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

Thanks Allen this will be something to watch for if and when I upgrade my camera.

Mike

Thank you for trying, but you are NOT looking at the fabric I am drawing attention to. As I said I am choosing to post process to try to get all of the ISO samples to match the 5DmkIII sample as I feel this is the best sample of the bunch. You think that there is a favor nod going to the 7DmkII and if I were to use this clip I would some what agree. I want you to look at the fabric in question at ISO 100. I think you have agreed with me some what but I would like you to try to PP each sample to get the best from each body. Showing equal levels of processing for two entirely different approaches by Canon is less than ideal.In fact it is unfair.

In other words the Canon 7D RAW straight out of the camera looks bad but after some balanced PP it is almost the same @ ISO 100 as the 5DmkIII. Where as the 7DmkII and the 70D are wacked.
For the record the rest of the image looks very similar, it is when we look at the red issue on the fabric and the lower detail as well that it becomes VERY clear.

I can look past the subtle lighting changes over time and the different shooting postion/angle too. I seriously doubt these are why the red fabric is wacked. As I said before I think it is a DPP issue. But if it is not then Canon has a bigger issue on their hands.

I also find USM to work FAR BETTER on the 7D files than the basic sharpness filter. You seem to disagree. I am using a calibrated NEC monitor it is 1920 X 1200 and 26".

I also think you used a high ISO image I am using ISO 100 only. This is comparing, for now, best case images of all bodies. If the detail and accuracy is not here at ISO 100, it certainly won't be found at higher ISO's.
Using your picture crop and a ISO 100 image you can see in the wood bar more detail in the 7D image than the 7DmkII's.

I have also found that when I tweek an individual area for maximum detail and noise it is not balanced when I move to another area. So to be fair I must consider the entire image for a balanced approach and not favor one area over another. If one uses too much sharpening you will see white outlines around objects and the sharpening must be reduced to eliminate this artifact. When I do this the 7DmkII is less sharp than the 7D every time.

I think in order to get a less noisy image Canon has elected to soften the image. And I can see this as long as take care to work with the entire image and not try to prove my point in one small section.

2n10
i cannot help but feel that IF Canon is messing with the red channel that it will also impact the rest of the image. John please down load the ISO 100 RAW's of the 5DmkIII and the 7D and the 70D and the 7DmkII. You will see what I mean. Then wait for Imaging Resource to update the 7DmkII images with a production camera's files. If it is gone then you can decide if you want one or not. If it is there and the image overall is softer too you may find the 7D to be the better camera and price.

Really the 7DmkII is all about features and IMO they are very impressive but if at the end of the day iimage quality has a important place too, then the 7D may be the better body. Since I already own the 7D for me the features are more than fine. I will for now keep my $$$ in my wallet. YMMV

Mike glad it was of benefit to you.

AllenF wrote in post #17227775
2n10
i cannot help but feel that IF Canon is messing with the red channel that it will also impact the rest of the image. John please down load the ISO 100 RAW's of the 5DmkIII and the 7D and the 70D and the 7DmkII. You will see what I mean. Then wait for Imaging Resource to update the 7DmkII images with a production camera's files. If it is gone then you can decide if you want one or not. If it is there and the image overall is softer too you may find the 7D to be the better camera and price.

Really the 7DmkII is all about features and IMO they are very impressive but if at the end of the day iimage quality has a important place too, then the 7D may be the better body. Since I already own the 7D for me the features are more than fine. I will for now keep my $$$ in my wallet. YMMV

I'll need to look at them at home. I am wondering if the lens has anything to do with it since it is new?

I will admit I have not seen any issues on any of the real world samples nor seen anything about the issue you have found.

I am reading a LOT of posts reacting to the thread title that seem to be uninterested in reading what the author posted.

???

EDIT, Sorry on to page two. I'm the one that needs to read the whole THREAD before replying!

IMHO, the only conclusion we can draw from Alans test, is that currently DPP is having a problem with the red channel from the 70D/7DII sensor.

If in fact DPP is rendering these reds with less detail than from other bodies, I would not conclude that this = "worse all together"

I haven't noticed odd red channel results from my 70D but I don't use DPP. I use Lightroom 5 and Photoshop CC along with the Nik Collection.

OK, I redid my samples and turned everything off. I selected Neutral picture style and selected unsharp mask. Auto light optimizer is off or unchecked. The 7D (70D is too) is checked by default you must uncheck it. I applied the following to the 3 different images all ISO 100 raw from the 5DmkIII,7D,7DmkII. the following settings are

5DmkIII USM
Strength (1)
Fineness (10)
Threshold (0)

7D USM
Strength (3)
Fineness (10)
Threshold (0)

7DmkII USM
Strength (3)
Fineness (10)
Threshold (0)

Overall they look very similar with the 5DmkIII having the most detail followed by the 7D and then the 7DmkII but now much closer. However there still is the red fabric issue although now much less so than before.

Noise is the best in the 5DmkIII but to be fair it is pretty close between the 5DmkIII and the 7DmkII the 7D is a bit behind. If I were to guess I think the noise removed from the 7DmkII is what is also removing the detail. I could sharpen the 7DmkII image more but it would add a subtle halo that would be wrong.

What about the 70D?
Strength (3)
Fineness (10)
Threshold (0)
With these setttings it is the least sharp. It is better on the red fabric than the 7DmkII but worse than the 5DmkIII or the 7D. Now seeing these RAW's with the fully zeroed out DPP settings I have to say that I think the 7DmkII in final production form will be pretty close to the 7D but a tick behind in detail. It will have better noise handling but at the expense of detail. I hope Canon can fix the red channel issue in DPP. If not, it to me is a deal breaker. If Canon can match the detail Of the 7D AND fix the red channel issue I might be tempted to pull the trigger but it is an $1800 might.

I have to thank teamspeed for helping me see that even in RAW DPP is really mucking up the picture quality. I assumed that RAW processing was clean by default. Boy was that a mistake. You have to go in and select neutral picture style and zero out all the sharpness settings first and then proceed to use USM to get a high quality result. Using the 5DmkIII file to compare against is extremely helpful. Actually it is a must IMO.

I guess there is still one question I need to find an answer to. What lens is Imaging Resource using on each of these bodies? I would hope it is the same one perhaps the 100 macro or it's L version. I think that would be my choice.

CyberDyneSystems wrote in post #17227873
IMHO, the only conclusion we can draw from Alans test, is that currently DPP is having a problem with the red channel from the 70D/7DII sensor.

If in fact DPP is rendering these reds with less detail than from other bodies, I would not conclude that this = "worse all together"

In the title I said for now. I fully understand that PP is a fluid moving target. FOR NOW there seems to be an issue. I hope Canon fixes it. And IF I had a hand in getting it done so much the better

I also think folks here would like to know this as well and not plunk down $1800 and find out after the fact.

With all this effort to dissect a yet to be released camera, I can't imagine what the proctology exams and reports are going to look like in a mere 9 days.

AllenF wrote in post #17227941
OK, I redid my samples and turned everything off. I selected Neutral picture style and selected unsharp mask. Auto light optimizer is off or unchecked. The 7D (70D is too) is checked by default you must uncheck it. I applied the following to the 3 different images all ISO 100 raw from the 5DmkIII,7D,7DmkII. the following settings are

5DmkIII USM
Strength (1)
Fineness (10)
Threshold (0)

7D USM
Strength (3)
Fineness (10)
Threshold (0)

7DmkII USM
Strength (3)
Fineness (10)
Threshold (0)

Overall they look very similar with the 5DmkIII having the most detail followed by the 7D and then the 7DmkII but now much closer. However there still is the red fabric issue although now much less so than before.

Noise is the best in the 5DmkIII but to be fair it is pretty close between the 5DmkIII and the 7DmkII the 7D is a bit behind. If I were to guess I think the noise removed from the 7DmkII is what is also removing the detail. I could sharpen the 7DmkII image more but it would add a subtle halo that would be wrong.

What about the 70D?
Strength (3)
Fineness (10)
Threshold (0)
With these setttings it is the least sharp. It is better on the red fabric than the 7DmkII but worse than the 5DmkIII or the 7D. Now seeing these RAW's with the fully zeroed out DPP settings I have to say that I think the 7DmkII in final production form will be pretty close to the 7D but a tick behind in detail. It will have better noise handling but at the expense of detail. I hope Canon can fix the red channel issue in DPP. If not, it to me is a deal breaker. If Canon can match the detail Of the 7D AND fix the red channel issue I might be tempted to pull the trigger but it is an $1800 might.

I have to thank teamspeed for helping me see that even in RAW DPP is really mucking up the picture quality. I assumed that RAW processing was clean by default. Boy was that a mistake. You have to go in and select neutral picture style and zero out all the sharpness settings first and then proceed to use USM to get a high quality result. Using the 5DmkIII file to compare against is extremely helpful. Actually it is a must IMO.

I guess there is still one question I need to find an answer to. What lens is Imaging Resource using on each of these bodies? I would hope it is the same one perhaps the 100 macro or it's L version. I think that would be my choice.

Turn the fineness slider down to 1 or 2.. And the threshold up to 1 or 2. Huge difference. More so on the crop files.

speedync
I did as you suggested. It was better for noise but the detail suffered quite a bit. For detail I am sticking with what I have.
Thank you for your input. I look forward to everyone's ideas here.

Lt Colonel wrote in post #17227991
With all this effort to dissect a yet to be released camera, I can't imagine what the proctology exams and reports are going to look like in a mere 9 days.

The files we have available now aren't going to be any different once the rest of the globe gets their copy in a week.

CyberDyneSystems wrote in post #17227852
I am reading a LOT of posts reacting to the thread title that seem to be uninterested in reading what the author posted.

???

EDIT, Sorry on to page two. I'm the one that needs to read the whole THREAD before replying!

That's it, it's timeout for you!