras. The issue of moire and aliasing can be fixed, in theory, without compromising stills by over sampling. My prediction is that in the near future, say 3-5 years, we'll have DSLRs without any major visual artifacts in video. Maybe even RAW video. The only major gripes will be audio and form factor.

DSLRs have lots of room to improve, RED does not. RED is already pretty maxed out on quality and functionality. However, as DSLRs improve, their costs will say in place. The performance gap is closing but the price gap will remain.

RED is already Maxed out? Do you keep up with RED progress, at all? Have you shot RED???

 

[NoxNoctus]

So ... what again are people waiting for?

Excuses. They're the greatest inhibitor of productivity.

My brother is getting into photography. When I got a D300, I gave him my D70. When I went with the 7D, he bought my D300. What's he done with it? not much. He's waiting until he can buy some ED glass instead of going out there and learning with the cheapos and paying his dues. I never had ED glass, never had a $5000 top of the line editing suite for photoshop, you do what you can with the tools you can afford. Creativity doesn't come from owning a $5000+ camera system.


anyway, Congradts on the SXSW entry Barry!

 

[USLatin]

Congrats on SXSW Barry!!!

I'd add one more thing to the motion being as good as stills, and that is dynamic range. We don't want just all digital stills have to offer but also all film has to offer. Cameras like the F35 and M-X gets us a bit closer to that but there is still a bit more to go.

Michael, you forget the audio I/O module, a bigger and better audio I/O module, wireless remote, wireless remote with follow focus, bigger screens, more screens, EVF, bigger media, better power solutions, handles, and don't forget the lens mounts (PL, Red, electronic Canon, electronic Nikon). You can start small and grow with your kit in many different directions. Most of those will remain usable and rentable through at least two brain/sensors, and several more than two.

It gets really hard to compare VDSLRs to anything other than the Scarlet non-fixed though, and money wise you want to compare it to Scarlet Fixed. So we are not going to get anywhere with this really. Bottom like VDSLR's rock for some stuff, and Red plays a whole different tune.

 

[J Davis]

If anything the vDSLR's are preparing us for using the DSMC form factor

wanna go handheld? get rails!

 

[shaocaholica]

They are fundamentally attached to cameras that emphasize stills. So until a majority of photographers also become videographers/cinematographers/filmmakers and demand good video from their HDSLR, we won't get it.

Status Quo is not the same as fundamental. Just because thats what we have at this instant does not mean its a fundamental. You yourself say its doable which contradicts classifying it as fundamental.

To do otherwise would be a poor business decision!

Some would say putting 1080p24/25/30 on a APS-C interchangeable lens camera for $800 is a poor business decision when that market could me milked for much more money.

Why spend a bunch of money making Product A have Feature B if only 5% of Consumer A uses it? Especially when you have Product B that already does Feature B? And both are selling well?

That doesn't work when Product A and Product B are made by different companies with different agendas and goals. Unless I misread that.

Indeed they can! And it can come cheap - $5000! But for Full Frame its going to cost you.

Lets go back 2 years and show ourselves footage from these currently flawed DSLRs and ask how much we would price them. I'm sure it would be a hell of a lot more than $800 or even $2500. So where are you getting this $5000 figure?

To do this well on a Full Frame sensor requires a LOT of processing power. And, even using a great code like REDCODE, a LOT of storage space.

If we limit DSLRs to 1080p max for a while, the processing power in these cameras will increase over time while the resolution will remain constant allowing more data to be read. Yeah, the sensor will need some updating but all things in small steps, but small steps happen pretty quickly in this industry. Storage is constantly getting denser and faster. Even RED is going to have to use off-the-shelf CF cards which will be the same cards DSLR owners will have access and compatibility with.

Sooner than that. RED will be shipping DSMC (Digital Still and Motion Camera) within the year. And add one handle essentially transforms the Scarlet or Epic body into something close to DSLR format. In fact, their prototype "Tatoo" cameras are heading out to beta testers soon. If other camera manufacturers don't follow suit with something similar they'd have to be crazy, in my opinion, or really counting on the "current state of the industry" to hold them up.

Yeah, I can't wait but the price difference is still vastly different between the DSMC and T2i.

I beg to differ and RED begs to differ. They are undoubtedly pushing a new direction (and so are pretty much on the edge of quality and functionality [imho]), but they have always said that there will be much more for them to do. Once 2/3" and S35 come out, there is FF35. Then 617. Then 645. And all along there will undoubtedly be progress made to sensor function (Mysterium -> Mysterium-X -> Monstro -> ??), hardware design (processors, ASICS, etc.), and software (REDCINE-X, RED4K).

I'm glad RED is putting these things on the table but common logic puts DSLRs way behind RED which also means they have more room to improve in comparison. Again, the performance gap will close because DSLRs have a lot of low hanging fruit left while the price between DSLRs and RED will remain quite large. Thats not to say RED is failing, just that a new market for cheap(er) vDSLRs is growing and many of its members may want RED but won't be able to afford it so all they have left to count on are the vDSLR makers.

 

[Barry_Green]

Sounds good Barry. What is the name of the film?

More info here: http://www.dvxuser.com/V6/showthread.php?t=199546

 

[shaocaholica]

RED is already Maxed out? Do you keep up with RED progress, at all? Have you shot RED???

Perhaps maxed out was a bad term to use. My point is that vDSLRs have more room to improve than RED because RED is already so far ahead.

 

[shaocaholica]

No, but those issues *are* fundamentally attached to the way a still camera has to read its chip out to deliver video.

I guess we have different definitions of "fundamental". When I say "fundamental", I mean unchangable or else it would totally change the object. At least within known science. If its something that can be addressed with the same technology that is currently used, then its not a fundamental limitation IMO.

Yes, but it doesn't happen at video frame rates, not at this price point. The whole reason we have this awful aliasing/moire problem is because the only way they could read out the chip at video frame rates is to skip 90% of the chip, reading out only 10%.

I never said in the present. Look around you at how much technology is progressing.

Five years? I'd buy that it'd be possible they could have this all sussed out properly within five years.

Thats my point. And in 5 years, there will still be people who can afford a T2i or 7D equivalent but not much more so they are hoping that things get better which they always will.