Pages: 1 ... 3 4 [5] 6 7   Go Down

Author Topic: D800E -Truly Amazing Low-Light Performance  (Read 47781 times)

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10385
Re: The value of SNR @ 18%, and maybe @ 1% or @ 6 stops below maximum
« Reply #80 on: November 22, 2012, 05:52:36 am »

Although there are situations where extreme DR is important (like scenes of unusually high Subject Brightness Range or forgetting to set the ISO speed appropriately when photographing moths at dusk), I would say that the vast majority of photography involves not much more that the traditional "normal" SBR of about three stops either side of mid-tones.

You're probably right in that respect, BJL. The vast majority of photography is now done with an iPhone or iPad, followed by the so-called P&S camera which in turn is followed by the cropped-format DSLR. The full-frame DSLR seems to be a relative  rarity.

Unfortunately perhaps, I'm not one who bases his standards on the practice of the majority. The nature of much my photographic activity is to strive to capture as much detail as possible in the scenes I'm photographing, so that I have a wider creative scope when later processing the image back home on my calibrated system.

Each to his own. Most people seem to be most interested in photographing their friends and loved ones standing in front of any object or scene that's worth photographing in its own right, hence the development of automatic face detection in autofocussing.

I'm more interested in taking notes of a scene that interests me, and a high resolution, high DR, high SNR camera is the best note-taker that I know.

As an example, I recently attended a Thai boxing performancen (I'm currently travelling in Thailand), even though I'm not much interested in the activity of two men attempting to inflict brain damage upon each other.

I did my best to get some interesting shots of a gloved fist smashing into a face, using my D800E with attached flash unit, but it was very much a case of 'hit and miss'.

When I later reviewed my shots after downloading to my Dell Notebook, I noticed that quite often the activity of the boxers was less interesting than the expressions on the faces of the audience on the other side of the ring. I saw expressions of joy, elation, disgust and total boredom, right next to each other, as seen through the V shape of a boxer's legs.

However, and it's a big 'however', such faces were partially beyond the reach of the flash. They were significantly underexposed. But never mind, or 'Mai Bpen Rai' as they say in Thailand. I have a D800E. I can make something of those crops through the boxer's legs because my camera captures detail in the shadows. If I were still using my much revered Canon 5D, I'd be stuffed; deleted image.

Another source of confusion I'd like to address is the notion that 'engineering DR specifications' are not relevant to practical photography, and that the 14EV DR claims of certain Nikon camera, from DXO measurements are no more than engineering specifications.

It's surely not difficult to appreciate that any camera, whatever its DR specification, will not produce 'nice' images within the region of the limits of its dynamic range.

All cameras, whether iPhone, P&S, or DSLR will produce crap photos at the engineering limits of their dynamic range.

The D800, 14 stops down from an ETTR exposure at base ISO, produces crap images, far worse than my underexposed shot of the world's largest moth, I have no doubt.

The issue is this, if I can use an analogy with my only other functioning full-frame camera, the Canon 5D. The 5D has a claimed 11 stop DR by DXO. The D800E has more than 14 stops, at print. Let's call it a 3 stop difference.

Surely it's understood that both the D800 in its 14th stop, and the 5D in its 11th stop produce unacceptable image quality. There's not much point in complaining or arguing that the 14th stop, in the case of the D800, and the 11th stop in the case of the 5D, is of little photographic quality. We all should know that.

The point is, if the scene has an 11 stop dynamic range, how do the cameras compare?

In the 11th stop, the 5D image is crap because that's its engineering limit. However, in the 11th stop the the D800 image is very presentable because that's 3 stops down from its enineering limit. Got it?
Logged

thierrylegros396

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1947
Re: D800E -Truly Amazing Low-Light Performance
« Reply #81 on: November 22, 2012, 12:56:24 pm »

Almost right Ray, but some cameras clip better than others.

And some exhibit banding, others not !

I found my old G10 or my EOS400D give better images than my RX100 when clipping occurs in pools or rivers for example.

But RX100 clips better @ISO125 than @ ISO80.

That's why I always set it @ISO125 or auto ISO (125 to 800) when using it in JPG.

Thierry
« Last Edit: November 25, 2012, 02:34:43 pm by thierrylegros396 »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: The value of SNR @ 18%, and maybe @ 1% or @ 6 stops below maximum
« Reply #82 on: November 22, 2012, 03:19:07 pm »

Hi Ray,

I don't think that anybody here says that DR is unnecessary or undesirable, just that most subjects fit in a narrower tonal range. Obviously there are cases where extended DR is helpful.

I don't see any contradiction between good SNR18% and good DR. My impression is that the D800 excels in both areas. To that you can add 36 MP and the choice of OLP filter or not.

Best regards
Erik


You're probably right in that respect, BJL. The vast majority of photography is now done with an iPhone or iPad, followed by the so-called P&S camera which in turn is followed by the cropped-format DSLR. The full-frame DSLR seems to be a relative  rarity.

Unfortunately perhaps, I'm not one who bases his standards on the practice of the majority. The nature of much my photographic activity is to strive to capture as much detail as possible in the scenes I'm photographing, so that I have a wider creative scope when later processing the image back home on my calibrated system.

Each to his own. Most people seem to be most interested in photographing their friends and loved ones standing in front of any object or scene that's worth photographing in its own right, hence the development of automatic face detection in autofocussing.

I'm more interested in taking notes of a scene that interests me, and a high resolution, high DR, high SNR camera is the best note-taker that I know.

As an example, I recently attended a Thai boxing performancen (I'm currently travelling in Thailand), even though I'm not much interested in the activity of two men attempting to inflict brain damage upon each other.

I did my best to get some interesting shots of a gloved fist smashing into a face, using my D800E with attached flash unit, but it was very much a case of 'hit and miss'.

When I later reviewed my shots after downloading to my Dell Notebook, I noticed that quite often the activity of the boxers was less interesting than the expressions on the faces of the audience on the other side of the ring. I saw expressions of joy, elation, disgust and total boredom, right next to each other, as seen through the V shape of a boxer's legs.

However, and it's a big 'however', such faces were partially beyond the reach of the flash. They were significantly underexposed. But never mind, or 'Mai Bpen Rai' as they say in Thailand. I have a D800E. I can make something of those crops through the boxer's legs because my camera captures detail in the shadows. If I were still using my much revered Canon 5D, I'd be stuffed; deleted image.

Another source of confusion I'd like to address is the notion that 'engineering DR specifications' are not relevant to practical photography, and that the 14EV DR claims of certain Nikon camera, from DXO measurements are no more than engineering specifications.

It's surely not difficult to appreciate that any camera, whatever its DR specification, will not produce 'nice' images within the region of the limits of its dynamic range.

All cameras, whether iPhone, P&S, or DSLR will produce crap photos at the engineering limits of their dynamic range.

The D800, 14 stops down from an ETTR exposure at base ISO, produces crap images, far worse than my underexposed shot of the world's largest moth, I have no doubt.

The issue is this, if I can use an analogy with my only other functioning full-frame camera, the Canon 5D. The 5D has a claimed 11 stop DR by DXO. The D800E has more than 14 stops, at print. Let's call it a 3 stop difference.

Surely it's understood that both the D800 in its 14th stop, and the 5D in its 11th stop produce unacceptable image quality. There's not much point in complaining or arguing that the 14th stop, in the case of the D800, and the 11th stop in the case of the 5D, is of little photographic quality. We all should know that.

The point is, if the scene has an 11 stop dynamic range, how do the cameras compare?

In the 11th stop, the 5D image is crap because that's its engineering limit. However, in the 11th stop the the D800 image is very presentable because that's 3 stops down from its enineering limit. Got it?

« Last Edit: November 22, 2012, 03:51:31 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

jeremypayne

  • Guest
Re: D800E -Truly Amazing Low-Light Performance
« Reply #83 on: November 22, 2012, 03:42:08 pm »

RX100 clips better @ISO125 than @ ISO80.

That's why I always set it @ISO125 or auto ISO (125 to 800).

Thierry

125 is the base ISO of the RX 100.  80 and 100 are 'fake' ISOs.

... Mine got wet and has died ... :-(
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10385
Re: D800E -Truly Amazing Low-Light Performance
« Reply #84 on: November 23, 2012, 12:10:50 pm »

125 is the base ISO of the RX 100.  80 and 100 are 'fake' ISOs.



That doesn't appear to be true, Jeremy. If you check the DXOMark graphs, you'll see that at ISO 80 all specs are highest for the RX100. At ISO 100 DR is about 0.25 stops worse and at ISO 200 over 1/2 a stop worse. Same applies to SNR, Tonal Range and Color Sensitivity.

I can't think of any reason why Thierry would think that the RX100 clips better at ISO 125, unless he's using in-camera jpeg processing, or some RAW converter that is imposing its own peculiarities.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10385
Re: The value of SNR @ 18%, and maybe @ 1% or @ 6 stops below maximum
« Reply #85 on: November 23, 2012, 12:33:54 pm »


I don't think that anybody here says that DR is unnecessary or undesirable, just that most subjects fit in a narrower tonal range. Obviously there are cases where extended DR is helpful.


Hi Erik,

Glad to hear it, but I sense some confusion here. Are you saying that most subjects before they are photographed fit in a narrower tonal range, that is, they have a fairly low SBR, such as indoor scenes, or outdoor scenes on a dull day; or do you mean that most subjects are photographed in such a way, and subsequently processed in such a way, that they are made to fit in a narrower tonal range, whether on print or screen?

The very fact that many camera models now have a built-in HDR feature, and image processing programs like Photoshop include a Merge to HDR facility, and many stand-alone programs exist that are designed specifically for creating HDR images, would tend to suggest that a lot of scenes do not fit into that fairly narrow tonal range that you and BJL describe.

My own experience also suggests that most subjects do not fit into that fairly narrow tonal range, but that may be because I'm an outdoor type of person and live in a country which is blessed with many days of sunshine throught the year. Those who live in the UK, or some northern European country where the sun shines for only two days each year, may be quite satisfied with the low DR capability of their camera. (I may be exaggerating just a bit. I believe St Petersberg has on average as many as 65 days of sunshine per year.  ;D )

Having been involved in photography as an amateur since I was a youngster, I recollect frequently over the years seeing blown skies in the photos people have shown me.

When the first reasonably affordable DSLRs appeared on the market, I recall much discussion about the limited DR of these new cameras compared with film. I recall figures such as 5 to 6 stops of DR for slide film, 7 to 8 stops for color negative, and 9 to 11 stops for B&W film. I recall the Canon D60 had a DR not much better than that of slide film.

In those days we didn't seem to use the concept of defining DR in relation to a normalised, fixed size image or print. Was that DR specification of B&W film a property of the average size of the partcles of light-sensitive chemicals on the film surface, analagous to a pixel? If so, there would be a huge difference in DR between the average 35mm format of the day and the 8"x10" format preferred by Ansel Adams, using the same B&W film type.

If we make an assumption that the best, finest grain, B&W film of the day, in 35mm format, had a DR of 11 stops, then a large-format field camera holding the same type of film would surely have had a significantly higher DR than 11 stops.

Whether that's true or not, I simply don't know. It may be the case that only SNR at 18% followed that rule of approximately one stop improvement for each doubling of film area, and that any increases in DR with film format size are compromised by some other limitation.


Logged

thierrylegros396

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1947
Re: D800E -Truly Amazing Low-Light Performance
« Reply #86 on: November 23, 2012, 12:52:53 pm »


That doesn't appear to be true, Jeremy. If you check the DXOMark graphs, you'll see that at ISO 80 all specs are highest for the RX100. At ISO 100 DR is about 0.25 stops worse and at ISO 200 over 1/2 a stop worse. Same applies to SNR, Tonal Range and Color Sensitivity.

I can't think of any reason why Thierry would think that the RX100 clips better at ISO 125, unless he's using in-camera jpeg processing, or some RAW converter that is imposing its own peculiarities.

Just have a look, here http://www.dpreview.com/reviews/sony-cybershot-dsc-rx100/11

Completely different of DXOMark graphs !

Who to trust in ?!!
Logged

jeremypayne

  • Guest
Re: D800E -Truly Amazing Low-Light Performance
« Reply #87 on: November 23, 2012, 01:07:08 pm »


That doesn't appear to be true, Jeremy. If you check the DXOMark graphs, you'll see that at ISO 80 all specs are highest for the RX100. At ISO 100 DR is about 0.25 stops worse and at ISO 200 over 1/2 a stop worse. Same applies to SNR, Tonal Range and Color Sensitivity.

I can't think of any reason why Thierry would think that the RX100 clips better at ISO 125, unless he's using in-camera jpeg processing, or some RAW converter that is imposing its own peculiarities.

Check the specs:
http://www.dpreview.com/reviews/sony-cybershot-dsc-rx100
http://store.sony.com/p/Sony-Cyber-shot-DSC-RX100-Camera/en/p/DSCRX100/B#specifications

Also, the in-camera menu identifies 80 & 100 as "expanded" ISOs that are different than 125 to 6400.

Lastly, the specs on the DXO site:

http://www.dxomark.com/index.php/Publications/DxOMark-Reviews/Sony-DSC-RX100-Review/Preview

Everything I have seen first hand and on the web indicates an ISO range of 125-6400.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: D800E -Truly Amazing Low-Light Performance
« Reply #88 on: November 23, 2012, 02:16:33 pm »

Hi,

DPreview data is JPEG and DxO-mark is raw data.

Best regards
Erik


Just have a look, here http://www.dpreview.com/reviews/sony-cybershot-dsc-rx100/11

Completely different of DXOMark graphs !

Who to trust in ?!!
Logged
Erik Kaffehr
 

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10385
Re: D800E -Truly Amazing Low-Light Performance
« Reply #89 on: November 23, 2012, 11:21:31 pm »


Everything I have seen first hand and on the web indicates an ISO range of 125-6400.


Everything except the actual measurements of the RAW data from a bunch of meticulous scientists with PhDs, at DXO.

I think Erik may have hit the nail on the head. In-camera jpegs have always, unavoidably and necessarily, been processed in accordance with some esthetic, or merely practical decision made by the designers of the camera.

Just recently during the furore over comparisons between the Canon 5D3 and the D800, Canon fanboys, feeling a bit let down by the possibility that Nikon had produced a better value and better performing camera, would display out-of-camera jpegs from the 5D3 which showed noticeably less noise than similar shots from the D800 at the same high ISO, thus demonstrating that Canon's jpeg engine applied more noise reduction than Nikon's.

For those who shoot in RAW mode, the measurements at DXOMark are a remarkably accurate guide as to what to expect from a camera.

Nevertheless, it's possible that DXOMark may occasionally publish mistakes. I'm still a bit puzzled, just a bit, as to why their graph for the ISO Sensitivity for the Canon 50D implies that ISO 100 on the 50D is an expanded ISO and that the real base ISO is 200. I noticed the figures at DXOMark after I'd been using the camera for some time, often compromising on shutter speed or DoF by using ISO 100 for the sake of better SNR, rather than ISO 200 which I now realise would have been a better choice.

Could it be what  DXOMark is really saying is that, although ISO 100 is the official base ISO for this camera according to Canon, our tests indicate that the differences in performance between ISO 100 and 200 are so small that they lie within the margin of error of our measurement techniques.

I'm assuming this is probably the case because my own tests comparing identical, high contrast scenes, shot at ISO 100 and ISO 200, using half the shutter speed for an ETTR at ISO 200, indicate no discernible difference in DR (or shadow detail), and only the slightest difference in SNR at 18%, noticeable at 400% on monitor or print, but invisible at 100%.

In other words, for all practical considerations, ISO 200 is the true base ISO for the Canon 50D, despite what the Canon manual may imply, and that is something which is useful to know. The only reason to use ISO 100 would be for the sake of a slow shutter speed.

Logged

thierrylegros396

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1947
Re: D800E -Truly Amazing Low-Light Performance
« Reply #90 on: November 24, 2012, 03:39:11 am »

Thanks for those infos, Ray  :)

I'm also puzzelled by ISO Sensitivity graphs of almost all MF cameras at DXOMark.

Horizontal  ??? ::), what does it really mean in thoses cases ?!

Have a Nice day.

Thierry
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: D800E -Truly Amazing Low-Light Performance
« Reply #91 on: November 24, 2012, 04:15:00 am »

Hi,

DxO describes the ISO standard here: http://www.dxomark.com/index.php/About/In-depth-measurements/Measurements/ISO-sensitivity

Please note that the ISO standard gives only one figure for ISO and that figure is based on sensor saturation.

Some cameras have adjustable gain. All Canons and many Nikons belong to that category. Another solution is to not have variable gain just adjust exposure but tag the raw image with exposure correction. I guess that is exactly what "ISO-less" cameras do.

If you check out the DR graphs on DxO mark you can see that some cameras have essentially a straight line, with DR dropping linearly with ISO. That essentially means ISO-less design. An ISO-less design  means that you can achieve the same result with say three stops of underexposure as using 8X times higher ISO.

If all this sounds gibberish, that's OK. The descriptions at DxO are not complete, in my humble opinion.

Best regards
Erik

 


Thanks for those infos, Ray  :)

I'm also puzzelled by ISO Sensitivity graphs of almost all MF cameras at DXOMark.

Horizontal  ??? ::), what does it really mean in thoses cases ?!

Have a Nice day.

Thierry
Logged
Erik Kaffehr
 

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: The value of SNR @ 18%, and maybe @ 1% or @ 6 stops below maximum
« Reply #92 on: November 24, 2012, 05:08:53 am »

Glad to hear it, but I sense some confusion here. Are you saying that most subjects before they are photographed fit in a narrower tonal range, that is, they have a fairly low SBR, such as indoor scenes, or outdoor scenes on a dull day; or do you mean that most subjects are photographed in such a way, and subsequently processed in such a way, that they are made to fit in a narrower tonal range, whether on print or screen?
I assume that he is talking about the "subjective DR" of the scene. I.e. the distance between the darkest interesting things in a scene and the brightest interesting things in a scene.

Quote
The very fact that many camera models now have a built-in HDR feature, and image processing programs like Photoshop include a Merge to HDR facility, and many stand-alone programs exist that are designed specifically for creating HDR images, would tend to suggest that a lot of scenes do not fit into that fairly narrow tonal range that you and BJL describe.
I think that your argument is good. I will provide an alternate angle nonetheless.

When photographers talk about "HDR", they often mean two things at once, but occasionally one or the other: 1)Synthesizing an image from exposure brackets, 2)Tonemapping an image to reduce its dynamic range. The popularity of HDR _could_ be partially interpreted as a desire among photographers to produce that "dramatic look" of tonemapping, irrespecitve of the scene DR or their cameras DR.

This is similar to how the CD format allows for a DR of 96dB, yet most popular music exploit only a small fraction of that - mastering producers squeeze and clip away all of the life from the music in order for it to sound loud or "catchy" or whatever. The irony is that some music lovers are turning to vinyl - not because of its inherent qualities but because the content put onto vinyl may tend to be less damaged by hearing-damaged producers.
Quote
My own experience also suggests that most subjects do not fit into that fairly narrow tonal range, but that may be because I'm an outdoor type of person and live in a country which is blessed with many days of sunshine throught the year. Those who live in the UK, or some northern European country where the sun shines for only two days each year, may be quite satisfied with the low DR capability of their camera. (I may be exaggerating just a bit. I believe St Petersberg has on average as many as 65 days of sunshine per year.  ;D )
I believe that partially cloudy/sunny scenes (like we have in my part of the world) can have a substantial subjective DR.
Quote
I recall the Canon D60 had a DR not much better than that of slide film.
Yet people shot slide film?

I think that it is hard to compare the DR of film vs digital. It has been tried many times, and the worlds internet users cannot seem to agree.

-h
Logged

allegretto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 660
Re: D800E -Truly Amazing Low-Light Performance
« Reply #93 on: November 24, 2012, 07:40:23 am »



If all this sounds gibberish, that's OK. The descriptions at DxO are not complete, in my humble opinion


I am so glad YOU said this. It almost seems to me as I read it that I have new questions that go unanswered, and some of it seems difficult to discren meaning. sensorgen.info seems to make sense to me in terms of interpretation of how the whole camera actually works, but isn't quite as up to date as some of DxO's info.

What are your thoughts on sensorgen?
Logged

jeremypayne

  • Guest
Re: D800E -Truly Amazing Low-Light Performance
« Reply #94 on: November 24, 2012, 10:00:00 am »

Everything except the actual measurements of the RAW data from a bunch of meticulous scientists with PhDs, at DXO ...
... Nevertheless, it's possible that DXOMark may occasionally publish mistakes.

Strange.  Hadn't seen that data. 

But something weird is going on here ... They don't seem to test 125, but they do 80 and 100 ... BUT nobody refers to those settings as anything other than "extended" ISOs.  They aren't included in Program mode or Auto ISO.

This may be one of 'those' cases where the science just wasn't convenient for the marketing folks.

I will care more if Amex decides to give me purchase protection on my unit that got soaked ...
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: D800E -Truly Amazing Low-Light Performance
« Reply #95 on: November 24, 2012, 10:47:25 am »

Hi,

The info given by DxO is probably a good description of the ISO standard. The problem is that the standard described is based on sensor saturation, also known as the full well capacity (FWC). So what DxO says is that ISO is based on sensor saturation. But, sensor saturation would only occur at base ISO. When you increase ISO, underexposure will result, but how are you calculating ISO then?

Some MFDBs show normal ISO curves while some don't.

If you compare Hasselblad and Phase One curves enclosed it is obvious that DR curves are quite a bit similar but ISO curves are very dissimilar.

I'm pretty sure that the underlying data is consistent but I have difficulty in making sense of the presentation of ISO, especially if DxO-s discussion of the ISO standard is taken into account.

Best regards
Erik




I am so glad YOU said this. It almost seems to me as I read it that I have new questions that go unanswered, and some of it seems difficult to discren meaning. sensorgen.info seems to make sense to me in terms of interpretation of how the whole camera actually works, but isn't quite as up to date as some of DxO's info.

What are your thoughts on sensorgen?
Logged
Erik Kaffehr
 

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8915
Re: D800E -Truly Amazing Low-Light Performance
« Reply #96 on: November 24, 2012, 11:22:21 am »

Hi,

The info given by DxO is probably a good description of the ISO standard. The problem is that the standard described is based on sensor saturation, also known as the full well capacity (FWC). So what DxO says is that ISO is based on sensor saturation. But, sensor saturation would only occur at base ISO. When you increase ISO, underexposure will result, but how are you calculating ISO then?

Hi Erik,

Cutting the exposure signal inhalf, and applying double the ISO setting, should result in the same Raw signal level when ISO controls the ADC gain (within the statistical probability of shot noise+readnoise). Any deviation in the Raw signal level is recorded as ISO deviation.
For truely ISO-less sensors (no ADC gain control by the ISO setting) that doesn't work, the signal will be half as well.

Quote
Some MFDBs show normal ISO curves while some don't.

That means that some are not truely ISO-less.

Cheers,
Bart
« Last Edit: November 24, 2012, 11:24:47 am by BartvanderWolf »
Logged
== If you do what you did, you'll get what you got. ==

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 18123
  • When everyone thinks the same, nobody thinks
    • My website
Re: D800E -Truly Amazing Low-Light Performance
« Reply #97 on: November 24, 2012, 11:24:47 am »

... It almost seems to me as I read it that I have new questions that go unanswered, and some of it seems difficult to discren meaning...

DXO is French. They think in French, they speak it, and then it is translated to English. As the saying goes, something always gets lost in translation. Me thinks (in English... I think) ;)

thierrylegros396

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1947
Re: D800E -Truly Amazing Low-Light Performance
« Reply #98 on: November 24, 2012, 11:25:29 am »

Strange.  Hadn't seen that data. 

But something weird is going on here ... They don't seem to test 125, but they do 80 and 100 ... BUT nobody refers to those settings as anything other than "extended" ISOs.  They aren't included in Program mode or Auto ISO.

This may be one of 'those' cases where the science just wasn't convenient for the marketing folks.

I will care more if Amex decides to give me purchase protection on my unit that got soaked ...

I've made a lot of tests this afternoon, and now I've understood almost everything.

So, I've made manual shots (@f3.5) of a colorchecker, 1st 1/4s @ISO80, 2nd 1/5s @ISO100, 3rd 1/6s @ISO125.

All shots are properly exposed (ETTR) and need only a slight correction (max 0.1Ev) to obtain exactly same exposure in LR4.2.

Results: RAW with same exposure exhibit more noise @ISO100 and even more @ISO125, so DXOMark DR results are correct !

JPG doesn't need any exposure correction, exposure really identical, but it's clear that ISO125 is progressively compressed in the highlights, but neither ISO80, nor ISO100, so DPReview results are correct too !

So, if you shoot RAW, use ISO80 if it's possible.

Also noticed that the 2 last patches of my colorcheker were blinking (@ISO80 and 100), but RAW data were not clipped !!!

Have a Nice Day.

Thierry
Logged

thierrylegros396

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1947
Re: D800E -Truly Amazing Low-Light Performance
« Reply #99 on: November 24, 2012, 12:32:21 pm »

From the Full SNR curve, you can read DR and SNR18%.
Is that right  ???

Another question, with LR or DXO, what's the SNR possible improvement (in dB) without significant loss of resolution ?!
« Last Edit: November 24, 2012, 12:38:19 pm by thierrylegros396 »
Logged
Pages: 1 ... 3 4 [5] 6 7   Go Up