What is your take?

[Georg R. Baumann]

Wow....Bloody Hell! This shows exactly the quality that I would like to achieve and print. <slurp!>

http://www.maxlyons.net/maroonbells.htm

http://www.tawbaware.com/maxlyons/cgi-bin/image.pl?showFileName=ST_PICT4196_4201_Tree_Plane.jpg&gallery=8

http://www.tawbaware.com/maxlyons/cgi-bin/image.pl?showFileName=ST_IMG_2391_IMG_2401_Badlands_Startrail.jpg&gallery=8

Now take that first one, with 230 tiles, I think I might be right to consider my future Mac to be equipped with at least 8 GB Ram, I even think of 16 GB.

 

[Bart_van_der_Wolf]

Pheww.... now there is more to consider of course, the optics.... so what you guy sare saying is that for stitching a fixed focal prime is more advisable, which means I have to find out what optics are most advisable for that purpose on a Nikon, the latter seems to become more and more my camera of choice at the moment.

Because the field of view is essentially controled by the number of tiles (and the amount of overlap), you can concentrate on getting high(er) quality optics than most wide-angles can possibly offer. Normal to medium-tele focal lengths offer very well corrected optics for decent prices. The preferred FL is more or less dictated by the output size requirements, and to a lesser extend by the number of tiles needed. If HDR sets need to be assembled, one might lean towards fewer tiles to reduce postprocesing time and storage volume. I find that for my outdoor work I use my TS-E 45mm most, and the TS-E 90mm when I can't get close enough or need extreme resolution. Only for very confined spaces, interiors, I use a short FL to reduce the number of tiles.

What is the total weight of your UPOPP <smile> Did you order it in the states, I would have to find out import duty as well.

I started out by replacing my tripod head with the BH-55 without clamp. The 'UPOPP' comes with a PCL-1 clamp that can be mounted on the BH-55 (I specified that I wanted to mount the PCL-1 to the BH-55 to the RRS people, and they added the screw and tools needed).
So I almost always have my BH-55+PCL combination (1.042 Kg) on my tripod to allow mounting the camera, which requires a Camera specific plate or L-plate. I use an L-plate to allow portrait or landscape orientation without compromising balance, because the centre of gravity remains centered. The remaining parts of the UPOPP only add 0.856 Kg., and come in their own Neoprene pouch.

RRS only sell direct, through their website. For some countries there is a limitation to the insured amount when shipping, so check their website for specifics. Import duty and VAT will be added by customs because of importing into Europe.

Heck, you guys got me hooked. <smile> I already plan to do my first multirow stitch when I have all that gear and chose a location, the highest seacliffs in europe.

It's good to have such a goal, but you may want to choose something more modest to start with. You'll need to develop some feeling for the amount of overlap required for the different subjects, and play with some different projections for different subjects. You'll develop exposure tactics for variable lighting, and subject movement, and try out the latest developments in (HDR or tonemapped) stitching software.

Now there is a challenge, I need to find a way to deal with the water movements, eventual clouds, and on top, I want to shoot this as a HDR with 5x exposure, which brings me to a point. Did you shoot HDR multirows as well?

Water and cloud movement is usually not a problem with modern blending algorithms. Tall grass in the foreground moving in the wind may require a bit more manual intervention in the otherwise mostly automatic stitching process. You also have choices to make with regards to short exposure times (where a tilt and shift lens can help to use wider apertures), or long exposure times which will blur difficult detail due to motion. HDR multirows are not much different from single rows, but the issue of realistic HDR tonemapping remains.

HDR plus tonemapping is going through lots of development as well, and may pose separate challenges, so try to master stitching first, then tackle HDR. Again, here it pays to having done the homework first, which output size is needed, and how many/few tiles do you want to generate to minimize potential issues later on.

I just read about some new software developments of stitching software and blenders, so even better postprocesing tools will become available as time goes by. Also Photomatix is being updated, with improved features, and Autopano Pro will support much improved HDR control in a newer version.

These are exciting developments, but we still need to take proper images/sequences first and good shooting gear helps.

Bart

 

[Bart_van_der_Wolf]

Wow....Bloody Hell! This shows exactly the quality that I would like to achieve and print. <slurp!>

http://www.maxlyons.net/maroonbells.htm

Now the challenge shifts to finding a printer that can handle 154 x 104 inches @ 300ppi output ...

Bart

 

[Georg R. Baumann]

Thanks Bart, I gues that RRS is a no brainer if one wants to be light on foot.

I seriously think about leasing a monster such as the 64" 11880 from Epson.

This here was my very first humble attempt on the subject last october. I used a strong ND grad, 3x exposure bracketing 0.5s, 1s, 2s per shot, all together 27 shots = 81 pictures -> ACR -> Photomatix -> stitch -> final processing.

I really enjoyed that. I took a while to observe the rhythm of the waves and tried to shoot in sync as best as possible, kinda counting the seconds inbetween shots. The post processing took me the better part of 2 hours. As for tall gras moving in thge foreground, I even like that effect to a degree.

 

[Michael Fontana]

HDR plus tonemapping is going through lots of development as well, and may pose separate challenges, so try to master stitching first, then tackle HDR.
Bart

Another methode - a bit more time consuming, but a solution for very difficult light situations is to stitch all the bracket shots first with the help of a template, and blend these in PS, later. That methode saved my life, when the autoblending-routines of Photomatix and friends produced a unpleasant stitch.

One of the problems with pano's with big angles: the contrast can be sometimes much bigger due to the larger image angle than a singleshot would have.

If someone can chose the good moments only, its not a problem; but I have sometimes to shoot in lousy light conditions and still get good results.

Therefore different methods are quite usefull.

OT: Georg, I didn't knew you speak russian - Octobersky °grin°

 

[Michael Fontana]

Michael, the pano with the fence and the log cabin really is intruiging, a real beauty, what printsize did you choose from the Lab? My first thought when I saw it was that I would revisit that spot over and over again until I get a rainbow in the very distant of that landscape.

On the Skyscraper I see artefacts on teh edges of the building, I guess this is due to the small jpeg and the original does not show such, would that be right?

I suppose there are ways to adapt lenses like the Zeiss distagon onto the Nikon, without loss of any quality I hope.

Bear, the °fence° will be printed something like 150/80 cm....
from here, its about a 4 h drive; plus a 2 hours walk.... so the rainbow might be gone, until I'm at the place ;-)

See, panos look bad on a screen, unless it's a zoomify or QTVR. One has to look at 100% to see all these details - and the entire image at the same time. THat's part of the stitching game, too.

Yes, the tower has in 100% no jaggies; it's the inner strcture, behind two glasses what looks like artefacts: here's just a screenshot as webjpg:

You can't adapt a Zeiss to a Nikon, due to a shorter flange-distance (??) of the Nikonmount/system. But a decent Nkon lens will be fine as well.

 

[Georg R. Baumann]

Another methode - a bit more time consuming, but a solution for very difficult light situations is to stitch all the bracket shots first with the help of a template, and blend these in PS, later. That methode saved my life, when the autoblending-routines of Photomatix and friends produced a unpleasant stitch.

One of the problems with pano's with big angles: the contrast can be sometimes much bigger due to the larger image angle than a singleshot would have.

What template are you referring to?

Do you mean that you process each of the exposure brackets say 10 pictures with each 3 exposure, eb1_1 to eb1_10, eb2_1 to eb2_10, eb3_1 to eb3_10.

Then you stitch eb1-10, then stitch eb2-10, then eb3-10 and blend the three stitches?

OT: Georg, I didn't knew you speak russian - Octobersky °grin°

My ancestors breed borzijs in Russia *grins*

I read up a lot in tha past 2 nights on a variety of technology available from MF to LF, from hasselblad to Phase one P45, the latter would be something I REALLY WANT TO TRY! <slurp>

However at 400Euro plus VAT rent per day for the P45 alone, I mean.... ok scratch that.

It appears to me that in terms of "clean files" the MFDB has a huge advantage, a P45 generates a 112 MB 8bit TIFF and this has to count for something.

But if we would take two 16 MP and could compare them, say the 1DsMKIII and this one for example:

http://www.robertwhite.co.uk/product.asp?P_ID=540&PT_ID=392&P=Hasselblad-503CWD

Would I be right to assume that the latter gives the much better results? I do not know what size the Canon Sensor is, but I would assume the Canon packs more pixels on a smaller sensor.

When I look at the quality of glamour shots Frank Doorhof exhibits, I wonder whether such systems would not make sense for Landscapes as well, or Mamiya for example.

When I look at Alain Briots work, I also wonder, why does he not use the largest DB's available instead of shooting film, money ain't the issue for him. Then there are others like Charles Cramer and so many others, who also do shoot with DB's.

So in terms of "cleanest file possible" and higher contrast range I conclude that the DB's give me a much better quality for print output than stitching from a smaller sensor. Coming to think, I wonder what camera/cameras Max Lyons uses for his stitches.

May be leasing a MFDB is a better option than buying into a Nikon and Optics....sigh.... I need a coffee now....

P.S.:

Interesting comparison and comments by Frank Doorhof and Rainer Viertelboeck, P21, P30 and P45 vs. 1DsMKIII:

http://luminous-landscape.com/forum/index.php?showtopic=21794

P.P.S.

Mhhh, multirow stitching with that 16 MP Hassy and 40mm Zeiss f/4 CFE IF... I guess that would be ....wow!

Can you use T/S lenses on a Hassy?

Can you use a Hassy on a Pano head?

 

[Georg R. Baumann]

Exercise

Friends,

I spoke with Robert White again todayt and I expect a formal quotation on a basic system that consists of a arca swiss mono rail and digital back with schneider lenses.

Once I have the offer, I will post it here and ask you kind gentlemen what your thoughts are on this system and whether I can spare cash in choosing different but equally functional gear for such a basic setup.

I have to say, this is an exercise for me to look into the initial capital investment shooting digital with a view camera. I asked to include a training day to get to grips with the basic functionality.

My basic idea is to invest rather into such a system than into a 64" printer. Then may be get a decent 17" or 24" inch printer for my own printing and have larger formats printed externally.

It will be very interesting to compare what the specific system would cost me in the USA.

Again, thanks so much for your patience and your valuable advise!

 

[Georg R. Baumann]

here we go:

19.055 Sterling +VAT

I included the CF-39 for I find the price difference to the CF22 ridiculous and look forward to learn what you think about it.

Dear Georg

Thank you for the email. Sadly I am currently unable to offer a 31MP digital back option for you. Please find prices below. Please note that all prices quoted exclude VAT and shipping.

Arca Swiss 6x9cm F Metric compact with Micrometric Orbix £2825.00 plus VAT
Arca Swiss 6x9cm to Hasselblad V adapter plate £254.00 plus VAT
Hasselblad CF-22 digital back £13,000.00 plus VAT
Hasselblad CF-39 digital back £14,495.00 plus VAT
Hasselblad V adapter plate (fits back to Arca) £625.00 plus VAT
Schneider 60mm f4 Apo Digitar lens in copal shutter £506.00 plus VAT
Arca Swiss 110mm lens panel copal 0 £45.00 plus VAT
Sekonic L758D light meter £305.00 plus VAT

 

[Georg R. Baumann]

Well, I asked them to give me more details on the deal.

He mentioned that all gear comes with 1 year warranty only. Astonishingly, there is no option to extend the warranty for the Hasselblad DB to three years, something which I am not comfortable about at all.

Additional costs of hefty 300 Sterling plus VAT plus expenses per training day have to be considered.

At this point in time, I can not see myself dishing out the equivalent of a Honda CRV to get a decent start in LF photography. Let aside additional considerations such as learning curve and a total different approach to getting the right shot.

My gut feeling tells me to stick with plan "A", upgrade my DSLR gear and stitch, which is a compromise on the final quality that I would like to achieve of course, and the artist in me doesn't like compromises.

Then again, I have to test the market first.

 

[Michael Fontana]

Hi Bear

a 60 mm digitar is far away from beeing wide enough with a MFback - for landscapying.

Templates:
When stitching you can save a template of the stiching process, and apply it to other files, taken from the exactly same position.

So that other methode, not using Photomatix but blending - in PS - 3 different stitches (bracket shots) together, works like that:

Making the bracket shots, lets say: + 1.5 f/0/ - 1.5f - for all the panoheads position's: shot A, B, C, D, E.We got 15 shots.

Then, instead of using Pmatix to blend the bracket shots into 5 files, we first stitch the 3 panos, with 1.5 f difference, each: a bright, a °normal and a dark stitch. Usually, I'd do the °normal° stitch first, save the template, and apply it to the other 2 stitches, to make sure, the stitches will be aligned perfectly.

Now, you blend it in PS, with layermasks.
I don't say this being the fastetst methode, but it can save your live!

 

[Michael Fontana]

Stitching is cool!

A actual, real world example:

a architect build a little library, surrounded by other buildings in the old town; extremly difficult to get a good shooting position, for the outside shots.

Finally, I spotted a toilet, on the neightbour house, with a window, at 2 meters distance from the libraray's facade, with > 7 metres of height! :

Okay, the day shots were unusable, but it gave me a starting point, even I realised, that, due to the window's bars, etc, it wasn't possible to mount the panohead.
I went back in the evening and redid 3 shots with the 17 mm and a homemade tripod:

 

[Michael Fontana]

Well, I knew, that the alignement of the lens - the NNP or entrance pupil - was worse than bad; shift, pitch, yaw und roll, etc were all different in the 3 shots.

But due to the small window, the windows bar's, its position at 2.5 meters higher than the floor, and bloody hand dryer conflicting with the ladder, etc... there wasn't any choices.

I bite the apple, and tried two stitches; here's the better attemp:

Okay, it needs some retouching, but isn't it great?
Yes, the perspective is a bit overdoing, but still!

 

[Georg R. Baumann]

Yeah... impressive!

 

[Asher Kelman]

Okay, it needs some retouching, but isn't it great?
Yes, the perspective is a bit overdoing, but still!

This Michael, carries the essence of good photography, that is bulldog drive to get the position right. Your inventive opportunism clearly pays off. Impressive!

Asher

 

[Georg R. Baumann]

How many cameras do you have Asher?

.... 25?

<grins>

 

[Georg R. Baumann]

<slurp> )

http://www.sinar.ch/site/index__gast-e-1776-50-1897.html

....my precioussssss.... hehehe

 

[BradleyGibson]

I also think the colour palette in this image is gorgeous. Your decision to be patient and come back later really paid off--well done!

-Brad

 

[Asher Kelman]

Well Georg,

What are your thoughts now for large landscape?

Have you considered mapping out what you might want with your digital Olympus on a few visits and then returning with a 4x5 film camera to take the real image and then scan it?

Asher

 

[Georg R. Baumann]

Hi Asher,

I did not have much time recently to go further on that part, however, the discussion s here helped me a lot to gain some more understanding.

My thoughts are, although the entire market for manufacturers in the digital realm of photography is miniscule when we think of the possible impact they can make on the semiconduictor industry compared to the rest, thinking about digital backs here, I am somewhat convinced that the development in this area will be en par with the resolution of film pretty soon.

For me personally, yes, at some point in time, after the startup buzz has eased, and out of curiousity, I will treat myself with a largeformat view camera and probably also have one build by Hugo Zhang.

However, the economics, as far as my understanding goes is a different cattle of fish. I did not safe the exact details, but it was like a 6x8 processed and scanned at around 100 euro a pop for a professional service somewhere in England. Considering the capital costs of a system with a 33MP digital back, well, you get there pretty fast when you shoot with a viewcamera.

My short to midterm term investement plan has a Sinar Hy6 system, which gives me the best of both worlds, on priority one.

 

[Asher Kelman]

Forget getting scans of film like that. The Epson 750 will give you perfect scans and if you are careful and scan for highlights and then for shadows and combine this will give you a file that will simply amaze you. Even a used epson for $200 or a new one for $100 would still be mind blowing. Don't get raz-mataqzzed by theory. Just try as I have said and you will be amazed.

Camera $250 used

Epson Scanner $500 approx

That's it!

Asher

 

[Ben Rubinstein]

I have the Camera Fusion adaptor for LF which allows flat stitching to silly amounts of megapixels (about double what the lenses actually give but that's another story). When I finally get round to shooting some worthwhile stuff with it rather than just learning how to use it I'll post a full writeup/blog here.

As far as I know these adaptors are not made for WA lenses, I can use a 135mm (44mm FF equivelent) with a 1:2 ratio but nothing wider is a possibility outside the studio. As such your landscape options are limited.

To be honest I think stitching is the way to go. Although I find the distortion from pano stitching to be unliveable with for close up architectural work*, for landscape it's pretty much perfect, you can get stupid amounts of resolution and thereby huge print sizes, you get to use a nice sharp and distortion free normal lens (50mm or so) which resolves far far more in real terms per tile than any LF equivelent and it's far faster and easier to do in the field. You can have made a full multi layer stitch before I've even begun to start making tiles! You get AF too which is nice.

Keep in mind that you don't need to shoot pano's to stitch, the ease of the shooting means that you can literally heavily overshoot tiles and then when they are joined, crop to whatever ratio or perfected composition that you require. It's very hard to be exact with stitching in your composition, you usually overshoot along the sides and top and bottom, but hey with digital why ever not? Remember that the final print is an expression of your emotions when viewing the scene, you don't have to get it perfect in camera composition wise for the above to be true.

*See the link on Projections that Bart brought on page 3:

"First - a word of warning: If you are looking for a projection, that will map a spherical (even partial) panorama on a flat surface without bending lines: This won't work!"