Pages: 1 2 [3] 4 5 ... 10   Go Down

Author Topic: Moving from 35mm to Medium Format. Need guidance (Sorry for yet another thread!)  (Read 46736 times)

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

Good points from Steve as usual. I would say that trying to max out performance from your D800 is a good advice.

Regarding color:

- Photograph a gray card in same light the subject is illuminated with and adjust WB for white card
- Consider to build a profile using a color checker card, or some of the alternatives. A color checker card and Adobe DNG Profile Editor works just fine and even allows you to tweak your rendition

Regarding sharpness:
- Focus using live view at maximum magnification and use MLU and self timer or cable release

All this applies to MF, too, if applicable.

Here you see same P45+ exposure with different DNG Profiles on LR and also C1 processing:

http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/Color/ColorTuning/Samples/3750_FULL/
http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/Color/ColorTuning/Samples/3750_Persons/
http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/Color/ColorTuning/Samples/3756_FULL/
http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/Color/ColorTuning/Samples/3756_flower/

Color balance was on the gray card in the last shot.

Best regards
Erik

Synn -

Based on the way you work, medium format seems a viable choice for you.

The D800 is such an interesting camera as I have heard both raves and complaints about the file. Perhaps it's just post processing - getting the color right. Hard to say - but certainly it might be worth seeking out some counsel on getting the most out of the Nikon file and then see if you still feel it doesn't have that certain something for you.

I'm not surprised you are swayed by the Leaf skin tones.

If you do end up pursuing a medium format system, there's no reason to not consider a Leaf Aptus-II 7 or 8. I have generally felt like the Aptus-II 7 would be a better choice for portraiture, the larger sensor and fewer pixels providing more of a different feel than 35mm, but I've also not had any complaints from anyone who moved downstream to an Aptus-II 8 (or P40+) from a larger sensor like an Aptus 22 (or P25+). In fact, they've been universally pleased. I don't feel the extra 7 megapixels is significant, but it's true the Aptus-II 8 will be less prone to moire and have a slightly cleaner result at ISO 100/200/400 (if you would use those).

If you imagine shooting with a Leaf digital back and loving the files, then the real question becomes the camera. Since you like the 24mm equivalent - your only option there is 35mm on a sensor the size of the Aptus-II 7, which is closest, translating to a 25mm. The 28mm becomes 20mm - very wide. On the Aptus-II 8, the 28 becomes 22mm, maybe also too wide. And the 35mm becomes 28mm, maybe not wide enough. So - key question - how important is that focal length to you? If it's real important, then that also means if you end up with an Aptus-II 7/35mm combo, that your only option would be a Hasselblad H camera with the 35mm lens, since no other 35mm leaf shutter lens exists.

A lot to think about - but I think first it is real important to make sure you're getting the most out of your D800.

Also, not having a local Hy6 dealer doesn't mean you can't get support on that system.  ;)


Steve Hendrix
Capture Integration
Logged
Erik Kaffehr
 

synn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1235
    • My fine art portfolio

Thank you steve for that very comprehensive response. I have dropped you a PM.

Erik: All that is done and done. I shoot with a grey cards, use a custom profile, everything. The samples I have posted are all shot this way. Even after all that, the D800 output lacks the subtle tonality that I see in files from a Leaf back.
Logged
my portfolio: www.sandeepmurali.com

bcooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1520

Rightly or wrongly I've owned and/or used  about every brand of digital camera made and since I shoot 99.999% people in still and motion, I can list in varying order the best out of camera skin tones, but it is kind of a mute point as everyone works with different lighting, subjects, even back end processing.

My first suggestion would be process in C-1 for most cameras, or Iridint Developer.   C-1 is good out of can on most problematic files, Iridient is great if you go under the hood and set some single channel parameters.

Overall I think all digital is usually very subject dependent, but for commercial work I've shot 75% of what I do with Canons, 1ds I, II, III and now the 1dx.  The 1dx being the worst of the bunch and I find very Nikon like which to me is very global and very red/orange warm.

If your going medium format I strongly suggest testing in your lighting and style and very much do this with many multiple of subjects, especially one image with all subjects together.

That will tell you as much as anything.

But, my bottom line is this.   If I shot with direct light or flash and didn't have to tether, I'd shoot with a M-8 and/or M-9 (and buy two of each due the crazy s__t they do), but the look is worth it.

If I could shoot mostly with medium format I'd buy a 40mpx Hasselblad, as those files really produce beautiful skintones.  Our studio was asked by a Asian client to process and retouch a session from a Chinese photographer using a 40mpx blad and I was really surprised at how pretty the skintones were.   The Chinese photographer  shot was a duplicate in lighting style, props and subjects of the Ad series we started so I had a direct comparison to my cameras.


IMO

BC

P.S.  and if your buying medium format I'd call Steve Hendrix and he will give you the truth even if it works against him.    My next call would be to Doug Peterson.  Both these guys know their stuff and won't sell you just to sell you.

« Last Edit: October 16, 2013, 05:45:20 am by bcooter »
Logged

synn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1235
    • My fine art portfolio

Hi Bcooter,

Thank you very much for that comprehensive response and my apologies for responding almost a month late (Work, travel and whatnot). I have been in touch with both gentlemen that you have mentioned and they have helped me understand  so much about the medium format landscape. I am evaluating the various options now and should be able to make my decision by the end of the year.

I have also been practicing on Capture One and have finally managed to get it to deliver some nice colors out of my Nikon files. They now look a bazillion times better than they ever did in Lightroom (Even after custom profiles) and now I can viably use my D800 as the second camera in my setup.

Here's an example:

http://www.sandeepmurali.com/p234558723/h726e47e7#h726e47e7

That image looked "Dead" after the LR conversion (Posted a few pages back). I am much happier with this version.

Funny thing is, out of fun, I was working on some very old files of mine, from my erstwhile D70s. Ran them through Capture One and finished in Photoshop. I am so surprised at how smooth the skintones were from that camera. In a way, more analog and less plasticky than the current gen cams, including the D800. I am not a physicist, but I get a feeling that in the quest of stupid high ISOs, camera makers now are losing focus on base ISO color reproduction (Pun intended).

Here's an example:

Logged
my portfolio: www.sandeepmurali.com

TMARK

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1841

The D70 was a CCD camera, as I recall, just like the digital backs and Leica M8/9/S2.  CCDs seem better at color deliniation and separation, so a caucasian model's blush really looks like blush, not a melding of magentas and reds in the model's skin with the blush.

My experience with the D800 is similar to yours.  D3x was better by a bit.  There is a global yellow/red bias, and particular hues of red have a strong bias or shift.  I tortured myself working through this.  I would get a somewhat neutral file and fix it in PS with a layer of Hue/Saturation, then some curves in the seperate red/yellow channels.  Then the color is great.

That being said, I think the 1ds3/5d2 had better skin out of the box. 

Hi Bcooter,

Thank you very much for that comprehensive response and my apologies for responding almost a month late (Work, travel and whatnot). I have been in touch with both gentlemen that you have mentioned and they have helped me understand  so much about the medium format landscape. I am evaluating the various options now and should be able to make my decision by the end of the year.

I have also been practicing on Capture One and have finally managed to get it to deliver some nice colors out of my Nikon files. They now look a bazillion times better than they ever did in Lightroom (Even after custom profiles) and now I can viably use my D800 as the second camera in my setup.

Here's an example:

http://www.sandeepmurali.com/p234558723/h726e47e7#h726e47e7

That image looked "Dead" after the LR conversion (Posted a few pages back). I am much happier with this version.

Funny thing is, out of fun, I was working on some very old files of mine, from my erstwhile D70s. Ran them through Capture One and finished in Photoshop. I am so surprised at how smooth the skintones were from that camera. In a way, more analog and less plasticky than the current gen cams, including the D800. I am not a physicist, but I get a feeling that in the quest of stupid high ISOs, camera makers now are losing focus on base ISO color reproduction (Pun intended).

Here's an example:


Logged

eronald

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6642
    • My gallery on Instagram

I think anyone who wants to buy MF should evaluate *exactly* the intended solution in his studio.

On a related note, I picked up a Canon 1Ds3 on a swap for my 5D2, and find the focus on the big Canon is excellent (unlike the 5D2) and the files very nice. Forum members  might find this camera current and of interest for "people" use.
Edmund
Logged
If you appreciate my blog posts help me by following on https://instagram.com/edmundronald

jduncan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 434

I think anyone who wants to buy MF should evaluate *exactly* the intended solution in his studio.

On a related note, I picked up a Canon 1Ds3 on a swap for my 5D2, and find the focus on the big Canon is excellent (unlike the 5D2) and the files very nice. Forum members  might find this camera current and of interest for "people" use.
Edmund


The auto focus system in the new Canon  1dx and  5D3 is one generation ahead from Nikon's and the older Canons. I because for my type of shooting any modern autofocus will do, so I use Nikon. But using the Canon is eye opening.

Best regards,

J. Duncan
Logged
english is not my first language, an I k

BobDavid

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3307

The skin tones of Hassey files right out of the box are beautiful. Weirdly the RAW files produced by the Sony a850/a900 are very comparable to the look of Aptus 22 MP files. I am using a Nikon d800 now. I like the camera a lot. My work requires the pixel density and the selection of lenses that Nikon offers. It is possible to tame the flesh tones coming out of the Nikon, however it helps to have an accurate monitor, a good eye for color, and knowing the camera really well. When you compare a Nikon d800 file to a 22mp MFB file you will see the difference. Though CCD chips have less dynamic range than the current generation of CMOS chips, MF files look smoother, punchier, and more nuanced. I really love the look of the files rendered from the ancient Ixpress 96c MF back that I used to use on occasion back around 2005. The RAW developer, Flexcolor was a bit retro 1995, but it was powerful.
Logged

Phil Indeblanc

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2017

Hi Sandeep,

I read your post and some of the replies....

I have to agree with folks that are suggesting to FIRST look into your post processing. Take on a differnet approach to it. Also I think you underestimate the "negative" impact the AA has. I am surprised you didn't get a D800E?
(I know the moire). I don't do a lot of fabric, and willing to do the post if I come across the issue. for 10 years it has happened a hand full of times that I was easily able to manage.

Years back,before I purchased a DB, I had a 1Ds and I also had a Kodak SLRc. The Kodak had no AA, and the files were 3D MF like. The 1Ds was soft. (The "bite" has a lot to do with the AA)... Except I had such trouble with the way the Kodak handled lighting, specially if any came in direction of the lens, I would get magenta overcasts and such, so I put that down and went the MF route. Shame about the Kodak, it had a great software tool set, ISO 06!.  I would guess the same sensor was what made the Kodak Db so successful for a while. 

I have been meaning to rent a D800E to see how it would do with my studio shoots. I would never consider a camera with a AA filter for my studio needs.
I also think the way lighting is setup gives a great impression on the overall impact of an image. I have seen well balanced light from a 35mm look very MF tonally. Before you plunk down any money, I would do as others have suggested and see what you can do about exposing in the lighting, and post processing. (Sometimes slight fill lights can do wonders in giving you range). 

The problem with the above info I give is that I have only compared a P25MFdB  to my AA-soft 5Dm2 as a equal mpixel size. The Kodak was 14mp, and the 1Ds was 11mp.  So if the NonAA D800E is anything like a Kodak SLRc at a 36mp resolution, this SHOULD bypass a P25 MFdB ???  There is a 2 bit difference between most top end Dslr vs MFdb, 14/16 bit, 2 channels are for noise... Erik had written something about it a while back.

I have been using this back for some time, and I do love the results. I do want more resolution and more DOF, so that was my reason to look for the D800E. But I don't see myself doing anything about it.
I do rememebr sopmething about the P30 that I would avoid. Something about the lens over sensor aliagnment, is not suited for Tilt/Shift if you ever do go toward tech cam.

Also, I LOVED this on the Mamiya RZ. I have a 35mm lens which is a bit fishy, but an amazing lens. The glass is really good, and overall a great setup. Beautiful large view(I switch the gg glass to brighter). I would also caution about the Hass software RAW converter, If C1 has support for Hass, then no prob. But I had bad experience with it when testing years back(surely they must have improved).  I think the P40 is a top contender also for long exposure, and the 45 all around. I was looking to upgrade to the P45, and see them online for about $7-9k. If I had a great gig and justify it I would get one of these.

Sometime back, the MFDB makers were adamant about the photosite size being the dominant trait compared to 35mm sensors(reason the 1Ds files are beautiful /11microns), as some are 12microns large. Now with newer MFDBs having 6 or so, and surely technology has gotten better with less gap microlenses, better SN ratio, light fill factor, etc... did this bridge the gap? and if so, why couldn't 35MM with the same range photosites be on par with the MFDBs? Maybe if Foveon would license the sensor there maybe some interesting designs in the future?

I hope this helps.

Logged
If you buy a camera, you're a photographer...

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

Here are some images from a recent shoot, with mixed cameras (Sonya Alpha 99 + Hasselblad V/P45+), downscaled to 4000x4000 pixel frame: http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/RealWorld/

Here are some comparable images from 24MP FF and the Hassy/P45+:

http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/MFDB_VS_DSLR2/
http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/MFDB_VS_DSLR/

Here are some more samples:

http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/Samples3/
http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/Samples2/
http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/Samples/

The Hasselblad lenses are old designs, new lenses on a Phase One camera would be more modern designs.

The P45+ uses a Kodak sensor while most Aptus backs and later generation Phase One backs have Dalsa sensors. The impression I got is that Dalsa sensors may have better color.

The reason that the P30 did not work well with shifts is that it uses microlenses, so does the IQ180 while the IQ160 has no micro lenses, as far as I know.

Best regards
Erik
« Last Edit: November 09, 2013, 10:08:28 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

Phil Indeblanc

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2017

If you can do yourself a favor and not look at what camera took these and come up with true reasons why you like one over the other, and convince yourself somehow that some post processing wont make the satisfactory adjustment it needs, then you have yourself a winning choice. ....
....Not saying you cant eliminate a few off the bat for SOOC reasons.
Logged
If you buy a camera, you're a photographer...

ondebanks

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 858

However, where that DR is placed is of importance to me. The D800, while having industry leading DR, places most of it in the deep shadows. Great when one is lifting the shadows, but the highlight spectrum contains nowhere as much information. No matter how I vary my lighting, whatever modifier I use, the way the mids move to the highights is simply too... for the lack of a better word, plasticky. Even the samples I have seen from a 22MP back has better "bite" in it and much more graceful gradation from mids to highlights.

Hi Sandeep,

What you describe is real, and a direct result of the sensor physics/electronics.

The D800 and other recent CMOS cameras can be crudely categorized as "low signal, low noise" while the MFD units - especially the older backs with larger, fewer pixels - are "high signal, high noise" in comparison. They do not achieve their high signal because they are more light sensitive (on the contrary, they are typically one stop lower in quantum efficiency), but because (1) they are set one stop lower in base ISO, making you double exposure time to receive more signal, and (2) they have larger pixel surface areas. And they have larger pixel well capacities to accomodate that larger signal. 

Analogy - consider a basin and a coffee mug. I leave them both outside in a rain shower. If I cover half of the top of the basin beforehand, it will catch only half of the water falling on it - this is like the one stop lower q.e. of the MFD sensors. Nevertheless, it still has a larger open area to the rain than the mug, so it will still catch more raindrops in a given time - this is like the pixel size/surface area difference. If I additionally take the mug inside halfway through the shower, the basin gets a further doubling of its catch - this is like the 1 stop lower ISO and corresponding longer shutter speed. Then I take the basin inside as well. I carefully pour the contents of the mug into a measuring jug, measuring all but a few drops that adhere to the sides of the mug. I try to do the same with the basin, but I slosh and spill some water - this is like the readout noise difference. I reckon that in both cases, I have measured 99% of the rainwater collected, or to put it another way, the volume I measured was 99 times larger than the estimated error - this is like the dynamic ranges being the same.

Anyway, if (as I have done) one models the signal to noise curves for the individual noise components in a sensor, and for their composite effect, you get quite different trends for these different types of sensor. The MFD units are killed by high readout noise at the shadow end (which is why they are poor at high ISO, when the shadows are pushed to become mid tones), but in the base ISO mid tones and highlights, where readout noise slips into insignificance behind shot/Poisson noise, they have better S/N due to their larger signals and lower percentage shot noise - as you put it, "much more graceful gradation from mids to highlights". The D800 and its ilk, on the other hand, will keep discriminating stop after stop of shadow detail thanks to its low readout noise, but as it is doing so with mere handfuls of signal photons at the bottom end, it can look rather quantized compared to the thousands of distinct shades of equally low S/N scuzz at the bottom of the DR range for MFD images.

In short, as I have said often, the quality along the range covered by the DR is very important...few people realize this and they look only at the overall quantity of the DR.

Ray
Logged

ondebanks

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 858

Years back,before I purchased a DB, I had a 1Ds and I also had a Kodak SLRc. The Kodak had no AA, and the files were 3D MF like. The 1Ds was soft. (The "bite" has a lot to do with the AA)... Except I had such trouble with the way the Kodak handled lighting, specially if any came in direction of the lens, I would get magenta overcasts and such, so I put that down and went the MF route. Shame about the Kodak, it had a great software tool set, ISO 06!.  I would guess the same sensor was what made the Kodak Db so successful for a while. 

Phil - no, that Kodak SLRc had a completely different sensor tech to the Kodak DBs (and even the earlier Kodak DSLRs based on the Nikon F5 and Canon 1n). The 14n, SLRn, and SLRc had a Fill Factory CMOS sensor, while all the other Kodaks had Kodak's own CCD sensors.

I would agree however that "the "bite" has a lot to do with the AA". Big pixels without an AA filter have much better MTF than ones with an AA filter, even smaller ones. I see this all the time in my Canon 5DII vs my Kodak DCS645M back. The Kodak wins. It also wins on colour.

Sometime back, the MFDB makers were adamant about the photosite size being the dominant trait compared to 35mm sensors(reason the 1Ds files are beautiful /11microns), as some are 12microns large. Now with newer MFDBs having 6 or so, and surely technology has gotten better with less gap microlenses, better SN ratio, light fill factor, etc... did this bridge the gap? and if so, why couldn't 35MM with the same range photosites be on par with the MFDBs?

Well the 1Ds has 9 micron pixels, not 11 microns...but you are onto something here. There are Phase One and Leaf 5.2 micron pixels with no AA filter, and there are Nikon and Sony 4.9 micron pixels with no AA filter. On a per pixel comparison, the gap has been bridged to a large extent. Several differences remain however: colour response, noise characteristics along the DR range, and above all, MFD will always have larger sensors...

Ray

Logged

ondebanks

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 858

The reason that the P30 did not work well with shifts is that it uses microlenses, so does the IQ180 while the IQ160 has no micro lenses, as far as I know.

Hi Erik,

I always thought that too: no microlenses for both the P65+ and its successors the H4D-60, IQ160 and Credo 60. It's the natural conclusion when you see a DB with a low base ISO of 50 - it can't have high q.e., so surely it can't have microlenses! Further evidence came from the P65+ not having a bad reputation for lens cast with tech cams...not nearly as problematic as trying to use a microlensed Kodak sensor on a tech cam.

And everytime I said "no microlenses in the 60 MP backs" on forums like this, with Phase One, Leaf, and Hasselblad reps/dealers present, no-one contradicted me. So it appeared to be fact.

But then Dalsa belatedly (in 2012) released a datasheet for the FTF9168C used in these 60MP backs, and there it was...there are microlenses!
I was surprised to see that this CCD is a scaled up version of the 48MP FTF6080C (which would have been a twin/competitor to Kodak's 50MP CCD if any company had decided to use that Dalsa chip). The FTF6080C datasheet had been kicking around for a few years. Microlenses, yes. High q.e., not really...midway between Kodak non-microlensed and microlensed CCDs. So, modest q.e., together with Dalsa's trademark deep pixel capacities...that's why it has a base ISO of 50.

Dalsa have a quite fascinating conference proceedings paper online, from 2008: "A 36×48mm^2 48M-pixel CCD imager for professional DSC applications". Let me know if you cannot access the full text PDF version; I can email it to you. This describes the FTF6080C and compares it to their previous 33MP sensor, and Kodak's 39 MP sensor. They mention things like:
"Improvement of QE by integrating an anti-reflective nitride layer on the poly-silicon gate electrodes;
Application of gap-less microlenses on top of the RGB color filter pattern, optimized for wide angular response and low color cross-talk"


Combined, these steps did result in considerable q.e. improvements over earlier, unlensed Dalsa sensors...but it has to be pointed out that those earlier sensors were already well behind the unlensed Kodak sensors.
So the q.e. order is (worst <-> best): Dalsa unlensed, Kodak unlensed, Dalsa lensed (48 & 60 MP), Kodak lensed (18, 31 and 40MP).

The stuff in that paper about how they optimized their microlenses for wide angular response is great. It explains why the big brother of that sensor in the P65+ did not have a bad reputation for lens cast with tech cams.

Further improvements were that "the readout noise at 25MHz pixel frequency was decreased from 17e- noise to 11e- noise after CDS". That pips the best Kodaks (12.5e-).

Nothing was done however to improve the rather awful dark noise in long exposures...not until the re-engineered miracle that is the IQ260!

Ray


« Last Edit: November 11, 2013, 11:08:35 am by ondebanks »
Logged

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com

And everytime I said "no microlenses in the 60 MP backs" on forums like this, with Phase One, Leaf, and Hasselblad reps/dealers present, no-one contradicted me. So it appeared to be fact.

As I recall I once did corrected this notion.

But usually when I notice a minor technical error that does not mislead a reader on the actual practical use of the equipment I bite-my-lip and ignore it.

In other words it's just as useful for a user to hear "no microlenses on the 60mp sensor" as it is to hear "the micro lenses are there but are less prohibitive of light coming in at a strange angle" when it comes to picking a back which works well on a tech camera.

I'd imagine it's even possible I've used "no microlenses" once or twice myself in search of a succinct way to categorize backs for people as either movement-on-tech-camera compatible or not.

If you look back at my description over the years of the P30+ sensor you'll notice I used adjectives like "strong" or "deep" microlenses to describe the reason that back cannot be used on a technical camera with any significant movements with normal/wide lenses. At least I think that's how I've described it, hard to remember everything you've written after this many posts :).
« Last Edit: November 11, 2013, 11:24:21 am by Doug Peterson »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

Just as a general comment, I would suggest that the main (or rather only) benefit of MFDs come from the larger area.

Best regards
Erik


Phil - no, that Kodak SLRc had a completely different sensor tech to the Kodak DBs (and even the earlier Kodak DSLRs based on the Nikon F5 and Canon 1n). The 14n, SLRn, and SLRc had a Fill Factory CMOS sensor, while all the other Kodaks had Kodak's own CCD sensors.

I would agree however that "the "bite" has a lot to do with the AA". Big pixels without an AA filter have much better MTF than ones with an AA filter, even smaller ones. I see this all the time in my Canon 5DII vs my Kodak DCS645M back. The Kodak wins. It also wins on colour.

Well the 1Ds has 9 micron pixels, not 11 microns...but you are onto something here. There are Phase One and Leaf 5.2 micron pixels with no AA filter, and there are Nikon and Sony 4.9 micron pixels with no AA filter. On a per pixel comparison, the gap has been bridged to a large extent. Several differences remain however: colour response, noise characteristics along the DR range, and above all, MFD will always have larger sensors...

Ray


Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

I made an experiment recently, the idea being that large pixels with no OLP produce fake detail. I just took a feather I found on a beach and photographed it at home. I got a lot of artifacts both with my Sonnar 150/4 and the Planar 80/2.8 using the same shooting position, so I guess this is something we always see.

After shooting with the "Blad" I shot from the same position with the Sony Alpha 99 that has somewhat smaller pixels with an OLP filter. The Alpha 99 had less color artefacts but still a lot. Finally I shot an image with my Alpha 77, which has smaller  pixels (3.9 microns I think). The Alpha 77 has also an OLP filter, I think, the Alpha 77 image is best.

These images were taken at 150 mm setting at fixed distance. The images were scaled down P45+ dimensions using ImageMagick which uses Lanzcos I think.

What I see is that small pixels with OLP filtering may be the best solution. Please note, in this setting I ignored sensor size, I just looked at the image quality at fixed magnification of the sensor image.

http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/FakeDetail/comparison1.html

Best regards
Erik


Phil - no, that Kodak SLRc had a completely different sensor tech to the Kodak DBs (and even the earlier Kodak DSLRs based on the Nikon F5 and Canon 1n). The 14n, SLRn, and SLRc had a Fill Factory CMOS sensor, while all the other Kodaks had Kodak's own CCD sensors.

I would agree however that "the "bite" has a lot to do with the AA". Big pixels without an AA filter have much better MTF than ones with an AA filter, even smaller ones. I see this all the time in my Canon 5DII vs my Kodak DCS645M back. The Kodak wins. It also wins on colour.

Well the 1Ds has 9 micron pixels, not 11 microns...but you are onto something here. There are Phase One and Leaf 5.2 micron pixels with no AA filter, and there are Nikon and Sony 4.9 micron pixels with no AA filter. On a per pixel comparison, the gap has been bridged to a large extent. Several differences remain however: colour response, noise characteristics along the DR range, and above all, MFD will always have larger sensors...

Ray


« Last Edit: November 11, 2013, 02:35:13 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

eronald

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6642
    • My gallery on Instagram

Ray,

 I your analysis you assume the sensors are linear in the high range; but AFAIK  some new CMOS sensors use the antiblooming to get a (tuneable) shoulder in the highlights. I checked this assumption with Aptina and they concurred, said this had been going on for some time.

Edmund
Logged
If you appreciate my blog posts help me by following on https://instagram.com/edmundronald

Phil Indeblanc

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2017

Nice samples to look over Erik.
Without looking at what camera they were, subjectively I thought the A99 looked nice, while the 77 had less moire, or "filler" info.
So you used the same lens on all?

I would love to see this test on harder or reflective subjects, perhaps using a Schneider 180 HM or 120 macro digital glass. Or even the Rodenstock perhaps?

What I did sense from my smaller files was that in post production, retouching a MFdb file was always meatier in having more information. I havn't shot anything that I really edit hard other than the MFdb files, so I don't have much to compare. 14bit makes a difference, perhaps the 16bit still may hold a meatier file? But to what degree?

Thanks for correcting my micron size on the 1Ds, and good to know about the Kodak difference. I guess I forgot. I try to keep the tech stuff to a minimum, and work on the process and art of the shoots, but  I am a techy to some degree:-) Having said that, the SLRc Kodak made some amazing files. I still have it, just not had much reason to do any testing, specially with so many new camera since that, and FINALY Nikon out (besides Foveon) with the first pro level DSLR without a AA!(as far as I know) .
Logged
If you buy a camera, you're a photographer...

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

Regarding the aliasing samples I was not clear enough.

I used the Sonnar 150/4 on the Hasselblad and my 70-400/4-5.6 at 150 mm on the Alphas. I could have used the Sonnar on the Alphas to, with an adapter, but decided not to. MTF-wise the two lenses are pretty close, with the Sonnar having the edge. The images were shot from tripod at the same distance. In the set conditions, smaller pixels make a larger image, but I scaled down all images using a decent scaling method.

I don't have Schneider or Rodenstock lenses. On the other hand the Sonnar 150/4 is pretty good and a better lens is causing more artifacts in this kind of test.

What this essentially told me was that smaller pixels are good.

Best regards
Erik


Nice samples to look over Erik.
Without looking at what camera they were, subjectively I thought the A99 looked nice, while the 77 had less moire, or "filler" info.
So you used the same lens on all?

I would love to see this test on harder or reflective subjects, perhaps using a Schneider 180 HM or 120 macro digital glass. Or even the Rodenstock perhaps?

What I did sense from my smaller files was that in post production, retouching a MFdb file was always meatier in having more information. I havn't shot anything that I really edit hard other than the MFdb files, so I don't have much to compare. 14bit makes a difference, perhaps the 16bit still may hold a meatier file? But to what degree?

Thanks for correcting my micron size on the 1Ds, and good to know about the Kodak difference. I guess I forgot. I try to keep the tech stuff to a minimum, and work on the process and art of the shoots, but  I am a techy to some degree:-) Having said that, the SLRc Kodak made some amazing files. I still have it, just not had much reason to do any testing, specially with so many new camera since that, and FINALY Nikon out (besides Foveon) with the first pro level DSLR without a AA!(as far as I know) .

Logged
Erik Kaffehr
 
Pages: 1 2 [3] 4 5 ... 10   Go Up