D500/D610/D4s/D5 at ISO 6400

DX cameras need to be compared at ISO settings that are one stop lower. In actual use, a DX camera with a specific lens model, is nearly equivalent to an FX camera with the same lens plus a TC-14.

For example, I typically shoot figure skating with the D5 and 70-200 VR II plus TC-14E wide open at f/4. If I used a D500, I'd omit the TC, use f/2.8 and set ISO one stop lower to obtain the same shutter speed: Makes a D500 sound attractive.

The FX advantage at higher ISO settings is only real when equivalent lenses are not available. For example, an FX camera with a 200 f/2 would be equivalent to a DX body with a 135 f/1.4, which is not obtainable. Also, I can use the D5 with the 80-400 VR at f/5.6, but there isn't an equivalent lens (55-270 or 60-300 f/4) for a DX camera: Doesn't make the D500 sound so attractive.
 
No, if you are printing or outputting to the same final complete-image size.

Yes, if you are outputting to the same dpi (larger print from higher-resolution camera), and the dpi value is low enough to see the individual-pixel contributions.

--
Source credit: Prov 2:6
- Marianne
Excuse my ignorance . I have read that the bigger the pixel the more photons it is capable of taking in without saturation coming into effect which produces noise .
Saturation does not produce noise, rather it is just the point at which highlights clip. Noise is present at all signal levels. In fact, pushing the sensor pixels into saturation removes the shot noise (leaving just the fixed-pattern noise).
Example : both d90 and d700 have the same amounts of pixels but the pixels on the d700 are bigger being it a FX sensor . These pixels are capable of intaking much more light photons without saturation
You do not need to consider pixel size. It is just the area of the sensor. Both cameras, given similar sensor technology, can collect about the same number of photons per square millimeter of sensor area without saturating, but the FX camera has 2.25x as many square millimeters of area.
So the D700 for a given amount of light produces less noise then a d90 . Am I right ?
It depends on how you measure "amount of light," i.e., whether you are talking about light emitted by the subject, or total light received by the sensor.

If you take a photo with both cameras at the same settings, for example, 1/100 shutter speed, f/2.8, 50mm lens, same subject and lighting (but different subject distance), then the FX sensor gathers more light and the overall image SNR (signal/noise ratio) is better.

But if you change the lens for the DX camera to 35mm f/2 so that the subject distance is the same as for the FX camera, then both sensors receive the same amount of light and the images have the same SNR.

Ultimately, if you are working at base ISO on both cameras, the FX sensor will be able to collect more photons without saturating, and there you are correct that the best FX image will be better than the best DX image. However, again, you do not need to consider the size of the pixels.
Is this noise I am talking about the same noise as shot noise or are they different types ?
At high signal levels, the shot noise dominates. Shot noise is proportional to the square root of the number of photons received, or is the same as the square root of the number of electrons collected if you want to use electrons as the measuring unit. Thus the ratio of signal to noise is also proportional to the square root of the number of photons collected, i.e., SNR improves as you collect more photons.

For example, the D5 can register about 130,000 electrons at base ISO. The SNR for a maximum-brightness highlight will be about 360:1. At mid-gray, which is about 3.5 stops below that, the electron count is 11,500 and the SNR is 107:1. If you are using the camera at ISO 10,000, the electron count for mid-gray is only 115 and the SNR drops to only 10.7:1 (at best; read noise hasn't been included in this figure).
Again , excuse my ignorance , just trying to learn
--
Source credit: Prov 2:6
- Marianne
So all those high iso noise test between FX and DX can be misleading if you don't change the parameters be it aperture or subject to lens distance . Both cameras in order to test them correctly must be recieving the same amount of of light . But then again they should have same SNR if they are gathering the same amount of light , right ?
 
So all those high iso noise test between FX and DX can be misleading if you don't change the parameters be it aperture or subject to lens distance .
Yes, it is not always correct to assume the FX camera will produce a better image.
Both cameras in order to test them correctly must be receiving the same amount of light .
Using equivalent lenses on each camera, e.g., 300 f/2.8 on FX and 200 f/2 on DX, does mean each sensor receives the same amount of light, i.e. same number of photons. That is one usage case.

Sometimes, however, if one is limited in choice of optics, providing each camera with equivalent lenses may not be feasible. For example, if all you have is a 24-70 f/2.8 zoom, and you take a photo with an FX camera and that lens at 50mm f/2.8, then with the DX camera at 33mm f/2.8, the DX camera will be at a disadvantage. That is another usage case.
But then again they should have same SNR if they are gathering the same amount of light , right ?
Yes. It is worth pointing out that when both sensors receive the same amount of light, the DX camera will be set to an ISO that is one stop lower in order to produce the same brightness in the camera JPEG image and the same values in the RAW file.
 
Just try to compare the high ISO images with low light setting in the "Image Comparaison Tool", not with the "Daylight setting". Then you could really see somme differences and you could really see the ISO advantage of FX vs FX (just at sensor level).

Not much differences with Daylight setting.

1 - But, to compare really FX vs DX, you need to compare with equivalent lens focale and aperture to have the same DOF.

Example :

FX : ISO 400, 1/100e, F/4, 50mm lens

DX : ISO 200, 1/100e, F/2.8, 35mm lens

You have the same picture, same DOF, same exposure, but ISO is higher on FX. Here is the advantage of DX vs FX, no need to crank the ISO higher to have sufficient DOF.

Where the FX wins, it's at aperture near 1.8, 1.4 (wide open) because there is no equivalent lens for DX ( remember : aperture / 2).

Advantage DX (but have a look at point 3)



2 - Resolving power of lenses


One advantage of FX, because of the resolving power limitation of the lens, you've got better results with a FX sensor vs DX sensor (at the same resolution).

DX sensor, because of the pixel density, needs very good lenses too resolve the full definition and most of the time, the lens is the limiting factor.

FX sensor have the same restriction, but it appears latter !

Advantage FX



3 - The third point, is about shutter speed for action picture
.

To achieve a perfect result with DX, because of the higher pixel density, you need to choose a higher shutter speed to freeze the movement at the pixel level. And to achieve that, you need too increase the ISO then you loose the point 1 advantage.

Here, the FX wins, again.



--
Arnaud Marie
France
http://arnaudmarie.zenfolio.com
 
Last edited:
I really apreciate your help and the time you take to explain and answer my questions , just one more thing , reviewers , when doing the testing , are the taking these things into consideration . I asume they should know what they are doing , no ? Or are they as confuse as I was .
 
Oh wow! I just realised... this camera takes photographs as well! Amazing... see you later! (exits stage right to explore the possibilities)
Only 11 posts in, and I just might have the "honor" of being the first to ignore you. I don't think I'll be the last.
 
I really apreciate your help and the time you take to explain and answer my questions , just one more thing , reviewers , when doing the testing , are the taking these things into consideration . I asume they should know what they are doing , no ? Or are they as confuse as I was .
From what I've seen, it is rare for reviewers to compare DX and FX cameras, using equivalent lenses.

Most simply assume ISO 5000 on FX is the same as ISO 5000 on DX, and never think about how to use a DX camera in an equivalent manner.

One exception is Stany Buyle, who did post some DX/FX comparisons recently that included examples with the DX camera at a 1-stop lower ISO setting.
 
Wish more people would adopt this mentality. The digital age has certainly ushered in a level of performance measurement that a lot of people waste untold hours measuring performance that exceeds their abilities and needs.
 
I really apreciate your help and the time you take to explain and answer my questions , just one more thing , reviewers , when doing the testing , are the taking these things into consideration . I asume they should know what they are doing , no ? Or are they as confuse as I was .
From what I've seen, it is rare for reviewers to compare DX and FX cameras, using equivalent lenses.

Most simply assume ISO 5000 on FX is the same as ISO 5000 on DX, and never think about how to use a DX camera in an equivalent manner.

One exception is Stany Buyle, who did post some DX/FX comparisons recently that included examples with the DX camera at a 1-stop lower ISO setting.

--
Source credit: Prov 2:6
- Marianne
Totally agree ! I just modified my previous post (512) to explain this.... I didn't take the time the first time I wrote.

--
Arnaud Marie
France
http://arnaudmarie.zenfolio.com
 
Last edited:
So if i get it right, the so called 'FX sensor is better'-hype, is not so much because of sensor design, pixel size, etc, but more so because DX camera's (often) have inappropriate lenses to compete with the FX counterparts, and because of that loose the stop of light (that they should have gotten extra), to compensate for being DX rather then FX, yielding the same end score if, such lens is found.

Like FX 24-70mm F2.8, 70-200mm F2.8

Two incredible popular workhouses.

On FX they are what they are, no chance in parameters.

On DX they are 'darker' (or rather do not yield enough light for the sensor) and to much tele to yield the same image, thus are not comparable.

16-47mm F2 and 47mm-133mm F2 lenses for DX would basically mean the advantage FX 'is so hyped about' suddenly (mostly) falls away?

What about colour accuracy? Ppl Like CMOS colours (unless medium format) more then CCD (unless a few ppl that want old skin tones). People like FX colours more then DX (see D700 vs D300). Is this just myth? Or is there some truth in the fact that FX (can - not always) produce better colour (and thus better image)?

Like My D800 files are much much much better then the ones from D80, no matter what i try, D800 files always are preferred, for MP, sharpness, colour, noise, etc. Am i falling into some myth here? Cause D80 in your (Marianne) theory should be close to the FX in noise, colour, etc. Unless I understand wrong.

What about (theorical) lens designs? I mean the F2 zooms i talk about above here (like 47-133mm F2). Would they (in theory) have the same size as their FX counter part? I'd think not, Looking at the few exeptions thrown at us, like Sigma F1.8 art zooms (very very big lenses, with fewer zoom then Fx 2.8 zooms). I think size, portability for DX F2 zooms is a bit problem.

Also the D5S(R) tests (photozone.com) that are appearing now, absolutely show that 50 megapixels (FX) is a massive stress tests for lenses borders and corners. Even the very well received Canon 35mm 1.4 II falls quite big behind there. Seeing how (arguably) some of the best lenses in the world, like the Canon, Sigma art lenses etc, are struggling to get sharp corners, wouldn't this play in favor of FX rather then DX? the D500 has aprox 46MP if produced in FX format (same pixel density). But because D5 is bigger sensor it needs less pixels per square inch, leading into same MP, with less penalty cause corner sharpness drop. Perhaps lens design limitations of this era, are making FX slightly more preferable, cause FX has a little more headroom for resolutions that are high enough to satisfy people (like D8x0, D750 etc), with with still enough sharpness transmitted from the better lens design (not being punished by the already now shown D5S(R) megapixel lens punishment).

In other words: DX has to few lenses now. Most people use fitting FX sensor capable lenses instead. This yields incredible DX sharpness (cause center- wich obviously is best of lens). But because you have to push more pixels on same density, if they want to compete with FX, they already would need 36-50 megapixel DX sensors. And extrapolating results of photozones 5DS tests, current lenses barely can produce that much sharpness (mind you that would equal pixel density of FX sensor of more then 100 megapixels, more then double of the already stress test proving 50MP fx result). In other words, since the current market mostly has FX capable lenses, the FX sensor can spread the pixel better on the lens surface with a lower drop in sharpness compared to DX sensor with same MP count. Only if specialized DX lens appears, this can be countered (Sigma Art zooms seems like the only, but very good so, solution). But the big lack of these kind of lenses, makes FX the more capable system overall?
 
Last edited:
Like FX 24-70mm F2.8, 70-200mm F2.8

Two incredible popular workhouses.

On FX they are what they are, no chance in parameters.

On DX they are 'darker' (or rather do not yield enough light for the sensor) and to much tele to yield the same image, thus are not comparable.
FX lens used on DX sensor and the images are darker? It should be the reverse way, DX lens used on FX sensor and the corners of the images are darker.

FX lenses are designed to yield enough light for FX sensor, and of course DX sensor will get enough light from FX lenses.
 
I have taken a look at the ISO 6400 sample pictures. So for me the quality is useless. To much noise. At the moment I am using a Nikon D90 Camera. It has only 12 mpixels. Maybe I would like the D610. It has more mpixels. Ofcourse is a fullframe. After reviewing some sample pictures and seeing the useless ISO 6400. I'm considering the D700. FullFrame will give me more depth control. Ofcourse i'm not an expert. But considering the D700, D610 and D810 the D610 must be a bad choice. To many pixels to take advantage from a D90.
 
I have taken a look at the ISO 6400 sample pictures. So for me the quality is useless. To much noise. At the moment I am using a Nikon D90 Camera. It has only 12 mpixels. Maybe I would like the D610. It has more mpixels. Ofcourse is a fullframe. After reviewing some sample pictures and seeing the useless ISO 6400. I'm considering the D700.
I presume you mean the D750, rather than the D700 which is several generations old?
FullFrame will give me more depth control. Ofcourse i'm not an expert. But considering the D700, D610 and D810 the D610 must be a bad choice. To many pixels to take advantage from a D90.
How many pixels do you consider "too many?"

Do not fear pixels. They are not as scary as "24 million" or "36 million" might suggest.

--
Source credit: Prov 2:6
- Marianne
 
Last edited:

Keyboard shortcuts

Back
Top