Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: michele on October 15, 2008, 09:01:14 am
-
Hi folks,
I had the chance to use for a while (30 minutes) the Hasselblad H3DII-39 with the HCD 28mm. The camera was in a very heavy and stable tripod, mirror lock-up, cable release... ISO 50, f/11. The file was opened with Phocus, no sharpening just DAC corrections... I hope these sampels could help someone, there are the total and two crops, one of the very extreme left corner and one of the focus point...
Bye, Michele
-
If it pleases the court I will add my own.
Images shown are from an H3D-31 at 200 ISO, 1 second exposure. Taken by myself and colleague last year. HCD28 lens of course.
Two full views are shown with Lens Correction on and off. You might like to layer them in PhotoShop to see the difference.
Details from centre and the left hand side.
David
-
Hi David,
I'm asking to myself why your images look better then mine... I can see more details in yours, also in the corner yours are better, and you took the photos with the 31 megapixels back, and it supposes to have microlenses, and I guess it should have some bad artefacts in the corners with the 28mm like the P30 with the 28mm mamiya...
Maybe the 39 mpx back outresolves the resolution of the HCD 28mm? I didn't yet say wellcome to you in the forum...so wellcome
Michele
-
Hi David,
I'm asking to myself why your images look better then mine... I can see more details in yours, also in the corner yours are better, and you took the photos with the 31 megapixels back, and it supposes to have microlenses, and I guess it should have some bad artefacts in the corners with the 28mm like the P30 with the 28mm mamiya...
Maybe the 39 mpx back outresolves the resolution of the HCD 28mm? I didn't yet say wellcome to you in the forum...so wellcome
Michele
Hi Michele,
Well without seeing your raw file it is hard to say! But it looks like you might have clipped the histogram hence some of the posterisation that is happening. Like in the back of the chair for example. I would NOT expect to see this happening at all.
The 31MP has the same size pixels as the 39MP so they both resolve to the same extent. The 31 will however have better higher ISO performance.
You are welcome to send me the 3F file and I can take a look.
Thanks for the welcome.
David
-
......also in the corner yours are better, .......
Michele,
One of the contributing factors in the corner performance differences you notice is that since your reference image is taken with a
''more full frame' back than the sample from David you are using more of the image circle available with the 28mm.
Sometimes, when working at the limits of an image circle there is compromised sharpness in the extreme corners
Mark
-
Michele,
One of the contributing factors in the corner performance differences you notice is that since your reference image is taken with a
''more full frame' back than the sample from David you are using more of the image circle available with the 28mm.
Sometimes, when working at the limits of an image circle there is compromised sharpness in the extreme corners
Mark
Yes, Mark, but I have seen several Mamiya 28mm images posted on the web that were taken with the cropped sensor of the Phase P30 and the corner performance looked, to my eye, very soft. David's sample does not.
-
Yes, Mark, but I have seen several Mamiya 28mm images posted on the web that were taken with the cropped sensor of the Phase P30 and the corner performance looked, to my eye, very soft. David's sample does not.
Mamiya and Phase. What' you expect :>))
I was just responding to Michele to point out one generic reason why a cropped vs a fullish sensor might exhibit different corner sharpness characteristics.
Mark
-
Michele,
One of the contributing factors in the corner performance differences you notice is that since your reference image is taken with a
''more full frame' back than the sample from David you are using more of the image circle available with the 28mm.
Sometimes, when working at the limits of an image circle there is compromised sharpness in the extreme corners
Mark
To show that the HCD28 performance is just as excellent on a 49mm sensor I have found one of my shots from an H3D39. Same deal as before, with and without DAC, left hand side crop and a centre crop. Not the most enthralling of captures but it delivers the point!
This image was also shot with our GPS accessory. (See meta data image)
On a side note there was a small amount of Moire on the bicycle basket which was successfully removed with our new Moire filter, also increasing the detail (not blurring) as well.
HCD28 Lens, f8, 1/200s, 100 ISO.
David
-
Just a thought...
For those of you who would prefer some level of Vignetting, there is a a control for this in the lens corrections...
David
-
Thanks to All,
actually I know that the 31 megapixels is 33x44mm and the 39megapixels is 36x49 (about), but as hcubell says I saw the 28mm Mamiya with thw P30 producing strange behaviour in the corners... Just to say, go to captureintegration.com and you'll see some test...
Overall, i think the difference I can see between my files and the David's files it's first of all in the different subject, and secondly due to the fact that I'm not used to the Hasselblad system, I'm pretty sure that I can do more after have studied and tested the enitre system. But I guess it's useful to see what Hasselblad can do in a beginners hand. I think Hasselblad has a great potential, but I wish I could use the H3DII with a PhaseOne back... Hassy is a very great platform but, and it's just a feeling a mood, I don't like very much their backs...
-
Now, I confess I use Hasselblad backs (at this moment). Before that Leaf and even for a blue monday Phase. How could you state you don't like their (or anyone elses) backs without having used them for at least a while (thinking months not hours)? From the backs I have used sofar I can say you can get to work everything in the way you like if you take the effort to get to know the system and how to work it.
All 4 manufacturers have pretty similar products that are different in small, but sometimes not insignificant, areas. What makes you think you don't like Hasselblad backs? Only a mood or a feeling? I can understand feelings for a family member, a pet, maybe even a preferred camera system or way of working but isn't this a form of 'group induced sentiment'?
I also confess I had my doubts before using the Hasselblad system and I still have, I probably will always have it which keeps me very critical. But... it does work, it does deliver very good results, it has few hickups. It is one of those tools that work but generate little love which indeed can be confusing.
-
It is one of those tools that work but generate little love which indeed can be confusing.
Well, it's because it works that I have grown to love mine. Often, with expensive items that have a big reputation, it can be the other way round. Every time I zoom in to 100% and see what this system (H3d39II) is capable of when I get it right (not always!) I love it a little more. Michele, you do have work much harder with MFDBs - please don't see that as a patronising or superior statement. I've been at it more years than I care to admit but it took me quite a while to get the best from the system. And I agree with others above, the backs are very similar in performance.
-
Same here Stuart, exactly the same. Even the same way of growing in love maybe I should have phrased the sentence as; "It is one of those tools that work but generate little love at first which indeed can be confusing'.
Everytime you go over your images at a 100% it is a joy to see what the results are which are better than expected in most cases. If you are producing shitty images it most certainly was not because the tool you were using
I do have other tools that are less forgiving, harder to produce the same kind of results under some circumstances but still feel better. Maybe not entirely fair because it is like comparing apples & oranges. The thing I talk about (this is just my case) is my Digiflex. The ability to use excellent glass at f1.4 and use it handheld at 1/30th while still generating sharp images with a digital back. A body that is big but smaller than the H. These are all things my H isn't and probably never will be. Maybe the fact that I take the Digiflex to occassions where I am not working for someone else has got to do with it as well. Anyway still comparing apples & oranges. Maybe I just like apples a bit more but the H is definitely one of the finest oranges I ever tasted
-
Thanks to All,
actually I know that the 31 megapixels is 33x44mm and the 39megapixels is 36x49 (about), but as hcubell says I saw the 28mm Mamiya with thw P30 producing strange behaviour in the corners... Just to say, go to captureintegration.com and you'll see some test...
Overall, i think the difference I can see between my files and the David's files it's first of all in the different subject, and secondly due to the fact that I'm not used to the Hasselblad system, I'm pretty sure that I can do more after have studied and tested the enitre system. But I guess it's useful to see what Hasselblad can do in a beginners hand. I think Hasselblad has a great potential, but I wish I could use the H3DII with a PhaseOne back... Hassy is a very great platform but, and it's just a feeling a mood, I don't like very much their backs...
Hi
Here's a test shot I did with the 28mm Mamiya and a P30+ a couple of months ago at my local dealer. The fall off as soon as you leave the center was quite a disappointment to me. It would be interesting to see how good C1 pro is at correcting this I'm very happy with my 28mm hassy lens, which is 1k US less than the Mamiya one, and I see a big difference in quality. Should I need to correct anything, the lens corrections in Phocus is producing 100% magic results, no matter at which aperture and distance the shot is done!!
I own a P25 for my Contax as well and like both systems with each it's advantages and disadvantages. The P25 is easy controlled from the back by four visible buttons, the H3DII back however can be controlled directly from the camera, and has a larger and better display in my opinion.
As earlier said here at this forum, I'm interested in hearing how good the new C1 Pro is at doing lens corrections and also if the P65+ is available for my H2.
Cheers
David
-
The ability to use excellent glass at f1.4 and use it handheld at 1/30th while still generating sharp images with a digital back.
Now you're talking dirty! (But I love it.) It's often a bummer for me that I can't handhold the Hass (or any MFD) with better results unless I'm outside on a really bright day. But that's what the 35 DSLRs are for I guess.
-
Ha ha
-
To show that the HCD28 performance is just as excellent on a 49mm sensor I have found one of my shots from an H3D39. Same deal as before, with and without DAC, left hand side crop and a centre crop. Not the most enthralling of captures but it delivers the point!
This image was also shot with our GPS accessory. (See meta data image)
On a side note there was a small amount of Moire on the bicycle basket which was successfully removed with our new Moire filter, also increasing the detail (not blurring) as well.
HCD28 Lens, f8, 1/200s, 100 ISO.
David
Show me a shot with the H3DII-39 with the HCD 28mm, leveled up with vertical lines at the edges of the image. A true, informational image. Something like that would catch my attention.
-
Hi David,
when you use DAC, do you lose image "real estate"? As far as I know DxO delivers around 96 % of the original image and Canon DPP only 92 %. And if, does it depend from the lens?
Best,
Johannes
Hi Johannes
Well if it's 100% is hard to say but to me they for sure look 100% real estate, no matter which lens I use! I've read that Phocus uses a huge matrix of more than 50.000 numbers to correct for all possible distances and apertures and that even the HTS1.5 will be supported. Maybe David from Hasselblad can join in here and possibly post us some images from the HTS?
I still haven't been able to test C1 Pro. Someone here that could post us some images before/after or just tell about their experiences?
Cheers
David
-
To Hasselblad's credit, in David's bicycle example photo where mustache distortion is very evident and strong for the 28mm, DAC makes it a perfect lens. That's quite impressive.
Yes there are more than 50 000 lookup tables in Phocus.
Here's a quick DAC primer.
No lens is perfect. Every lens suffers from some degree of distortion, chromatic aberration, or vignetting. By shooting test charts at various apertures, focal lengths (think zoom), and combinations of extension tubes, you can mathematically describe these characteristics.
The CRUCIAL bit of data however is the subject to sensor plane distance, without this data you CANNOT make accurate corrections only guesstimates. This is why some software asks you to specify wether the shot was taken at near or infinity focus. As far as I'm aware the H system is the only system that passes subject to sensor distance back to the software.
The system is able to work with the HTS because the HTS has sensors that record the degree of tilt/shift/rotation and pass it back (you can see it in meta data), again this is the only tilft/shift system I know of that does this.
Nick-T
-
Show me a shot with the H3DII-39 with the HCD 28mm, leveled up with vertical lines at the edges of the image. A true, informational image. Something like that would catch my attention.
We have that of course.
Image with and without DAC (distortion only, then vignetting. Remember - Vignetting correction can be controlled in terms of density but is not simply a generic amount - it varies depending on the lens data) plus the same corrections 'attempted' in Photoshop. You can see it is in no way comparable. Plus timewise - all you do is tick a box.
This becomes very interesting with the HTS (Tilt and Shift) as of course if you apply some shift then it is no point having the Vignetting correction centralised for example - it will need to be offset. This is done by movement data being recorded by the HTS and sent to the image file.
[attachment=8983:NoDac.jpg]
NO corrections at all.
[attachment=8984:distortion.jpg]
Correction for Distortion only applied
[attachment=8985:distvign.jpg]
Distortion and Vignetting applied. Vignetting at 100%
[attachment=8986:pshop.jpg]
The same image with correction attempt in PhotoShop
David
(H3D39 at 0.35m)
-
Hi David,
when you use DAC, do you lose image "real estate"? As far as I know DxO delivers around 96 % of the original image and Canon DPP only 92 %. And if, does it depend from the lens?
Best,
Johannes
Yes you are correct in assuming it will depend on the lens. Distortion is non existent in the longer lenses so real estate is 100%.
Obviously the HCD28 requires the most amount of correction. I post two images from the HCD28 so you can layer them in PhotoShop and see the difference.
[attachment=8991:NoDAClo.jpg]
This image has no distortion correction.
[attachment=8992:DAClo.jpg]
This image has the distortion correction.
David
-
David,
In studying Elizabeth Carmels work on her web site, and I understand she is a H3D user, I notice she is consistently getting DOF that you would normally get with a 4x5 and tilt = from foreground to infinity appear in sharp focus.
Do you know what technique she uses to accomplish this?
Helicon Focus was my guess but I'd like to know.
I've just purchased the H3DII-39 Kit with the optional 28mm D-Lens. Moved from Phase P45+ on H2.
-
Hi Jack,
Congrats on your new purchase!
I haven't met Elizabeth so I am not sure. I have a feeling she could be a member on the Hasselblad Digital Forum, so you could ask her yourself!
www.hasselbladdigitalforum.com
But my guess would also be Helicon Focus (now part of CS4 I believe.)
David
-
Is that so? Did Abobe take over Helicon and has it integrated Helicon focus into CS4? I knew there was focus blending in CS4 which I find very interesting.
-
Having recently bought Helicon, and yesterday CS4, with my record on timing such matters, it wouldn't surprise me at all if Adobe have incorporated it!
-
We have that of course.
David
(H3D39 at 0.35m)
Thank you very much.
Mark
-
Is that so? Did Abobe take over Helicon and has it integrated Helicon focus into CS4? I knew there was focus blending in CS4 which I find very interesting.
I have Helicon Focus, but I just purchased an upgrade to CS4 in the hope that Adobe has done its own focus blending code and it is better than Helicon. When I made the switch from shooting 6x7 film to an H3D-39 two years ago, I was quite frustrated by how difficult it was achieve acceptable depth of field with the H3D-39 compared to my Pentax 67. In part, this was due to the fact that I was advised not to stop down below f/16 with the H3D. What I think was more significant is that the resolution of the H3D-39 with the HC lenses is so amazing that the elements that are in perfect focus in a frame make the elements that are not look more out of focus than I found in my 6x7 film. In short, the depth of field is just less as a practical matter. As a result, I have taken to bracketing for focus on a regular basis and a good software solution for blending is very important to me. Helicon is quite good, but not always perfect. It will be nice to be able to do it within PS with all of the PS tools at your disposal.
BTW, Elisabeth Carmel has a blog and you can ask here there whether she uses Helicon.
-
Yes you are correct in assuming it will depend on the lens. Distortion is non existent in the longer lenses so real estate is 100%.
Obviously the HCD28 requires the most amount of correction. I post two images from the HCD28 so you can layer them in PhotoShop and see the difference.
[attachment=8991:NoDAClo.jpg]
This image has no distortion correction.
[attachment=8992:DAClo.jpg]
This image has the distortion correction.
David
Hi
I managed to find a C1 4.5 and Mamiya 35mm lens correction test, posted by Guy Mancuso on forum.getdpi.com:
http://forum.getdpi.com/forum/showthread.p...3692&page=2 (http://forum.getdpi.com/forum/showthread.php?t=3692&page=2) (Go to post #40 to see the images)
Is it me or doesn't the corners look like 4.5 is adding USM as an attempt to improve resolution?
/David
-
Hi Michele, I am wondering if someone walked through the room during the exposure. In that first shot I see some blurry milky looking band right through the image from approx. that easel going right. Looks like flare or some kind of blur or ghosting. Any idea?
-
Hi Michele, I am wondering if someone walked through the room during the exposure. In that first shot I see some blurry milky looking band right through the image from approx. that easel going right. Looks like flare or some kind of blur or ghosting. Any idea?
Hi Lisa,
yes you are right, It was my dealer that was walking into the room... I didn't mention it because it doesn't effect the corners of the image, and I'm concerned about that not the center... Afeter many attempts, I got a very good TIFF from the RAW, and I'm sure that the HCD 28mm it's a very impressive lens
-
Yes there are more than 50 000 lookup tables in Phocus.
Here's a quick DAC primer.
No lens is perfect. Every lens suffers from some degree of distortion, chromatic aberration, or vignetting. By shooting test charts at various apertures, focal lengths (think zoom), and combinations of extension tubes, you can mathematically describe these characteristics.
The CRUCIAL bit of data however is the subject to sensor plane distance, without this data you CANNOT make accurate corrections only guesstimates. This is why some software asks you to specify wether the shot was taken at near or infinity focus. As far as I'm aware the H system is the only system that passes subject to sensor distance back to the software.
The system is able to work with the HTS because the HTS has sensors that record the degree of tilt/shift/rotation and pass it back (you can see it in meta data), again this is the only tilft/shift system I know of that does this.
Nick-T
Small correction!
DAC data tables are not constructed via shooting test charts. We use mathematical data from the optical design software to calculate the corrections. This is better and more accurate practice than shooting test charts. Which is how lens corrections in 'other' software would have to work as they do not have access to the optical design.
We are now close to 57,000 data tables with the latest being Ultra Focus data for the new 35-90 zoom lens and of course the HTS. VERY many tables!
Best,
David