You wait years for CMOS backs to arrive, and then... Disappointment.

[eronald]

I find the PDN review surprisingly honest. They have clearly shot with the thing. Here are two telling excerpts.

--
Through trial and error—and timing—Matter was able to nail down the focus and capture crisp photos. If you're considering moving up from a full-frame DSLR to a medium-format system like the IQ250, this sort of adjustment is something to take into consideration. Despite many improvements on this back and with Phase One's 645DF+ camera body, shooting medium format is still more challenging than using a DSLR.
--
We used the IQ250's display to give us a sense if the timing was right on a shot but, despite its better than average resolution, it didn't give us a good indication of sharpness when we zoomed in.
--

I think a $250 Canon Rebel with a couple of fast primes would do a better job than the P1 on those shots, and give good separation - but then I'm not a professional photographer, and even less an "artist"

Edmund

[Steve Hendrix]

The biggest objection that holds our clients back from the Pentax 645D is the uncertainty over the commitment Pentax has for the platform, and their lack of development for MFD before (and after) the 645D . Most of our clients are (at least somewhat) familiar with the products from Phase One/Leaf/Hasselblad, or have purchased MFD products before and will again. They know that Phase One is a proven brand for someone who wants MFD. They have been producing solutions for MFD since the 1990's, and they've continued to develop technology for the platform. They have a legacy and an impressive recent resume of accomplishment to go along with it. For someone ready to invest substantially into MFD, Pentax remains un-proven regardless of how nice some of the aspects of the 645D are.

Secondly, when someone is interested in a Pentax, they frequently choose to purchase a pre-owned Hasselblad H4D-40 from us instead. The price difference is small and they are similar solutions since they seem to use the same sensor, but Hasselblad has a much larger lineup of lenses and system accessories, much better software, etc. And they have a legacy and a tradition of MFD development. There isn't that much of a difference in the profit made from sale of a pre-owned H4D-40 compared to a Pentax 645D.

So, I have to say - any lack of "push" on our part for Pentax is as much a symptom of the same concern our clients have, which is, where is Pentax going with MFD, and how much future development will they invest in? When they build a legacy and prove that buyers can enjoy rapid development, not to mention scale, of products, then perhaps our clients will look harder and longer at the platform.

Until then, I feel the best success for Pentax comes from new users to MFD, or users who cannot or will not commit funds toward the higher cost options from Phase One/Hasselblad/Leaf/Leica, etc.


Steve Hendrix
Capture Integration

[BernardLanguillier]

Your observation #2 is absurd. I know and have shot with a lot of non-pro Phase One shooters and I can absolutely positively say that is total nonsense. (I'm being polite at this point).

Now I would deny that there are folks who simply want the best regardless of cost but that is NOT the same as buying things just because they are expensive.

Graham,

I hope you have noticed that my point was that the 645D is IMHO not positioned high enough in price to appeal to some people who look for the best, simply because there is a natural tendency for them to see a strong correlation between "the most expensive" and "the best". I could elaborate on this, it is not just the price of the 645D but the equity of the Pentax brand in terms of high end photographic performance.

I agree that spending the biggest amount of money is not their goal per se, but using "the best" or what is widely accepted as "the best" is a clear goal in that it makes such a purchase more satisfactory than any other for them.

I have never said that this applies to all non pro phase shooters. It does also apply at different levels in terms of segments, price brackets,...

Among those people who behave this way, many of them are not willing to openly admit that their decisions are driven by such motives because they realize that they are hard to defend from a pure rationale standpoint, and those are typically people who have gotten where they are today because of their higher than average ability to act rationally.

Let's face it, when you have a certain level of income, spending 20,000+ US$ on a Phaseone back is not a big deal. At least is it not that bigger of a deal than spending 7,000 US$ on a 645D. So you might as well not take the risk of going cheap and spend the extra money because you know, based on different inputs starting with peer opinions, that you will get something good enough that your comfort that you have the best is not going to be challenged. That will be money well spent.

Anyway, I am not here to convince you, please ignore this if you feel that it doesn't make sense.

Cheers,
Bernard

[eronald]

Steve,

 Thank you. You bring an interesting perspective.

Edmund

The biggest objection that holds our clients back from the Pentax 645D is the uncertainty over the commitment Pentax has for the platform, and their lack of development for MFD before (and after) the 645D . Most of our clients are (at least somewhat) familiar with the products from Phase One/Leaf/Hasselblad, or have purchased MFD products before and will again. They know that Phase One is a proven brand for someone who wants MFD. They have been producing solutions for MFD since the 1990's, and they've continued to develop technology for the platform. They have a legacy and an impressive recent resume of accomplishment to go along with it. For someone ready to invest substantially into MFD, Pentax remains un-proven regardless of how nice some of the aspects of the 645D are.

Secondly, when someone is interested in a Pentax, they frequently choose to purchase a pre-owned Hasselblad H4D-40 from us instead. The price difference is small and they are similar solutions since they seem to use the same sensor, but Hasselblad has a much larger lineup of lenses and system accessories, much better software, etc. And they have a legacy and a tradition of MFD development. There isn't that much of a difference in the profit made from sale of a pre-owned H4D-40 compared to a Pentax 645D.

So, I have to say - any lack of "push" on our part for Pentax is as much a symptom of the same concern our clients have, which is, where is Pentax going with MFD, and how much future development will they invest in? When they build a legacy and prove that buyers can enjoy rapid development, not to mention scale, of products, then perhaps our clients will look harder and longer at the platform.

Until then, I feel the best success for Pentax comes from new users to MFD, or users who cannot or will not commit funds toward the higher cost options from Phase One/Hasselblad/Leaf/Leica, etc.


Steve Hendrix
Capture Integration

[ErikKaffehr]

Hi,

I don't know. It seems like a real shooting in real life conditions with bad light.

I have checked out the 6400 ISO image, see actual pixel crops below. It is not bad for 6400 at 1:1 view, and they could probably be improved in processing.

Best regards
Erik

That PDN review is painful.  The samples look like they go with another piece entirely.

[EricWHiss]

Can anyone say how the PDN samples compare to D800's ISO for ISO?    I am wondering if there might be some updates to C1 coming that will improve the rendering of these files?  Seems too early to start discounting, but wow those PDN sample files …. ugh!   Maybe when the RAW processing, profiles, etc are further developed, Phase will tell the photog's it's okay to start taking sharp shots now...

[ErikKaffehr]

Hi,

Those images are pretty much unmanipulated, AFAIK. In this case the photographer probably had some intention to use motion induced unsharpness as an artistic expression (if it fails or not is another question).

Something I note is that there is significant colour bookeh on the out of focus background. So I guess the lenses are not as well corrected as the Otus.

The boots in the attached image are almost sharp, there is still some motion unsharpness. That is the 400 ISO shot. A static studio shot would be sharper.

Regarding the 6400 ISO image I would say it is very clean. I don't see patchy colours, and blacks are pretty clean. There is clearly quite a bit of noise reduction, either in the image or applied in post. A bit to much for my taste. Pixel for pixel I guess it is pretty much what I have seen from Sony sensors.

Best regards
Erik

Can anyone say how the PDN samples compare to D800's ISO for ISO?    I am wondering if there might be some updates to C1 coming that will improve the rendering of these files?  Seems too early to start discounting, but wow those PDN sample files …. ugh!   Maybe when the RAW processing, profiles, etc are further developed, Phase will tell the photog's it's okay to start taking sharp shots now...

[EricWHiss]

Erik,
Quit kidding around okay?  This is serious stuff here we are talking about.     Next time take a real careful look at the files before posting okay? 

No let's wait to see how these look with the C1 update….

but you might be right about at least one thing - not all of what's visible (in not a good way) in those PND samples can be due to the sensor - some might be from the lens (CA) and DF combo. Might but can't say for sure.

[ErikKaffehr]

Hi,

The CA is definitively axial chroma, also known as LoCa or spherochromatism. Almost all lenses have it at full aperture. Otus has almost eliminated it.

Best regards
Erik

Erik,
Quit kidding around okay?  This is serious stuff here we are talking about.     Next time take a real careful look at the files before posting okay? 

No let's wait to see how these look with the C1 update….

but you might be right about at least one thing - not all of what's visible (in not a good way) in those PND samples can be due to the sensor - some might be from the lens (CA) and DF combo. Might but can't say for sure.

[torger]

Did not find the link to PDN IQ250 review in the thread, so here it is:

http://www.pdnonline.com/gear/Phase-One-Unveils-Fi-10189.shtml

It has full-size large jpegs, casually shot images of some dancers in natural and artificial light, some are smaller than max resolution but I assume it's due to cropping rather than downscaling.

The ISO 100 image looks fine although the subject is not too good for evaluation (and the image is not sharp). The high ISO pictures are obviously noise-reduced quite heavily, and frankly I don't like the look at all. You can't judge colors when lit with those artificial lights, but the ISO images are heavily smeared, no way near 50 megapixel detail coming out of those. As the shots are not sharp the detail loss is a bit hard to judge though.

Already at ISO800 it looks pretty bad.

My guess is that processing is somewhat to blame rather than the sensor. Crude noise reduction = detail and color destruction.

One of the advantages of my old CCD sensor which does not have the best dynamic range by modern standards is that its noise quality is nice and random, film-grain like, so I rarely do any noise reduction at all. I'm hoping this sensor has the same quality, but I have not yet seen any samples without noise reduction.

I should say though that I don't shoot high ISO very often so I'm no expert on how good it can get. Are those images close to what you would see when pixel-peeping a D800 or Sony A7r?

[ErikKaffehr]

Hi,

Attached is an 800ISO shot from my Alpha 99, actual pixels of course, my default processing in LR5.3

Best regards
Erik

Did not find the link to PDN IQ250 review in the thread, so here it is:

http://www.pdnonline.com/gear/Phase-One-Unveils-Fi-10189.shtml

It has full-size large jpegs, casually shot images of some dancers in natural and artificial light, some are smaller than max resolution but I assume it's due to cropping rather than downscaling.

The ISO 100 image looks fine although the subject is not too good for evaluation (and the image is not sharp). The high ISO pictures are obviously noise-reduced quite heavily, and frankly I don't like the look at all. You can't judge colors when lit with those artificial lights, but the ISO images are heavily smeared, no way near 50 megapixel detail coming out of those. As the shots are not sharp the detail loss is a bit hard to judge though.

Already at ISO800 it looks pretty bad.

My guess is that processing is somewhat to blame rather than the sensor. Crude noise reduction = detail and color destruction.

One of the advantages of my old CCD sensor which does not have the best dynamic range by modern standards is that its noise quality is nice and random, film-grain like, so I rarely do any noise reduction at all. I'm hoping this sensor has the same quality, but I have not yet seen any samples without noise reduction.

I should say though that I don't shoot high ISO very often so I'm no expert on how good it can get. Are those images close to what you would see when pixel-peeping a D800 or Sony A7r?

[torger]

I'd say that your example looks better, but it's difficult to compare. Safe to say is that the PDN review sample images does not make the best to show the IQ250 from it's best side. As it looks there it might be that some would prefer sensor plus on an IQ260 for high ISO shooting.

However, as it's Sony that's made the sensor I would expect that the high ISO is very similar to other Sony cameras, possibly slightly worse due to bayer design optimized for color at base ISO, but that would be fractions of a stop.

The raw converter can make a huge difference to the look of a high ISO image. Noise reduction settings of course, but also the demosaicer. I'd love to have a raw file to play with

Comparing a 135 camera with a MF will be a bit difficult though, it's probably better to compare with say IQ260 sensor plus to make a valuable comparison, or why not the current Pentax 645D. I think the 135 cameras will be much better in practice as you can have wider aperture. Ie instead of an 80/2.8 you might be using a 55/1.4 and gain more than 1 stop just in the glass, and if you look at the journalist flagships you gain say one extra stop on the sensor compared to the high resolution CMOS (Sony Exmor was never the best concerning high ISO, unparalleled base ISO DR is its thing), so in a typical high ISO use case comparison with a Canon 1DX or Nikon D4 I think the IQ250 would look like a joke.

I don't think this model is to compete with 135 though (obvious from the price if nothing else), but rather to be a evolutionary improvement of high ISO in medium format space. Remains to be seen who will need it. It seems to me that the best rational way to go would still be to keep your current CCD back and get a flagship DSLR if you need high ISO shots at as good quality as possible. When you're spending this amount of money on gear there should be no problem to have that extra DSLR on the side.

[LKaven]

[...] ...(Sony Exmor was never the best concerning high ISO, unparalleled base ISO DR is its thing), so in a typical high ISO use case comparison with a Canon 1DX or Nikon D4 I think the IQ250 would look like a joke.

The Exmor has mainly one enemy, and that's thermal noise.  Except for thermal noise, which can be mitigated with a manual black-frame subtraction, the sensor in the D800 does as well as my D4, and in some respects better. 

In very dim lighting, there is often some scene feature that is better lighted than the rest, and that area in the scene is rendered with noticeably more detail by the D800.  A small well-lighted face can pop out of an otherwise grainy image.  Since the IQ250 does black-frame subtraction full time automatically, it might produce some very good mood shots at ISO 12800-25600 equiv.  We shall see if that turns out to be the case or not.

[torger]

Detail looks murdered in the high ISO shots shown so far, but with other processing settings it may turn out well. If the D800 can do well, the IQ250 should be able too, minus the smaller aperture optics penalty.

[bcooter]

Detail looks murdered in the high ISO shots shown so far, but with other processing settings it may turn out well. If the D800 can do well, the IQ250 should be able too, minus the smaller aperture optics penalty.

It's more than sensor, or body or lens, or lighting.  It's everything.

I've shot a lot of 800 to 1000 iso with every camera and no matter how you cut it you need light and you have to know the complete system to make it work.

These cameras are like film, where it takes a while to zone in on what works, what doesn't, it's not just point and shoot.

That's the problem with most of these early samples, they're point and shoot, thinking it's going to work and well we see the results.


I'm amazed that the little olympus, which is a 43 sensor and tiny compared to the Phase shoots such an amazing 1000 iso file, but it took me a month to really learn what works, on what seting.

I think this is where these samples fell through, that and somebody thought they could hide noise in post processing, which really screws things up, unless you do the noise reduction in layers.

Hi,
I don't know. It seems like a real shooting in real life conditions with bad light.

I think there is just a lotta bad all the way around.


IMO

BC

[synn]

800 ISO with somewhat decent available light. Credo 40.
Frankly, I don't need any higher ISO performance. If it does, that's probably a shot I am not interested in publishing anywhere.

[torger]

Here's a 100% crop from a ISO800 shot of my ancient Leaf Aptus 75, no noise reduction, basic unsharp mask sharpening. For being a back announced in 2005 I think it looks rather good, and a good amount of detail is kept. It will only work if you have a well-exposed image though and you don't need to push anything, because dynamic range at this ISO is very poor. And color fidelity is bad, so you get a certain "look" which is quite unlike what you get at base ISO, the loss of color fidelity rather than noise and DR is mainly what makes the use case narrow from my point of view. But still, images can be made .

Modern sensors don't lose as much color fidelity at high ISO but it's an aspect that sometimes is forgot when looking at the results. Engineer types like me is often focused on noise and forget about color. Careless noise reduction also kills color.

[ndevlin]

--
Through trial and error—and timing—Matter was able to nail down the focus and capture crisp photos. If you're considering moving up from a full-frame DSLR to a medium-format system like the IQ250, this sort of adjustment is something to take into consideration. Despite many improvements on this back and with Phase One's 645DF+ camera body, shooting medium format is still more challenging than using a DSLR.
--
We used the IQ250's display to give us a sense if the timing was right on a shot but, despite its better than average resolution, it didn't give us a good indication of sharpness when we zoomed in.
--

"EXACTLY!!" he proclaims, he with the exalt of vindication.

- N.

[torger]

"EXACTLY!!" he proclaims, he with the exalt of vindication.

I read that too, and it feels a bit odd that the display and image rendering on it (or live view) would not be good enough for sharpness check. I suspect that they're not used to it yet and when you get used to it you can identify what's sharp and not. Otherwise it would be a major problem, but if so it's probably fixable through firmware.

[ndevlin]

@torger, I haven't met a single camera where LV works for critical manual focus. It's a total fail on the D800e.  Maybe Canon is better, but who cares.

- N.