fredmiranda.com Login
This topic is in the 2015 archive. You can read it and like its posts here. It is read only.

Archive 20155DSR has more DR than 5DIII

  
 
curious80
Offline
Details
p.2 #15DSR has more DR than 5DIII
molson wrote:
curious80 wrote
molson wroteIn some real-world applications, I've found the 5D SR to have more dynamic range than the D810, but the extra latitude available is in the highlights, not in the shadows like the Nikony sensors.

This is a single exposure, and no ND grad filters were used:
There is no such thing as highlight latitude vs shadow latitude as far as the actual raw sensor data is concerned. The sensor DR is sensor DR, whether it gets assigned to shadows or highlights is just a matter of camera metering and jpeg tone curve.
That's a semantic (some might say pedantic) argument, ....


Not really. All cameras support different jpeg profiles. You can change a Nikon tone curve to look more like Canon or the other way around, after which this artificial distinction goes away. Bottom line is that the exposure that you need to avoid blowing the highlights would be the same whether you are using a Canon or a Nikon.


With the 5D SR, I can expose for the midtones/shadows without worrying about noise or banding, and all the detail is still recoverable in the highlights. Since the lowest EVs have the fewest bits of color data, the Canon approach should (and apparently does) produce better results, because it lets you expose really far to the right (RFTTR™) where most of your bit data is located.


It is actually the opposite. If Canon is preserving more highlights then it is under-exposing compared to Nikon, in which case it is not doing RFTTR (not that RFTTR is always a good strategy).
Jul 07, 2015 at 02:57 PM
Cliff L.
Offline
Details
p.2 #25DSR has more DR than 5DIII
curious80 wrote:
molson wrote
curious80 wrote
molson wroteIn some real-world applications, I've found the 5D SR to have more dynamic range than the D810, but the extra latitude available is in the highlights, not in the shadows like the Nikony sensors.

This is a single exposure, and no ND grad filters were used:
There is no such thing as highlight latitude vs shadow latitude as far as the actual raw sensor data is concerned. The sensor DR is sensor DR, whether it gets assigned to shadows or highlights is just a matter of camera metering and jpeg tone curve.
That's a semantic (some might say pedantic) argument, ....
Not really. All cameras support different jpeg profiles. You can change a Nikon tone curve to look more like Canon or the other way around, after which this artificial distinction goes away. Bottom line is that the exposure that you need to avoid blowing the highlights would be the same whether you are using a Canon or a Nikon.

Earlier quoteWith the 5D SR, I can expose for the midtones/shadows without worrying about noise or banding, and all the detail is still recoverable in the highlights. Since the lowest EVs have the fewest bits of color data, the Canon approach should (and apparently does) produce better results, because it lets you expose really far to the right (RFTTR™) where most of your bit data is located.
It is actually the opposite. If Canon is preserving more highlights then it is under-exposing compared to Nikon, in which case it is not doing RFTTR (not that RFTTR is always a good strategy).


I guess I'll continue doing it the wrong way, because the results are a noticeable improvement over the Nikony way, and I no longer need ND grad filters or HDR bracketing...
Jul 07, 2015 at 03:03 PM
curious80
Offline
Details
p.2 #35DSR has more DR than 5DIII
molson wrote:
curious80 wrote
molson wrote
curious80 wrote
molson wroteIn some real-world applications, I've found the 5D SR to have more dynamic range than the D810, but the extra latitude available is in the highlights, not in the shadows like the Nikony sensors.

This is a single exposure, and no ND grad filters were used:
There is no such thing as highlight latitude vs shadow latitude as far as the actual raw sensor data is concerned. The sensor DR is sensor DR, whether it gets assigned to shadows or highlights is just a matter of camera metering and jpeg tone curve.
That's a semantic (some might say pedantic) argument, ....
Not really. All cameras support different jpeg profiles. You can change a Nikon tone curve to look more like Canon or the other way around, after which this artificial distinction goes away. Bottom line is that the exposure that you need to avoid blowing the highlights would be the same whether you are using a Canon or a Nikon.

Earlier quoteWith the 5D SR, I can expose for the midtones/shadows without worrying about noise or banding, and all the detail is still recoverable in the highlights. Since the lowest EVs have the fewest bits of color data, the Canon approach should (and apparently does) produce better results, because it lets you expose really far to the right (RFTTR™) where most of your bit data is located.
It is actually the opposite. If Canon is preserving more highlights then it is under-exposing compared to Nikon, in which case it is not doing RFTTR (not that RFTTR is always a good strategy).
I guess I'll continue doing it the wrong way, because the results are a noticeable improvement over the Nikony way, and I no longer need ND grad filters or HDR bracketing...


Who said anything about right or wrong way. I am sure Canon 5DR is capable of producing as good or better results compared to Nikon D810. I am just objecting to your assertion that 5DR has more highlight room than Nikon which doesn't make any sense for a digital sensor. There is no way that in the hands of an experienced D810 shooter who understands its working, it will produce any less highlight detail than Canon 5DR.
Jul 07, 2015 at 03:17 PM
curious80
Offline
Details
p.2 #45DSR has more DR than 5DIII
kevindar wrote:
good to see. it seems like it may be just quarter stop better than the last Canon DR champion, 6d, and only 1/3-1/2 stop worse than the 360 dollar Nikon d5200 at base iso, and of course easily beating it at higher ISO.
....


You can't really do a straight DR comparison of an FF camera with an APS-C camera. Sure D5200 might have higher numerical DR, but that only tells you that in the deepest of the shadows it has less noise than 5DR. It hides the fact that over much of the "regular" shadow region, 5DR would have less noise than D5200.
Jul 07, 2015 at 03:25 PM
Cliff L.
Offline
Details
p.2 #55DSR has more DR than 5DIII
curious80 wrote:
There is no way that in the hands of an experienced D810 shooter who understands its working, it will produce any less highlight detail than Canon 5DR.


I will look forward to seeing your examples, then...
Jul 07, 2015 at 03:39 PM
Ziffl3
Offline
Details
p.2 #65DSR has more DR than 5DIII
curious80 wrote:
molson wrote
curious80 wrote
molson wrote
curious80 wrote
molson wroteIn some real-world applications, I've found the 5D SR to have more dynamic range than the D810, but the extra latitude available is in the highlights, not in the shadows like the Nikony sensors.

This is a single exposure, and no ND grad filters were used:
There is no such thing as highlight latitude vs shadow latitude as far as the actual raw sensor data is concerned. The sensor DR is sensor DR, whether it gets assigned to shadows or highlights is just a matter of camera metering and jpeg tone curve.
That's a semantic (some might say pedantic) argument, ....
Not really. All cameras support different jpeg profiles. You can change a Nikon tone curve to look more like Canon or the other way around, after which this artificial distinction goes away. Bottom line is that the exposure that you need to avoid blowing the highlights would be the same whether you are using a Canon or a Nikon.

Earlier quoteWith the 5D SR, I can expose for the midtones/shadows without worrying about noise or banding, and all the detail is still recoverable in the highlights. Since the lowest EVs have the fewest bits of color data, the Canon approach should (and apparently does) produce better results, because it lets you expose really far to the right (RFTTR™) where most of your bit data is located.
It is actually the opposite. If Canon is preserving more highlights then it is under-exposing compared to Nikon, in which case it is not doing RFTTR (not that RFTTR is always a good strategy).
I guess I'll continue doing it the wrong way, because the results are a noticeable improvement over the Nikony way, and I no longer need ND grad filters or HDR bracketing...
Who said anything about right or wrong way. I am sure Canon 5DR is capable of producing as good or better results compared to Nikon D810. I am just objecting to your assertion that 5DR has more highlight room than Nikon which doesn't make any sense for a digital sensor. There is no way that in the hands of an experienced D810 shooter who understands its working, it will produce any less highlight detail than Canon 5DR.


I find more head room in the high light on canon than I do nikon ... About 1/2 of stop. It is noticible in editing.

-mark
Jul 07, 2015 at 03:43 PM
Monito
Offline
Details
p.2 #75DSR has more DR than 5DIII
A camera can have more highlight headroom than a different camera, when midtones get the same exposure and produce the same values. Just like a camera can have more shadow room. Dynamic range does not automatically get centered.

There two kinds of right exposure: A) Correct for midtones and going straight to JPEG. B) Correct for the widest range of important tones so that the Raw image can be pulled or pushed in processing.

A and B might differ on the same camera and A might differ between two cameras as well as B.

Note that B includes the key word "important". This means that the photographer chooses which tones are important and which can be sacrificed and then computes the exposure on that basis. If the scene is low contrast you might even "over" expose and pull it down in processing, though often 'straight' exposure is best then.

I'm not sure whether the 5DS has more room in the highlights or not. I do know that it is still easy to blow out whites. When an over-exposed tone is blown in one or two channels modern recovery software can make it seem like there is more headroom. For example, a sunrise has predominantly yellow and orange tones in the blown area. But a white ship in brilliant sunlight blows out the whites with no recovery.

It was ever thus.
Jul 07, 2015 at 04:08 PM
SoftImage
Offline
Details
p.2 #85DSR has more DR than 5DIII
All manufacturers overstate ISO. It is possible that at ISO100 real 5DSR ISO is lower than that of D810. So, for same exposure settings at ISO100 5DSR image will be less bright compared to D810 and hence will 'preserve' more highlight. That will put it in a disadvantageous position in shadow areas though. We will know more when DXO tests for real vs stated ISO.


molson wrote:
In some real-world applications, I've found the 5D SR to have more dynamic range than the D810, but the extra latitude available is in the highlights, not in the shadows like the Nikony sensors.

This is a single exposure, and no ND grad filters were used:


Jul 07, 2015 at 04:27 PM
Schlotkins
Offline
Details
p.2 #95DSR has more DR than 5DIII
Nice to see. The gap between the 6D and 7R is much bigger with this data (25%) than Dxo (16%). The a7r data seems really high - it's above the D810, which seems wrong.

Anyway, I'm guessing the 5Ds R will end up being about 13 stops on dxo, which is pretty good. Interested to see noise and and color comparisons. I have a a7rII on order, but for 1 stop of DR, I'd rather have a 5Ds R and use my current a7r as an accessory.

Chris
Jul 07, 2015 at 05:05 PM
Cliff L.
Offline
Details
p.2 #105DSR has more DR than 5DIII
SoftImage wrote:
All manufacturers overstate ISO. It is possible that at ISO100 real 5DSR ISO is lower than that of D810. So, for same exposure settings at ISO100 5DSR image will be less bright compared to D810 and hence will 'preserve' more highlight. That will put it in a disadvantageous position in shadow areas though. We will know more when DXO tests for real vs stated ISO.

molson wroteIn some real-world applications, I've found the 5D SR to have more dynamic range than the D810, but the extra latitude available is in the highlights, not in the shadows like the Nikony sensors.

This is a single exposure, and no ND grad filters were used:


It's amazing how so many people who have obviously never used a digital camera still feel compelled to post their astute comments on these threads...

Did DPReview just kick a bunch of people out of their forums or something?
Jul 07, 2015 at 05:40 PM
EB-1
Online
Details
p.2 #115DSR has more DR than 5DIII
5DSR has more DR than 5DIII

What has that to do with the D810 or other Nikon bodies

EBH
Jul 07, 2015 at 06:01 PM
kevindar
Offline
Details
p.2 #125DSR has more DR than 5DIII
curious80 wrote:
kevindar wrotegood to see. it seems like it may be just quarter stop better than the last Canon DR champion, 6d, and only 1/3-1/2 stop worse than the 360 dollar Nikon d5200 at base iso, and of course easily beating it at higher ISO.
....
You can't really do a straight DR comparison of an FF camera with an APS-C camera. Sure D5200 might have higher numerical DR, but that only tells you that in the deepest of the shadows it has less noise than 5DR. It hides the fact that over much of the "regular" shadow region, 5DR would have less noise than D5200.


This generally is certainly true, though hopefully neither has much noise at iso 100.

Owning a sony a6000 (crop), and a canon 5d3 the sony can take a bit more abuse in the shadows. though I agree when it comes to pure Lumi noise, at 100% crop, the 5d3 still has a bit of smother image.

I am happy canon is making progress, and as I said DR is not the end all be all. I have discovered many times I am perfectly fine shooting in dualiso mode using magic lantern, which gives me my 14 stops of DR at the slight sacrifice of resolution. I have also discovered that nothing beats bracketing when possible in terms of image quality. shadows pushed three stops, at best, are equivalent to ISO 800 (if there was no read noise amplification, in a perfect world), and are still inferior to shadows at iso 100 when is what you get when you bracket.

All said, the improvement in DR for canon sensor has been very very incremental over the last 5 years, lagging the other manufactueres full frame bodies but 2 stops.
Jul 07, 2015 at 06:08 PM
Herb
Offline
Details
p.2 #135DSR has more DR than 5DIII
dmcphoto wrote:
Ziffl3 wrote
big country wrotei do not have the R, but i have been shooting the 5DS at weddings, engagements and some birds w/ the 5DS. honestly, if it were 6-8 fps, and had a bigger buffer, and could go a little higher on the native ISO this would be pretty close to my ultimate camera.
I would resemble this remark..... 6-8fps ... I am okay with 5fps, but 6-8 would be cool.
The buffer ... there is a place that could use a bump. This would be my 1 gripe.

Native ISO .... I am okay with the 6400.

The other spot is battery life. Need the grip for weddings... but this could be me.
I am not in the mood to carry additional batteries when shooting all day.

I need to get into the last wedding I shot to understand the colors.

-Mark
I have the 5DSR and I'm finding the battery life to be surprisingly short. I am not sure how it compares to the 5D3 as my other cameras are 1DX and 1DS3, but some anecdotal playing around with the camera does suggest that battery life could be a problem. Both cameras (5DS and 5DSR) should be identical in this regard.

Doesn't the 5DIII battery grip fit this camera as well?
Jul 07, 2015 at 06:18 PM
Sneakyracer
Offline
Details
p.2 #145DSR has more DR than 5DIII
SoftImage wrote:
Bill Claff has just completed his measurement of 5DSR DR. You can compare it with other cameras.

http://home.comcast.net/~NikonD70/Charts/PDR.htm


Nice!

So basically above iso 400 the venerable and king of DR the D810 has no more DR than the 5DSR.



Jul 07, 2015 at 06:22 PM
mttran
Offline
Details
p.2 #155DSR has more DR than 5DIII
molson wrote:
SoftImage wroteAll manufacturers overstate ISO. It is possible that at ISO100 real 5DSR ISO is lower than that of D810. So, for same exposure settings at ISO100 5DSR image will be less bright compared to D810 and hence will 'preserve' more highlight. That will put it in a disadvantageous position in shadow areas though. We will know more when DXO tests for real vs stated ISO.

molson wroteIn some real-world applications, I've found the 5D SR to have more dynamic range than the D810, but the extra latitude available is in the highlights, not in the shadows like the Nikony sensors.

This is a single exposure, and no ND grad filters were used:
It's amazing how so many people who have obviously never used a digital camera still feel compelled to post their astute comments on these threads...

Did DPReview just kick a bunch of people out of their forums or something?


hmm, you are posting on his thread
Jul 07, 2015 at 06:23 PM
SoftImage
Offline
Details
p.2 #165DSR has more DR than 5DIII
molson wrote:
SoftImage wroteAll manufacturers overstate ISO. It is possible that at ISO100 real 5DSR ISO is lower than that of D810. So, for same exposure settings at ISO100 5DSR image will be less bright compared to D810 and hence will 'preserve' more highlight. That will put it in a disadvantageous position in shadow areas though. We will know more when DXO tests for real vs stated ISO.

molson wroteIn some real-world applications, I've found the 5D SR to have more dynamic range than the D810, but the extra latitude available is in the highlights, not in the shadows like the Nikony sensors.

This is a single exposure, and no ND grad filters were used:
It's amazing how so many people who have obviously never used a digital camera still feel compelled to post their astute comments on these threads...

Did DPReview just kick a bunch of people out of their forums or something?


Interesting response.

Going back to what I said. I really don't know what 5DSR's real vs stated ISOs are and that's why I said 'it is possible that....'. There is a reason the thought came to my mind as a possibility. For any given exposure settings my 5DIII images are brighter than my 5DSR images, in other words 5DSR images are less bright. It maybe that the difference is due to ACR's rendering. I am not the only one noticing this difference though. If you go to this page and scroll down to Dynamic Range you will see how the luminosity is less in both the brightest and darkest part for 5DSR compared to 5DIII.

http://www.martinbaileyphotography.com/2015/06/30/canon-eos-5ds-r-digital-slr-camera-review-podcast-478/

Jul 07, 2015 at 06:56 PM
mttran
Offline
Details
p.2 #175DSR has more DR than 5DIII
Congratulation Canon !!!! you now match D3X DR, 7 years after Dec 2008. Still less than D5200 DR but we happily take it.
Jul 07, 2015 at 09:49 PM
gdanmitchell
Offline
Details
p.2 #185DSR has more DR than 5DIII
dmcphoto wrote:
I have the 5DSR and I'm finding the battery life to be surprisingly short. I am not sure how it compares to the 5D3 as my other cameras are 1DX and 1DS3, but some anecdotal playing around with the camera does suggest that battery life could be a problem. Both cameras (5DS and 5DSR) should be identical in this regard.


I think I'm finding it to be shorter than my 5DII, but I also have to say that I have not yet actually run a battery all the way down — leaving the possibility that it isn't less battery life but a different curve to the remaining charge readout.

Dan
Jul 07, 2015 at 10:01 PM
EB-1
Online
Details
p.2 #195DSR has more DR than 5DIII
Are you using the camera for multiple shots, i.e., bracketing/HDR or continuous 5FPS, or just one shot and then wait for power off?

At normal temperatures the 5D III is good for over 2000 frames with plenty of power remaining. Will I get about 1500 (75%) or less on the 5DsR under the same conditions?

EBH
Jul 07, 2015 at 10:20 PM
SoftImage
Offline
Details
p.2 #205DSR has more DR than 5DIII
mttran wrote:
Congratulation Canon !!!! you now match D3X DR, 7 years after Dec 2008. Still less than D5200 DR but we happily take it.


You are talking about base ISO. Canon actually has done pretty well above ISO400. 5DSR seems to have improved base ISO DR as well as DR above ISO400. I looked at cameras Bill Claff has tested and only two cameras are better than 5DSR DR above ISO400, Sony A7s and Pentax 645Z.

In any case I agree with Kevindar's comment above. Lifting shadows with any camera is poor man's bracketing. We need much better sensors to replace multi shot bracketing, fill flash or filters.

Jul 08, 2015 at 02:51 AM
FM ForumsCanon ForumJoin Upload & Sell
  
 

You are not logged inLog inRegister