P65+ to be announce on monday by phaseone, 60.5 mpg, sensor size 53,9x40,4, pixel pitch 6x6 micron, 50-800 iso, 1 fps, no info about sensor manufacturer. EC price 26.900 euros +tax. delivery supposed to start in Q4 of 2008.
this is all i know, peter

Yawn.....................!

D.

Stay tuned, or if this boars you, go back to sleep. More soon.

Michael

I posted in May I'd heard it was going to be 70mp, more interestingly that it could also be shot at half the pixel count, for speed and maybe noise. That would be an interesting feature if it turns out to be true.

Kevin.

I would hope that it could be shot at lower resolution. 380 meg 16 bit tiff, anybody? I work with scans that big and bigger, and its a pain on my Mac Pro with 7 gigs and a 10,000 rpm drive.

the real estate is, however, good news.

I received an e-mail from a a friend of mine, a Phase shooter, who asks simply:

"who told Phase we needed 60 megapixels?"

Well, who was it? Which one of you told Phase 60 was a great idea and what the market demands?

I am however, super-psyched about the fuller (almost full) Frame size. That is a huge positive.

Hi
If true this will be less expensive than the H3D-50 with a larger pixel count + a larger sized chip. To me making the chip larger ( physical size ) is more important than putting more pixels in the same sized sensor. This also makes the lenses closer to 1:1. With the Mamiya AFD film magazine the actual film size is Image Size on film size is 41.5 x 56 mm. The new sensor is 53.9 x 40.4 mm, if true?

Thanks Denis

The only trouble is I'm not sure if those older Mamiya lenses are able to resolve such details.

Hi
What do you mean by older? The new zooms look very good. Maybe the 35mm & 55-110 could have issues. I think getting MF closer to FF 35 digital is a big step. Plus there are new lenses to come from Mamiya & maybe from 3rd party manufactures to in the future.
Denis

I know everyone moans about more pixels, I do wonder though if you were shooting with 60 mpx how many would miss the 60 if you had to go back to 30mp. 60 could possibly give better skin gradation in difficult light. I have often wondered with bayer pattern sensors how many green pixels are being used to make a blue sky or blue pixels to shoot yellow cloth etc. Just a thought from someone shooting a tiny 21 mpx on 35 mm.

Kevin.

sorry, just heard that the price my be sightly higher (still under 30.000.- euros), peter

How much is a P45+ in EU at the moment?

They can be had for 30.000 US$ including the body in the US, which is less than 20.000 Euros if I am not mistaken.

Would the P65+ be 50% more expensive than the P45+, or is it another case of EU citizens being ripped off?

Regards,
Bernard

Given the earlier multiplications I find that the new sensor has a multiplication factor of only 1.06

I think this is close enough to full frame to justify the term

It might not seem much, but the same calculation gives me a multiplication factor of 1.17 for my existing Aptus back which is more than the figure I'd heard in the past of 1.1.

An extra 10% along a dimension is fairly significant when wide angles are a little crushed.

For example, a 45mm lens on this chip would give the equivalent of a 40mm in FOV with the Aptus 75. A 55mm on this chip would be equivalent to a 47 on the Aptus 75.

There would be a fairly significant advantage to eg: a 35mm lens such as a digitar with good coverage as it would give a similar fov to a 28mm lens.

If only this didn't have too many darn pixels it would be great!

I think that a P25 is the perfect back -- not because it is the one I have : )-- since it has the larger size -so far- with the biggest sensel

P 45 6.8 x 6.8
P 25 9 x 9
P 30 6.8 x 6.8
P 21 9 x 9
P 20 9 x 9

When you get a camera with sensel size 6, are you upgrading or downgrading (from size 9)?

Are we falling in to the consumer trap of going for the pixel count as in "more is better"?

What about diffraction problems?

At least it is probably safe to say that a P 25 will remain a "pure" digital back, and we can tell the client that he is getting one of the best IQ available on a compact file size.

On the other side, this new Kodak sensors have a lot of improvements in the way they are made, if they only had a model with 9 size sensel, full -or almost full- frame that would be what? 30MP ?

That would be the one to upgrade to...

P45+ = 26.990.- euros + tax (20% in austria), no body included.
peter

Word up Leonardo.

if thats really the best price you can get, you guys should come over here to the US and get one for the same price in $ and save oh, about 10,000 Euros...

I have a P30+ and the last thing I miss when shooting a 5d is file size. Most of the shooters I know don't want or need bigger files. The qualification is that we all shoot people for magazines and other print collateral. In truth, we could all be using 5d's, for the most part. So 60 megapix seems a waste to me. Hell, 30 seems a waste. As to tones, only high end printing really shows the better tones and color. Even with the most beautiful file ever shot, some hack in a prepress department will butcher your image. Just look at this month's Lucky.
All that being said, if the pixels can be binned to 30, that's pretty cool. And for people making large prints and landscape guys, 60 will be welcome.

One thing you have to consider is that pixel sizes may be getting smaller, but alot of development has gone into them to make them better. If quality decreased, what would be the point of going bigger?

The Kodak 22mp (KAF22000) sensor was developed something like 5 years ago so to suggest that the only progress we have made since then is making the same pixel quality smaller i think maybe somewhat shortsighted. If you actually look at the cross section of the pixels between the 9 micron 22mp and the 6.8 micron 39mp chip you see that the the actual pixel well where the light is captured is roughly the same size. The 6.8 micron pixel has significantly less "deadspace" around each pixel well. Not to mention that it has a greater ability to produce color gamut. compare images of the same scene with a 22mp and 39mp back and you will see a real difference.

That being said the P25 really is a fantastic back and for that reason it is still on the market and doing quite well. but I do think that given the development time for sensors to improve over the last 5 years, I would certainly hope that the new CCDs from Kodak and Dalsa will do much more than just up pixel count of the same quality.

Just wondering how significant the differences are between 22mpx multishot (in 4-shot mode) and the new 60mpx backs when it comes to resolution and colour accuracy in still-life and art-repro applications?

(Just thinking out loud)
Cheers,
J

Getting better sensor uniformity would be one important thing to me, reducing colour aberrations.

Because "The Kodak 22mp (KAF22000) sensor was developed something like 5 years ago", could we see a new generation of 9 microns 22MP (or 30MP) CDD (for those thinking it is enough) ?

For the work I do, medium format really needs to write some presets or change their color response.

All of digital is somewhat flaky compared to film in the way it picks up ambient color and all of digital is so tilted towards holding the highlights and shadows even the dslrs look somewhat flat, but this (and I'm not a color scientist) makes for a lot of color contamination (or better put it digital sees everything and the medium format backs see too much for what I do).

Film was different in the fact it was kind of dumb, it didn't seem to see everything. With film I was always adding fill and with digital I would rather start in black and work my way to white.

The medium format files are the most problematic. When they are on they are completely amazing and if your working with controlled lighting under the conditions they seem to be designed for they hold detail, highlights, shadows like no dslr can, but when you are working with mixed lighting, like sunlight and hmi fill, or your working with very hard direct light in studio, they cast and require a lot of post work to get skin colors to the look you remembered in film.

Little things like high key, extreme low key and back light going to flare seem to make digital go crazy.

I assume that whoever writes the software/firmware/grey balances . . . whatever, is shooting color charts and vegtables but they really need to shoot under a lot of conditions and skin tones in almost every kind of light.

Last night I was putting together some pre production materials for a project and went onto the servers and pulled down images from the 1ds1, 1ds2, A-22, p30 and p30+.

Some were amazingly on, others were almost impossible to get that great skin color without working in photoshop. Once again when the medium format was ON it was amazing, but when off it requires selecting and almost painting the skin.

The most uniform of all the cameras was the original 1ds1. It had it's issues, but is was the most consistent.

As far as 60mpx, well that's up to whoever buys it. Personally I think it's just easier and sound bites better to claim more megapixels than it does to mention stable software, or show beautiful skin tone color, but these companies will do what they want and I guess there is value in bigger is better.

Regardless, why none of these companies have not hired some kid out of Art Center, send him on a week long shoot of 5 skin tones, all types of lighting, clothes and locations and had him shoot their digital backs, next to film and compared the results is beyond me.


JR

Hi,

Just adding that we have diminishing returns. Going from 11 MPixels to 22 the amount of information is doubled, so we can make 40% larger prints. Going from 39 to 50 Mpixels we have a 20% improvement, so we can make 10% larger prints. Do we have a visible difference? Probably not. If we can see an improvement it is probably more related to improved sensor design and improved signal processing.

Best regards
Erik

Well, you are right about that, for sure. The DSLR's have the advantage of a much smaller sensor to get right.

What I was referring to was the way that at least the Dalsa MF sensors are 'printed' in sections. As a result, the chips surface is divided into a number of rectangles that appear to have slightly different characteristics. As a result, with non-retrofocal lenses you can get patches of greenish or magenta hue distributed over the picture. This is then overlaid by subtle 'italian flag' colour shifts, particularly with wide angles which are shifted.

Lets hope that the chippers manage to improve on the production process.

Only Dalsa has to reveal their hand now, presumably they have something in the works too.

Amen to that.

As someone who cannot get ACR4 to save out TIFFs (Hasselblad 223Mb files that were first processed in Flexcolor) without constantly getting the dreaded, "There was not enough memory......" error, I can totally agree with you. I'm running Vista 64bit/8Gb RAM and one of the fastest processors available and I still get the error.

As has been suggested, landscapers might like this, but no one else needs it, IMHO.

D.

I think you must have something wrong in your Photoshop setup. I'm converting my P45+ files to 16-bit TIFFs (same file size as what you're working with) on a Mac Pro with 4 Gbytes of RAM and I've never seen the out of memory message. In fact, I'm stiching up to five 8-bit TIFFs (total open image size of almost 600 Mbytes, with no memory problems.

Mort,

Just PM'ed you.

D.

I wrote some software to calibrate a back, and did some testing around January. The difference in mixed light is incredible with my P45+. But raw converters cannot pick up the calibration info at the moment. My Matlab code is not exactly suitable for end-users like James It's possible that we could get a hacked version of dcraw to do what we need for the general public, but at the moment the only suitable version I know of is in the hands of Xrite and they have IP issues releasing it.

I proposed to the ICC DP working group that we have templates to drop the calibration info the files so that all Raw converters can see it. There is now on the agenda for the ICC Portland meeting in the fall a vote, that a request may be made via our liaisons to the TIFF/EP and DNG committees to incorporate the requested mods to their standards, at which point in a year or so things might improve for end users.

I will try to get out some end-user software myself later this year, I have some work to finish first before the Portland meeting.

Edmund

If you do that though, Phase in EU won't want to know you. So hope you don't need support, or that it can be resolved by phone, at mutually convenient times. Phase have been quite offensive to me on this matter!

Hope this helps

Interesting Mort,

I recently shifted back to OS X from XP64, because I was getting the same error with stitches of even 3 extracted P45 files (ca 200-230Mb each).

Never had the problem once in OS X - even with 6 shot stitches.

EDIT: Oh sorry, with 8Gb memory.

There is a lot of things that medium format does that is good. Tethering, detail, lack of an aa filter, the format for vertical pages and some of them have open files and stable software.

Then again there is no buzz in talking about stability or batch processing, but double the size, double the area, that's an easy one to advertise.

Most of the people I knwo that have shot digital for a period have gone through a lot of stages, from the first stage of yippeee it's finally as good as film, to the final responses when you just kind of shake your head and go on with life.

I don't know what any of these new annoucements mean, other than bigger files and for some the ability to print over a wall, but for most of us that work with these cameras under pressure, there is a whole lot of things we'd like to see addressed before the file sizes grow.

Then again, I've already pruchased and I'm probably not the exact target market.

I know the camera companies want to move product and from the dslrs to the smaller medium format companies, they are more predisposed to listen to good replies to what they offer, vs. non positive replies, (I guess that means negative).

We all want that in life, but then again if everyone keeps telling us we're great, we have little incentive to get better.

I'm curious as how these new cameras will be received. Will they be the must have addition, or just a variation on a current theme?

I don't sell cameras so I don't know, but I am positive once they hit the streets the process and the buzz (good and bad) will start all over again.

JR

Pretty unlikely .....

During Hasselblads conference call this week we were told that Kodak has discontinued the 22mp chip and stock is only what is on hand ...

So I would soon expect to see just about everyone's 22mp offerings fade into history with the possible exception of Leaf since they are the only back manufacturer to use the Dalsa chip in the 22mp offering.

Matt LaPointe
Dodd Camera - Chicago
2840 W. Armitage
Chicago, IL 60647
(773) 227-3633
mlapointe@doddcamera.com

go multi layers, do something with different plugin filters, sooner or later you'll encounter this(200m+ original files)

Somebody, in this or some other thread asked a very good question: "What do the product managers listen to when they come up with these products" (approximate quote). I am in the same camp of people that say 60mp is too much for most people, but could it be that the major MFDB companies are listening to what the clients want? Have any of you received a survey about what you would like to see?

No. It is definitely something wrong with the setup or hardware to get memory problems.
On my Powerbook G4 with 2Gig of RAM, I made the texture file that was printed onto vinyl and applied to the Honda F1 Car last year... ("Earthdreams" world map).
That was the biggest file I have ever worked on... approx 16 Gigabyte photoshop file with dozens of layers printed at 300ppi all over the car.
Yes it was slow to work with
but I didn't ever get a memory problem.
The secret was a firewire Raid Disk attached to the laptop as the photoshop scratch disk.
So believe me... a piddley little 200Mb file should never cause a problem.
All the best
Andrew

And don't forget Sinar - eMotion 54LV uses the Dalsa 22MP sensor....

And it also was just reduced to $17,695, which means you can pick up a 22MP Hy6 Kit (minus lens) for less than $22,000.

Steve Hendrix
www.ppratlanta.com/digital.php

"And it also was just reduced to $17,695, which means you can pick up a 22MP Hy6 Kit (minus lens) for less than $22,000.

Steve Hendrix
www.ppratlanta.com/digital.php"

MF price war!!!! *g*

For $17,999, H3DII/31 still looks like the best deal..I'm tempted.

Steve

I'm in the "60Mp? No thanks" camp myself. However, if you could choose to shoot at lower resolution with faster speed AND get very clean 1600 iso @ 30Mb I would consider upgrading in the future. But like many have said before me, my P25 still blows me away every time I use it so I feel no rush whatsoever.

not total 200mb, but each layer 200mb, and I easily end up with 30-50layers, see attached file for example...
I'm not a hardware or software expert, so I really don't know which is the problem. well I guess it's the plugins, coz everytime I run out of memory I'm trying to do something with the plugins. Good info of your scratch disk secret:) I'll try that.

cheers,
dustblue

Why is each layer 200MB? It appears that each layer covers a small percentage of the whole image area, e.g. a plane or a man. This does not result in 200MB layers if some of the layer is transparent. If you are superimposing full images on each other but masking 90% of each layer out, then you can create much smaller files by deleting most of the unused parts of each layer. Hope that makes sense.

Yep. That's probably the problem.
Unfortunately all or most filters rely on RAM to apply the effect, and then this gets saved to the
scratch disk... so a big scratch disk may not help your particular filter problem.
Good luck
A

yes of coz, sorry I should elaborate more. the final psd file which contains the big original file generally end up with the size of 3-6G. 8bit half that size(which I have to usually do on behalf of speed...)

Yes, that's true, but only after you have flattened the image (or merged visible) to remove all the duplicated areas on the image. I have intermediate images of around 1.5-2Gb, which (seem to) require a scratch file of some 4-5Gb (don't ask me why, but that's what I'm seeing in the scratch files sizes in the status bar). With extra layers, for pp, I've even got over the 4Gb max file size limit, and had to complete the processing, rather than saving and finishing later, even after using a compressed save algorithm. A set of 60Mpixel files would be interesting(!) to work on.

This is when using PS CS3; I have ptgui, but haven't investigated whether it might be more conservative in it's use of ram, during this sort of process.

Oh, and I'm not saying that it is *necessary* to always work with files of this size!

Hope this helps

I don't know about you but I was very tuned-in to my film back in the day, and I shot only a few types because I knew what I wanted and what would deliver..now if someone handed me a back or holder with something else and I shot it and got it back from the lab I would have not liked it..

my point is that it took a lot of shooting and processing to dial in the desired film look. So too do I have to shoot and tweek to get the look I need with a digital workflow.. It was my experience with film, that unless I was shooting the same stock, the same lights, and the same set, all parameters would be different.. I never did a big shoot with out running some film first.

The biggest difference now is that I know a lot more about post and prepress..I know what will print and what won't..before I'd hand them a lovely(to my eye) velvia transparency and say "have fun matching this" and know that it was't my problem..and knew as well that there were masters in the prepress industry who could tweek their system and even improve on what I gave them. It has been a lot of work to try to get up to their knowledge level as all of us need to be delivering a finished image.. not prepress ready, but damn good so it can be converted by joeblow print guy. Do I wish I could go back to dropping the film at the lab and editing on the light table.. sometimes. However, I now deliver a better, printable file than I did a sheet of film 5 years ago. Hey we were spoiled then, kodak, fuji and the lab determined everything, our curves,hue/sat, levels,iso, etc...we just shot into their predetermined conditions.
I wouldn't mind if the next back I got had a preset for astia100, then I could tweek from there if needed. However I am happy now shooting a bit flatter & processing a flat raw file and adjusting to the look I want, more work but worth it.

Regarding flakyness of digital with mixed lights..I tend to think that my small digital camera which I use in mixed light situations handles it better than transparency film ever did.

We don't prior test digital like we did film but there isn't a lot of point. It's not like your going to add 10cc of cyan over the lens, or even start gelling the lights, especially with mixed daylight, hmi.

I guess we could, but we all know it will fix in post.

The point is I do shoot a lot of different work and I see a big difference in how all of these cameras react.

Shooting a Nikon, Canon, Phase and Leica and put them in their native processor, or even a 3rd party parocessor and you will see a great deal of difference in the way they respond to colors, the the difference is not global, where adding some cyan cleans it up.

Some of these cameras just respond funky. I just shot a hmi, daylight situation and the beautiful darker toned models went magenta red on some of the cameras, fine on others (no reason to say which), but it wasn't a global red/magenta it was in the curves.

I fixed, heck we're always fixing it, but it would be easier if they started out more stable or at lease were more adjustable before it came out of camera.

JR

I guess I have simplified to the point of using one manufacture, well at least a combo of kodak/hasselblad on both my small (leica) and larger (hasselblad) cameras .. so I don't see the variety of backs you do..or skin tones! with all there was and is to learn digitally I had to keep the variables down to a minimum.. I don't even what to think about a 60mp back at this point,
unless it had much larger sensor (real estate)

As well as Mamiya.

Other than that look at what Nikon did with the D3. They decided to stick to a low pixel count to provide optimal DR and noise... some brands seem to understand our needs better.

Cheers,
Bernard

Only that phase has the option to give us a back with a larger sensor which can have both sides. More px and lower noise.

Well, Phase gives you more and more pixels for sure; however, they are far from the noise level of the D3. Apparently their customers appretiate the pixel count higher than DR.