Jump to content


  • Posts

  • Joined

  • Last visited

Everything posted by Policar

  1. [quote name='EOSHD' timestamp='1344102996' post='14956'] In the review it means 28mm equivalent, not 28mm focal length lenses on both cameras. It is 8mm on the Nokia. The field of view was matched. [/quote] Score one for Nokia then (and one for the iPhone's video--wow!). Just got a survey from Canon regarding dSLRs, btw. Emphasis seemed to be on what kind of user needs dSLR video and why (power user vs casual photographer; web video vs. narrative vs. videography, etc.). Primary emphasis seemed to be on to what extent Canon should further hybridize its dSLRs rather than splitting video off into a separate camp (obviously Canon is trying both strategies to some extent), but lots of questions about image sharpness as a concern. They're listening, just not responding.
  2. A test with two 28mm lenses is ridiculous; the FOV for the 5D would be much, much wider. A modern point and shoot with a 28mm lens will outresolve any dSLR with a 28mm lens, but at a matched FOV it will be another story. The camera is still pretty impressive.
  3. [quote author=RichST link=topic=884.msg6468#msg6468 date=1340587876] I also think it's the latter, binning to a single 1920x1080 bayer grid would take less time to read off the sensor and could make use of the camera's debayering and processing chip(s) [/quote] I know next to nothing about this, but I agree just based on how similar 5D Mark III footage looks to a JPEG still zoomed to 100%. It's too bad; the camera is per-pixel sharp (kinda) when raw files are processed in DXO mark and optionally sharpened with deconvolution (when I use my ultra-sharp lenses!).  I remember all those sample 5D Mark III JPEGS were published and everyone thought they were terrible, and they were--but it had more to do with the camera's JPEG and sharpening engine than it did with the resolution difference between the D800 (whose sample photos were processed from RAW) and the 5D III.  There are so many reasons this camera should be better than it is, but I assume debayering is low-level code.  Oh well.  It's still not bad.
  4. Is the 5D III selling that much worse?  From what I understand they're both selling very well.  The D800 does look particularly great for stills. I'm just not sure Canon will lose that many customers.  And lens sales matter more for profitability than do body sales (presumably, most D800 buyers are upgrading from D700s, not 5D Mark IIs).  As for video, the non-standard lenses available for the GH2 and black magic camera and their non-standard workflows put them in high-end enthusiast owner/op territory exclusively (and you really need to buy special-purpose lenses around video, pretty exclusively, for either camera, which is a significant expense not mitigated by versatility).  And high-end enthusiasts isn't as big or lucrative a market as say rentals, wedding videographers, journalists who want to shoot video and take stills, more casual hobbyists, etc.  Most hobbyists (the biggest market by far) are interested in ease of use, which is the real advantage with Canon cameras, that and lens availability (just look at any rental shop's inventory); most professionals are interested in reliability and compatibility over image quality, and the Canons and Nikons excel there.  There are still huge markets for which "good enough" but easy is great. The 5D III is a pretty boring jack-of-all trades camera, but for a professional shooter or hobbyist (both of whom want versatility) it does everything well enough and the Canons are exceptionally reliable.  It fixes aliasing and bad AF, the two most common complaints with its predecessor.  I don't think Canon will gain any market share this generation, but they won't lose much either.  Canon's biggest miscalculation was making the 5D III so incrementally better than its predecessor that most shooters (stills or video) won't upgrade, but Mark II owners still buy Canon glass….  If the C300 does fail, that might be more of a legitimate wake up call than the 5D III's middling success, but the progeny of that are years off, at least... The D800 does look great though, doesn't it?
  5. Canon is selling tons of 5D Mark IIIs (even if most are primarily used for stills) and the C300 is apparently selling fine, too.  I'm not sure when this rude awakening is going to occur, if ever.  If you like the black magic camera (hard to know since no one has used it), get it.  Imo the internal battery makes it useless except as a test bed, but based on most of the replies here all anyone's shooting is resolution charts.  Granted, the 5D Mark III is quite soft (no worse than the 7D), but it's good enough as an A cam for HDTV and web and a B cam for theatrical; if that's not enough, the C300/F3/FS100/FS700 is a dirt cheap rental. I don't think the 5D is intentionally crippled, at least in terms of IQ (its lack of focus peaking is infuriating, though).  The 5DIII seems to be binning before readout.  Either the pixels are binned at 4x4 to three channels of 1440X810 and added using a scheme similar to the C300 or they're binning to one 1920X1080 bayer grid that's then debayered.  My guess is the latter.  Either way you can expect about 75% of the linear resolution you want, while other cameras are oversampling quite a bit....  There's this artifact that looks like aliasing or stair stepping in high frequency saturated red areas and next to Alexa or Epic footage it just jumps out.  Canon has bad debayering and sharpening algorithms, which is what makes it look even worse.  The JPEGS out of camera look terrible at 100% and so do the 1DX's.  Canon would have to rewrite their debayer algorithm and sharpening algorithm entirely to get better video since the binning appears to be done in hardware.  And the read noise (also a hardware issue) is horrible on the 5DIII.  I don't think it's crippled; it's just not very good (it is, however, the best low light video camera available except that the noise is ugly). Presumably the 1DX is skipping lines (worse for low light, but presumably the read noise and quantum efficiency make up for it), but oversampling.  So it's sharper but more aliased.  I doubt it will be a popular camera for video.  The video looks sharper but still not great and it suffers in more important areas (lack of headphone jack; ergonomics; etc.).  I don't think it was designed for video and, paradoxically, that's probably why it has sharper video. I'm still waiting on a 7DC, but not holding my breath.
  6. [quote author=HurtinMinorKey link=topic=649.msg4848#msg4848 date=1335667323] I think it would make more sense if the binning was done by software.  I'm guessing it's just a matter of  telling the processor to read more info for each frame. [/quote] The sensor can't read out fast enough for the binning to be done in software.  The 1DX's sensor has twice as many readout channels to achieve twice the fps at full frame (twice as many pixels), and four times at 1.3 crop (also roughly twice as many pixels read out per second).  Obviously no one has publicly confirmed how the binning is done, but if the sensor were fast enough to read out full raster, why would Canon have twice as many readout channels on the 1DX for twice as many pixels per second?  The c300 seems to bin in camera, though.
  7. I don't think they can improve resolution, and I do think there's some kind of binning that's done in hardware and that will be impossible to effectively circumvent, but if they can either (ideally) improve the sharpening algorithm to something less completely terrible or even increase the ALL-I bitrate so you can sharpen it without macroblocking we'd be in business. The resolution isn't as bad as most people claim (it's significantly better than 720p and better than the 5DII), but the apparent softness is even worse.  We need some better in-camera sharpening, though most people are going way overboard in post.  For CUs the camera looks amazing as is.  But look at wide shots and compare them with the Alexa...not even close.
  • Create New...