Ad
Ad
Ad
Pages: « 1 [2] 3 4 5 »   Bottom of Page
Print
Author Topic: Comparisons From DT's 250 and 260 testing in the Library  (Read 7171 times)
gazwas
Sr. Member
****
Offline Offline

Posts: 385



« Reply #20 on: February 12, 2014, 05:48:50 PM »
ReplyReply

I sold my P65 as I was never 100% happy with the results when shooting interiors, especially considering all the hoops you had to jump through. Doug's test confirmed my feeling that while the CCD's deliver beautiful results in well illuminated conditions, in challenging lighting they never lived up too their price IMO.

The IQ250 has shocked me at how good it is and Doug's test images look wonderful. Over sharpened yes, but that is easily addressed. Just a big shame the cost of the IQ250 is still so high especially considering the smaller (which does't bother me) size chip.
Logged

trying to think of something meaningful........ Err?
BJL
Sr. Member
****
Offline Offline

Posts: 5170


« Reply #21 on: February 12, 2014, 06:08:27 PM »
ReplyReply

I've never heard that a specific lens design can have more DoF than another, seems to defy physics. But as DoF is subjective ...
Different lens designs can produce somewhat different perceived DOF at the same focal length and f-stop; I understand that this is mainly due to different position of the aperture diaphragm.  Some lenses produce a "circle of confusion" (out-of-focus blurring of the light from a single point) that is brightest at its center, with brightness decreasing distinctly towards the edge of the CoC, whereas others give more uniformly brightness, or perhaps are even brightest at the edge of the CoC. The former give a sense of more DOF (and generally nicer, softer bokeh) while the latter is allegedly better for the sharpness of in-focus details, so macro lenses tend to go in that direction.

The standard DOF formulas just look at the radius of the CoC for an object at a given distance in front of or behind the plane of focus, not the brightness distribution within that disk. (Yes, the "circle" of confusion is properly called a disk!)


Back on topic: thanks for all concerned for the tests and the careful scrutiny and discussion of them; it seems that a lot of myths and misconceptions about sensor types are being dispatched.
Logged
BJL
Sr. Member
****
Offline Offline

Posts: 5170


« Reply #22 on: February 12, 2014, 06:16:42 PM »
ReplyReply

That is a bit of revisionist history. CCD used to be better than CMOS. There is a reason all the high end astrophoto cameras were CCD. ...
I am not sure who you are accusing of revisionist history, but I agree with your basic point: the early CMOS sensors were used as a cheaper (and less power hungry) but inferior alternative to CCDs, mostly in small, low cost cameras. But the development of the active pixel approach and Canon's refinements already pushed CMOS ahead in most respects (while NOT being cheaper for DSLR-sized sensors!), and further developments like on-chip column parallel ADCs have opened the performance gap.

Note also that one reason for many specialized applications staying with CCDs is that it apparently is (or was) less expensive to put a new, low volume design into production with CCD that with active pixel CMOS. But astronomy and other custom-designed sensors are moving towards CMOS; check out Teledyne-Dalsa's website.
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7888


WWW
« Reply #23 on: February 12, 2014, 11:52:36 PM »
ReplyReply

Hi Stefan,

You are an engineer, so you are not supposed to understand this…

Best regards
Erik


This needs to settle down a bit, but in my brain it already forms a BIG question: if the CMOS is SO much better, who will stil buy a CCD ?


Greetings from Germany
Stefan
Logged

torger
Sr. Member
****
Offline Offline

Posts: 1590


« Reply #24 on: February 13, 2014, 01:18:21 AM »
ReplyReply

CCDs are still larger. But when full-frame CMOS arrives I think those that want the best will want that, but Phase One has clearly indicated with the current IQ250 pricing that full-frame CMOS will be more expensive than the current full-frame CCD.

So those that want to save some money and don't need the superior CMOS performance may buy a CCD back instead.

Then we have the relatively poor tech cam performance due to poor wide angular response, I think this test masks the color cast issues a bit since it's hard to see slight shifts in the colorful ceiling but I'm quite sure they're there. Just compare the colors in the ceiling of the IQ260 32 HR image with the IQ250 32 HR image. Current CCDs are clearly better in this aspect, but I don't know if it's an inherent problem of CMOS or if it's just about pixel size. If large CMOS gets the backside illuminated manufacturing process we could see a drastic improvement in angular response. I hope for that.

My guess from the current result is that the 32HR is barely usable if you need color stability when shifted, but that the 40mm Rodenstock could be okay. So a full-frame CMOS with the current sensor technology and pixel pitch would not be able to go wider than 40mm with shifting. For me as a landscape photographer that would be okay, but an interior photographer I would guess would want the 32 to work well too.

Color will also be a factor, there's a lot of mythology and mystery around the magic of "CCD color", if the CMOS CFAs + profiling won't turn out as pleasing as the current CCDs, CCDs may still have a chance, but then mainly in portrait I think, not in tech cam space. At some point CMOS will take over of course, just as it has taken over in DSLR space. I hope it's sooner rather than later so it can trickle down the second hand market to my hands within reasonable time :-).

(I'd rather want a 48x36mm sensor that works with the SK35mm, I find that to be a more balanced system, I think about more values than just resolving power, but I don't think we'll see such a thing, as I think resolving power is the main selling point for these type of cameras.)
« Last Edit: February 13, 2014, 02:33:49 AM by torger » Logged
Paul2660
Sr. Member
****
Offline Offline

Posts: 1967


WWW
« Reply #25 on: February 13, 2014, 08:01:12 AM »
ReplyReply

Question to me is when a full frame CMOS Arrives will it be by Sony?  What a lot of people seem to think is the results are just because it's CMOS vs CCD.  Other companies current CMOS chips if sized to MF I feel you would be seeing a lot more noise in the shadows along with color splotches and harsh  noise in the shadows.

My point being Sony has a excellent track record with CMOS and high DR since the D800.  If when the full frame comes out next year or later this year, it may not be a  Sony chip.  Sony is hard at work on their 54 Mp chip for 35mm and Sony has commented that they are planning on on 2015 release.  This is no rumor it's in several Sony press releases. 

If the next chip comes out and it a Dalsa then it's up in the air as to what they will supply.  They are a huge fab company but I don't know if they have anything in CMOS in the 35mm or larger market.  Lots of things may different like how the current live view on the Sony works, which us as good as it gets but Sony does know a bit about live view.  Can Dalsa or another company get to the level of where Sony is? 

It will be interesting.

Paul C





Logged

Paul Caldwell
Little Rock, Arkansas U.S.
Photography > http://photosofarkansas.com
Blog> http://paulcaldwellphotography.com
Rob Whitehead
Jr. Member
**
Offline Offline

Posts: 71



« Reply #26 on: February 13, 2014, 10:46:48 AM »
ReplyReply

I sold my P65 as I was never 100% happy with the results when shooting interiors, especially considering all the hoops you had to jump through. Doug's test confirmed my feeling that while the CCD's deliver beautiful results in well illuminated conditions, in challenging lighting they never lived up too their price IMO.

The IQ250 has shocked me at how good it is and Doug's test images look wonderful. Over sharpened yes, but that is easily addressed. Just a big shame the cost of the IQ250 is still so high especially considering the smaller (which does't bother me) size chip.

Yes, but I reckon your P65 is just fine for shooting landscape exteriors! Just don't try a 10 second exposure in low light like I did this evening - ughh. Definitely CMOS territory.

 Wink
Logged

Carpe lucem
MrSmith
Sr. Member
****
Offline Offline

Posts: 918



WWW
« Reply #27 on: February 13, 2014, 10:54:46 AM »
ReplyReply

"So those that want to save some money and don't need the superior CMOS performance may buy a CCD back instead."

hang on? if you have been drinking the kool aid like everyone else then surely CCD is the pinnacle of imaging fidelity?

 Roll Eyes
Logged
henrikfoto
Sr. Member
****
Offline Offline

Posts: 708


« Reply #28 on: February 13, 2014, 11:04:01 AM »
ReplyReply

So the IQ250 is just a bigger D800e...
With the new profile in Capture One we will get the same results pr. pixel from the Nikon?
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7888


WWW
« Reply #29 on: February 13, 2014, 12:02:17 PM »
ReplyReply

Hi,

Yes, the small size of the sensor may be a bit painful. On the other hand, I would think small size may be interesting for Harblei HCam and Alpa FPS shooters who can use Canon's T&S lenses. In some situations live view may be very helpful and clean shadows are also nice.

The price tags is pretty high for a small sensor, but as you say, it may push some decent size CCDs in the second hand market.

Best regards
Erik


CCDs are still larger. But when full-frame CMOS arrives I think those that want the best will want that, but Phase One has clearly indicated with the current IQ250 pricing that full-frame CMOS will be more expensive than the current full-frame CCD.

So those that want to save some money and don't need the superior CMOS performance may buy a CCD back instead.

Then we have the relatively poor tech cam performance due to poor wide angular response, I think this test masks the color cast issues a bit since it's hard to see slight shifts in the colorful ceiling but I'm quite sure they're there. Just compare the colors in the ceiling of the IQ260 32 HR image with the IQ250 32 HR image. Current CCDs are clearly better in this aspect, but I don't know if it's an inherent problem of CMOS or if it's just about pixel size. If large CMOS gets the backside illuminated manufacturing process we could see a drastic improvement in angular response. I hope for that.

My guess from the current result is that the 32HR is barely usable if you need color stability when shifted, but that the 40mm Rodenstock could be okay. So a full-frame CMOS with the current sensor technology and pixel pitch would not be able to go wider than 40mm with shifting. For me as a landscape photographer that would be okay, but an interior photographer I would guess would want the 32 to work well too.

Color will also be a factor, there's a lot of mythology and mystery around the magic of "CCD color", if the CMOS CFAs + profiling won't turn out as pleasing as the current CCDs, CCDs may still have a chance, but then mainly in portrait I think, not in tech cam space. At some point CMOS will take over of course, just as it has taken over in DSLR space. I hope it's sooner rather than later so it can trickle down the second hand market to my hands within reasonable time :-).

(I'd rather want a 48x36mm sensor that works with the SK35mm, I find that to be a more balanced system, I think about more values than just resolving power, but I don't think we'll see such a thing, as I think resolving power is the main selling point for these type of cameras.)
Logged

torger
Sr. Member
****
Offline Offline

Posts: 1590


« Reply #30 on: February 13, 2014, 12:02:57 PM »
ReplyReply

So the IQ250 is just a bigger D800e...
With the new profile in Capture One we will get the same results pr. pixel from the Nikon?

No. The color filter array (CFA) is different in the IQ250 sensor. It's supposed to be optimized for color fidelity, while it's said that Nikon D800 has compromised it a bit to get better high ISO. And the color profiles in Capture One are different, probably with different design goals.

And I just noticed you get the MF sensor centerfold(tm), but it's horizontal rather than vertical on this sensor (it's weak though, so don't worry) Smiley
« Last Edit: February 13, 2014, 12:10:22 PM by torger » Logged
henrikfoto
Sr. Member
****
Offline Offline

Posts: 708


« Reply #31 on: February 13, 2014, 12:20:30 PM »
ReplyReply

Ok. Thank you Torger!

Would be interesting if someone could test them against each other using the same profile.
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7888


WWW
« Reply #32 on: February 13, 2014, 12:37:15 PM »
ReplyReply

Hi,

Maybe, may be not. I have not seen any real evidence supporting this. As far as I can recall, DxO data is a bit to the contrary. I am a bit skeptical, but I have no experience with Nikon, only Sony and Phase One P45+.

Comparing the two, I know the Sony Alpha 99 I have is more accurate on the ColorChecker than the P45+ I have, and I also know that the P45+ with C1 is the least accurate on the ColorChecker of the converters I have tested. That is what I know. But I also know that accuracy on the ColorChecker is not the whole truth. My SLT 99 (with LR 5.3) has twice the accuracy of my P45+ with Capture One 7.1.3.

I would also argue that if you cannot reproduce a ColorChecker, your colour reproduction is not accurate, but it may still be nice, good or pleasant.

Digital Translation does repro work and they have created according to their info they have created profiles for one of the Phase One cameras they use with a DeltaE (average) of around 2.0, almost twice as good as my SLT99 with LR 5.3. So it is absolutely clear that Phase One backs can do very good reproduction, using the right profiles.

Just to make clear, I enjoy my P45+. It is a good MF digital back and I enjoy working with it. But I don't think it is a silver bullet.

Best regards
Erik

It's supposed to be optimized for color fidelity, while it's said that Nikon D800 has compromised it a bit to get better high ISO.

« Last Edit: February 13, 2014, 01:05:41 PM by ErikKaffehr » Logged

torger
Sr. Member
****
Offline Offline

Posts: 1590


« Reply #33 on: February 13, 2014, 01:05:59 PM »
ReplyReply

Maybe, may be not. I have not seen any real evidence supporting this. As far as I can recall, DxO data is a bit to the contrary. I am a bit skeptical, but I have no experience with Nikon, only Sony and Phase One P45+.

DxO shows less good values for Kodak sensors, a bit better for the Dalsa. Many think the Dalsa has better color. Sony has a few models that score very well, the A99 and the A900 I think it was, others are in the middle of the pack.

I have also not seen any real evidence on that these tradeoffs are made, but I think Doug's article on the matter is correct, that Phase One had a choice when they ordered the sensor from Sony how to optimize the CFAs. What can be seen on DxOmark is that the same sensor technology can have quite different color responses, so CFA filters can change while the underlying sensor is the same. So that it will be different color response than the D800 is guaranteed, but if it's better and will please current Dalsa users is a different question...
Logged
Paul2660
Sr. Member
****
Offline Offline

Posts: 1967


WWW
« Reply #34 on: February 13, 2014, 01:15:20 PM »
ReplyReply

The Phase one profile for the IQ250 at first glance does fine on the d800e files I tested it on. I am hoping to do a bit more detailed work this weekend. 

I never liked the phase one default profile for the d800e.

Paul C
Logged

Paul Caldwell
Little Rock, Arkansas U.S.
Photography > http://photosofarkansas.com
Blog> http://paulcaldwellphotography.com
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7888


WWW
« Reply #35 on: February 13, 2014, 01:29:17 PM »
ReplyReply

I might just have the impression that Phase One put little effort on a competitors file?

If IQ250 CFA design is different from D800 CFA design the IQ 250 profile should not work well with the D800. Either way it seems that C1 did a bad job on the D800.

Best regards
Erik




The Phase one profile for the IQ250 at first glance does fine on the d800e files I tested it on. I am hoping to do a bit more detailed work this weekend. 

I never liked the phase one default profile for the d800e.

Paul C

Logged

torger
Sr. Member
****
Offline Offline

Posts: 1590


« Reply #36 on: February 13, 2014, 02:23:14 PM »
ReplyReply

I might just have the impression that Phase One put little effort on a competitors file?

 Grin

It does come into mind.... but on the other hand, what would I do if I where Phase One? I wouldn't cry if the D800 profile didn't turn out too well.

The accuracy probably not too good if using the IQ250 profile on the D800 due to the different CFAs. However the global sensitivity of the channels are compensated for through the white balance constants embedded in the raw file which should even out the differences, so it's not too suprising that good results can be had anyway. And while I'm sure the CFAs are different, the difference maybe is smaller than they would have been if the sensors came from different manufacturers.
Logged
Paul2660
Sr. Member
****
Offline Offline

Posts: 1967


WWW
« Reply #37 on: February 13, 2014, 02:38:11 PM »
ReplyReply

When viewing 2 image, one with the Capture One D800e and the other with the IQ250 on a D800e file, the differences are subtle, but for me the sky has a better look.  This is a quick side by side two image taken with seconds of each other during a hand held pan.

left image I used the IQ250 profile right image I used the Generic D800e profile.  They are very close and neither has serious problem, i.e. purple shadows, or color miss matches.

Paul C
Logged

Paul Caldwell
Little Rock, Arkansas U.S.
Photography > http://photosofarkansas.com
Blog> http://paulcaldwellphotography.com
AreBee
Full Member
***
Offline Offline

Posts: 138



WWW
« Reply #38 on: February 13, 2014, 03:54:44 PM »
ReplyReply

Quote from: Paul2660
When viewing 2 image, one with the Capture One D800e and the other with the IQ250 on a D800e file, the differences are subtle...

As is the difference between an "Outdoor Daylight" IQ250, IQ180 and IQ280 profile when applied to a D800E landscape photo NEF, in my opinion. The IQ180/IQ280 profile are one and the same?
Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7888


WWW
« Reply #39 on: February 13, 2014, 05:09:30 PM »
ReplyReply

Hi,

I wouldn't call that difference subtle…

Best regards
Erik

When viewing 2 image, one with the Capture One D800e and the other with the IQ250 on a D800e file, the differences are subtle, but for me the sky has a better look.  This is a quick side by side two image taken with seconds of each other during a hand held pan.

left image I used the IQ250 profile right image I used the Generic D800e profile.  They are very close and neither has serious problem, i.e. purple shadows, or color miss matches.

Paul C

Logged

Pages: « 1 [2] 3 4 5 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad