7D mk II image quality RAW worse than 7D for now

I have been doing a little comparing and thought this was interesting. I down loaded from Imaging Resource their RAW sample test images. These were the ones with zero noise reduction applied. I got 5DmkIII, 7D, 7DmkII, and 70D samples. This would allow me to compare them against each other. I then updated Canon's DPP to the latest version so as to be able to process the RAW's from all 4 cameras.

I used the samples from the 5DmkIII as the standard by which the others must meet knowing that in all likely hood they would not get there. No surprise there but at ISO 100 the 7D was really pretty good. In fact it was very close in detail. I also found the 5DmkIII would go from ISO 100-800 with zero lose of detail and even 1600 was almost too close to call. IMO ISO 3200 is perfectly fine for the absolute perfectionist.

Overall the image quality was very nice from all of the bodies. However there appears to be a flaw in the DPP software where the 70D and the 7DmkII are concerned. If you look at the fabric sample wheel and in particular the red fabric with the darker leaf pattern you will see that as the software is processing/loading the image, it starts out fairly sharp but once the image is fully loaded it just pops on the red fabric into a very blurry mess. It does it for both the 70D and the 7DmkII. However the 5dmkIII and the 7D are sharp and detailed. Also look at the fabric texture, that is the the weave of the cloth. The detail lost in the pink fabric as well as the dark green fabric on the 70D and 7DmkI is also noticeable. You can also compare the settings for shadow levels on the black fabric with the lighter curling pattern. These fabrics and their subtle patterns allow you to set shadow and highlight levels and also contrast. The white fabric with the leaf patter that looks like it is embossed is good for the highlight level. Just be sure to leave it a bit blown out as I think this looks a bit more natural. Season to taste.

I hope it is a DPP software glitch. But even when the 70D/7DmkII image is loading, and it starts out sharper and then goes blurry, it is still less sharp then the 7D.
I think the 7DmkII could be sharper but DPP is not letting this happen so we will have to wait.

This is the issue with software needing to catch up to the hardware.

That being said the 70D RAW should be absent of this issue as it has been around for a while now. I have not compared this update to the previous one so perhaps it is a recent development in the latest software version I don't know.

Please try it for yourself before you pile on me. It is super easy to see.

Many would perhaps say that I am being far to critical. I disagree. The red channel issue is major IMO. I also recognize that we all will be doing some cropping. It is the nature of wildlife and sports and also Motorsports. I love airshows. Cropping is a must for me. Nature/wildlife too. So lost detail will become more noticeable on a cropped image, and corps close to 100% are very likely. With the 7D I do not have the issues I am seeing in the 70D/7DmkII. Until Canon or someone else figures out how to overcome these issues I am out of the market.

One thing I have learned from this test is that I can take these test images and establish a repeatable formula to apply/set all the major parameters for each ISO and know with reasonable certainty that the image will be dialed in. Using the 5DmkIII as my best case standard helps me compare other bodies to what is the best and this allows me the opportunity to tweek them and get them to be their best and also to better understand their weaknesses as well.This establishes reasonable expectations of how things will turn out.

You can also see how Canon has messed with the sharpness a bit to reduce noise. IMO RAW should not be messed with like this. I should be the one to decide how much messing I want to do. Canon leave the RAW alone. If you must, play with the JPEG's but again at zero noise reduction you should leave even the JPEG alone.

Perhaps Adobe and other software will be better. I am not spending money on software for only a test. I will leave this to others to try and report back.

In conclusion I have come away with a deeper appreciation for my 7D. I have found that properly dialed in, and where I must do so, ISO 800 is very good. I still think ISO 400 is where I am absolutely sure I will not be disappointed.

I found it interesting that Canon did not tell us one word of the 7DmkII having improved image quality. For once they know to be quiet. IMO, as of today, the IQ of the 7DmkII is slightly to fairly worse than the 7d. The red channel is wack . If they can get a handle here then I will need to re-due the test samples and compare. This in fact could lead me to a different conclusion.

For ME and MY needs if the image quality is a tie or worse than the 7D I will not be getting a 7DmkII. I don't need 10fps. I don't need better build as I have had zero issues with the current 7D build quality. I don't need a bigger buffer either. I even don't need 64 AF points over the 19 I have now. In other words, what I do need is better, not softer, higher ISO performance.

I think the only way I will get this is in a 5dmkIII or perhaps 5DmkIV.

For now I can afford to wait.

If I get the urge to upgrade I think it will be glass. I am really wanting a 300 2.8 L telephoto. Just too far out of reach and the 7D is not as resolving of detail as this lens deserves. But it is better than the 100-400L I have now.

Popcorn in hand, I look forward to your INFORMED comments.

I just think it is premature to make this level of analysis prior to the camera shipping. Some very interesting considerations for when the camera is available for more intensive reviews though.

Yes I see your point but the 70D which the 7DmkII is based on has a serious red channel issue with the current DPP software too. I agree that the finished body in the hands of folks will tell the real tale but the 70D is a real concern. Don't you think?

I believe that is the first time someone has said they thought the 7D2 images mentioned were worse than the 7D images.

Please down load the Image Resource images yourself and tell me what you see. All you need are the ISO 100 RAW with zero noise reduction and the latest version of Canon's DPP. That is 4 ISO RAW's and you will know what I am seeing.

There is no way for there to be NO NR applied to the raw, if you used DPP. You have to go to the NR tab and zero them out for all samples. Just because the camera setting says "Disabled" for NR, there are still NR values being set for each camera, and each model is set to a different 2 values. What you are seeing in that "gap" from initial load to final view is DPP applying all the filters, picture styles, and NR based on the raw values.

Or go to your DPP preferences and set the NR option there to no honor raw camera settings...

I have already done all of this in one of my mini reviews (link in my signature).

TL;DR: You had unrealistic expectations that a 7D mark 2 would be as good as a 5D3 and you made conclusions about image quality based on unsupported RAW developers and flawed tests.

tkbslc wrote in post #17227340
TL;DR: You had unrealistic expectations that a 7D mark 2 would be as good as a 5D3 and you made conclusions about image quality based on unsupported RAW developers and flawed tests.

DPP supports the 7D2, Canon released that on 9/29.

David Arbogast wrote in post #17227315
I just think it is premature to make this level of analysis prior to the camera shipping. Some very interesting considerations for when the camera is available for more intensive reviews though.

The cameras have already shipped, there would be no way for final retail outlets to be able to ship on 10/30 to customers without all the cameras already being in various supply chain stages, air or ship transport, etc.

Have the cameras changed from when IR did their samples? I doubt it, but one cannot be 100%.

Folks PLEASE down load the RAW samples and THEN comment, if possible. If I did it wrong you will be able to see and correct me from fact not hunches. I hope I am wrong but I tried everything to get the 70D/7DmkII images to be better and the issue would not go away. The fact that the 5DmkIII and the 7D were fine speaks volumes IMO. Also I tweeked the images in DPP so as to get a real world not a straight out of the camera pic. I PP every pic I take . The 7D OOTC is not pretty. But with a gentle massaging in PP it is very nice indeed.

AllenF wrote in post #17227357
Folks PLEASE down load the RAW samples and THEN comment, if possible. If I did it wrong you will be able to see and correct me from fact not hunches. I hope I am wrong but I tried everything to get the 70D/7DmkII images to be better and the issue would not go away. The fact that the 5DmkIII and the 7D were fine speaks volumes IMO. Also I tweeked the images in DPP so as to get a real world not a straight out of the camera pic. I PP every pic I take . The 7D OOTC is not pretty. But with a gentle massaging in PP it is very nice indeed.

PLEASE read my post. There is NO WAY to get zero NR in the raw file using DPP unless you either change your preferences, or change the sliders on the 3rd tab. I know a bit about DPP and raw files...

Go back, change your DPP preferences to ignore NR (or default to 0,0), then redo your raw files.

Here is my post where I did 2 different samplings, produced a large image with the different ISO comparisons, then later in the post, I show the initial work for my NR actions for the 7D2 and what it can do for the higher ISO values. The bigger difference with the IQ from the 7D vs 7D2 was really in the sharpening methods employed by the different generation cameras/DPP. Zero out the NR and zero out the sharpening for both, and check out the differences, that is where the 7D2 shows virtually no difference from the 7D, but that would raise a stink.

https://photography-on-the.net …305&highlight=m​ini-review

The samples I used were the pre-release images. Once Imaging Resource puts up production camera images I will look at these. My contention is with the 70D image doing the same thing as the 7DmkII. I have to think as I noted in my initial comment/observation that if we see it in the 70D and the 7DmkII then we have at the very least and issue with the DPP software. At worst we may have a bigger issue.

TeamSpeed wrote in post #17227346
The cameras have already shipped, there would be no way for final retail outlets to be able to ship on 10/30 to customers without all the cameras already being in various supply chain stages, air or ship transport, etc.

Have the cameras changed from when IR did their samples? I doubt it, but one cannot be 100%.

By "shipped" I meant actually in buyers/users hands, not shipped to retailers. As one who buys gear solely on line all gear has to get shipped to me as the user. I totally forget that some people actually buy from brick and mortar retailers.

AllenF wrote in post #17227369
The samples I used were the pre-release images. Once Imaging Resource puts up production camera images I will look at these. My contention is with the 70D image doing the same thing as the 7DmkII. I have to think as I noted in my initial comment/observation that if we see it in the 70D and the 7DmkII then we have at the very least and issue with the DPP software. At worst we may have a bigger issue.

Again, the NR for the 7D2 is different than the 7D image, and you are seeing the impacts of DPP performing noise reduction, and also different sharpening techniques. Read my post, I already went through the analysis. The 7D employs simple sharpening, the 7D2 employs USM, 2 different algorithms, and the NR sliders at the same ISO levels when NR is set to disabled show different values from each other.

Teamspeed
Thank you for the suggestion. I did as you asked and while an improvement occured the image compared to the 7D was still messed up. This is the case for the 70D AND the 7DmkII.

Please try it for yourself and then report back here. If I did screw up perhaps you will show me where and all will learn from my putting my foot in my mouth.