APS-C Image Quality (7d vs nex7)

[briansquibb]

Converted in Capture One 6

what we know is that the 7D needs the user to apply some intelligence and discernment in his choice of conversion software

I have to agree RAW + Capture One 6 is like switching to L lens all over again. Simply amazing the improvement.

Is capture one really that awesome?? im using adobe PS5 and camera raw and loving it, but if there is something im missing in raw conversion please explain. I never understood DPP anyways unless you didnt have photoshop..

I do most of my RAW processing in DPP. It is a surprisingly capable piece of software - weakness is that it hides many of the usefull functions.

Capture One seems to have been designed along the same lines as DxO - basically a workflow/batch app.

 

[neuroanatomist]

neuro Thank you so much. I could have sworn you were dead wrong but I just tested it and paid attention to what you said and indeed, it's the re-framing (even though it's not that extreme of a physical move) that gives the artistic change I love in the 16-35 (going from wide to standard visually). I used a 16-35 and tried without any re-framing, holding the camera stationary, and there was no change in perspective distortion, it was just closer or further away. I was under completely wrong impression. Thank you very much, that will be useful to have a better understanding of what is actually happening. cheers

It's a misconception that most people have, actually - so, you were in good company.

 

[neuroanatomist]

I'm curious about this as well. I'm not sure I want to pop for another piece of software, especially since I haven't found Lightroom to be all that beneficial over Photoshop (both use the same RAW processing, just different interfaces) Can someone who has used both please quantify what, if any, advantages other RAW processing software might have over Photoshop? (I'm not talking about image editing, I'm talking about RAW processing and some concrete examples would be helpful)

Great question - and one that gives me an excuse for a simple test.

I expect the primary difference in RAW converters is how they handle noise and color, and secondarily, how they handle lens corrections. For example, one reason I prefer DxO to ACR is that the former bases the lens corrections (distortion, vignetting, etc.) on controlled lab testing, whereas ACR's lens profiles are sometimes based on user-submitted results (subject to variation in lighting, etc.).

For a comparison, I picked a test image with a fair bit of noise - an ISO 3200 shot from the 7D. This shot simulated real-world high-ISO use, i.e. I used ND filters to reduce the illumination, rather than a fast shutter speed in bright light (as is commonly used in ISO noise testing, and results in lower read noise that isn't necessarily reflective of real-world shots). Shot was with the 7D and EF 70-200mm f/2.8L IS II @ 85mm, 1/60 s, f/5.6.

The same RAW image was converted to JPG with Canon's own DPP, DxO Optics Pro, Adobe Camera RAW (in CS5), and Apple Aperture 3. The appropriate camera and lens correction modules were used where available as part of the converter (i.e. all except Aperture 3, where you need a separate plugin).

Unfortunately, different RAW converters use different algorithms and the relative settings mean different things in different programs - so, there's really no way to meaningfully compare across settings. The approach I took was to assume that the developers of the software know their software best, and chose default settings optimal for the camera and ISO setting - obviously, they're going to be biased to what the developers think a good image looks like, but at least it's a baseline for comparison. For DPP default, I used the Standard Picture Style (the 'no adjustment' image used Neutral).

Here's what the output looks like (click for larger, up to 1600 pixels with View All Sizes, which when viewed at full size are 100% crops):



Personally, my order of preference is DxO > DPP > Aperture > Adobe Camera RAW. ACR looks grainy and oversaturated to me, and with Aperture the contrast is on the low side. Again, these are just the default settings - with specific, image-dependent adjustments better results are certainly possible with any of the packages.

 

[briansquibb]

Personally, my order of preference is DxO > DPP > Aperture > Adobe Camera RAW. ACR looks grainy and oversaturated to me, and with Aperture the contrast is on the low side. Again, these are just the default settings - with specific, image-dependent adjustments better results are certainly possible with any of the packages.

Interesting - I use DPP for RAW processing and DPP seems to have least noise - and I have been saying that the 7D at 1600 is fine, whereas others have been complaining of noise.

 

[neuroanatomist]

Personally, my order of preference is DxO > DPP > Aperture > Adobe Camera RAW. ACR looks grainy and oversaturated to me, and with Aperture the contrast is on the low side. Again, these are just the default settings - with specific, image-dependent adjustments better results are certainly possible with any of the packages.

Interesting - I use DPP for RAW processing and DPP seems to have least noise - and I have been saying that the 7D at 1600 is fine, whereas others have been complaining of noise.

It's a quality thing. DPP's noise is coarser/blotchier, and comes at a cost of more reduction in sharpness.

 

[briansquibb]

It's a quality thing. DPP's noise is coarser/blotchier, and comes at a cost of more reduction in sharpness.

I guess it is a compromise thing then - take a really sharp picture and lose a little of that for some visual reduction in noise.

For smaller images (up to A4 print) I would guess then the loss of sharpness would not really show, whereas ugly noise would. If the image with worse noise was put through NR software then there would be loss of IQ. So bottom line is that it probably makes little difference which path is taken in the real life situation - just shows the importance of good glass.

 

[dr croubie]

http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/%28appareil1%29/736|0/%28brand%29/Sony/%28appareil2%29/619|0/%28brand2%29/Canon (that link should bring up a nex7 vs 7D direct comparison)

The results are in, thankyou DXOMark.
Noise at 100% ('screen') is neck and neck, nex wins in low, 7D wins in mid, nex wins at high. But only just. I wouldn't decide between them on that factor alone. Resizing to the same size, ie 'print', nex has the slight advantage over 7D for the whole range, but still only just. It's nowhere near 5D2 quality (which you wouldn't expect from an aps-c anyway).
Sony have always been better at the DR, nex wins easy in the low and high ends, in the middle-iso the 7D is equal at equal sizes, better at 100%.
Tonal range is too close to call at 100%, nex just wins for print sizes (again, the advantages of shrinking higher MP).
Colour sensitivity the nex wins across the board at print sizes, and at high/low iso at 100%, the 7D just touches it at mid-range ISO.

Certainly interesting, if this gives the indication of performance of the D300s replacement, and the A77 (does the a77 have fixed pellicle mirror? If so that'll reduce the usable iso because of light-loss in the mirror). The nex may not be able to touch the 7D for fast contrast autofocus, but in a DSLR or DSLT it can have a fight on its hands.
(now i'm very interested in a nex, seeing as it can take a whole lot of other rangefinder lenses like Leica M that can't go on a dslr. Too bad i can't afford leica M lenses either though...)

Still, more competition = good. Bring on the 7D2 - nex7/d400/a77-killer