Cost vs Performance: IQ250 vs D800e

[bjanes]

Phase One has finally managed to offer a CMOS sensor with the IQ250 and DXO (if they test the camera) will rank it first in their sensor database, above the D800e which has held the championship title since it became available.

The IQ250 has 50MP, but this is not significantly greater than the 36 MP of the Nikon. The IQ250 sensor at 33mm x 44 mm is less than medium format full frame but offers 1.68 times the sensor area of the 24mm x 36mm sensor of the D800e. From these dimensions, I calculate that the IQ sensor is 6124 x 8165 pixels, or 1.24 times the linear picture height of the D800e. The aspect ratio of the IQ250 would result in less wasted megapixels for a 16 x 24 inch print. If Sony obtains the same performance from the new sensor as with the D800e sensor, this should result in a gain of 0.75 f/stops. The engineering DR of the D800e is 13.4 EV and the predicted DR of the IQ250 would be just above 14 stops, consistent with the DR stated in the PhaseOne announcement.

The IQ250 will cost US $34,900 and the D800e lists for $2,997 at B&H. Does this increased performance warrant the 1160% cost differential between the two cameras? It will for some perfectionists with a large pocketbook, but probably not for most users.

Discussion is welcome.

Bill

[torger]

Would be interesting to know what the sensor pricing is. Since Sony is selling to so many players maybe someone could get hold of some semi-public figure. My guess is that the cost of the sensor itself to the manufacturer is $2500-$3500 or so.

As I said in a different thread Phase One is not about providing good price/performance, it's about providing the best and then charge a lot for it for the customers that can afford it. If you want medium format CMOS there's actually no other alternative at this point, and Phase One has been good at being first with the latest, and also provide quite sweet upgrade deals with better pricing. So if you want to stay in the forefront of image quality, Phase One is the system to invest in. That's what they want you to think, and I think they do a quite good job in delivering.

[synn]

Judging imaging technology by numbers will get one nowhere.
Why not shoot two pictures with each tool by yourself and reach your own conclusions?

[bjanes]

Judging imaging technology by numbers will get one nowhere.
Why not shoot two pictures with each tool by yourself and reach your own conclusions?

Some of my favorite quotes from Lord Kelvin:

    "To measure is to know."

    "If you can not measure it, you can not improve it."

    "In physical science the first essential step in the direction of learning any subject is to find principles of numerical reckoning and practicable methods for measuring some quality connected with it. I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely in your thoughts advanced to the state of Science, whatever the matter may be." [PLA, vol. 1, "Electrical Units of Measurement", 1883-05-03]

    "The true measure of a man is what he would do if he knew he would never be caught."

I would say that physical measurements can tell quite a lot about the performance of imaging systems. However, the final arbiter must come from examination of large prints. The studies should be double blind so as to avoid subjective impressions, perhaps embellished by proponents of the two systems.

I do not have access to the IQ250 but perhaps others can perform such testing, but I am not holding my breath for such a study to appear anytime soon. Of course, the PhaseOne proponents will say that the IQ blows away the D800e just as they previously stated that MFDBs have an additional 5 stops DR over FF 35 mm.

Regards,

Bill

[Bart_van_der_Wolf]

I would say that physical measurements can tell quite a lot about the performance of imaging systems. However, the final arbiter must come from examination of large prints.

Hi Bill,

I agree, and have therefore added the IQ250 to the pre-selectable sensors in my (DOF) Output Quality Planning Tool. The settings should probably also allow to already predict Hasselblad (not in multi-shot mode) and Pentax image quality.

I do not have access to the IQ250 but perhaps others can perform such testing, but I am not holding my breath for such a study to appear anytime soon. Of course, the PhaseOne proponents will say that the IQ blows away the D800e just as they previously stated that MFDBs have an additional 5 stops DR over FF 35 mm.

Not to mention 'CCD color' and 3-dimensional rendering by CCDs. This is going to be fun...

Cheers,
Bart

[Vladimirovich]

just as they previously stated that MFDBs have an additional 5 stops DR over FF 35 mm.

6 stops, 6 !!!

[OliverM]

    "To measure is to know."

To know what to measure is a good measure of what we know.

[RichDesmond]

Some of my favorite quotes from Lord Kelvin:

    "To measure is to know."

    "If you can not measure it, you can not improve it."

    "In physical science the first essential step in the direction of learning any subject is to find principles of numerical reckoning and practicable methods for measuring some quality connected with it. I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely in your thoughts advanced to the state of Science, whatever the matter may be." [PLA, vol. 1, "Electrical Units of Measurement", 1883-05-03]...

As an engineer by trade, I am naturally drawn to measurement, precision and accuracy. And these things can tell us a lot. But...it is also true that not everything important is measurable, and not everything that's measurable is important...

[ErikKaffehr]

Hi,

That is a relevant question to ask. In a way I have been asked the same question regarding why I keep my Hasselblad and my P45+ as I don't find it really superior to my Sony cameras? That is a good question and I don't have a good answer to it. It is about me liking the camera and I'm pretty sure I'd miss it.

Interesting enough after 6 months and something like 1500 frames I still don't know what I feel about it, it is still a learning experience. I would definitively not buy the Hassy again, but it is also inprobable I would sell.

I am quite impressed by the IQ250, based on paper form and what Doug says. There are a couple of reasons I am not going to buy it. The foremost reason is I really cannot afford it, the other one is the crop factor, but that may be less of a factor if there is an adequate spectrum of lenses.

Would I have something like 50-60 k$US to spend I would buy it, with a technical camera and 4-5 lenses. But, for me as for many of us money is a finite resource. So no IQ250 for me.

Best regards
Erik

Phase One has finally managed to offer a CMOS sensor with the IQ250 and DXO (if they test the camera) will rank it first in their sensor database, above the D800e which has held the championship title since it became available.

The IQ250 has 50MP, but this is not significantly greater than the 36 MP of the Nikon. The IQ250 sensor at 33mm x 44 mm is less than medium format full frame but offers 1.68 times the sensor area of the 24mm x 36mm sensor of the D800e. From these dimensions, I calculate that the IQ sensor is 6124 x 8165 pixels, or 1.24 times the linear picture height of the D800e. The aspect ratio of the IQ250 would result in less wasted megapixels for a 16 x 24 inch print. If Sony obtains the same performance from the new sensor as with the D800e sensor, this should result in a gain of 0.75 f/stops. The engineering DR of the D800e is 13.4 EV and the predicted DR of the IQ250 would be just above 14 stops, consistent with the DR stated in the PhaseOne announcement.

The IQ250 will cost US $34,900 and the D800e lists for $2,997 at B&H. Does this increased performance warrant the 1160% cost differential between the two cameras? It will for some perfectionists with a large pocketbook, but probably not for most users.

Discussion is welcome.

Bill

[bjanes]

Hi Bill,

I agree, and have therefore added the IQ250 to the pre-selectable sensors in my (DOF) Output Quality Planning Tool. The settings should probably also allow to already predict Hasselblad (not in multi-shot mode) and Pentax image quality.

Not to mention 'CCD color' and 3-dimensional rendering by CCDs. This is going to be fun...

Cheers,
Bart

I used your tool to determine the maximal print size at very high quality for the IQ250 and my D800e. The IQ250 produces such a print at 11.38 x 8.529 inches, whereas I can get only 10.08 x 6.28 inches from my Nikon

Bill

[Bart_van_der_Wolf]

I used your tool to determine the maximal print size at very high quality for the IQ250 and my D800e. The IQ250 produces such a print at 11.38 x 8.529 inches, whereas I can get only 10.08 x 6.28 inches from my Nikon

Hi Bill,

Yes, from a resolution point of view it's not a huge difference, but anything beyond 10% can still be considered significant. There is of course more to it than resolution alone, but that is harder to plan (other than buying another camera). Changing a lens and stitch if additional resolution needed, is easier than changing the Bayer CFA characteristics ...

Cheers,
Bart

[Brian Hirschfeld]

In my opinion there is no comparison between medium format and a full-frame 35mm camera. The Nikon D800 is a beast of a camera. Initially I was against it but now see its merits simply from a resolution / cost ratio and as being a compact modern AF camera with a wide range of lenses and an extremely high MP count. However there is much more to medium format. Medium format, to me is as much about the physical size of the sensor (and of course the pixel's themselves) as it is about the megapixel count. I find that larger sensors produce images which are special and uniquely different from an aesthetic standpoint then those created with smaller-sensored cameras. When I shoot my Nikon D4 or my Leica's I'm expecting different things from the camera then when I am shooting with my IQ180.

Medium format is as much about the shooting experience as it is about the performance of the camera in my opinion. And I think this is backed up by the fact that if you look at the current medium format camera's on the merits of their performance (outside of the imaging sensor) they are vastly inferior to some of the most basic sub-FF DSLR's! People still shoot them and pay for them. Clearly there is a market. Most of this market is probably due to the performance of the sensor and the caché (and of course technical attributes but thats not my point). A 50mp medium format sensor is catering to a vastly different market then any 35mm DSLR. period. If the IQ250 is as versatile as it seems to be on paper, then it will open up new worlds of possibility for medium format photographers, who again, are vastly different then 35mm full-photographers. There is no market convergence and the PhaseOne IQ250 and Nikon D800 are not competing.

[RichDesmond]

I used your tool to determine the maximal print size at very high quality for the IQ250 and my D800e. The IQ250 produces such a print at 11.38 x 8.529 inches, whereas I can get only 10.08 x 6.28 inches from my Nikon

Bill

I think you're making the classic mistake of taking source data (or assumptions) that have fairly low resolution (and I'm speaking here much more generally than sensor resolution) and then popping out some numbers at the end that imply much finer resolution than the source data supports.

In this case, exactly what is the definition of "very high quality" for these prints? What's the difference between "very high quality" and "high quality"?  What assumptions go into that definition? Would the average person be able to detect the difference in a Nikon print vs. the IQ250 if they were printed the same size? If they could, is it the resolution, or something else?

DPI is a very small part of what makes up "very high quality" print, in my experience at least.

Going back to my earlier post, there's a third statement: Things that are easily measured are often over valued.

[langier]

Test and measure to your heart's content. For me, what I use isn't key to my work or my success, though at times, it does help. The most important aspect to me is the end product--web, print or publication.

If one is happy with the results and is successful, the method and the tools to get there are irrelevant. Few ever cared the brush used by Rembrant or the etching press used by Picasso. It's the final artwork that counts!

[ErikKaffehr]

Hi,

If you check Bart's description you see that normal quality corresponds to 20/20 vision. He gives pretty good descriptions of the parameters used and those parameters are normally accepted standards for human vision. Properties of human vision are quite well known.

Best regards
Erik

I think you're making the classic mistake of taking source data (or assumptions) that have fairly low resolution (and I'm speaking here much more generally than sensor resolution) and then popping out some numbers at the end that imply much finer resolution than the source data supports.

In this case, exactly what is the definition of "very high quality" for these prints? What's the difference between "very high quality" and "high quality"?  What assumptions go into that definition? Would the average person be able to detect the difference in a Nikon print vs. the IQ250 if they were printed the same size? If they could, is it the resolution, or something else?

DPI is a very small part of what makes up "very high quality" print, in my experience at least.

Going back to my earlier post, there's a third statement: Things that are easily measured are often over valued.

[Bart_van_der_Wolf]

In this case, exactly what is the definition of "very high quality" for these prints? What's the difference between "very high quality" and "high quality"?  What assumptions go into that definition? Would the average person be able to detect the difference in a Nikon print vs. the IQ250 if they were printed the same size? If they could, is it the resolution, or something else?

Hi Rich,

The tool is rather verbose about which assumptions are used for the different qualifications. The differences are significant enough to see the difference, of course assuming otherwise identical circumstances. For instance, the difference between "higher" and "very high" when viewed from 3 feet distance, is 60.68 x 45.48 inch output @ 136.5 PPI versus 34.67 x 25.99 inch output @ 238.9 PPI with identical visual resolution / quality from the same file, or 75% higher resolution / quality with the same output size.

When different cameras are being compared, then many other factors can become deciding factors. Skill in postprocessing different types of camera files is not unimportant for comparisons either, but the tool will neutralize the comparisons as far as resolution is concerned. One down, several more to go.

DPI is a very small part of what makes up "very high quality" print, in my experience at least.

Fuzzy prints can be as bad as fuzzy concepts, depending on the intended use of the image.

Going back to my earlier post, there's a third statement: Things that are easily measured are often over valued.

Or underestimated in their usefulness ...

Cheers,
Bart

[bjanes]

DPI is a very small part of what makes up "very high quality" print, in my experience at least.

Going back to my earlier post, there's a third statement: Things that are easily measured are often over valued.

Rich,

From your misuse of DPI (a printer term), I surmise that you are not a tech guru even though you may be a successful photographer. 

Best regards,

Bill

[bjanes]

Medium format, to me is as much about the physical size of the sensor (and of course the pixel's themselves) as it is about the megapixel count.

Brian,

I look forward to your reporting on the new camera, merely to be informed about what high end tools are available and their functionality, even though I will never be in the market for MFDB.

BTW, what is so special about a MFDB pixel? Noise, DR, and color response can all be measured and what else are you evaluating? How does a MFDB pixel compare to that of your D4 and what differentiates them at the pixel level?

Bill

[Bart_van_der_Wolf]

BTW, what is so special about a MFDB pixel? Noise, DR, and color response can all be measured and what else are you evaluating? How does a MFDB pixel compare to that of your D4 and what differentiates them at the pixel level?

Hi Bill,

We'll wait for Brian's report after he has actually used it himself, but one of the benefits of a physically larger sensor array is due to the improved MTF and resolution caused by a larger magnification factor for the same field of view (with a required longer focal length). Lens quality differences may mitigate or increase the distinction though.

Cheers,
Bart

[pier64]

The price is saying two things: image quality is going to be at the same level than CCD; Phase One does not want to cannibalise its current customer base but is leaving the door open for someone else to substantially expand the current digital medium format market. This is potentially a big mistake.

Pier