Marbuel
Leading Member
Sorry about "dramatical" ... In Czech we can use such word to describe "huge" or "significant", "unexpected" etc. ... I did not think about the right English equivalent 
Marek
Marek
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
25 MP APS-C coming?
- Thread starter wildkat2
- Start date
MightyMike
Forum Pro
lets think of it this way... with the 5x crop factor a super zoom that goes to 576mm equiv. is really at that point a 115mm lens on FF or approx. 77mm on an APS-C...
Now if you crop the shot with an 77mm lens on 16mp APS-C to 576mm (1.5mp crop) does it resolve the same amount of detail as the 12mp 1/2.33 sensor? well since the P&S has 8 times the pixels it'll likely win, however if the APS-C had 8 times more pixels approx. 130mp then it should match the P&S with the same amount of crop.
This is all assuming the lens is up to the job, but is a P&S lens can be made up to the job then a larger format lens should be able to be made up to the job
--
Mike from Canada
'I like to think so far outside the box that it would require a telephoto lens just to see the box!' ~ 'My Quote
'
http://www.airliners.net/search/photo.search?sort_order=views%20DESC&first_this_page=0&page_limit=180&&emailsearch=mighty_mike88%40hotmail.com&thumbnails=
Winder
Senior Member
Then there would be no reason for Nikon to make the D3s. They would just make a D3x and everyone could downsize the image.
I looked at replacing my Canon 5D with an A900 and worked with a lot of A900 files trying to match the Nikon D700 output at high ISO by downsizing. It does not work, you can get close. The advantage is that the A900 has smaller photosites and as a result the noise that does appear is finer and less noticeable. The A900 is only slightly better than my Canon 5D at higher ISO, but the noise is less noticeable and less objectionable because it appears as a finer grain.
Contax 645, Canon 5D, Olympus E-3 (sold), Playing with a K-7
I looked at replacing my Canon 5D with an A900 and worked with a lot of A900 files trying to match the Nikon D700 output at high ISO by downsizing. It does not work, you can get close. The advantage is that the A900 has smaller photosites and as a result the noise that does appear is finer and less noticeable. The A900 is only slightly better than my Canon 5D at higher ISO, but the noise is less noticeable and less objectionable because it appears as a finer grain.
--
Contax 645, Canon 5D, Olympus E-3 (sold), Playing with a K-7
That's not correct. The two sensors create market separation.
Also, the higher resolution needs more CPU power, which costs money.
The higher-MP sensor's wafer-yield rate might be lower, too.
Well, you picked a camera known for its poor high ISO.
You should have compared the D3x or the 5D MkII to compare with the D700.
For such a comparison to be meaningful you have to apply noise reduction until the higher-MP shot at the resized size has the same level of detail as the native low-MP sensor.
I'm not sure whether you did that. Your comment about the finer noise of the A900 seems to indicate that you did not.
--
'Well, 'Zooming with your feet' is usually a stupid thing as zoom rings are designed for hands.' (Me, 2006)
'I don't own lenses. I pwn lenses.' (2009)
My Homepage: http://www.JensRoesner.de
CharlieDIY
Veteran Member
When I was a kid, shortly after dragons roamed the earth, experts opined that no drag racer could reach more than 144 or 145mph. I haven't followed the drag racing much, but I've seen short track (half the usual track length) hit that a couple summers ago.
Things do change, often much more than any of us believe possible.
--
Charlie Self
http://www.charlieselfonline.com
Quartermile (usual track length) Top Fuel are now beyond 330mph
http://en.wikipedia.org/wiki/Top_Fuel
330 mph = 147 m/s = 530 km/h
speed of sound at 0m elevation is 340 m/s
--
'Well, 'Zooming with your feet' is usually a stupid thing as zoom rings are designed for hands.' (Me, 2006)
'I don't own lenses. I pwn lenses.' (2009)
My Homepage: http://www.JensRoesner.de
For god's sake get out of your cave. There are bunch of these self-proclaimed "experts" who are stuck on the idea that more MP is bad. It's not. More MP == better IQ. This comes straight from Eric Fossum (inventor of CMOS)
see
http://forums.dpreview.com/forums/read.asp?forum=1000&message=37409155
see
http://forums.dpreview.com/forums/read.asp?forum=1000&message=37409155
paulkienitz
Veteran Member
The D3s is built for pure speed. Part of that is to reduce the number of bytes that have to be pumped through the various buses and pipelines. My point is that there is nothing wrong in the finished image with using more megapixels. I don't mean there's no advantage to building a camera to compromise resolution for the sake of some other area of performance.
Yes, I got your point and agree.
Maybe I side-tracked a bit.
--
'Well, 'Zooming with your feet' is usually a stupid thing as zoom rings are designed for hands.' (Me, 2006)
'I don't own lenses. I pwn lenses.' (2009)
My Homepage: http://www.JensRoesner.de
Suhas Sudhakar Kulkarni
Forum Enthusiast
I guess this is a half-rumour.
I do believe Nikon d400 and sony a700 upgrade is in the wings but my guess is they will be about 20 mp.
Even if sony did develop a 24 mp sensor, I doubt they will bring it one board now. Otherwise the users will not be tempted to upgrade for next two upgrade cycles (approx 2 years each) because any upgrade from 24 mp will be too little.
--
Suhas
I do believe Nikon d400 and sony a700 upgrade is in the wings but my guess is they will be about 20 mp.
Even if sony did develop a 24 mp sensor, I doubt they will bring it one board now. Otherwise the users will not be tempted to upgrade for next two upgrade cycles (approx 2 years each) because any upgrade from 24 mp will be too little.
--
Suhas
Cideway, I have to admit that I don't have the thorugh understanding of the lens and sensor resolution, but data seems to indicate that:
Even if a lens does not outresolve an 8.5mp sensor gives even more resolution when it is paired with a higher resoltion sensor. See the data below, 18-55mm f/3.5-5.6 IS (the cheapest zoom lens by Canon) provides more resolution when it is put on a higher MP body (even though it was not outresolving the less MP sensor in the first place):
Even if a lens does not outresolve an 8.5mp sensor gives even more resolution when it is paired with a higher resoltion sensor. See the data below, 18-55mm f/3.5-5.6 IS (the cheapest zoom lens by Canon) provides more resolution when it is put on a higher MP body (even though it was not outresolving the less MP sensor in the first place):
GordonBGood
Veteran Member
Drawing general conclusions for the general OP question by using current results from specific current technologies, whether from Sony or Canon is meaningless, as a new 25 MP (or more) APS-C sensor would likely be of new technology. The Canon sensors as used in the Canon 7D, T2i, and 60D have a known problem with Dynamic Range (DR) at low ISO's and with red colour separation that affects their DxOMark scores, making such comparisons for the purpose of talking about pixel density image quality pointless.
However, NorthwestF's comment that the 18 MP Canon sensor's are still less than the resolution of good lenses and that their actual pixel density is almost that of a 24 MP APS-C sensor is valid.
Regards, GordonBGood
KentG
Senior Member
As I said during that original discussion the theoretical maximum capability of APS-C is closer to around 21 MP.
But the main issue is that few current lenses are going work well with that pixel density that would be equal to a 45 MP FF sensor. The equivalent line pair resolution of the optics would have to be upwards of 100 to work and few lenses fit that category.
Maybe they are going to come out with some new kind of optical technology to make it work.
Kent Gittings
But the main issue is that few current lenses are going work well with that pixel density that would be equal to a 45 MP FF sensor. The equivalent line pair resolution of the optics would have to be upwards of 100 to work and few lenses fit that category.
Maybe they are going to come out with some new kind of optical technology to make it work.
Kent Gittings
KentG
Senior Member
Yes but those figures are not particularly good. Better is relative as that lens is not super in any of the 3 although it is the redesign for the 50D and up sensors. A 50/1.4 lens from Canon should show around 80 or more in all three cameras at the proper aperture.
Kent Gittings
Kent Gittings
From dxomark --> our publication --> Dxomark insights --> "Contrary to conventional wisdom, higher resolution actually compensates for noise
" (due to formatting of dpreview forum text, hyperlink to the particular page does not work)
"For equal, normalized SNR, a high-resolution camera is still better than a low-resolution camera. While it is always possible to simulate a low-resolution camera using a higher resolution camera (since downsampling is easy), it is not possible to simulate a high-resolution camera using a lower-resolution camera other than by interpolating or inventing data. "
" (due to formatting of dpreview forum text, hyperlink to the particular page does not work)
"For equal, normalized SNR, a high-resolution camera is still better than a low-resolution camera. While it is always possible to simulate a low-resolution camera using a higher resolution camera (since downsampling is easy), it is not possible to simulate a high-resolution camera using a lower-resolution camera other than by interpolating or inventing data. "
Actually, that is exactly my point - even though the lens is so-so on a lower MP camera, it still shows more resolution on a high MP camera. And this is true for any lens (I don't have the mathematical expertise to explain this, but this is what is happening in reality). I am sure someone with better mathematical understanding can explain this.
GordonBGood
Veteran Member
Luis, the increasing computer requirements and data storage with increasing Mega Pixels (MP) may slow down once the number of MP reaches the maximum resolution of the best lenses at their optimum aperture at perhaps 50 MP for APS-C: I believe that once this happens and if MP keeps going up (for reasons of eliminating the need for the Anti Aliasing (AA) filter) to reduce the appearance of colour moiré patterns for certain high resoluiton detail patterns, the camera raw files will become less "raw" in that the camera will likely start to reduce them to a less effective resolution before output.
Regards, GordonBGood
GordonBGood
Veteran Member
Jeff, again according to my calculations, you would just start to see the effects of diffraction for a 16.4 MP APS-C sensor at about f/8, they would start to be obvious at f/11, and completely diffraction limited for about f/16.
Regards, GordonBGood
tko
Forum Pro
Anyone who believes imaginary marketing pitches deserves to fall off the edge of the ocean on that flat earth.
Does anyone really believe a quantum leap in efficiency is going to appear soon? How do you go from ISO's in the thousands to ISO's in the hundreds of thousands? Two order of magnitude improvement? Even the simplest of calculations will show this isn't possible. Do you also believe the gas company is hiding a secret car that can get hundreds of miles to the gallon?
Remember, gasoline only has so much energy, and light only has so many photons. Once you've captured every single photon with 100% efficiency there is no way to get better, no matter how much you'd like to believe.
Does anyone really believe a quantum leap in efficiency is going to appear soon? How do you go from ISO's in the thousands to ISO's in the hundreds of thousands? Two order of magnitude improvement? Even the simplest of calculations will show this isn't possible. Do you also believe the gas company is hiding a secret car that can get hundreds of miles to the gallon?
Remember, gasoline only has so much energy, and light only has so many photons. Once you've captured every single photon with 100% efficiency there is no way to get better, no matter how much you'd like to believe.
Keyboard shortcuts
- f
- Forum