"Downgrading" for a very specific reason

[drmikeinpdx]

I don't miss my 5D classic a bit. The LCD screen was tiny, the autofocus system sucked and it was very slow writing to the memory card. I did like the smoothness and color of the images, but I can get the same look from the Mark 3 with Lightroom when I want it.

It's really hard for me to make a valid image quality comparison with my 5D3, or with any two cameras, because so much is dependent on the software I use for post processing.

I've only been using Lightroom for two years. It's fun to go back and re-process the old images from the 5D classic and see what improvements I can make.

 

[memoriaphoto]

I don't miss my 5D classic a bit. The LCD screen was tiny, the autofocus system sucked and it was very slow writing to the memory card. I did like the smoothness and color of the images, but I can get the same look from the Mark 3 with Lightroom when I want it.

It's really hard for me to make a valid image quality comparison with my 5D3, or with any two cameras, because so much is dependent on the software I use for post processing.

I've only been using Lightroom for two years. It's fun to go back and re-process the old images from the 5D classic and see what improvements I can make.

I know what you mean. Handling is poor compared to todays standards... the LCD is really crappy on the original 5D and the AF is poor. But even so...the images are still gorgeous. I sold my old 5D Classic a couple of years ago but I miss it... I'm thinking about getting another one again, just for fun. Like having sex with your ex or something. And they are so cheap now anyway - the 5D, not my ex (at least not that I am aware of)

As for the lightroom editing part... I have never been able to get the original 5D look with my 5D3 files though. May I ask your workflow/settings in LR? It would be interesting to hear

 

[Pi]

Does anyone know where to find real-life RAW files (not testcharts/brickwalls) from the 1DS mark 3 for download so that I can play around a bit and get a general feeling?

Try http://www.imaging-resource.com/. I am not sure how "real life" you will find them.

 

[Pi]

In terms of color separation 5D>5d2>5d3. Of course, color separation cannot be measured with one number only, so those inequalities are a bit fuzzy. Color separation is related to noise, and in terms of noise, 5D<5D2<5D3.

 

[adhocphotographer]

All of these are very good points and interesting arguments....

I think if i were to start photography again, i would not go down the same path i did, but would buy an older second hand camera and a good lens. This thread is an argument for doing just that! To all the "what should i buy" guys out there... pick up a used 5D I, and a 50mm instead of the newest XXXD kit...

I've been saying this exact thing to everyone who asks me what camera is best to start seriously into photography. You can grow with a 5Dc, study with it and eventually make money with it. It still makes good files in all its ISO ranges in my tastes.

If anyone on this thread would like to contribute photos to the 5Dc thread under the camera bodies section, it would be nice to see more photos from it.

You have indeed! I knew someone has been, but i could not remember who, and was too lazy to search!

 

[Hannes]

I am beginning to wonder if people aren't just preferring the look of the old 12-bit files more in general compared to the 14-bit ones they have now. Those files are a bit more contrasty natively it seems.

My old 350D did produce great photos and the 400D I have sitting on my desk creates gorgeous shots. I even don't mind the ISO 1600 shots, they remind me a lot about 400 ISO film.

I also know what you mean with the 1D files being very nice to work with. I recently picked up a used 1D III for similar money to a new canon 700D. It is certainly a lot more camera, the files look wonderful and the 10fps makes you giggle every time you use it. I wouldn't mind if it had a few more megapixels and a little bit better high iso noise performance but on the whole I'm very happy with the purchase. It was basically half the price of a used 5D II and about a third more than a 5Dc

 

[privatebydesign]

Both the 1D MkIII and the 1Ds MkIII both have 14 bit output in 16 bit files.

 

[memoriaphoto]

For me it's not about the 12 vs 14 bit. It's just the old rendering of Canon sensors from the Digic II/III era. They must have used significantlly different color filters. I've read somewhere by a swedish sensor expert that Canon changed their RGB layout in the Bayer filter starting with the EOS 50D model, in some areas to the better but in daylight - sadly - worse. If this is true or not, I couldn't tell but there's a clear difference, I know that much.

Also, now after editing some 1Ds3 raw files I am absolutely convinced that they appear more pleasing to my eyes but also respond better in post processing.

 

[RustyTheGeek]

Yup, you're not imagining the difference. This topic has been touched on in numerous, often vociferous debates here in the past. It's likely the tighter bandwidth of the CFA on the older sensors.
I don't know how much the in-camera processing differs, as far as generating raw files is concerned, but even using the same PP workflow I prefer "the look" from Canon images from the Digic 3 or older generations. I sold off most of my newest Canon gear when switching to PentNikon but I retained my favorite Canon croppers; 40D, 450Ds and even 1000Ds. I still prefer the older 350D in some ways, it also has "that look" that's very similar to the 5DC. You can also find this difference, to a less extent, even in PnS cameras like the G-series. G3 is one of my favorite old compacts for IQ. Combined with the different low-iso noise character, those old cameras produced very pleasing results with minimal tweaks in post. HI iso is no comparison, modern wins.

too bad Ankorwatt/Mikael isn't here to have his, "I told you so" moment.

Couldn't agree with you more! Also, funny you should mention the 350D. I recently bought a used 450D for my mother who needed a light DSLR. Took it for a spin before handing it over, just to make sure everything was OK. When I loaded the RAW files in Lightroom I felt like ".yeah....these files...look....REALLY good!" Straight out of cam, normal light, normal ISO range, kit zoom... for a short second I thought to myself. ....man if only my 5D3 could look this good! Plus the other perfomance of course ;D But still...that was a 250 dollar used camera.

Ditto! I still have a lot of love for the 30D Digic 3 images I took and also the 40D. My favorite body that I still occasionally use today is the 5D. I've never gotten quite as close or connected with the 60D I own. In fact, I may buy another 40D someday. I really loved shooting the 30D and 40D. Great cameras, great images.

 

[memoriaphoto]

Allright...just got me one! Pulled the trigger on a mint condition 1Ds3 yesterday. So far it really shines! On lower ISO it is indeed better than the 5D3, hands down. Images are clearer and crispier and can handle post production work way better. And colors more balanced and pleasing...absolutely LOVE it.

The only thing I miss from the 5D3 is the AutoISO and silent shutter.... I thought the low-res LCD would be a potential issue, but it really isn't. Good enough...

What was it Obi Wan said? Taking the first step into a larger world? Nerdy reference but true nevertheless...

 

[Pi]

DXOmark measures the Sensitivity metamerism index for the 1DsIII to be 86 vs. 74 for the 5DIII (and 80 for the 5D2), all in daylight. The blue CFA filter is very "clean". I have this weird thought that the DXOmark data may be actually worth something...

Enjoy your new toy!

 

[memoriaphoto]

Yes, I saw that on DXO as well...seems the 1Ds3 is one of the most color accurate cameras out there. Bad score for the 5D3 even though I often prefer its image character compared to the 5D2.

What's really surprising is that the mighty 1Dx scored about the same as the 5D3 (actually even worse). Not sure what that means in reality but I would have thought it'd be at least on par with the 1ds3

Sh*t, I've promised myself not ever EVER to go to DXO mark and look for charts and numbers.

 

[Aglet]

I was in a time-wasting mood tonite so looked at over 140 different camera specs on DxOmark and specifically for the metamerism sensitivity ratings.
YUP, 1ds3 and 5d are at 86 and 84 respectively, and these are 2 bodies that are extolled for their image quality.
another good image camera, the 20d, is at 83.. a score it shares with the sx50 compact superzoom and G1x!

Of all the ones I've looked at, there are few bodies on DxoMark that score higher in this metric, 2 are Canons.. and there's a few Sonys. I'm not a Sony fan so not familiar with them but they're included at the bottom.

Would you believe...
coming in at 87 is the.... 50D! (my least favorite body because of it's high subjective noise levels)
and, in TOP SPOT, with a SCORING of 88 ... (Can we have some trumpets please?)
The lowly consumer body, LIKELY ONE OF THE CHEAPEST CONSUMER DSLRS EVER MADE...
THE 1000D!

Instead of everyone rushing off and wasting a lot more time to see how a bunch of Canon bodies fare (in daylight index), here's what I put together, below.
Generally, it seems that most cameras that score at least in the high 70s can provide pleasing color although we see that even 70 or lower can still provide a decent image, witness the 6D.
Combined with unobtrusive noise characteristics, better CFA results seem to be what's appealing to our visual sense of perceived image quality. (yet there's all those processing factors in there that should be able to make up for a good part of the difference?..)

Canon

88 1000d
87 50d
86 1ds3
84 5d
83 sx50
83 g1x
83 20d
82 1d4
81 400d
81 30d
80 550d pre-prod
80 5d2
80 450d
80 s100
79 s90
79 40d
79 350d
78 7d
78 500d
78 1100d
78 G16
77 70d
77 60d
76 g12
76 g11
75 700d
75 s95
75 m
74 5d3
73 650d
73 600d
73 100d
73 1dx
69 6d

Pentax

85 km
84 mx1
83 k52, k52s
82 k20
82 kx
81 q10
81 K10D
81 k7
78 k500
78 k5
78 k200
77 k50
77 k01
77 kr
76 q
76 k30
75 645D


Nikon

85 d60
85 d40x
85 d40
84 d300
84 d200
84 d70
84 d70s
84 d50
83 d700
83 d2xs
83 d2x
82 d90
81 1j1, 1v1,
81 p7700
80 d3200
80 d3000
79 1j3, 1v2,
79 coolpix p6000
79 d3s
79 d7100
79 d5200
79 d3x
79 d80
78 1s1, 1j2
78 p7800
78 d800
78 d7000
78 d5100
78 d3100
77 d800e
77 d4
77 d610
77 d600
77 d300s
77 d3
77 coolpix p30
76 coolpix a
75 d2h
73 p7100
71 p7000

Olympus

86 e520
84 410
83 pen ep2
83 510
82 e600
82 e420
82 e3
80 pen epm2
80 pen epl5
80 pen ep3
80 e5
79 pen ep5
79 omd em5
79 omd em1
79 pen epl3
79 pen epl2
78 pen epm1
78 pen epl1

Fuji

82 s5pro
81 s3pro
79 xf1
79 f800exr
78 f600exr
77 x100
76 xz1
74 xs1

I'm not a big Sony fan, but most of their stuff is pretty well placed in this area.

Sony

87 a900
86 a77, a230,
85 a99, a58, a450, a35,
84 a37
83 a33
82 a100
79 a200, a290
74 a700

 

[Sporgon]

^

I think what you have proofed here is that at least one of the DXo measurements is a load of b******s, and is misleading when taken out of context with all the other characteristics of the sensor.

 

[TrabimanUK]

I think what you have proofed here is that at least one of the DXo measurements is a load of b******s, and is misleading when taken out of context with all the other characteristics of the sensor.

Yep. Using another set of DXO scores for image quality, you get the 6D joint top with the 1Dx with a score of 82. As Mark Twain, who wrote a book or two once said, "lies, damned lies and statistics".

If you like the colours / image quality / look of it then go for the one that you like / suits you. One person's love for a 5Dc's image is differnt to someone elses love for a 7D or 1Dx's images.

Embrace your differences and maybe the world can be a better place (cue the New Seekers and "I'd like to teach the world to sing")

 

[Sporgon]

(cue the New Seekers and "I'd like to teach the world to sing")

;D

You're showing your age !

 

[Pi]

I was in a time-wasting mood tonite so looked at over 140 different camera specs on DxOmark and specifically for the metamerism sensitivity ratings. [...]
coming in at 87 is the.... 50D! (my least favorite body because of it's high subjective noise levels)

The metamerism sensitivity index is just one of the many possible measures of how close the color vision of the sensor is to the human one. It is not affected (much) by noise. DXO have a different way to measure what they call tonal range, which includes both noise and color separation. In that, I believe the 50D does not shine at all.

For good accurate color reproduction, you need CFA filters which project to a color space close to the human one but you also need low base ISO noise, and you need "good separation", which can keep the values of the color matrix under control (the conversion from Bayer to rgb creates its own noise). Canon's cameras mainly differ by the blue channel separation.

BTW, DXO has ellipse charts or something of that sort for color noise. The 1DsIII performs very well compared to any other Canon, almost as good as a Nikon.

 

[privatebydesign]

DxO's metamerism test is so irrelevant even they don't use it in their sensor scores. In the opening line explaining their test concept they point out that "Digital processing permits changing color rendering at will".

To put the various scores in perspective. Take a picture with auto white balance, now take another with manual WB 1,500K different, that would score you a 50 on their test. Which is fine if you couldn't adjust WB in post! As for jumping on a camera that shows a difference of a few points, grow up, the differences are miniscule even if they weren't totally adjustable. Most camera phones are tested in the 40's and colour is not forefront in most peoples analysis of their IQ.

Oh, and make some decent camera profiles...........

 

[Pi]

DxO's metamerism test is so irrelevant even they don't use it in their sensor scores. In the opening line explaining their test concept they point out that "Digital processing permits changing color rendering at will".

To put the various scores in perspective. Take a picture with auto white balance, now take another with manual WB 1,500K different, that would score you a 50 on their test. Which is fine if you couldn't adjust WB in post! As for jumping on a camera that shows a difference of a few points, grow up, the differences are miniscule even if they weren't totally adjustable. Most camera phones are tested in the 40's and colour is not forefront in most peoples analysis of their IQ.

Oh, and make some decent camera profiles...........

This is a gross misrepresentation of what the index means. They do say that this index has a small impact of the IQ but no adjusting in color rendering can compensate for loss of information. You left out the most important part of that sentence, and by doing that, you distorted the meaning of that sentence, and its raison d'être. Here is the original:

The sensitivity metamerism index (SMI) is defined in the ISO Standard 17321 and describes the ability of a camera to reproduce accurate colors. Digital processing permits changing color rendering at will, but whether the camera can or cannot exactly and accurately reproduce the scene colors is intrinsic to the sensor response and independent of the raw converter.

 

[privatebydesign]

Show me pictures from two cameras with a difference of 10-15 points where the difference is not adjustable to the RAW file with a decent workflow.

P.S. If it is such a critical aspect of sensor design why don't even they include it in their proprietary sensor scores?