Canon announces new EOS-1D X flagship

[KenPapai]

<p>There is a hilarious link on <strong><em>What The Duck's </em></strong>FB page to a video of the 1Dx in continuous shooting mode -- "Sounds like the register racking up your receipt when you pay for it..."</p>

<p>(it's in "inrsoul" 's Flickr photostream) </p>

<p>LOL!</p>

[hjoseph7]

<p>Hopefully some of those new features will trickle down do the 5D Mk III, and the 7D Mk II. I definately like the dual card slots, increased dynamic range and double exposure capabilites...</p>

[emiliogtz]

<blockquote>

<p>But apparently not you, Bob.<br>

As Daniel explains up the page, a higher pixel count does not in itself - now, or ever - equate to more noise, and in fact <em>all</em> the evidence so far points to noise <em>decreasing</em> as pixel count increases, other things being equal.<br>

Nikon D7000 vs. D300, Canon 7D vs. Canon 40D, 1D Mk IV vs. Mk III, 5D Mk II vs. 5D... See a pattern?</p>

</blockquote>

<p>I would not say that <em>other things being equal </em>applies here. Those cameras are from different generations you are comparing. Other things being equal would apply to the D3s and the D3x Nikons, which belong to the same generation. </p>

[danielleetaylor]

<p><em>Other things being equal would apply to the D3s and the D3x Nikons, which belong to the same generation.</em></p>

<p>These are not the same generation. The D3s is nearly one year newer. That's a long time in chip design and fabrication.</p>

<p>That said, these two bodies are probably the closest comparison we have to judge the issue of noise and pixel density. Looking at the Imaging Resource ISO 6400 test samples with the D3s images scaled to the same resolution as the D3x...</p>

<p>* With NR off the D3s appears somewhat cleaner than the D3x at 100%. This is hard to judge though because the D3x noise is sharper along with all the other detail. I would say that the D3x does exhibit a bit more noise, but that sharpness plays the primary role in the perception of difference. At 50% it's harder to tell a difference. A quick, automatic Noise Ninja pass makes the D3x image a bit cleaner yet no worse in sharpness / detail than the D3s.</p>

<p>* With NR Normal the differences diminish. There's nothing to see at 50%. At 100% again the D3x noise is sharper along with everything else in the image. The sharper noise is a bit more intrusive but I doubt there's significantly more in this case.</p>

<p>Based on these I'm confident that proper processing in RAW would yield a D3x ISO 6400 image that had roughly the same noise levels, but greater detail, than a D3s image. If this is the gain that can be expected from a lower pixel density, then those of us saying that pixel density is at most a minor issue in final image noise are proven correct.</p>

<p>This raises the question of why the D3x was capped at 6400. My guess is that processor bandwidth plays a major role in all of these decisions. At the time Nikon probably could not meet their requirements with the CPU available and a 25 MP sensor at higher ISOs. (RAW and "NR off" do not mean there's no processing of the sensor read out before writing the file. Even if it's significantly less, these guys are probably reluctant to introduce dramatically different response times and frame rates based on NR selection.) Canon has three CPUs in the 1DX and you can bet they would have no chance of 12 fps at a higher resolution.</p>

<p>I will add one more point: no one can guarantee that chip designers will not find a way to leverage a lower pixel density into superior high ISO performance, all other things being equal. But that certainly doesn't appear to be the case with Nikon. Until the 1DX is tested I have no idea if that's the case now. But all evidence to date suggests that pixel density is not a primary factor in image noise.</p>

[jamie_robertson2]

<blockquote>

<p>As Daniel explains up the page, a higher pixel count does not in itself - now, or ever - equate to more noise, and in fact <em>all</em> the evidence so far points to noise <em>decreasing</em> as pixel count increases, other things being equal.<br>

Nikon D7000 vs. D300, Canon 7D vs. Canon 40D, 1D Mk IV vs. Mk III, 5D Mk II vs. 5D... See a pattern?<br>

You won't find a single <em>credible</em> piece of evidence for the assertion that a higher pixel count has a noise penalty.<br>

 </p>

</blockquote>

<p>That's absolute nonsense I'm afraid. According to your theory, a 30MP EOS 1DX would have lower noise than the current 18MP version. Utter rubbish! You are comparing newer cameras to older cameras and forming a false conclusion.<br>

 <br>

If pixel density had no effect on noise, full frame DSLRs could use the same pixel density as compact cameras and we would have 300MP+ sensors. There is only ONE reason why DSLR pixel density has stayed within sensible limits so far... and that is to keep the noise under control.</p>

[zml]

<p><em>> There is only ONE reason why DSLR pixel density has stayed within sensible limits so far...</em></p>

<p>Yes. The fabrication process and its limitations.</p>

[jamie_robertson2]

<blockquote>

<p>Yes. The fabrication process and its limitations.</p>

 

</blockquote>

<p>No, acceptable noise levels is the reason. If Chinese camera manufacturers can put 15MP on a cheap tiny digicam sensor, Canon and Nikon could put 300MP+ on a huge FF DSLR sensor if they wanted to. They reason they don't is because image quality would be just as bad as a cheap compact and file sizes would be stupid. If very high megapixel FF DSLR sensors with low noise were possible to economically manufacture at present they would be on the market. You're kidding yourself if you think otherwise.</p>

[zml]

<p><em>> No, acceptable noise levels is the reason. If Chinese camera manufacturers can put 15MP on a cheap tiny digicam sensor, Canon and Nikon could put 300MP+ on a huge FF DSLR sensor if they wanted to.</em></p>

<p>No, fab process. Putting 15 mpixels of crappy pixels on a small sensor is child's play comparing to putting even 15 high quality megapixels on a 24x36 mm CMOS. Besides, quality (in cell phone terms, i.e. not that great in terms of real cameras) cellular phone camera sensors barely reach 8 megapixels (Sony, OmniVision. ) And no, even if Canon and Nikon rotated on their respective appendages till the cows come home, neither, and especially not Nikon, would be able to put 300 usable megapixels on a 24x36 mm sensor yet (you know...the limiattions of the fabrication process...)</p>

[john_bellenis]

<p>I'm not a camera or sensor manufacturer and I don't pretend to be one. I'm just a photographer, and what I know for sure is that the last thing I want is a 300mp camera - regardless of how good or bad the noise is</p>

<p>Storage of the RAW and TIFF files would be ridiculous and processing 16 bit TIFF files in Photoshop would be a joke. Sure storage is getting cheaper, but a 3T drive would be considered small, and think about the RAM needed to manipulate those enormous TIFFs in Photoshop (nearly 2 Gig each for 16 bit) and apply filters! 15 to 25MP would seem to be the sweet spot for me considering file size and final usage, storage, ease of use in Photoshop and the number of images on a CF card</p>

<p>The next question is who needs those huge files? For all my reproduction needs (full page and double truck magazine ads) you'd be throwing away pixels by the bucket load.</p>

<p>Maybe the reason we don't have 300MP DSLRs is simply that the manufacturers aren't completely insane?</p>

[craig_meddaugh]

<p>Nobody needs those huge files since as optical limitations of our lenses would ensure that the resolution would be wasted. Even at the &d's pixel pitch, my top-end superteles are hitting the limits, so 45-50MP would be the most that would make sense unless Canon wants to issue a lot of really good glasses. </p>

[zml]

<p><em>> 15 to 25MP would seem to be the sweet spot for me considering file size and final usage, storage, ease of use in Photoshop and the number of images on a CF card</em></p>

<p>I shoot PhaseOne 65+ back quite often and can assure you that handling of ca. 360 MB files is not that difficult with Windows, 64 bit OS, oodles of RAM, etc. :-)</p>

<p>The "no it is fab, no it's noise" is a bit tongue-in-cheek but very a propos nonetheless: to many shooters the noise characteristics of the new Canon 1Dx at high(er) ISO will be make it or break it.</p>

[john_bellenis]

<p>I totally agree Michael - I will be in the market for a 1D X as soon as the initial tests reveal there are no major first issue type problems. 18mp is great for my uses and if the noise is as low as it would appear, this could be the perfect camera for me. The rest of the features - AF and drive speed, increased video capabilities, twin CF cards, etc. etc. are just icing on the cake.</p>

<p>I also shoot MF digital every week, but we aren't talking about 360MB files from a 300mp camera - we are talking about 16bit TIFFs that would be around 1.8 Gig each and that's a storage and processing problem!</p>

[danielleetaylor]

<p><em>That's absolute nonsense I'm afraid. According to your theory, a 30MP EOS 1DX would have lower noise than the current 18MP version. </em></p>

<p>I don't believe Keith would claim that. Given the same level of technology and sensor surface area the total image noise would be roughly the same.</p>

<p><em>Utter rubbish! You are comparing newer cameras to older cameras and forming a false conclusion.</em></p>

<p>It is not false to conclude, from his examples, that technology is a much larger factor than pixel density in determining image noise.</p>

<p>Also note that the 25 MP D3x compares favorably with the 12 MP D3s at ISO 6400 despite the D3s being the newer camera.<em> </em></p>

<p><em>If pixel density had no effect on noise, full frame DSLRs could use the same pixel density as compact cameras and we would have 300MP+ sensors. </em></p>

<p>While there would be engineering and fabrication challenges to using the same density on APS-C and 35mm chips, this would otherwise be possible. Please note though that once you hit the densities seen in P&S sensors per pixel noise increases fall out of balance with per pixel magnification decreases. Past a certain point smaller pixels will contribute more noise for a given technology level. We're just not at that point with DSLR sensors. (I don't know if that would occur at 300 MP or some lower level like 200 MP. I would have to work out the math.)</p>

[danielleetaylor]

<p><em>18mp is great for my uses and if the noise is as low as it would appear, this could be the perfect camera for me.</em></p>

<p>18 MP is great for most uses. I've said many times that the 18 MP 7D makes excellent 24" landscape prints. Other subject matter is less demanding. A lot of my surfing shots would go well beyond 24" (and some have) despite the 7D's older, smaller sensor which is no doubt comparatively "noisy" against the 1DX. The 1DX will make a brilliant indoor sports camera and will cover the vast majority of subject matter / print size combinations people encounter, period. Most people just simply do not need to print larger than 18 MP can for any subject matter.</p>

<p>But some of us do want to see more MP for larger landscape and studio prints. MF remains out of my budget. I would be thrilled to see a 36 MP FF sensor from Canon, especially at the 5D price point. I can only hope we will see such a product from Canon.</p>

<p>As for the 1DX it is no doubt a fine camera. If I did professional indoor sports I would be looking forward to March. But it won't gain me anything over my 7D for landscapes or surfing. Indeed, in broad daylight the 7D might still be the better surfing camera for the reach. I suspect a lot of 5D mkII and 1Ds users will feel the same way as those bodies are typically used for landscapes and studio work. It's not that the 1DX is bad. It's a fantastic 1D upgrade. It just doesn't offer much for a large class of users.</p>

<p>Contrary to Canon's initial marketing which accompanies this PR, noise is not what typically limits enlargements from the 7D, 5D mkII, and 1Ds mkIII.</p>

[ed_avis2]

 

<blockquote>Contrary to Canon's initial marketing which accompanies this PR, noise is not what typically limits enlargements from the 7D, 5D mkII, and 1Ds mkIII.</blockquote>

<p>In a way, noise is what limits shutter speed and aperture. You might shoot at ISO 1600 with your older body because any higher than that, there is too much noise for your application. Now a newer body allows you to shoot at ISO 6400 with the same noise level, so you can have a shutter speed four times as fast. That lets you use a cheaper non-image stabilized lens, or leave your tripod behind. Alternatively you could 'spend' the extra sensitivity on a narrower aperture, so you can use a cheaper and lighter f/4 lens rather than some huge chunk of glass. That has to be worth something.</p>

<p>Whatever size enlargements you make, at some point there is a maximum noise level you will tolerate. That varies between individuals, but it exists: nobody shoots at absolute maximum ISO all of the time. Anything which lets you get a faster ISO while keeping within your personal noise threshold is a significant improvement.</p>

[craig_meddaugh]

<p>Ed, you are correct, but think about the market segment this camera is aimed at. 1D X shooters already have the big, fast, stabilized lenses. I guess I could be happy with this camera if Canon were to issue some high quality slower telephoto lenses and if it could AF at f/8 (I'm hoping that marketing overlooked the fact that this camera does AF at f/8... because if it doesn't that's a major mistake on Canon's part), but absent that, this seems to be aimed strictly at photojournalists and sports shooters.</p>

<p>I know wildlife is a small part of the market, so I don't expect Canon to cater to us, but for wildlife work, this camera is a bit of a disappointment. I guess they are trying to market to event shooters as well, but for the price difference between this camera and the 5dII you could get a hell of a flash setup and never care about high ISO performance ever again. </p>

[massimo_foti]

<p>I am with Ed on the ISO, less noise means more flexibility/choice.<br>

Looks like there are bad news about AF at f/8, read below:<br>

<a href="http://www.birdphotographers.net/forums/showthread.php/90630-Canon-EOS-1D-X-Digital-Camera-Body-Details?p=731924&viewfull=1#post731924">http://www.birdphotographers.net/forums/showthread.php/90630-Canon-EOS-1D-X-Digital-Camera-Body-Details?p=731924&viewfull=1#post731924</a></p>

[phule]

<p>[[Looks like there are bad news about AF at f/8, read below:]]</p>

<p>Are you claiming that you know for certain you cannot just tape the pins on the TC?</p>

<p> </p>

[massimo_foti]

<p>I am not claiming anything. In all sincerity, I don't even care, I am not into long telephoto lenses.<br>

I am just reporting information that I think may be valuable and looks like it's coming from a trustworthy source.</p>

[ed_avis2]

<p>I'm not a pro wildlife photographer, but I sometimes shoot wildlife with a 500/7.2 lens I bought cheaply. In blazing sunlight it's fine but otherwise not much good handheld. If I wanted to improve the number of good shots I get, I could spend a large sum on a 500/4 stabilized lens, or about the same amount on a 1DX which will, if Canon's marketing is to be believed, let me shoot at ISO 25000 or so and get decent quality.</p>

<p>Even if you already have the big stabilized lens, at dusk each minute brings the sun a little lower and the scene a little darker. An extra stop of brightness from your lens can cost a lot of money but that might only allow you another 20 minutes of shooting time before it gets too dark. If you shoot wildlife that gets active at sunset, a few stops of low-light performance in the sensor could be a great investment.</p>

[massimo_foti]

<p>Faster lenses aren't just about low-light, they are important to achieve thinner DOF and subject isolation. Low noise at high ISO is great and I am all for it, but it can't always compensate for slower lenses.</p>

[dan_south]

Just in time for the London Olympics! ISO 51,200 should provide some interesting new opportunities for freezing the

motion of fast athletes competing at night or in indoor venues.

[abbas_haider]

<p>Good day all Canon 1Dx is the upgrade of the MK4 and not the merge of the 1D series. Canon stopped the production of 1ds as they are making very good market on the 5DII<br>

The speed of the camera clearly shows that it is an upgrade of the MK4. though the 1.3 crop was awesome for the extra reach.<br>

On the other hand canon did increase the MP from 16 of Mk4 to 18 on 1Dx. and they are going slow on this cam to avoid noise at high iso<br>

They have just change the names for the marketing purpose</p>

[phule]

<p>[[ Canon 1Dx is the upgrade of the MK4 and not the merge of the 1D series]]</p>

<p>Except for the fact that the Canon EU page starts with:</p>

 

<blockquote>

<p>The EOS-1D X is the new flagship camera in the Canon DSLR range, taking the best of the EOS-1D and EOS-1Ds cameras and melding them into a single unit that gives both high resolution and high speed shooting. No longer do you need to choose between the two when both are available in one body.</p>

</blockquote>

[alan_bryant1]

<p>After pondering this for a bit, I am increasingly surprised that they dropped f/8 AF support. For the birding/wildlife crowd that's got to be a problem. Looking at the numbers from a pixels-on-the-subject standpoint, consider two rigs:</p>

<p>1D mk 4 with 600/4 and 2x extender, and</p>

<p>1DX with 600/4 and 1.4x extender</p>

<p>Imagine a distant subject. If the first combination can put, say, 1,000,000 pixels on the subject, the second would only be able to put 341,000 pixels on that same subject. This is because you're losing both pixel pitch *and* 1.4x extension.</p>