Old Canon 5D with "fat" pixels. Magic color?

Discussion in 'Canon EOS' started by ilya_khotimsky, Jun 23, 2016.

  1. Dear forum:
    Here I am talking about the original, old Canon 5D, a 12.8 Mpx camera. No "mark', just 5D. It was Canon's first full frame DSLR. 12.8 megapixels on the full-frame seem like a joke today, when even crop-sensor cameras have 16-18 Mpx or even more. However, I heard that this low pixel count on 5D actually has a certain advantage. It means that each pixel is larger, "fatter" than any pixel on a modern DSLR, so 5D is capable of creating the most amazing color and tone because of it. I know this is subjective, but I would appreciate if anybody who has had the old 5D can share their opinions about this. In other words, is old 5D worth buying today for the magic color of it's big pixels? Thank you in advance, Ilya
  2. I have owned a 5D almost since its introduction. And while it was the first reasonably priced FF camera, and
    was ground breaking when introduced, I don't find its color or tone "magical" compared with other cameras I
    currently use. I would certainly trade it for a Mk II in a heartbeat, although at low ISO it still produces very nice
    looking files.
    Landrum Kelly likes this.
  3. There is no doubt that a test comparison of 5D images compared to those from, say, a 20D will reveal the superiority of the 5D. The difference, however, is only obvious when the viewer knows which pictures came from which camera and sensor. If the labels are switched, so will the judgments be.
    In order to really see the "magic" of the 5D, you need to be looking for that.
    The 5D and 12 MP are adequate for the vast, overwhelming number of tasks asked of a camera.
    Frankly that is true even of my 1998 Kodak Professional DCS 560 (6MP) camera.
    Of course there are actual, physical differences that become obvious when your nose is figuratively pressed up against a single pixel, but these are of relatively little importance for most purposes.
    I went from a 5D and a 20D to a 5Dii and a 50D, and am quite happy, despite the siren call of 50+MP cameras.
    The 5D is fantastic and adequate to purpose. What more can you ask for?
    William Kahn likes this.
  4. Almost bought 5D, but the ISO's, FPS, etc. were....how shall I say this ? Deplorable!!! So I went with D700....and had my own magic :>).
  5. IME with the 5D (shot weddings with it for a number of years before upgrading to a 5D2), the output was quite 'magical' when compared with my 50D (I used the two together), although, in fairness, most of the 'magic' had to do with how creamy and beautiful the RAW files were - NOT any difference in color tones. At low to mid ISOs, the files needed very little if any NR to clean them up before exporting. The same can not be said for the 50D, which was frankly abhorrent (in my opinion) at anything over ~800 ISO.
    Of course, a substantial component of that 'magic' was the shallower DOF I could accomplish, which allowed portraiture to be stunning (where as the 50D's were 'eh' - even in good light). This of course is largely a result of the larger sensor. Even a 6D can easily do the same, though at low to mid ISOs, the 6D, 5D2, & 5D3 RAWS all need more post to become as buttery smooth as the 5D's RAWs were right off the card. Of course @ ~20+ MP, they can handle a LOT more NR without looking plastic fantastic - which yields good imagery much higher up the ISO scale.
    So while I wouldn't say that the 'magic color' of the pixels is a good enough reason to purchase by itself, if you are looking for a FF camera at a stunningly good price which is still capable of producing world class imagery (albeit more slowly, with a small dark (relative to modern cameras) LCD, and a sensor which requires regular cleaning) then the 5D may fit the bill.
  6. david_henderson

    david_henderson www.photography001.com

    I have used and still own 5D/%dii/5Diii. Can't say I've noted any image or colour superiority for the 5D. I've tended to use each of them mostly at ISO100/200/400. Clearly at higher ISO the more modern cameras win heavily.
    I suspect that if you perceive the 5D to have a colour/IQ advantage its probably because you want it to be so. Decent camera of course, though the need to clean the sensor and the relatively primitive autofocus persuade me that the Mark ii/iii are decidedly better cameras.
  7. Thank you for your responses. I am thinking of 5D not because it's the cheapest Canon's full frame at the moment, but because of it unique pixel pitch it might have some image quality advantage over the newer full frames, such as 5D mark ii, mark iii or 6D. If it does have the advantage, I would consider buying 5D simply because in 3-5-7 years there will be hardly any left in the working condition. So far, most people here think there isn't really any color/image quality superiority of 5D over the newer and more advanced full-frame Canons. let's see if anyone thinks differently.
  8. However, I heard that this low pixel count on 5D actually has a certain advantage. It means that each pixel is larger, "fatter" than any pixel on a modern DSLR, so 5D is capable of creating the most amazing color and tone because of it.​
    If shooting jpegs with incamera processing? That's a maybe. Shooting Raw? Not provable due to third party converter color engines.
    I say this maybe because incamera jpegs can have hue twists and uneven saturation boosts to produce a somewhat stylized color grade to captured scenes depending on white balance and exposure. My 6MP Pentax K100D with APS-C sensor has some large pixels but I can't connect that with its beautifully odd color rendering performance to sky blues, grass/leaf greens and pinkish orange warm pastels.
    So you'ld have to similarly dissect each memory color as I did to find out if the 5D's incamera jpeg has such a unique color look.
  9. I had the 5D and loved it.
    I loved the 5D II much, much more--better low-light performance, automated sensor cleaning, etc. I said to myself when I got the 5D II, "I will never need another camera," and in truth I never did. I wish that I had stopped spending right then and there.
  10. I also think the original Canon 5D has a superior image quality, color renditions, etc. don't know if "fatter" pixels contributed to that but it's possible
  11. I said to myself when I got the 5D II, "I will never need another camera," and in truth I never did. I wish that I had stopped spending right then and there.​

    I got a 5DII upon its release and have never felt the need to "upgrade," at least as far as DSLR's go. I decided that the only way I could substantially improve upon the 5DII was by going medium format, so I have recently acquired a 6x7 film camera.
  12. The idea that larger pixels deliver better results would be true, if all else is equal. Which it's not.
    In between the 5D and 5d mk.2, or in yellow land the D700 and D800, a lot happened in sensor development, in optimising the dynamic range, sensibility, the electronics around it. The later generations cameras can offset the "disadvantage" of the smaller pixels by the sheer progress made in their electronics. Any test can show that. The talk about better colours - shoot raw, get a better program, and that's where those better colours are.
    I still like my 12MP camera, and have no need for many more pixels but I do know technically, those newer cameras do simply improve on it, and quite a bit. Whether you want/need to spend the money for those improvements (and whether they're substantial enough for your uses) is a personal choice, of course.
  13. And just a reminder that sharpening and noise reduction quality can change levels of saturation which leads to hue and luminance shifts if judging color working with Raw and how the incamera processing applies it to jpegs. Also contrast induced saturation may provide different looks adjusting contrast within the camera processor verses applying it to the Raw or jpeg in a third party processor.
    I can't count how many times I've had to go back and boost saturation on a Raw file because the Detail slider in ACR/LR boosted the noise that made the overall image brighter (less saturated) when opened and downsized for the web in Photoshop.
    I would say resolution and pixel size has a lot to do with influencing color appearance in this regard.
  14. May be, I have 1D3 and 6D, somehow files from 1D3 looks better strait out of camera. I can match 6D files, but it takes some tweaking in post processing.
  15. I used a 5D as my primary camera for about 8 years, before "modernizing" with a 5DIII. The MK III is an improvement in almost all regards except two that stick out for me. I can only get a burst of 7 shots on the newer camera vs. 12 for the original (shooting RAW + small fine jpeg). And the AWB doesn't seem as good on the MK III under less than ideal lighting conditions (dull, overcast daylight). Of course the white balance can be tweaked in post, and I do that for any image that warrants it, but it was a little surprising that Canon got the AWB algorithm right for the 5D, but seemed to drop the ball with the MK III.
    All in all, those are rather minor complaints. The 5D was a dust magnet, and had no auto cleaning function like the newer cameras. I love the built-in level in the MK III, and the faster burst rate (even if it is shorter) can be handy too. Also, although it's difficult to know for sure without objective testing, I think the focus is more accurate too.
    I decided to upgrade when I did because I could get a reasonably good deal on a new MK III, and there was still some resale value left in the 5D (unlike the XTi I have sitting at home that's perfectly operational, but worth about as much as most consumer film cameras these days). The guy who bought it was ready to go full frame, and I think he's been happy with the purchase.
  16. I saw this video and it inspired me to track down a 5d. I went to B&H used dept. and tried to get one. They were out, so I mentioned a Mark II. They all agreed that Canon changed the sensor for a higher ISO and somehow, the color was not the same. So after finding one, I am sold on the color. The photos are amazing. The camera is really slow though. I tried both 50’s - 1.2 and 1.4. They both look great full open- after LR lens correction. The 1.2 makes it even slower. A6062721-42B5-429D-9455-D6B5B1E6BC0A.jpeg A9118593-F0A7-479B-BEF0-B84E19945BC6.jpeg 2BFDEDDD-59FD-4937-8EFB-A6930C255D0B.jpeg
  17. All that hype over old equipment. Why not tell those stories about old analogue camcorders? So silly.
  18. In the general sense, larger receptors (pixels) will produce a better image because the signal to noise ratio decreases. But, that assumes that the technology otherwise stays the same.

    It doesn't.
  19. I don't believe that's entirely true. A larger number of smaller photosites spread across the same area should capture the same total amount of light, so viewed at the image level, noise should balance out - although noise at the pixel level increases. (For example, from the Nikon world, the 12MP D700 is generally cleaner pixel-by-pixel than the 36MP D800, but the latter is significantly less noisy at the whole image level - there's more "grain", but the grain is smaller.) Per-pixel noise is mainly useful if you're trying to work out whether an increased sensor resolution will actually allow you to crop or enlarge to a greater extent, otherwise image-level noise tends to be a more useful metric.

    There are a few counters to that argument. Particularly in older sensors, not all the image area is available for photon capture - an amount goes to data transfer and other logic. "Gapless microlenses" help this significantly, although the reduced sensor area arguably still affects the well size. The more sensor sites you have, the more area is wasted on peripheral data. BSI sensors (and stacked sensors) also help here. Additionally, the more data you're transferring off the sensor and the more sites the sensor read-out has to pass, the more likely it is that this data will be corrupted. Modern high pixel count sensors are pretty good at this - the 40+MP Nikon D850 and Sony A7RIII may not quite match the 20MP Nikon D5 or 1DxII performance, but they're close (depending on how much Sony are baking their raw files), and they certainly keep the previous 16MP D4 generation honest. The D850 and 5DIV are basically a wash as ISO increases despite the D850's resolution advantage, although the D850 sensor is a little newer (the 5DIV does test better at higher ISOs than the older D810). Read into DxO's graphs what you will, obviously. More sensor sites also counters limitations in the raw file formats - Nikon only support 14-bit raw, for example, and if they made a modern 12MP sensor they may well find that the per-pixel sensitivity would be artificially limited by the output; effectively small sensor sites add "dithering".

    Older sensors did lose out appreciably as resolution increased, which is one reason resolution was relatively limited for a while. Still, the old Nikon 12MP D700 didn't test significantly better (and arguably worse) than the 5DII much of the time for image-scale noise, though both had improvements over the original 5D and the D3S was a significant step forward, still at 12MP. Technology certainly has more of an effect - although we're getting to the stage where there's not much more than engineering can do to improve the limitations of the physics, and we're running out of photons.

    I remember seeing the old 5D, and several friends had one. It has its place in history. But wow are those images noisy, and I'm assuming they're not at stratospheric ISO. :)
  20. -
    Last edited: Oct 5, 2018

Share This Page