December 20, 2014, 01:19:42 PM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - dgatwood

Pages: [1] 2 3 ... 59
EOS Bodies - For Stills / Re: Vertical curtain velocity
« on: December 19, 2014, 07:20:10 PM »
We don't read out CMOS sensors in the dark when using them for video, so obviously that's not essential.  I'm pretty sure the real answer is that it's cheaper to toss a leaf shutter in than to do it right....  :)

That's where you get the rolling shutter artifacts.

You get rolling shutter artifacts because Canon's CMOS sensors don't have a global electronic shutter.  The way a global electronic shutter works is pretty straightforward: in a single parallel operation, you shift the values of every pixel into a buffer.  Then, you read the buffered values at your leisure.  There's no rolling shutter problem, because the readout occurs simultaneously across all the pixels.

A global electronic shutter is significantly better than a focal plane shutter in every way.  Even focal plane shutters exhibit a small amount of roll, whereas a global electronic shutter behaves more like a traditional round leaf shutter outside the focal plane, exposing the entire sensor at once and ending that exposure at once.  And unlike either leaf or focal plane shutters, global electronic shutters are almost infinitely fast, so you can stop worrying about flash X-sync speed.

And, of course, they're more reliable....

As readout electronics and speed improve we will eventually get to a global electronic shutter for dSLRs, but we're not there yet.

We should be there.  There's no reason you can't do global electronic shutters with CMOS.  Several companies have been building CMOS sensors with GES for years.  It has nothing to do with speed, and everything to do with simply deciding that the GES is worth the extra sensor fabrication cost and complexity.

Also, as I stated previously, Canon dSLRs do not use leaf shutters.

Sorry, I meant focal plane shutter.  Terminology fail.  :)

Lenses / Re: Lens internal part falling off? its a refurb lens
« on: December 19, 2014, 06:57:39 PM »

My guess is that some piece of plastic didn't get trimmed perfectly cleanly, and the little bit of excess plastic sheared off the first time it rubbed against something.  Eventually, that ended up falling in just the right way to get out of wherever it was trapped, and it ended up between lens elements.

Depending on which elements, and depending on how that particular lens is designed, it could be an easy fix or a hard fix.  I'd definitely let Canon take care of it.  With that said, once they blow the plastic string out, I'd expect the lens to be fine.

EOS Bodies / Re: Noise - maybe it's good?
« on: December 19, 2014, 12:06:57 PM »
Noise; you are at 35mm focal length, the subject is not moving, and yet you are at 1/200s - Justify it with non-importance of noise but I won't hear you. To me, that is just a lack of technique.

Or a lack of a steady hand.  Not all of us hold cameras still enough for a slower shutter speed, you know.  :)

...I am undecided as to whether to trade the 70-300 in for one of the new 100-400s, keep both, or perhaps another lens such as the 300 f4 which might be useful on the 7D2 for campus animals (they let you get surprisingly close sometimes).

You aren't talking about coeds, are you?

Ever see deer on a college campus?  They get so used to people that you can walk right up to them.  They may or may not let you pet them, but occasionally you'll bump into one in the dark, and they'll look at you as if to say, "Look where you're going, moron," and then go right back to eating leaves or whatever.  :)

They also tend to cross at the crosswalks—more consistently than college students, too.

However, in some instances I also use non-rechargable Lithium 1.2V AA cells, typically Energizer Lithium or Varta (which may well be identical except for branding).

FYI, most non-rechargeable Lithium cells are nominally 1.5V, not 1.2V.

EOS Bodies - For Stills / Re: Vertical curtain velocity
« on: December 19, 2014, 10:19:09 AM »
The switch from CCD to CMOS, which need(ed) to be read out in the dark, necessitated a mechanical shutter.  dSLRs use a focal plane shutter, I don't know that Canon has ever used leaf shutters.

We don't read out CMOS sensors in the dark when using them for video, so obviously that's not essential.  I'm pretty sure the real answer is that it's cheaper to toss a leaf shutter in than to do it right....  :)

EOS Bodies - For Stills / Re: Vertical curtain velocity
« on: December 18, 2014, 05:58:08 PM »
I still have the original 1D, the APS-H CCD sensor with global electronic shutter which tops at 1/16000s and flash syncs at 1/2000.

Now a decade later we still can't have this luxury, sigh...

Yeah.  I really don't understand why camera tech took such a huge step backwards, going from electronic shutters to leaf shutters.  Global electronic shutters are clearly far better than anything you can do mechanically.  And if we had it on modern sensors with low enough noise, we could just stop caring about ISO entirely.  Sample the sensor several times, sum the values, and you're no longer limited by the bit depth of the ADC or the full well capacity of the sensors.  Instant high ISO DR.  Instead, we're stuck with these clunky mechanical things that fail whenever it is most inopportune....

EOS Bodies / Re: High Megapixel Camera Coming in 2015 [CR3]
« on: December 18, 2014, 12:02:33 PM »
The 7DII is chewing 20,2MP at 10fps and its buffer handles an infinite number of jpegs and about 30 RAW. Which translates to 4 fps and 12-13 RAWs in 50MP terms. But that is if a new camera is a copy of the 7DII for every thing else but resolution. Is that very likely? The 50MP camera comes a year later, will cost more and can justify more computing power, can use faster memory cards etc. etc.

The thing is, what's limiting Canon most at this point seems to be their USB bus speed:  Their SD card readers cap out right at the USB 2.0 maximum typical speed.  Canon needs to upgrade their SoC to a modern core and shift image processing functionality into dedicated GPUs.  If they did that, they'd be able to have two uncrippled UHS-II SD card slots writing at up to 512 MB/second—four times the maximum speed of CF, and fast enough to do 20 frames per second if you alternate between cards.  Of course, cards with such speeds don't exist yet—right now, you'd only be able to do about 10 fps continuous when alternating—but IMO it is critical that Canon stop shipping these dog-slow SD and CF card slots and move to proper UHS-II slots so that their reader will stop being the main bottleneck.

Canon General / Re: Canon USA Addresses the Gray Market
« on: December 16, 2014, 11:19:08 PM »
Multinational corporations would be well advised to embrace the globalized economy in all aspects. If they want to peddle their wares globally they better respect their customers right to purchase those products at the lowest possible price available. Wherever that may be ... globally. They should be grateful, we buy anything from them.

Not just that.  The lack of a global warranty means that if you're somewhere out of the country traveling, or if you move somewhere else, your U.S. warranty may not be valid.  IMO, that makes their warranty less useful than it otherwise would be.

I also take issue with them calling them "counterfeit" products.  Are they saying that a Canon-manufactured camera sold in Japan or elsewhere is not a genuine Canon camera?  That's just silly.

EOS Bodies / Re: Sony Sensors Coming to Canon DSLRs? [CR1]
« on: December 16, 2014, 05:38:51 PM »
"You keep using that word inconceivable innovation. I do not think it means what you think it means." :P

If you mean low ISO dynamic range, huge megapixel counts or a radical new sensor design, okay. In my book, though, DPAF is clearly in the "innovation" category.

It's certainly an interesting approach, but in my mind, innovation is more than just doing something in a new way.  Innovation is defined by the "aha" moment where the advantages to the new approach are immediately so obvious that you can't imagine having done it the old way.

The thing is, Fuji showed the first compact camera with on-die phase detection AF back in 2010.  Canon has been playing catch-up ever since.  Yes, DPAF has an advantage over dedicated focus pixels in that you don't lose the light that would otherwise fall on half of certain pixels, though that difference will matter less and less as resolution increases—but DPAF still feels more like a way to work around Fuji's on-sensor phase detection patents while still achieving the same benefits, rather than true innovation.

If Canon wants folks like me to see them as still innovating in the area of sensors, they should:

  • Start with a backside illumination design.
  • Etch both sides of the sensor, with vias for every pixel.
  • Put per-pixel buffers on the reverse side of the sensor die, thus giving you a true global shutter.
  • Put a sizable heat sink on the reverse side of the sensor to dissipate the heat from the back-side buffers, thus reducing thermal noise.
  • Use one or more on-die ADC circuits for maximum accuracy and minimum noise.
  • Take advantage of the global shutter to eliminate dynamic range limitations and remove the need for setting ISO values entirely.
That last one is the jaw dropper.  The benefits of a global shutter for video are obvious.  The benefits for stills are even bigger, though, and I don't think anybody is really taking advantage of that yet, which seems bizarre to me.

The entire reason image sensors have limited dynamic range is twofold: because the ADC can provide only a certain number of bits of precision, and because when the full-well capacity of a pixel is exceeded, that pixel cannot hold any more photons.  However, if you can get the read noise levels low enough, you can just sample the pixels several times per exposure, and sum the results in a wider register.  You can then make clever use of Huffman coding or run-length encoding to minimize the impact of all those extra zeroes, and you'll be able to accurately reproduce everything from a single photon all the way up to the brightest light.

That would be innovation.  Real innovation changes things in ways that are jaw-dropping and earth-shattering.  Using DPAF to do automatic AFMA might do that, and using DPAF to correct the slightly-off phase detect focusing results after the mirror goes up might do that, but DPAF by itself doesn't do that, IMO.  DPAF is clever, but it is far from pushing the limits of technology.

Just my $0.02.

EOS Bodies / Re: Sony Sensors Coming to Canon DSLRs? [CR1]
« on: December 16, 2014, 10:03:06 AM »
One shouldn't expect Dual Pixel sensors manufactured by Sony. No way.

If Canon handed back to Sony a variation of Sony's design with Canon's DPAF then why not?

They wouldn't even have to do that.  I don't think there's anything amazingly complex about dual-pixel from a design and fabrication point of view.  I'm pretty sure Sony would already be selling chips with DPAF capabilities if they didn't have to work around Canon's patents.

I shot a few hundred shots at a friend's stage event the other day, and... the camera's metering didn't fare too well, to say the least.  The meter was consistently pegged so hard to the left that I stopped even bothering to look at it.

Initially, I tried to use exposure compensation in evaluative mode, to see if that would be usable on the 6D.  Unfortunately, that was too inconsistent, and even with the exposure compensation cranked to the left as far as it would go, shots were still occasionally blown out, so I gave up and used full manual mode, with chimping on every single shot.  I would have killed for a usable Tv mode, because manually switching ISO settings and exposure times depending on whether I wanted to get motion blur or not is kind of a pain in the backside... but Tv just wasn't to be.

The problem is that when doing stage photography, all of the autofocus points are likely to be in focus, because you're shooting at or beyond infinity the whole time.  So Canon's evaluative metering tries to take into account the spot metering values for every single focus point, thus dutifully providing amazing contrast on the flat black stage background, while simultaneously ensuring that every single face is consistently blown out beyond recognition.  This is clearly not desirable behavior.

By now, I would have expected the evaluative metering system to recognize that, "Hey, there's no significant contrast near any of the in-focus areas except for the chosen AF point, and that spot is going to be massively blown out if I meter it this way," and then adjust its expectations (and its light meter readings) accordingly.  Unfortunately, Canon's AF system doesn't do this.

Of course, the problem is made worse by the lack of a spot-focus-follows-manually-chosen-AF-point mode, of course, but even if I had that, I'd still probably be swearing at it a lot.  There simply has to be a more sensible way to meter scenes like this.

Alternatively, give us a sensor with massively larger full well capacity—say 24 stops of dynamic range at high ISO—and I'll stop caring about the metering so much.  I'm not holding my breath for that one, though, as it probably won't happen until we get a proper global electronic shutter and can sum multiple successive samplings to produce the final image....

EOS Bodies / Re: Sony Sensors Coming to Canon DSLRs? [CR1]
« on: December 16, 2014, 12:26:45 AM »
I have received this information a few times over the years, but nothing has never come to fruition, so I’ve never lent much credence to it. I would imagine joint ventures on something as expensive as sensor manufacturing would help lower costs for companies and perhaps customers as well. It’s definitely possible, but I wouldn’t write it in stone yet.  If Canon does the manufacturing of the sensor can they put their own technology into a Sony sensor, such as Dual Pixel AF?

That would truly be a killer combination.  It would also mean that Canon would get money from every Sony sensor sold, including to Nikon et al, so it isn't entirely out of the question.

EOS Bodies / Re: Using full frame lens on crop body cameras ?
« on: December 14, 2014, 08:29:49 PM »
I agree with you about the 24-70 2.8 on a crop sensor. You won't be able to get as shallow depth of field with a crop señor as with a ff. I think that's the point he's tying to make that you geting more like a f4 depth of field compared to a 2.8 on ff. But yes you still get the light of a 2.8 lens on a crop sensor

Depends on how you look at it.  I'd argue that with the exception of the f-stop's effect on autofocus, you really don't get the light of an f/2.8 lens on a crop sensor.  I mean ostensibly yes, if you have two sensors with the same pixel size and you look at a pixel-sized crop, you would see the same amount of light, but that's not the way people use cameras in practice.

People typically use cameras by framing a shot, and then viewing it at screen size or printing it at a desired print size.  So the only truly interesting metric is the amount of light that makes up each square inch of output at a given size.  Using that metric, because a crop sensor sees light from only about 39% of the lens, per square inch of output, a crop body gives you an image produced with only about 39% of the light that you'd get shooting the same shot with a full-frame body (ignoring any differences in light caused by moving closer to the subject, which if included, would make the crop body look even worse by comparison).

When it comes to the actual image projected on the sensor, there's no meaningful difference between using a crop body and using a teleconverter on a full-frame camera—just a little bit of IQ loss caused by the quality of the TC's glass, and maybe a tiny bit of light loss from the glass itself.  And we say that using a 1.4X teleconverter makes a lens act like it is a stop slower.  By that same standard, using a glorified 1.6X teleconverter (a crop body) makes a lens act like it is 1.35 stops slower.  The only real exceptions to that rule are when either A. you'd be cropping the image on a full-frame to match the crop body (the reach-limited case) or B. you're talking about how the autofocus behaves.  But in the more general case, you really don't get the benefits of an f/2.8 lens.

Software & Accessories / Re: An Easy Magic Lantern How-To from CNET
« on: December 12, 2014, 10:29:42 PM »
I think you are wrong. I think it's clearly deliberate decisions. You can even see where they have even coded some more advanced things and then locked out the written code or removed it and if you listen in detail to all the things their guys say at trade shows and so on....

Bear in mind that such elisions could also mean that they tried it, but it didn't quite work as well as they had hoped, and they didn't have time to debug it further and make their ship date.

I mean just look at the games they played with AutoISO. Utterly trivial code. It's 1/100000000000000000000000th the complexity and time of tons of stuff in the firmware but they took FIFTEEN YEARS to dribble it out (and even now only for 7D2 and 1DX in reasonably complete form). And yet the basic code for it is so simple that you teach a newbie programmer (one who had never programmed before) how to code the basics of it (maybe not the modal synching into the knobs and dials, but the main code) in their first week and it would like one page of code.

Trivial, yes, but it is a feature.  That actually strongly supports my assertion that they're basically only making software changes that are driven by hardware changes (and only the minimal changes needed to support those hardware changes), while not taking the time to add even low-hanging-fruit features that are software-only.

Pages: [1] 2 3 ... 59