Analysis of RAW samples at Fred Miranda show weak DR

[Luds34]

Just to stir the pot a bit more...

Today was another local hands-on with PREPRODUCTION 6D II camera (just a single one unfortunately for 20 people, that was like tentacle hentai, really...). I've had a little chat with Canon rep. He was not allowed to discuss details (NDA is still going), but he basically admitted that 6D II should be on par with 5D IV, sensor wise. There were some hints, that he is not allowed to discuss certain goodies about sensor - that really intrigued me. But in the end, I was told to wait for retail samples, which is what we all are doing I guess.

So I really hope he was right and we're about to be pleasantly surprised after this preliminary mumbo-jumbo. And if he's wrong, then Canon will disappoint me gravely and I'll be the first here to admit I was wrong and will openly criticize Canon for such move in 6D line...

I suppose that is encouraging. But as others have said in this thread (and else where) already, all of Canon's latest cameras/sensors all appear to have moved to the new fabrication process. It doesn't make sense on any level for them to not do the same with this camera.

Either way, my pre-order is in place and I'm hoping it ships slightly earlier then expected and I arrives before my canoe trip at the end of the month. Com'on Canon, get some early stock out!

 

[fentiger]

In daytime, I'm usually in the 30-60 s range. It really depends on the scene and the people. At blue hour, around 3 min, but the biggest problem there is cyclists with lights.
[/quote]

Cyclist+Lights!!!! thats an oxymoron Here in England them two words in the same sentence don't exist, especially in Cambridge.

 

[ahsanford]

I didn't have time to play with my ND toys "in city", but it seems like it should take at least 4 minutes to completely eliminate people walking in the frame. How do you fight thermal noise then?

In daytime, I'm usually in the 30-60 s range. It really depends on the scene and the people. At blue hour, around 3 min, but the biggest problem there is cyclists with lights.

I've also heard you can use the Median Filter function in PS to eliminate folks with multiple exposures. Effectively, this method only outputs what is visible in every shot, so if you do it over an appreciable period of time (say 5 minutes), unless someone is sleeping or sitting/eating in frame, you'll only see the landmark and not the people around it. Apparently, it's a decent hack for really touristy areas where you never can pull everyone out of frame.

Of course, I'd imagine you'd need very consistent light to do that, so twilight/sunset or changing street lights could lead to some weird Median filter output. But I've never tried it myself -- has anyone here?

- A

 

[Jopa]

100% agree. Now you can even simulate shallow DoF in PP, so no need to use fast lenses anymore The only problem - it takes a lot of time to process the image correctly otherwise it'll look like cheap fake

Not true, just look at some of the tutorials on the link I provided. A ND Grad is a sledgehammer, great if you take pictures of railroad spikes, far too limited for much else.

Frankly I don't remember I said anywhere GND will cure AIDS, ebola and cancer It's doing what it's supposed to do - gradually reduce light. As long as it can also reduce processing time - it's a tool worth using. Same as camera DR, strobes, light meters, and a million of other photography tools. One can re-create and render a whole scene in 3ds max, but it will probably take longer compared to taking a picture.It's simply not correct to say GNDs suck only because they can't be used absolutely everywhere.

 

[Jopa]

I didn't have time to play with my ND toys "in city", but it seems like it should take at least 4 minutes to completely eliminate people walking in the frame. How do you fight thermal noise then?

In daytime, I'm usually in the 30-60 s range. It really depends on the scene and the people. At blue hour, around 3 min, but the biggest problem there is cyclists with lights.

Along with ND filters major photography stores should also offer traffic cones and signs "No Cyclists with Lights" and "Cycling? Turn off your lights!"

 

[ahsanford]

Frankly I don't remember I said anywhere GND will cure AIDS, ebola and cancer It's doing what it's supposed to do - gradually reduce light. As long as it can also reduce processing time - it's a tool worth using. Same as camera DR, strobes, light meters, and a million of other photography tools. One can re-create and render a whole scene in 3ds max, but it will probably take longer compared to taking a picture.It's simply not correct to say GNDs suck only because they can't be used absolutely everywhere.

Some people absolutely have no patience for compositing/masking, HDR tools, etc. while others hate fiddling with grads or shooting non-ideal horizons with them. I see both as tools in our toolboxes that can be brought to bear as circumstances dictate.

I'm neither a post-processing junkie nor a classic 'get it right in-camera' person. People should use what they like and feel more comfortable with.

Personally, though, I am a frantic day-to-day photographer as my family is in constant motion, so I never take enough time to do things properly (I have terrible habits like relying on Auto ISO, avoiding M mode due to constantly changing lighting, etc.). But landscapes are my delightful getaway from that, so I actually love how grads and high stop NDs force. me. to. slow. down.

- A

 

[Jopa]

Frankly I don't remember I said anywhere GND will cure AIDS, ebola and cancer It's doing what it's supposed to do - gradually reduce light. As long as it can also reduce processing time - it's a tool worth using. Same as camera DR, strobes, light meters, and a million of other photography tools. One can re-create and render a whole scene in 3ds max, but it will probably take longer compared to taking a picture.It's simply not correct to say GNDs suck only because they can't be used absolutely everywhere.

Some people absolutely have no patience for compositing/masking, HDR tools, etc. while others hate fiddling with grads or shooting non-ideal horizons with them. I see both as tools in our toolboxes that can be brought to bear as circumstances dictate.

I'm neither a post-processing junkie nor a classic 'get it right in-camera' person. People should use what they like and feel more comfortable with.

Personally, though, I am a frantic day-to-day photographer as my family is in constant motion, so I never take enough time to do things properly (I have terrible habits like relying on Auto ISO, avoiding M mode due to constantly changing lighting, etc.). But landscapes are my delightful getaway from that, so I actually love how grads and high stop NDs force. me. to. slow. down.

- A

LOL that's exactly my story

 

[stevelee]

As I have said in another thread, I'm upgrading from a T3i, which is still not a half-bad camera, but surely the 6D II is a significant improvement. If I had a 6D or even an 80D, I probably wouldn't be thinking about an upgrade at this point, and probably wouldn't be reading these messages. That is just me, though, and no reflection upon those who do want to upgrade.

And my DR concerns come mostly with church interiors when I travel, and I'm shooting with a G7X Mark II, so not even in DSLR territory. I want the interior not to be too murky, and I don't want the stained glass colors to wash out. I bracket and then stack in filmstrip mode in Adobe Bridge for HDR. Results don't look quite natural enough to suit me, but still usable. For example the shot in the Order of the Thistle chapel in St. Giles, Edinburgh is one I plan to print out and put in a frame in my hallway, since I'm changing a couple of the pictures there. I have a more straight-on version with corrected verticals, but I like this wonky one better. It looks more like it looked to me when I was looking up at it.

If you guys have cameras that can do that or better in one shot, I'm impressed.

 

[privatebydesign]

100% agree. Now you can even simulate shallow DoF in PP, so no need to use fast lenses anymore The only problem - it takes a lot of time to process the image correctly otherwise it'll look like cheap fake

Not true, just look at some of the tutorials on the link I provided. A ND Grad is a sledgehammer, great if you take pictures of railroad spikes, far too limited for much else.

Frankly I don't remember I said anywhere GND will cure AIDS, ebola and cancer It's doing what it's supposed to do - gradually reduce light. As long as it can also reduce processing time - it's a tool worth using. Same as camera DR, strobes, light meters, and a million of other photography tools. One can re-create and render a whole scene in 3ds max, but it will probably take longer compared to taking a picture.It's simply not correct to say GNDs suck only because they can't be used absolutely everywhere.

Dude, that is a pretty crappy misrepresentation of what I said.

Do what you want, I don't care and I hope you have fun doing it, I was, like most people here, just relaying my experiences (actually most people here just talk a load of sh!t that they know absolutely nothing about and have no experience in). Over time everybody finds ND grads are more and more limiting from a compositional and results point of view, further, I never seemed to have the optimal one with me, either grade or hard or soft graduations.

I don't care about in camera 'purity' or tools you need, I care about getting the results I want and am happy with. When I look through sites with photographs I respect and want to achieve virtually none use ND grads anymore, but that is me, you take what you want with whatever you want however you want. In a couple of years come back and tell me how much you are using those grads and if you are still happy with them, I was just trying to save you some time and I did it constructively by offering a cheaper and more flexible alternative that gives higher quality results. Clearly that was my mistake, sorry.

 

[ajfotofilmagem]

As I have said in another thread, I'm upgrading from a T3i, which is still not a half-bad camera, but surely the 6D II is a significant improvement. If I had a 6D or even an 80D, I probably wouldn't be thinking about an upgrade at this point, and probably wouldn't be reading these messages. That is just me, though, and no reflection upon those who do want to upgrade.

And my DR concerns come mostly with church interiors when I travel, and I'm shooting with a G7X Mark II, so not even in DSLR territory. I want the interior not to be too murky, and I don't want the stained glass colors to wash out. I bracket and then stack in filmstrip mode in Adobe Bridge for HDR. Results don't look quite natural enough to suit me, but still usable. For example the shot in the Order of the Thistle chapel in St. Giles, Edinburgh is one I plan to print out and put in a frame in my hallway, since I'm changing a couple of the pictures there. I have a more straight-on version with corrected verticals, but I like this wonky one better. It looks more like it looked to me when I was looking up at it.

If you guys have cameras that can do that or better in one shot, I'm impressed.

A church interior with detailed shadows, and keeping the stained glass windows colored, will not be possible with a DR of 14 stops, without multiple exposures and HDR treatment. It would take more than 20 stops to achieve this in a single exposure, and no camera today is capable of this.

Coming from a T3i, you only get a maximum of 2 HDR stops with the new models. I know it is a heresy, but a bounced flash would allow a better balance between internal and external lighting.

 

[stevelee]

A church interior with detailed shadows, and keeping the stained glass windows colored, will not be possible with a DR of 14 stops, without multiple exposures and HDR treatment. It would take more than 20 stops to achieve this in a single exposure, and no camera today is capable of this.

Coming from a T3i, you only get a maximum of 2 HDR stops with the new models. I know it is a heresy, but a bounced flash would allow a better balance between internal and external lighting.

No surprises there. I was being a bit cute when I suggested that you guys with more modern cameras perhaps could. Using flash there would have got me kicked out, perhaps even before I shot if anyone saw me in time. The G7X's little flash thingy does have a spring on it so that with sufficient dexterity, one can bounce its flash.

I have no idea how the DR of the G7X II compares with that of the T3i, newer technology vs. larger sensor. But two more stops DR than the T3i doesn't sound half bad to me, and maybe 14-bit RAW files vs. 10-bit(?) will give me more to play with in ACR. Usable ISO range is likely much greater in the 6D II, however it compares with existing and past FF bodies. And the T3i has obviously much lower noise than the older body I had before.

 

[ahsanford]

And my DR concerns come mostly with church interiors when I travel,

...

If you guys have cameras that can do that or better in one shot, I'm impressed.

'One shot HDR' inside of churches becomes this 'how much can I abuse this RAW shadows/highlights sliders' sort of exercise. It's a bad habit, but, sometimes yours truly [cough] a friend I know has been known to do it. :

If you are shooting handheld, FF will get you more dynamic range at the higher ISOs you'll probably be using in low lit interiors. But as ajfotofilmagem said, no camera I can think of can one-shot churches perfectly -- it's an absurdly difficult ask of any camera's latitude. So I'd peg FF as a 1-2 stop better helper, but not a silver bullet.

Given that, I'd honestly take the T3i + a tripod over the 6D2 without one in this unique instance, but I appreciate that's rarely allowed for these venues.

- A

 

[SecureGSM]

Private,
You are absolutely correct in all what you said. ND grads never worked 100% for me. There is not enough flexibility. I am not going to repeat all you have as you ar speaking from a solid, robust experience.
However, let me explain how I understand the situation:
Jopa enjoying the process. This what he is doing photography for. Hence loves his toys. Which is perfectly fine.
And for you getting from point A to point be in a most efficient and timely manner is utmost important since you are after results. Process is less important to you. Which is understandable again.
So let boys play with their toys and enjoy the process.
The passion for photography is what we all have in common and this is great.

100% agree. Now you can even simulate shallow DoF in PP, so no need to use fast lenses anymore The only problem - it takes a lot of time to process the image correctly otherwise it'll look like cheap fake

Not true, just look at some of the tutorials on the link I provided. A ND Grad is a sledgehammer, great if you take pictures of railroad spikes, far too limited for much else.

Frankly I don't remember I said anywhere GND will cure AIDS, ebola and cancer It's doing what it's supposed to do - gradually reduce light. As long as it can also reduce processing time - it's a tool worth using. Same as camera DR, strobes, light meters, and a million of other photography tools. One can re-create and render a whole scene in 3ds max, but it will probably take longer compared to taking a picture.It's simply not correct to say GNDs suck only because they can't be used absolutely everywhere.

Dude, that is a pretty crappy misrepresentation of what I said.

Do what you want, I don't care and I hope you have fun doing it, I was, like most people here, just relaying my experiences (actually most people here just talk a load of sh!t that they know absolutely nothing about and have no experience in). Over time everybody finds ND grads are more and more limiting from a compositional and results point of view, further, I never seemed to have the optimal one with me, either grade or hard or soft graduations.

I don't care about in camera 'purity' or tools you need, I care about getting the results I want and am happy with. When I look through sites with photographs I respect and want to achieve virtually none use ND grads anymore, but that is me, you take what you want with whatever you want however you want. In a couple of years come back and tell me how much you are using those grads and if you are still happy with them, I was just trying to save you some time and I did it constructively by offering a cheaper and more flexible alternative that gives higher quality results. Clearly that was my mistake, sorry.

 

[ahsanford]

Private,
You are absolutely correct in all what you said. ND grads never worked 100% for me. There is not enough flexibility. I am not going to repeat all you have as you ar speaking from a solid, robust experience.
However, let me explain how I understand the situation:
Jopa enjoying the process. This what he is doing photography for. Hence loves his toys. Which is perfectly fine.
And for you getting from point A to point be in a most efficient and timely manner is utmost important since you are after results. Process is less important to you. Which is understandable again.
So let boys play with their toys and enjoy the process.
The passion for photography is what we all have in common and this is great.

^^ This. This for days. ^^

Why we shoot, what we shoot and how we shoot rarely devolves down to a single best practice. Some folks are output driven, other folks are complexity driven, process driven, gear driven, efficiency driven, etc. and I've learned it's best to respect those differences.

- A

 

[stevelee]

'One shot HDR' inside of churches becomes this 'how much can I abuse this RAW shadows/highlights sliders' sort of exercise. It's a bad habit, but, sometimes yours truly [cough] a friend I know has been known to do it. :

If you are shooting handheld, FF will get you more dynamic range at the higher ISOs you'll probably be using in low lit interiors. But as ajfotofilmagem said, no camera I can think of can one-shot churches perfectly -- it's an absurdly difficult ask of any camera's latitude. So I'd peg FF as a 1-2 stop better helper, but not a silver bullet.

Some of the landscapes that I bracketed on the recent Britain trip I wound up not stacking. The "shadows/highlight exercise" worked just fine, sometimes using the mid-exposure, sometimes the one-stop-under. In the Rockies last fall, not long after I got the G7X II, I was taking pictures in the Garden of the Gods near sunset. Bracketing did help those pictures a lot, generally one exposure for the sky and one for the rocks, and one I didn't use. That helped me preserve the colors I had seen in each more than anything else.

In churches where the stones haven't moved significantly in eight hundred years and the windows have been in place at least from the nineteenth century, taking multiple exposures handheld is not that big a problem. I think Photoshop does an amazing job aligning them. I seem to have better luck with the HDR from ACR's filmstrip mode than with doing it directly in Photoshop, even though I have more experience with that.

My next trip of any consequence will be to Hawaii the end of the year, so almost for sure very different challenges. The 6D II will still be my new toy, so I might decide to take it instead of or along with the G7X II.

I am thinking about doing a staycation before then and renting a TS lens for a week, probably the 24mm.

 

[arbitrage]

Well, I'm not sure that I trust someone who analyzes unverified source material, admits those preliminary analyses are often incorrect, then proceeds to draw conclusions from those data.

All in all, I think it's rather sad that he's stooped to this level.

I guess I'm not disappointed or sad because this isn't a new thing. He's done it before, with the same caveats provided (or lack thereof if you translate them with your guide ) I don't expect him to wait until the camera ships because he doesn't if he has the chance to publish numbers early. It's hard to be disappointed if you already know Santa isn't real.

And I certainly don't blame him for people running wild with it. The world is full reactionary doom and gloomers. The whole business model of cable news is built on it.

+1
He does early analysis on every camera when RAWS are provided to him and labels it as such. He doesn't have any ads on his site so not like he is just trying to drum up clicks or ad revenues like CW is by posting the story.

Somehow I doubt we'd be seeing any of this criticism or doubt if these early RAWs showed a DR equalling the 5D4. Funny how people get so defensive when negative news comes out and doubt everything, seeing conspiracies everywhere. If these results showed 5D4 DR everyone would be raving about it, no one would be saying the RAWs are fake, the RAWs are preproduction, the RAWs are from a prototype sensor or what else.

If I could put money on it, I'd be laying down a lot on the bet that these early RAWs will turn out to be an exact representation of the production, consumer bought sensor. I will have no problem and be happy for potential 6D2 customers if somehow the DR is up to 5D4 levels.

The thing is, I have no interest in a 6D2 type body but if someone offered me to swap this 6D2 sensor (with low DR) into my 1DX2 without losing any other 1DX2 features (like FPS) I'd take it no questions asked simply for more MPs (5D4 would be even better of course). I don't use ISO 100-400 where the DR is different. After 400 all the cameras are almost equal (other than the D5 which dominates high-ISO DR as it should because the D5 is made for high-ISO shooting anyways and I wish the 1DX2 sensor was geared more like it sacrificing low-ISO DR for high-ISO DR and noise handling)

 

[Jopa]

100% agree. Now you can even simulate shallow DoF in PP, so no need to use fast lenses anymore The only problem - it takes a lot of time to process the image correctly otherwise it'll look like cheap fake

Not true, just look at some of the tutorials on the link I provided. A ND Grad is a sledgehammer, great if you take pictures of railroad spikes, far too limited for much else.

Frankly I don't remember I said anywhere GND will cure AIDS, ebola and cancer It's doing what it's supposed to do - gradually reduce light. As long as it can also reduce processing time - it's a tool worth using. Same as camera DR, strobes, light meters, and a million of other photography tools. One can re-create and render a whole scene in 3ds max, but it will probably take longer compared to taking a picture.It's simply not correct to say GNDs suck only because they can't be used absolutely everywhere.

Dude, that is a pretty crappy misrepresentation of what I said.

Do what you want, I don't care and I hope you have fun doing it, I was, like most people here, just relaying my experiences (actually most people here just talk a load of sh!t that they know absolutely nothing about and have no experience in). Over time everybody finds ND grads are more and more limiting from a compositional and results point of view, further, I never seemed to have the optimal one with me, either grade or hard or soft graduations.

I don't care about in camera 'purity' or tools you need, I care about getting the results I want and am happy with. When I look through sites with photographs I respect and want to achieve virtually none use ND grads anymore, but that is me, you take what you want with whatever you want however you want. In a couple of years come back and tell me how much you are using those grads and if you are still happy with them, I was just trying to save you some time and I did it constructively by offering a cheaper and more flexible alternative that gives higher quality results. Clearly that was my mistake, sorry.

I think it was a great discussion. Way better than talks about DR of an unreleased camera coming from an unknown source

I do like my results so far, here is a shot taken a few days ago, with a 10 stops ND and a GND, and it's also a blend of two images. Could I take it without a GND? Yes. Could I take it in one image? Yes. Could I take it even without a 10 stops ND? Why not... But it would be a different image. It's all art and there is no right or wrong, there is no "approved" or best technique that suites everybody. It's also not about what others do, that's about what I do, what I like and what I learn in the meantime. A few years ago I saw an article on SAR about Sony developing a sensor with variable exposure (unlikely true). Now imagine this kind of sensor in a consumer camera... If this ever happen - then yes, I will though all my GNDs away

 

[CanonCams]

If I could put money on it, I'd be laying down a lot on the bet that these early RAWs will turn out to be an exact representation of the production, consumer bought sensor. I will have no problem and be happy for potential 6D2 customers if somehow the DR is up to 5D4 levels.

I believe you will see returned units if its turn out to be true.

I feel somewhat bad for the 6D owners who sold their cameras anticipating an upgrade.

 

[Aglet]

... won't know until shipping product is tested by someplace like DxOmark...

DxO... https://m.androidcentral.com/editors-desk-dxomark-worthless

if all you look at is the final rating mark you've missed 97% of the highly valuable measurement data DxOmark publishes.
You can learn quite a bit from that published data, especially the SNR chart.

 

[Aglet]

It's all art and there is no right or wrong, there is no "approved" or best technique that suites everybody. It's also not about what others do, that's about what I do, what I like and what I learn in the meantime.

+1!