Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: DXO Pro built into raw converter = H3D...  (Read 7512 times)
narikin
Sr. Member
****
Offline Offline

Posts: 854


« on: October 20, 2006, 09:03:25 PM »
ReplyReply

Maybe I've missed it, (can be a bit slow on these things), but surely what Hasselbald is cleverly doing with the H3 is locking in the DXO pro type lens corrections into its Raw converter as a means of improving its apparent resolution/sharpness.

seems like a smart move as the camera/lens combo is locked down to few variables, and the raw converter will know what lens/aperture/focus was used and be able to apply correction appropriately. It will make the images appear a heck of a lot sharper in any trial, and mask the rather average H series lens performance well.

nothing to stop Phase doing the same in 4.0, but they need to buy a piece of DXO's action to get a jump start, and produce lens profiles for each basic camera / lens combo (Mamiya/ Hassy/ Contax) quickly. not that hard as there's not too many variables out there. If they do get deeply entwined with Mamiya its essential they put in this feature too, or the game is over, surely?
« Last Edit: October 20, 2006, 09:04:38 PM by narikin » Logged
Graham Mitchell
Sr. Member
****
Offline Offline

Posts: 2282



WWW
« Reply #1 on: October 21, 2006, 12:39:51 AM »
ReplyReply

It's not that big a deal. We are all using lenses which work without correction right now. What Hasselblad appears to be doing is designing a lens which RELIES on correction, so it requires a closed system which will apply the correction automatically.

There's no reason this same correction can't be applied by other software.
Logged

Graham Mitchell - www.graham-mitchell.com
narikin
Sr. Member
****
Offline Offline

Posts: 854


« Reply #2 on: October 21, 2006, 07:51:09 AM »
ReplyReply

please correct me if I'm wrong, but there is no correction software for MF lenses.

DXO pro doesnt make 'profiles' for anything over 35mm, so what could be used? the lens abberation corection in lightroom etc are way short of what DXO pro does.

a problem is that any side by side test is going to give the 39Mp H3 better results in sharpness over a 39Mp P45 + H2 because of the optical corrections, even with the exact same lens on each unit.

surely this is the future - with lens correction factors built into the raw processing software? its somthing Phase and Leaf just cant ignore.
Logged
Graham Mitchell
Sr. Member
****
Offline Offline

Posts: 2282



WWW
« Reply #3 on: October 21, 2006, 08:10:47 AM »
ReplyReply

There may not be any off the shelf solutions available today, but it's just a matter of making prfiles. If there is a demand, the profiles will come. You might be able to make your own, too.
Logged

Graham Mitchell - www.graham-mitchell.com
peterhorsley
Newbie
*
Offline Offline

Posts: 23


« Reply #4 on: October 21, 2006, 09:17:22 AM »
ReplyReply

This is the nub of the whole deal.  Lens designers no longer have limit themselves to delivering optically 'perfect' ray traces to the image plane of a camera.  The whole 'picture quality' game has a new dimension.

Software like  PTLens  can deliver 'distortion-free' images from any camera given the right processing information.  Of course there is nothing-for-nothing but with 39MPx of information the software changes are almost invisible.

So the Hasselblad people have leapt ahead: what is the best result for the combination of a 28mm lens/48mmx36mm sensor/H3 body/software-corrections possible?

The potential is fantastic. Imagine a tilting and shifting lens that produces rectalinear, chromatic aberration-free results on H3 medium format. Mind blowing. Truely large format flexiblity from a hand held camera.  Maybe a razor sharp 120mm/f1.4. What about a light weight lens? What about a 24mm?

The down side is that it is a front-of-camera to finished-file solution.  Throw away your V series 'blads and pre-2006 H series if you want the convenience of shoot it/print it.  

Today the ENTRY price is US$22,000 for a body/back/lens.  And tomorrow? (5 years time?)

For me, the idea is great; the marketing/long-term-commitment sucks.

Peter
Logged
Graham Mitchell
Sr. Member
****
Offline Offline

Posts: 2282



WWW
« Reply #5 on: October 21, 2006, 09:54:20 AM »
ReplyReply

Quote
Imagine a tilting and shifting lens that produces rectalinear, chromatic aberration-free results on H3 medium format.

Does Hasselblad even have a T/S lens for the H platform? In order for this to work, the shift amount and tilt angle would all need to be recorded for use in the correction software, as well as focal distance and aperture. I don't believe this will happen anytime soon!
Logged

Graham Mitchell - www.graham-mitchell.com
pss
Sr. Member
****
Offline Offline

Posts: 960


WWW
« Reply #6 on: October 21, 2006, 01:09:40 PM »
ReplyReply

i use mostly schneider lenses for my 6008 system (40,80,90,180) and none of them need any kind of correction...i don't even need to apply sharpening for the most part...
nobody was worried about lens correction when we shot film using the entire lens, now we use only the sweet spot, so this really should not be an issue...
of course i am talking about MF lenses here...canon wideangles (or probably nikon as well) are a different story, but they simply weren' built for the tasks trown at them nowadays....now people want ot shoot landscapes to be printed 30x40 with their 1DsmkIIs and 24mm....nobody would ever have tried that with an EOS and film....that was a job for MF or 4x5, which it still is...
Logged

Gary Ferguson
Sr. Member
****
Offline Offline

Posts: 522


WWW
« Reply #7 on: October 21, 2006, 03:05:37 PM »
ReplyReply

Quote
Does Hasselblad even have a T/S lens for the H platform? In order for this to work, the shift amount and tilt angle would all need to be recorded for use in the correction software, as well as focal distance and aperture. I don't believe this will happen anytime soon!

Interesting point.

I use DxO occasionally, mainly on certain Canon lenses like the 24-105mm and 14mm where there's complex distortion issues which DxO fixes more convincingly (and quickly) than doing the job myself in Photoshop. However, I'd really appreciate some DxO asistance with the three Canon T&S lenses, but no deal, as you say DxO can't handle the uncertainty of lens movements and say there's no fix in prospect. My Hasselblad dealer says that at least one H series T&S lens is in development (and has been for at least 18 months) so it will be interesting to see if they deliver a T&S optic with digital correction.

The trade off for retrofocus lens designers between distortion and resolution is clear. I use the Hasselblad V series with a Phase One back and a selection of wide lenses, the 38mm Biogon (non retrofocus so you get zero distortion, high resolution, but no lens movements and quite noticeable vignetting), the older 40mm Hasselblad lenses (moderate distortion, very low edge resolution, moderate movements possible on a Flexbody camera, moderate vignetting), and the new 40mm IF (very severe distortion, very high resolution, quite large movements possible on a Flexbody camera, moderate vignetting).

I guess you pays your money and you takes your choice, but if T&S digital correction ever becomes a reality then we could walk down both sides of the street at the same time!
Logged
marc gerritsen
Sr. Member
****
Offline Offline

Posts: 299


WWW
« Reply #8 on: October 21, 2006, 08:08:57 PM »
ReplyReply

What would be your suggestion to use on a 35mm Hasselblad lens, which has moustache distortion.
The lens correction software of Hasselblad has been due to come out any minute for the last couple of months, but I still have been fixing my architectural files with archaic PS software. I have sent Hasselblad a couple of raw files via send-it and they returned it perfectly straight. My problem is now that I know it can be done I want to use it on about 200 files ready for post production.
If this software is not available for download in the next comming days I will have to look at alternatives, so any suggestions?
Marc
Logged

narikin
Sr. Member
****
Offline Offline

Posts: 854


« Reply #9 on: October 22, 2006, 10:56:19 AM »
ReplyReply

Quote
i use mostly schneider lenses for my 6008 system (40,80,90,180) and none of them need any kind of correction...
[a href=\"index.php?act=findpost&pid=81529\"][{POST_SNAPBACK}][/a]


thats what we all think, until you see how much more there is to come when you do correct it... simply amazing what can be done.
« Last Edit: October 22, 2006, 10:56:30 AM by narikin » Logged
zzzone
Newbie
*
Offline Offline

Posts: 36


« Reply #10 on: October 22, 2006, 11:47:41 AM »
ReplyReply

Quote
What would be your suggestion to use on a 35mm Hasselblad lens, which has moustache distortion.
The lens correction software of Hasselblad has been due to come out any minute for the last couple of months, but I still have been fixing my architectural files with archaic PS software. I have sent Hasselblad a couple of raw files via send-it and they returned it perfectly straight. My problem is now that I know it can be done I want to use it on about 200 files ready for post production.
If this software is not available for download in the next comming days I will have to look at alternatives, so any suggestions?
Marc
[a href=\"index.php?act=findpost&pid=81569\"][{POST_SNAPBACK}][/a]


Hi Marc

There are ways and means.  I do very large scale panoramic images stitched them together using a 1Ds11 and a 5D.  In the process of doing this I have learnt much about lens distortion in able to stitch the images together satisfactorily.  One piece of software I have used is really cheap and supplied by www.kekus.com.  The downside there is a bit of a learning curve and you do need a panoramic tripod head to calibrate a lens.  But it can be done, though I haven't tried with MF!  The main man at Kekus Digital is most helpful but this is not a quick fix solution.  If you can take a week out to climb the learning curve it is a little steep at the start.

I use some Canon lenses with rather poor distortion characteristics that I put through my own PS actions which include the lensfix calibration supplied by Kekus.  The images are greatly improved.  Indeed I wrote to Canon years ago making them aware of the procedures I have adopted to correct their distortions.  There is the facility to correct chromatics as well within this software and you attach it to PS as a plugin.  I have been doing this for a long time way before DXO was ever on the market.  

My letter to Canon or email I think it was to CPS stated that Canon could improve so much with software.  Distortion, chromatics etc all could be sorted.  You carry on about 16 bit 22 million pixels etc etc but why don't they simply improve big time what they have already.  I saw the 28mm lens at Photokina on the Hasselblad stand and I spoke to the guys there about exactly this point.  At last something is happening.  Canon could make a fortune with software per lens or a most superb software suite to cover all.  Unfortunately I don't think they are interested because of the limited market.  Furthermore just look at there own software, nice raw conversion but otherwise poor.

This has to be the future.  It is not if but when.  Unfortunately for us it is when the mass market demands.  But it will happen I believe.

Regards


Steve Townsend
Logged
pprdigital
Sr. Member
****
Offline Offline

Posts: 422


WWW
« Reply #11 on: October 22, 2006, 12:10:32 PM »
ReplyReply

Quote
Hi Marc

There are ways and means.  I do very large scale panoramic images stitched them together using a 1Ds11 and a 5D.  In the process of doing this I have learnt much about lens distortion in able to stitch the images together satisfactorily.  One piece of software I have used is really cheap and supplied by www.kekus.com.  The downside there is a bit of a learning curve and you do need a panoramic tripod head to calibrate a lens.  But it can be done, though I haven't tried with MF!  The main man at Kekus Digital is most helpful but this is not a quick fix solution.  If you can take a week out to climb the learning curve it is a little steep at the start.

I use some Canon lenses with rather poor distortion characteristics that I put through my own PS actions which include the lensfix calibration supplied by Kekus.  The images are greatly improved.  Indeed I wrote to Canon years ago making them aware of the procedures I have adopted to correct their distortions.  There is the facility to correct chromatics as well within this software and you attach it to PS as a plugin.  I have been doing this for a long time way before DXO was ever on the market. 

My letter to Canon or email I think it was to CPS stated that Canon could improve so much with software.  Distortion, chromatics etc all could be sorted.  You carry on about 16 bit 22 million pixels etc etc but why don't they simply improve big time what they have already.  I saw the 28mm lens at Photokina on the Hasselblad stand and I spoke to the guys there about exactly this point.  At last something is happening.  Canon could make a fortune with software per lens or a most superb software suite to cover all.  Unfortunately I don't think they are interested because of the limited market.  Furthermore just look at there own software, nice raw conversion but otherwise poor.

This has to be the future.  It is not if but when.  Unfortunately for us it is when the mass market demands.  But it will happen I believe.

Regards
Steve Townsend
[a href=\"index.php?act=findpost&pid=81624\"][{POST_SNAPBACK}][/a]

The Hasselblad solution is not a software-only solution. There are micro-processors that have been implemented in the H3D that provide extremely detailed information - including alteration of the light path as it enters first the IR filter layer - then the sensor layer itself, both of which produce distortions regardless of the characteristics of the lens. This allows the software to produce a very accurate correction of the distortion, perhaps more accurate than with software alone.

Steve Hendrix
PPR Digital
www.ppratlanta.com
Dealer for Leaf, Hasselblad and Sinar
Logged

Steve Hendrix
zzzone
Newbie
*
Offline Offline

Posts: 36


« Reply #12 on: October 22, 2006, 12:25:24 PM »
ReplyReply

Quote
The Hasselblad solution is not a software-only solution. There are micro-processors that have been implemented in the H3D that provide extremely detailed information - including alteration of the light path as it enters first the IR filter layer - then the sensor layer itself, both of which produce distortions regardless of the characteristics of the lens. This allows the software to produce a very accurate correction of the distortion, perhaps more accurate than with software alone.

Steve Hendrix
PPR Digital
www.ppratlanta.com
Dealer for Leaf, Hasselblad and Sinar
[a href=\"index.php?act=findpost&pid=81628\"][{POST_SNAPBACK}][/a]
Steve

Who better than the manufacturer to get the best out of their hardware.  They know where the problems are better than anyone presumably.    It does seem to make sense and I am very pleased that they are trying hence my desire to test drive the new H3D.

Steve
Logged
Steve Kerman
Full Member
***
Offline Offline

Posts: 142


« Reply #13 on: October 22, 2006, 04:09:01 PM »
ReplyReply

Quote
The Hasselblad solution is not a software-only solution. There are micro-processors that have been implemented in the H3D that provide extremely detailed information - including alteration of the light path as it enters first the IR filter layer - then the sensor layer itself, both of which produce distortions regardless of the characteristics of the lens. This allows the software to produce a very accurate correction of the distortion, perhaps more accurate than with software alone.

Steve Hendrix
PPR Digital
www.ppratlanta.com
Dealer for Leaf, Hasselblad and Sinar
[a href=\"index.php?act=findpost&pid=81628\"][{POST_SNAPBACK}][/a]
Steve, as a guy who designs digital systems for a living, I have to say that I'm completely baffled by what you just wrote.  Everything you describe is a software operation.  And all a microprocessor does is execute software.

With no offense intended towards you, this sounds like something that came from a press release written by a PR guy who was employing the "Baffle 'em with b******t" principle in an atttempt to convice people that this can't be done in a Photoshop plug-in.
Logged
Photomangreg
Jr. Member
**
Offline Offline

Posts: 65


« Reply #14 on: October 22, 2006, 04:34:57 PM »
ReplyReply

Quote
Steve, as a guy who designs digital systems for a living, I have to say that I'm completely baffled by what you just wrote.  Everything you describe is a software operation.  And all a microprocessor does is execute software.

With no offense intended towards you, this sounds like something that came from a press release written by a PR guy who was employing the "Baffle 'em with b******t" principle in an atttempt to convice people that this can't be done in a Photoshop plug-in.
[a href=\"index.php?act=findpost&pid=81649\"][{POST_SNAPBACK}][/a]

Steve Kerman,

The way I understood it from the Hasselblad rep the other day, is that because of the integration between the lens, body and digital magazine, there will actually be the opportunity to change where the lens focuses, this will be a physical change and not a software change.  Both the IR filter and the protective glass filter distort how the image hits the sensor, obviously this wasn't a problem with film.  What HAsselblad has been able to do is map out the lenses at each F-stop and actually adjust where the lens focuses based on the amount of correction needed.  Sure it's powered by software and processors, etc., but it is physically changing where the lens focuses.  The more wide open you are shooting, the more, I guess difraction is the word, you get as the image passes through the two filters.
Logged
Steve Kerman
Full Member
***
Offline Offline

Posts: 142


« Reply #15 on: October 22, 2006, 04:47:34 PM »
ReplyReply

Quote
Steve Kerman,

The way I understood it from the Hasselblad rep the other day, is that because of the integration between the lens, body and digital magazine, there will actually be the opportunity to change where the lens focuses, this will be a physical change and not a software change.  Both the IR filter and the protective glass filter distort how the image hits the sensor, obviously this wasn't a problem with film.  What HAsselblad has been able to do is map out the lenses at each F-stop and actually adjust where the lens focuses based on the amount of correction needed.  Sure it's powered by software and processors, etc., but it is physically changing where the lens focuses.  The more wide open you are shooting, the more, I guess difraction is the word, you get as the image passes through the two filters.
[a href=\"index.php?act=findpost&pid=81653\"][{POST_SNAPBACK}][/a]
OK, I'll buy that explanation.  A change in focus point is clearly a before-the-fact hardware correction.

On the other hand, it would be pretty trivial to have an open system that supports this kind of correction, because the amount of information that needs to pass between the back and the camera to be able to do this is small, and easily documented.

I acknowldge, however, that this kind of technology is a valid proprietary market advantage that a company is within its rights to keep proprietary.
« Last Edit: October 22, 2006, 04:55:38 PM by Steve Kerman » Logged
eronald
Sr. Member
****
Offline Offline

Posts: 3946



WWW
« Reply #16 on: October 22, 2006, 05:38:14 PM »
ReplyReply

Quote
OK, I'll buy that explanation.  A change in focus point is clearly a before-the-fact hardware correction.

On the other hand, it would be pretty trivial to have an open system that supports this kind of correction, because the amount of information that needs to pass between the back and the camera to be able to do this is small, and easily documented.

I acknowldge, however, that this kind of technology is a valid proprietary market advantage that a company is within its rights to keep proprietary.
[a href=\"index.php?act=findpost&pid=81654\"][{POST_SNAPBACK}][/a]

Yes, the focus point changes with the diaphragm setting; this can be tabulated in a ROM and the computer driving the lens can move it a bit. Now, it wouldn't surprise me if our friends at Nikon and Canon are already doing this.

Edmund
Logged

Edmund Ronald, Ph.D. 
Graham Mitchell
Sr. Member
****
Offline Offline

Posts: 2282



WWW
« Reply #17 on: October 22, 2006, 06:28:13 PM »
ReplyReply

I don't buy this argument, but I'm happy to be proven wrong. The AF sensor is on a fixed plane which should be aligned perfectly with the back sensor. When the subject is in focus on the AF sensor it will be in focus on the DB sensor. Is Hasselblad somehow altering the laws of physics, or am I missing something?
Logged

Graham Mitchell - www.graham-mitchell.com
eronald
Sr. Member
****
Offline Offline

Posts: 3946



WWW
« Reply #18 on: October 22, 2006, 06:44:50 PM »
ReplyReply

Quote
I don't buy this argument, but I'm happy to be proven wrong. The AF sensor is on a fixed plane which should be aligned perfectly with the back sensor. When the subject is in focus on the AF sensor it will be in focus on the DB sensor. Is Hasselblad somehow altering the laws of physics, or am I missing something?
[a href=\"index.php?act=findpost&pid=81662\"][{POST_SNAPBACK}][/a]

May I repeat myself: When the diaphragm is stopped down the focus point shifts. It's the lens, happens with a sensor, film, even a ground glass. Has nothing to do with the AF sensor. Maybe someone here can point you at the right optics text or words to google, although it is usual to require  that the wizard  invoking the laws of physics research them himself

Edmund

Hint: A perfect lens would be a violation of the laws of engineering - all lenses have some design defects ...
« Last Edit: October 22, 2006, 07:46:57 PM by eronald » Logged

Edmund Ronald, Ph.D. 
Graham Mitchell
Sr. Member
****
Offline Offline

Posts: 2282



WWW
« Reply #19 on: October 23, 2006, 06:01:04 AM »
ReplyReply

ok, I missed the part about the diaphragm stopping down, so that is a potential difference between the AF's focal plane, and the sensor's focal plane. I'll look into this more. Plenty of MFDB users seem to be getting excellent results without this focal compensation. Perhaps it's present but insignificant? Marketing people like to sell snake oil. Just look at the world of high end audio!

edit: thinking more about this, clearly the effect won't affect wide open shooting, because the aperture setting doesn't change for the exposure. And when the camera is stopped down, the extended DOF would make the effect insignificant. I'm still googling to try and find out more. Is there a name for this effect?
« Last Edit: October 23, 2006, 06:33:43 AM by foto-z » Logged

Graham Mitchell - www.graham-mitchell.com
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad