So... there is a 7D now...

[Saul Rodgar]

These DSLRs that shoot video are interesting harbingers of what you get with a larger sensor -even though they are only reading every third line - still, you get a nice smooth HD image and great low light performance. The canon 5d mk2 has a still camera's "full frame" sensor that lets a lot of light in.

 

I think the problem here is the notion that a '"full frame" sensor . . . lets a lot of light in,' to use Jean's own words. This is completely untrue. If that were the case every light meter out there will be format dependent. So the exposure for 8mm would be "x," the exposure for 16mm would be "y" and for 35mm "z" (and so on) under the exact same light conditions. The only thing that lets "light in" to a camera is a lens, but exposure has nothing to do with the size of the gate or sensor used.

 

http://en.wikipedia.org/wiki/Exposure_(photography)

 

Generally speaking, bigger sensors have more resolution, true. As for DOF, a 50mm full frame 35mm format lens will have the same DOF whether it is used on S35 or S16. The FOV will definitely change (the dreaded and misunderstood "crop factor" everyone here so much hates).

 

http://en.wikipedia.org/wiki/Depth_of_field

 

I just completely disagree with the notion that bigger than S35 mm aperture (24 x 18 mm) sensors are necessarily better for every project out there, and that every camera released should have full-format 35mm sensors. DOF issues related to larger lenses used in 35mm would make it a pain to use for every project.

 

I have a feeling that once most people who are jumping to the "full frame" bandwagon realize the amount of work it takes to consistently get the image in focus for large scale projection, they will realize smaller formats aren't so bad after all, especially for low budget gigs. I am perfectly happy using 16mm and 2/3 inch HD formats for most of non-theatrical release work, but they also aren't right for every situation to be encountered in the real world.

 

What is wrong with having lots of DOF in one's pictures anyway? I realize it is largely personal taste, but some hacks like Ansel Adams seem to have embraced, perhaps to mediocre results ? . . :P

 

 

 

Thanks to John for pointing out my error in the S35 dimensions. Can't seem to get them straight yet.

Edited September 3, 2009 by Saul Rodgar

[Jean Dodge]

Without getting into the technical details, it's clear that the canon d5 Mk2 has an inherent strength when it comes to shooting in low light. Right out of the box, it also produces a grain free, gain free nice image at high ISO ratings that outperforms other DSLRs with smaller sensors. Yes, we can go around and around about the finer points of industry nomenclature, but since this is the first thread to discuss a new pro-sumer DSLR that shoots HD video, I figured I would keep the discussion rather general. Please excuse my generalizations and misuse of proper terminology.

 

As someone wise once said, there's not much point in arguing if the glass is half full or half empty, the real question is, what do you paln to do with the the water you have?

 

The 7D is part of a trend we are seeing where digital still cameras are starting to emulate certain aspects of motion picture cameras. I find it fascinating that the two worlds are combining. Both crafts have a lot to learn from one another. Aspects of presentation, distribution and audience are arising that affect our professions. People are staying home and watching blu-rays and canceling newspaper subscriptions, and where we once might have learned about a new fashion trend or political movement in the pages of LIFE MAGAZINE and from Walter Cronkite and at the local cinema, now we watch clips of Jon Stewart on our laptops and wait for Netflix envelopes to arrive with independently produced docs like TROUBLE THE WATER, or perhaps we check out still/motion picture essays like the UK Guardian's Dan Chung is producing instead.

 

But purely as a cinematic tool to be exploited for whatever advantage it might offer, it seems to me that the 7D is perhaps at best another choice for the budget-strapped indie or experimental film maker to use - and one that offers some interesting possibilities since the sensor is close to the size commonly used with cine lenses. A 7D with the proper lens mount, follow focus and good glass might enable a focus puller to do a decent job of handling a narrative project that could be made for modest means.

Edited September 3, 2009 by Jean Dodge

[Ram Shani]

and lets not forget all this at 1800$ include lense

 

is there ANY camera how can give you this?????

[Elliot Rudmann]

If one is going to compare the sensor in the XTi and the 5D on a size only basis, without taking into account other camera variables like sensor resolution, picture processing algorithms, encoding, actual sensor manufacturing technology, cost of the actual cameras being compared, etc then it is going to be like comparing apples and oranges.

 

At least one has to compare two cameras that are in the same ballpark.

 

You asked what's wrong with smaller sensors. While I did gloss over some marginal technical variables, I gave the most realistic, concise example I could think of to illustrate one of the key differences when it comes to image quality. If Canon ever made two cameras with the *exact* same specs with the only variable being chip size, I would have used that as an example. Unfortunately it's not that simple.

[Saul Rodgar]

You asked what's wrong with smaller sensors. While I did gloss over some marginal technical variables, I gave the most realistic, concise example I could think of to illustrate one of the key differences when it comes to image quality. If Canon ever made two cameras with the *exact* same specs with the only variable being chip size, I would have used that as an example. Unfortunately it's not that simple.

 

I'm sorry, I have no idea how an entry level DSLR and a higher tier DSLR offering even begin to compare.

 

The idea that a smaller sensor is mostly what keeps the XTi from competing with the much more advanced in overall design and production 5D is, well, pretty tenuous.

[Saul Rodgar]

Without getting into the technical details, it's clear that the canon d5 Mk2 has an inherent strength when it comes to shooting in low light.

 

Yes, but that has nothing to do with the size of the chip, as you claimed. HD (and raw) video cameras, regardless of their chip size are generally better at low light handling than SD video cameras and film. Just like film is generally better at handling highlights than video.

 

Look, I can understand why everybody is going gaga about "bigger" sensors. You claimed, however, that bigger chip cameras were allegedly the way to accomplish a "step forward" in camera design and production, and that they "let more light in," both statements which I completely disagree with --despite your efforts to come up with a reason as to why said bigger sensors are the "way forward" in camera technology, and why a bigger sensor camera is better at low light handling "without getting into the technical details."

 

Until anyone can provide me with a perfectly detailed and reasonable explanation why bigger than S35mm sensors are absolutely essential to take the proverbial "step forward" in camera technology, I will remain a non-believer.

[Satsuki Murashige]

Until anyone can provide me with a perfectly detailed and reasonable explanation why bigger than S35mm sensors are absolutely essential to take the proverbial "step forward" in camera technology, I will remain a non-believer.

I think what Jean is getting at is that the full frame sensor of the 5D is more sensitive than the smaller APS-C sized sensors of it's competitors (provided that they both have a similar number of pixels) because the 5D's pixels are physically larger, thus allowing more photons to hit each pixel.

 

If the smaller sensors were designed with pixels of the same size as the 5D then there would be little to no difference in low-light performance (aside from the quality of the signal processing and noise reduction built into the cameras). But that would mean that the smaller sensors would necessarily have a lower pixel density (say, 10MP rather than 18MP) to make up for the space on the sensor taken up by the larger pixels. So, if you want a camera system that has greater low-light sensitivity than the current offerings while at the same time providing the same amount of resolution, eventually you just have to make the sensor bigger (or find something other than silicon to make the sensor out of).

 

I'm not saying this is an "essential next step in camera design" or anything, but you have to admit some of these low-light images taken with the 5D lately could not have been done with the current filmstocks or with any of the current HD cameras. For that reason alone, the 5D will continue to get used until something similar by one of the major camera manufacturers comes out.

 

BTW, I spoke to a Fujifilm representative at Cinegear this year about them possibly making a super high speed film stock (like around 2000 ASA) to compete with the Genesis for available light night exteriors, but he was not receptive to the idea at all. So I don't think we'll ever be using film for this type of work... :(

[Saul Rodgar]

I'm not saying this is an "essential next step in camera design" or anything, but you have to admit some of these low-light images taken with the 5D lately could not have been done with the current filmstocks or with any of the current HD cameras. For that reason alone, the 5D will continue to get used until something similar by one of the major camera manufacturers comes out.

 

I could agree with the pixel count and size make it arguably better at low light image acquisition, there is got to be something in the internal processing of the 5D that makes that possible though, not just the sensor.

 

Thus, I think refinements in internal camera processing and particularly in encoding could make a huge difference with the current crop of sensors, without having to go to much bigger sensors.

 

I remain completely against the notion that by making larger sensors alone a camera will be better at low light or that going to medium format (6x4.5 and up) sensors will magically achieve the "next step of digital cinematography."

Edited September 5, 2009 by Saul Rodgar

[Satsuki Murashige]

Thus, I think refinements in internal camera processing and particularly in encoding could make a huge difference with the current crop of sensors, without having to go to much bigger sensors.

Well, it's all about trading off the given design parameters to achieve a certain goal, isn't it? If the goal is to make a more light sensitive camera, then you'd be playing with the type of sensor and materials used, the size of the sensor, the pixel density of the sensor, the signal processing, and probably a bunch of other things that I'm not aware of. When all the other factors besides size have been improved to the point of diminishing returns, then there's nothing else to do but to make the sensor bigger.

 

I think though that it's probably easier for camera manufacturers like Canon to just add a larger sensor than it would be for them to develop a new black silicon sensor or something. They already have full frame sensors after all.

[Saul Rodgar]

Well, it's all about trading off the given design parameters to achieve a certain goal, isn't it? If the goal is to make a more light sensitive camera, then you'd be playing with the type of sensor and materials used, the size of the sensor, the pixel density of the sensor, the signal processing, and probably a bunch of other things that I'm not aware of. When all the other factors besides size have been improved to the point of diminishing returns, then there's nothing else to do but to make the sensor bigger.

 

When companies like RED are creating 2/3 inch cameras (Scarlet), test footage from which that I have seen looks stunning in terms of resolution and noise levels in somewhat lower light situations, I hardly think that "there's nothing else to do but to make the sensor bigger" to make improvements to camera technology.

 

Again, this is a situation where most people would think the next logical step is to go "next size up" when the answer may lie elsewhere.

 

Not because I may not find a solution to a given problem, does it mean that the solution does not exist, sometimes well within striking distance. ;)

Edited September 5, 2009 by Saul Rodgar

[Satsuki Murashige]

When companies like RED are creating 2/3 inch cameras (Scarlet), test footage from which that I have seen looks stunning in terms of resolution and noise levels in somewhat lower light situations, I hardly think that "there's nothing else to do but to make the sensor bigger" to make improvements to camera technology.

Are you referring to the iguana shot? That's the only Scarlet test footage I've seen so far. It does look very clean but then we have no way of knowing how much light was used. That's my one reservation so far about the 2/3" Scarlet - how fast will the camera actually be with 3K stuffed onto that tiny sensor? I guess we'll find out!

[Saul Rodgar]

Are you referring to the iguana shot? That's the only Scarlet test footage I've seen so far. It does look very clean but then we have no way of knowing how much light was used. That's my one reservation so far about the 2/3" Scarlet - how fast will the camera actually be with 3K stuffed onto that tiny sensor? I guess we'll find out!

 

Yeah, the iguana shot is one. I remember seeing more, but I can't remember clearly now. Perhaps I am thinking some of the Lumix GH1 low light tests (also a tiny sensor, comparatively to the 5D) which looked very, very promising.

 

My understanding of how the extended sensitivity of a sensor works is not so much the amount of light it can be applied to it (it eventually reaches the limitations of the lens' light transmission capabilities [f stop]) but how much electricity it can be applied to it to boost its native sensitivity. In the past, this has come at a high cost, noise. Apparently, newer sensor technology and processing minimizes the apparent grain in the images at higher light sensitivity levels, regardless of sensor size.

 

Also, as Jean had pointed out, if the 5D with its huge 21 megapixel sensor is having to shed huge chunks of resolution to conform with 2 megapixel video (1080), then what is the point of having such a large sensor anyway? That is a 10:1 loss of resolution ratio!

 

Sure, the shallow DOF look of the lenses needed to cover the full frame 35mm image area is great. But, there is absolutely no need for that huge a sensor when so much information from it will be ultimately lost --even if one wanted to keep the same DOF look, as S35 could use the same lenses (same DOF) than full frame still photo 35mm.

 

Even for 5 megapixel motion picture images (which RED claims in its EPIC line), there is still arguably no need for such a large sensor (roughly 4:1 loss of resolution ratio with a 21 mp sensor).

 

My educated guess is that once the engineers at Canon, Panasonic, RED, etc realize how to truly maximize the potential of the newer crop of sensors, then much better cameras (in terms of resolution, compression and low light handling) will be produced with somewhat inexpensive technology, as they currently are -- without having to go to much bigger sensors at all. We shall see.

Edited September 6, 2009 by Saul Rodgar

[Saul Rodgar]

Sure, the shallow DOF look of the lenses needed to cover the full frame 35mm image area is great. But, there is absolutely no need for that huge a sensor when so much information from it will be ultimately lost --even if one wanted to keep the same DOF look, as S35 could use the same lenses (same DOF) than full frame still photo 35mm.

 

Should read:

 

Sure, the shallow DOF look of the lenses needed to cover the full frame 35mm image area is great. But, there is absolutely no need for that huge a sensor when so much information from it will be ultimately lost --even if one wanted to keep the same DOF look, as S35 lenses theoretically have the same DOF than full frame still photo 35mm lenses.

[Peter J DeCrescenzo]

Just an FYI if you haven't already seen it:

 

Yesterday, Phil Bloom posted some "late prototype" 7D footage of Dublin night life:

http://vimeo.com/6475938

[Saul Rodgar]

Just an FYI if you haven't already seen it:

 

Yesterday, Phil Bloom posted some "late prototype" 7D footage of Dublin night life:

http://vimeo.com/6475938

 

Sweet, thanks for posting. No, I hadn't seen it. Those Dubliners sure know how to party.

 

At the risk of being unfair because of the quality of the clip, I would say that low light capabilities of this camera seem to equal the 5D's from where I am . . .

[Satsuki Murashige]

I would say that low light capabilities of this camera seem to equal the 5D's from where I am . . .

Yeah, it does look pretty darn good, huh? Now I'm wondering if someone like Illya Friedman will make a PV mount for the camera so I can put some PV adapted Lomo square front lenses on it.

 

Do it Illya, I'll buy one! :)

[Saul Rodgar]

Yeah, it does look pretty darn good, huh? Now I'm wondering if someone like Illya Friedman will make a PV mount for the camera so I can put some PV adapted Lomo square front lenses on it.

 

Do it Illya, I'll buy one! :)

 

Wondering who will make a PL mount for these as well (if they don't already, though I can't seem to find any info on that).

[Satsuki Murashige]

Wondering who will make a PL mount for these as well (if they don't already, though I can't seem to find any info on that).

PL won't work because the flange depth is too short. PV is slightly longer and will clear the mirror.

[Saul Rodgar]

PL won't work because the flange depth is too short. PV is slightly longer and will clear the mirror.

 

Interesting. I speculatively raised this point in the past in a Canon DSLR PL mount thread to no avail. I feel a little vindicated. :P

 

Perhaps manually raising and locking the shutter up before the lens is put in would work? I know the D90 lets one lock up the mirror before changing lenses, dunno about theses Canon cams. Something less crude and not as permanent than gluing it up there?

 

This certainly would be a poor man's fix, but until Canon gets on the ball, that may be the only solution to get PL mount glass on a Canon DSLR, temporary and flimsy as it may be . . . ;)

 

Not that PV isn't good enough, but PL mount glass can be more readily accessible.

[Oliver Christoph Kochs]

The Canons let you lock up the mirror for sensor cleaning.

[Saul Rodgar]

The Canons let you lock up the mirror for sensor cleaning.

 

Right, I would favor something like a real lock though (maybe even operated from the inside), because if the mirror shutter was accidentally released while a lens that does not clear it is in and the mirror shatters on impact, the sensor would be showered with shards of glass big and small. Even if it the sensor were protected, it could still potentially spell disaster for the electronics.

 

That is one benefit of the Lumix GHD1 (no mirror shutter), though not having an optical viewfinder is rather unappealing in its own right.

 

Obviously, the best solution is if Canon would design the camera so that PL mount lenses were easily adaptable. But by that point it would be better to just get the camera right for film production and start from scratch. Sort of like a XL1 with large HD sensors. They would have to be smarter about it, as I think the XL series were never quite what they could have been from a camera operator's point of view, ergonomics are really bad.

Edited September 13, 2009 by Saul Rodgar

[Oliver Christoph Kochs]

I'd also love to put my PL mount lenses in front of a 5d2/7d but that's not gonna happen. Canon makes lots of money selling their lenses. They are also experienced in manufacturing cine-lenses so why should they switch to PL instead of delivering their own.

 

Speaking of DSLR cine cameras: Who wants the mirror anyway? Get rid of it! Give us uncompressed HDSDI instead and we are happy! For proper monitoring and to mount a Kipro. Thanks Canon.

[Peter J DeCrescenzo]

... Give us uncompressed HDSDI instead and we are happy! For proper monitoring and to mount a Kipro. Thanks Canon.

 

Even full-time, live 4:2:0 1080p HD via HDMI would be extremely useful, and we're so close with the 7D, but unfortunately, not quite there yet:

http://philipbloom.co.uk/2009/09/15/using-...ith-the-5dmkii/

http://www.dvxuser.com/V6/showthread.php?t=183539

http://www.dvinfo.net/forum/canon-eos-7d-h...monitor-7d.html

 

The good news is that Tramm Hudson, of Magic Lantern 5DM2 firmware hack fame, is working on it! Stay tuned ...

 

EDIT 9/17/09: This just in:

http://www.dvxuser.com/V6/showthread.php?t=183750

[Peter J DeCrescenzo]

FYI: Apparently this prototype adapter will allow _some_ PL lenses to be mounted on unmodified Canon EOS cams (5DM2, 7D, etc.):

http://fdtimes.wordpress.com/2009/09/22/canon-pl/

[Michael Maier]

I'm not sure I would hold my breath for a useful PL adapter for the Canons. If you noticed, the lens he is using on that prototype is a 300mm. Most likely because wider lenses protrude beyond the rear flange, hence their fine print warning:"Remember: this mount works with PL lenses that do not protrude beyond the rear flange. " Even Panavision which has a much longer flange distance, if I'm not wrong, only some zooms and telephoto primes will fit. Without the ability of using wider lenses these adapters won't be really useful.