5D3, 6D and D600 true RAW (dcraw) comparison

Home · Register · Search · Software · Hosting · Software · Join Upload & Sell

FM Forum Rules
Canon SLRs, primes, and zooms lenses reviews

FM Forums \| Canon-mount SLRs \| Join Upload & Sell
1 2 3 4 5 end

cgarcia
Offline
•
Upload & Sell: Off

p.2 #1 · 5D3, 6D and D600 true RAW (dcraw) comparison

D. Diggler wrote:

I wonder if CGarcia could put the 1DX up there against the 6D.

Yes, I also had that idea and fortunately www.focus-numerique.com has tested the 1DX. Here it is:

6D vs 5D3 vs 1DX

Despite slightly ahead of 5D3, until ISO 1600 it is not clear if 6D is cleaner or not compared to 1DX. After seeing this I'm sceptical about the improved dynamic range of 6D, taking into account the 1DX results at DxO. The lower noise at low ISO could continue being more related to the higher pixel size than to readout noise improvements. At high ISO, 1DX is clearly better than 6D, albeit by a little margin. The good news is that 6D seems more close to 1DX than to 5D3, despite it should be "just in the middle" due to the pixel size. Well, Canon has still a lot of work pending, but 6D buyers will get one of the best sensors Canon currently makes (the couple of megapixels seems the only "crippling" chosen, fortunately not the image quality) from the point of view of signal to noise ratio (we are not talking here about sharpness, etc).

I include bellow 1DX in the full size conversions, adding also a 6400 picture for those more interested in (very usable) high ISO rather than clean low ISO. The only parameter passed to dcraw is -a (for the white balance):

5D3 ISO 100
6D ISO 100
1DX ISO 100
5D3 ISO 6400
6D ISO 6400
1DX ISO 6400

Dec 04, 2012 at 02:46 PM

Wahoowa
Online
• • •
Upload & Sell: On

p.2 #2 · 5D3, 6D and D600 true RAW (dcraw) comparison

Thank you for the comparisons.

Dec 04, 2012 at 02:53 PM

cgarcia
Offline
•
Upload & Sell: Off

p.2 #3 · 5D3, 6D and D600 true RAW (dcraw) comparison

Michaelparris wrote:
The circuit board and the white fabric and the book in some areas seem to have greater detail (not by much) on the 5D MKIII......Just an observation.

Apart from focusing differences, both sensors may differ in the antialiasing filters. Dcraw uses exactly the same algorithms for all cameras. In truth, we don't look at the raw image (we can't!) but at least perhaps we see the most "rude" conversion.

Talking about sharpness, all pictures I have seen from 6D are soft, in my opinion specially those from dpreview's preview (despite using 24-70 II and other sharp lenses). I hope that this will be due to the jpeg engine default settings, operator error or environment/subject condition.

The advantage of dcraw, compared to other conversion tools, is that it really doesn't does anything by default to reduce the noise, despite the initial stages of conversion likely being a good place for that. Thus, instead of comparing detail (the only result after noise reduction) we can concentrate in compare the noise itself. It is very difficult to detect a small difference in noise between two sensors (like 1/4 or 1/3 fstop) after looking at cleaned jpegs using the "exchange rates" between detail and noise...

Dec 04, 2012 at 03:11 PM

StillFingerz
Offline
• • • •
Upload & Sell: On

p.2 #4 · 5D3, 6D and D600 true RAW (dcraw) comparison

Wahoowa wrote:
Thank you for the comparisons.

+1 Excellent efforts, example images are quite helpful, well done, thanks

Dec 04, 2012 at 03:12 PM

weezintrumpete
Offline
• • • •
Upload & Sell: On

p.2 #5 · 5D3, 6D and D600 true RAW (dcraw) comparison

The 6D's sensor looks like a winner

Dec 04, 2012 at 03:48 PM

cgarcia
Offline
•
Upload & Sell: Off

p.2 #6 · 5D3, 6D and D600 true RAW (dcraw) comparison

MayaTlab wrote:
I believe that for this comparison to be even fairer, it might be necessary to saturate more the red channel in the D600 files.

Yes, you are right. Dcraw applies some kind of color correction after demosaicing (its source code states that most values are taken from Adobe DNG Converter) but unfortunately still has no entries for D600. I tried D800 values, with no apparent change in the results. D600 noise figures as shown in this test are not entirely comparable to the other cameras (and any postprocessing will not be totally fair, since no exactly same edit applies). It takes some time until dcraw adds support for new cameras (in fact, D600 borders are "dirty" because neither it does properly determines the image size).

I suspect that D600 has more noise than the one reported in this comparison, but anyway I also suspect that it catchs (if not betters) the 6D as well in high ISO performance, specially at image level. But 6D (and 1DX / 5D3) have sensors good enough for great Canon glass, which is what really matters... by the moment

(Canon, please wake up!).

Dec 04, 2012 at 04:04 PM

cgarcia
Offline
•
Upload & Sell: Off

p.2 #7 · 5D3, 6D and D600 true RAW (dcraw) comparison

mmurph wrote:
Thx! And welcome to FM!

Michael

Yes, I apologize for having forget saying hello in my first post on this forum!. I'm in vacations with lots of time for toys and feel the urgency to share my findings, once I managed to run dcraw on the 6D RAWs.

I'm a amateur... and keep reading this forums for some time, which are very interesting. Currently I own a 400D (how old!), a EF-S 10-22 and the 70-200 F4L IS. I plan to upgrade soon to FF, mainly for the 17mm TS-E (many times I shot in wide angle a building, I calculate how much it could improve by shifting certain amount the frame up!). I also would love having another lenses, specially a fast prime (perhaps the Sigma 85/1.4?) but I decided a lot of time ago not to buy any more FF lenses without having a FF camera: I had a bad experience with a 10-20 sigma, with a soft side, and a crop camera only shows 38% of the FF image circle (plus the focal length crop factor changing the intended focal lengths). In addition, I decided not to invest in crop lenses once I realized that I will end in FF. Now, I desperately need a FF camera!

Another lens I'm interested is the 24-70 F4L IS, once the reviews come up. Despite the cold reception (well, specially in sites like canonrumors, not so much here in FM) such a lens is a one a lot of time ago I wished (now that stupid Canon has readed my mind!)

A balance in weight and image quality (if properly built). Equivalent in all respects to 17-55 F2.8 IS in crop cameras (depth of field, effective light gatering, weight) and with a better range for me. I tried a 17-55 a lot of time ago and didn't like it mainly because too much flare with every night light (and at least that copy not too sharp at 17mm). The 24-105 is a too much ambitious design, and the 24-70 2.8 is heavier and anyway not so fast as a 1.4 or 2.0 prime. I think that 24-70 F4 IS could be a good landscape and general purpose high quality lens... am I right?. In Nikon there is a nice 16-35 VR for landscapes as well as interiors, and between 24-28mm seems to match primes stopped down (Canon 17-40 has no IS, and using polarizers or when entering at church/museums its absence is a minus). Oh yes, I hate tripods :-).

Dec 04, 2012 at 05:38 PM

Ralph Conway
Offline
• • • •
Upload & Sell: Off

p.2 #8 · 5D3, 6D and D600 true RAW (dcraw) comparison

Nice. Looks like the D600 overdrives even my fanboyish expectation ...

Edited on Dec 04, 2012 at 08:26 PM · View previous versions

Dec 04, 2012 at 06:10 PM

abqnmusa
Online
• • •
Upload & Sell: On

p.2 #9 · 5D3, 6D and D600 true RAW (dcraw) comparison

not sure this proves anything about the cameras.
It does show how the program is coded to convert RAW images from each camera.

Dec 04, 2012 at 06:19 PM

Ralph Conway
Offline
• • • •
Upload & Sell: Off

p.2 #10 · 5D3, 6D and D600 true RAW (dcraw) comparison

Uh, I forgott:
Thank you very much. You did a great job!
Ralph

Dec 04, 2012 at 06:46 PM

Wahoowa
Online
• • •
Upload & Sell: On

p.2 #11 · 5D3, 6D and D600 true RAW (dcraw) comparison

abqnmusa wrote:
not sure this proves anything about the cameras.
It does show how the program is coded to convert RAW images from each camera.

Not sure what you mean, but these are not jpg files from the cameras.

Dec 04, 2012 at 06:54 PM

carstenw
Offline
• • • • • •
Upload & Sell: Off

p.2 #12 · 5D3, 6D and D600 true RAW (dcraw) comparison

cgarcia wrote:
Another lens I'm interested is the 24-70 F4L IS

[...]

Equivalent in all respects to 17-55 F2.8 IS in crop cameras (depth of field, effective light gatering, weight)

I don't see how an f/4 lens on FF is "equivalent in effective light gathering" to an f/2.8 lens on 1.6x crop? Light is light, no?

Dec 04, 2012 at 07:03 PM

Marsu42
Offline
•
Upload & Sell: Off

p.2 #13 · 5D3, 6D and D600 true RAW (dcraw) comparison

Fyi: There are some good comparisons of 5d3 vs 6d noise over at CR with a pre-release raw converter of CaptureOne ... the 5d3 shows more chroma noise, but since that's easily removed there really only a noise difference @iso12800+ ... see here: http://www.canonrumors.com/forum/index.php?topic=11279.msg202200#msg202200

And the bad news is that the high iso advantage of the 6d might be a tradeoff against softer images (look closely!) because Canon starts denoisesing raw files at lower iso settings than on the 5d3... that'd be the explanation for the chroma noise difference.

About more dynamic range on the 6d I'm very sceptical, if it'd be that easy to increase dr through slower readout and less fps Canon would have surely done that before?

Edited on Dec 04, 2012 at 07:22 PM · View previous versions

Dec 04, 2012 at 07:11 PM

JohnBrose
Offline
• • •
Upload & Sell: On

p.2 #14 · 5D3, 6D and D600 true RAW (dcraw) comparison

I think cgarcia was talking about the dop is similar from 2.8 on full sensor ot f4 on cropped sensor?

Dec 04, 2012 at 07:21 PM

kapytalyst
Offline
• • •
Upload & Sell: Off

p.2 #15 · 5D3, 6D and D600 true RAW (dcraw) comparison

The definitive test of whether or not the 6d has a better sensor than the d600 will come from from DxO Labs.

If the Nikon is better, we will have the results from them momentarily.

If the Canon is better, we will have the results from them sometime in April.....

Dec 04, 2012 at 07:27 PM

mmurph
Offline
• • • •
Upload & Sell: On

p.2 #16 · 5D3, 6D and D600 true RAW (dcraw) comparison

cgarcia wrote:
Another lens I'm interested is the 24-70 F4L IS, once the reviews come up.

That looks like it will be a very nice lens, assuming it performs to expectations. Canon has a very good track record with their recent lens updatyes. In additiona to the higher optical and build quality than previous models, the focusing is also much more precise on the newer camera bodies.

If you are able to get it in a kit at a good price it would definitely be a good purcahse. I think Lensrentals.com will likely have the first "authoritive": say whenm they get their copies.

My version I 24-70 is my favorite lens ever. I am still hoping to get a version II at a good price when I cvanb afford it.

Otherwise you might look to buy a copy of the 24-70 4.0 IS that is coming from a kit that gets split (by enbd user or eBay seller), or you might have to wait 6-9 months to find it at a reasonable price.

There is a cost penalty for being an early adopter, though it is much less on a lens than on a camera.

Thanks for sharing your comparison on the 6D! Whaterver the feedback you get here, it is something alot of people having been anxious to see.

You did a very, very good job in format and presntation. I know that was a lot of work on your part! Thanks again!

Best, Michael

Dec 04, 2012 at 07:46 PM

gdanmitchell
Online
• • • • •
Upload & Sell: Off

p.2 #17 · 5D3, 6D and D600 true RAW (dcraw) comparison

One thing that it proves is that all three cameras produce excellent results and with normal post processing workflow one can shoot at high ISOs and make lovely big prints with images from any of them. :-)

Dec 04, 2012 at 07:53 PM

MayaTlab
Offline
•
Upload & Sell: Off

p.2 #18 · 5D3, 6D and D600 true RAW (dcraw) comparison

kapytalyst wrote:
The definitive test of whether or not the 6d has a better sensor than the d600 will come from from DxO Labs.

DXO is a useful tool, but it certainly isn't foolproof. Three illustrations, as examples :
1) : they rate my D700 as being better at high ISOs than my 5DIII. I've got only one answer for that : ha ha ha. Even at 100%, disregarding the increased mp count, the 5DIII is slightly better than the D700.
2) DXO doesn't measure amplification noise, or colour shifts at higher ISOs, which are a pain to correct. If you were to look at their graphs, you'd believe the D800 is as good as the D4 at high ISOs. That's not the case : above 6400 ISOs, it's got quite a bit of amplification noise, while the blacks on the D4 remain black.
3) It doesn't tell you anything about banding.

At least it attracts attention to the importance of DR at low ISOs, and for that, I thank them. I also salute the effort. But it's not very precise.

in my opinion, opening files made in reasonably similar circumstances and looking at them on your own computer probably is the best way to make up your mind about sensors' performances.

Dec 04, 2012 at 07:58 PM

Ralph Conway
Offline
• • • •
Upload & Sell: Off

p.2 #19 · 5D3, 6D and D600 true RAW (dcraw) comparison

MayaTlab wrote:
in my opinion, opening files made in reasonably similar circumstances and looking at them on your own computer probably is the best way to make up your mind about sensors' performances.

+ 100

Dec 04, 2012 at 08:07 PM

StillFingerz
Offline
• • • •
Upload & Sell: On

p.2 #20 · 5D3, 6D and D600 true RAW (dcraw) comparison

gdanmitchell wrote:
One thing that it proves is that all three cameras produce excellent results and with normal post processing workflow one can shoot at high ISOs and make lovely big prints with images from any of them. :-)

+1

Dec 04, 2012 at 08:10 PM

1 2 3 4 5 end
FM Forums \| Canon-mount SLRs \| Join Upload & Sell 1 2 3 4 5 end