Hi milotimbol,

i shoot always iso 100, large diaphragm(?), always eternity on focus with 8mm sigma..RAW

make tifs fullsize then stitch, render fullsize and after go to photoshop.

Cheers
Tuur

Hi,

in small spaces i still use that, untill now, for the 8mm i think the 'eternity starts at 13 cm..(?) i was thinking about bringing the laptop so i can see fullscreen about the sharpness...

do you have a nodal ninja?? or pano head thing??

f 5.6-8 - 11
not the smallest ones depends a little on the light but for the sharpness i suggest as high as is comfortable..

I'm not that fancy photographer aswell so i have to learn a lot too..



i have now this but i'm also unsure... any body??




in photoshop ...most of the time i spend like 30 minutes on a picture (sphere) retouche, high sharp overlay mask, saturation and stuff highlights/shadow.. etc.. make smaller to jpg 11 size aprox 3500/4500 pix width..

also just for now, anybody better settings for going non-multiresolution???... when you go bigger you can go multi res and have fun and bigfiles

Cheers

tuur

[quote='milotimbol',index.php?page=Thread&postID=4043#post4043]Thanks tuur! How about small spaces? You still shoot focus on infinity? When you say large diapraghm, do you mean a large f-stop? How large f-22? max?[/quote]

Milo F22 is a really small hole you would need to leave lens open longer. Smaller F = larger hole. Larger F smaller hole. i sent you a PM with my tricks.

Tuur in your code:

Fps can't be 120, or well maybe it can, but it will suck out performance. 30-40 fps is normal. 60 fps for high end. It handles the smoothness of the pano. If the number is too high, it will actually have negative effects for the panorama.

Details, depends. For landscapes it can be 16. But if you have precise architecture or infrastructure, 32 is nice.

tessmode="auto" I don't know if auto is a genuine value, the documentation states you should put a -1 for auto (0 for spheres and 3 for cube)

to gain some fps, you should set the movequality and movequality10 to LOW. Because when you're panning, you suck up alot of cpu usage, and you don't really need to say every detail in the landscape. If people want to see something, they'll stop and look anyways.

I did some testing the other day. I had to make a tour more smooth. These were my old settings:

Ok mind you, I have 3d rendering workstation (high cpu, high ram, low graphic card). I had a 22-28 fps when turning a panorama @ 1680x1050. When turning around my cpu usage went from 20% (not moving) to 65-70%!

I tweaked some and got this:

Now when turning around I get 38 fps and my cpu usuage goes from 20% to 38-40% (moving). Thats a 30% boost! And an untrained eye can't really see it. But they will notice the fps increase. Between 30-40 FPS is a good number (above is not really noticeable, under it, the stuttering annoys.)

I've turned off flash 10, because of my poor graphic card. Most people don't have that strong of a graphic card. Im waiting the day that krpano can do a quick fps check to determine which version to use.

Interesting fact:
http://www.adobe.com/products/player_cen…enetration.html

98.6% Flash 9 penetration
75.3% Flash 10
as of March 2009.

My fov is usually between30 and 90. Above distorts the image to much with architecture. Andddd I work with spherical photographs 6000x3000, converted to tiles (progressive load --> waiting for the loading) , 1 level. I try to aim the total files size not to be bigger then 5mb.

thanx!!! zephyr, gonna study on this..

Somebody got more photo tips??


Thanx

Tuur

I shoot on a homemade panohead, 18 shots at 10mm. ISO 100, F8.

I use autopano pro to stitch to an equirectangular and then photoshop to clean up the layers. I use the superrune photoshop plugin to edit the zenith and nadir if needed, but it doesnt work on my mac... How do you convert it back into a sphere after using the ktransform plugin to edit the nadir?

kcube2sphere.exe or

CUBE to SPHERE droplet.bat

appear to do the same action, converting cubes to a sphere .tiff file

However, from what I have read on this board, cubes display better if you are using flash 10. I assume these tools function the same way on a mac as a PC.??

to patch the zenith and nadir I found a great trick. take you equirectagular image at 2:1 and flip the image 180 upside down, then go to Filter -> distort -> polar cooridinates select square to polar and apply. it will apply polar cooridiantes to your image and the hole at the bottom ends up and exact circle and can be clone stamped in photo shop very easily now. it also format the iamge so if you have a nadir cap or logo you just drag in over the hole then undo the previoos steps. Filter -> distort -> polar cooridinates select polar to square and apply. rotate back 180 and your hole is patched neatly.

Great thread I hope klaus makes a section for this topic,I like this forum I hate it when I have to go and read about panophoto tips on other websites.

I am in the process of buying a nodal ninja 5 as I was told the nodal ninja 3 blocks the up and down shot on some camera becuase its short.

I am also looking to buy the sigma 4.5mm F2.8 EX DC Circular Fisheye HSM most people use 8mm F3.5 EX DG Circular Fisheyeo


I am looking for a 3 shot solution for inside yatchs and close spaces.


I own www.yatch360.com and www.yatchs360.com soon to be developed websites.


tim2420

Quoted from "milotimbol"

Tuur,

My gear is composed of Canon 450d, Sigma 8mm fish eye, nodal ninja 3.

Hi Tuur

I also own a Canon 450d,

I have been creating partial panos because I do not own a fish eye lens or a nodal ninja,its great to learn krpano and the more you play with it the more you understand how it works.

Hi,

i have canon 50 D ninja 3 and 8mm sigma also 10-22mm canon etc etc..

Tim: in small spaces i do just the same as always, no need for 4.5 mm i think.

Milo: duplicate layer, select, filter -other- highsharp- set aprox 2.6 (try to play with that a bit.) in the layer box you see "normal" somewhere in a scroll menu there make it "overlay" or "soft light"...

play with some settings. flatten and go on with color and stuff.. (i have cs but my new CS4 bundle is waiting on my desk to install as leopard is..) and then we have a new learning curve


Cheers
Tuur

Hi everybody,

pros and cons to diffenrent shooting and pano techniques:

For commercial use i ran through different workflows in the past, to find out:
1) what pano-system is flexible and precise enough (at shooting time) to be prepared on different shooting situations
2) what stitching software delivers fast (time is money) AND controllable results and (if needed) top stitching quality
3) what pano/tour-software to use for my projects (time fleets, products come and go, but i want to continue for years without switching ten times!)

I first have to tell you:
I have a small company doing graphic work - and therefore i already owened good photo equipment before i started to complete parts for pano photography, so for

1) equipment
a) camera/lens equipment: To be flexible means to have Highresolution during "standard" pano-shooting because you can sell sth. more than others if needed (HiRes- printouts) AFTER your work is done. The customer cant imagine, what this looks like until you showed him and if you then have to reshoot...
Optical precision is also needed to have such prints but is also needed for your tour via internet: I´ve seen many panos they suffer from low sharpness, from chroma issues, low contrast etc.

My standard setup: Canon EOS 1DS MkII + Canon 15mm fullframe fisheye
The 15mm is old in construction but delivers much better contrast and sharpness than 8mm Sigma. With the Sigma I did 3-4 pictures (overlap!) - with the 15mm Canon i do 7 (6 around + zenith). This is an acceptable raise of shootingnumbers - therefore (fullframefisheye) it delivers a 90MPixels pano - pretty much for "standard" - example (no nadir, no zenith), example2 (this is HDR shot...).
If needed i shoot an additional nadir pic (leafing the 1DS on my pano head) with my 2nd body, a 5D MkII - suitable lens attached... ... this body is always with me becaus of shooting (later clickable) detail pictures without deattaching and reattaching a camerabody to the panohead again and again.
I also tried: Canon EOS 1DS II + Canon 14mm 2.8 L superwideangle (no fisheye distortion!!) - more pictures would be nessecary and most stitchingsoftware has problems with this combination...
I dont know customers needing Gigapixel Panos one by one, so this is not my mai concern, but i did some steps to try: Using the 5D body (21,6MPix) plus 85mm 1.2 L USM (btw.: optically superb) for a landscape pano delivers (32!! pic for a single row) nearly 500MPix Pano for one row. Extraordinary static type scenery as precondition limits this application and you have a lot of data to access. btw: the number of pixels in your camera-chip is the MAIN limiting FACTOR on the way to a multigigapixelPano. At the moment the only way to achieve MPixel results is to put a telelens on your camera (this multiplies the number of pictures you have to take and so the number of pixel you get) but this comes with a very small depht of focus and so is limited for static landscape scenery. Today you can find panos shot inside a fridge or even inside a bottle of cola - but you wont find gigapixel types of it!

b) pano head: my first choice was a Manfrotto SPHxx? - very flexible multirow head, but not precise enough with heavy bodys like the 1DS; i still use it sometimes but (and i recommand so) with smaller bodies (5D is OK, EOS 50, 450, etc...). For standard i now use a "precision" head from 360precision . Pro: It is very precise and and quick - no "setup" needed, because it is ONLY for one combination of body/lens (no missalignment possible too!). contra: it is an expensive part; you have to fix up the combination for your order, no post-adaption is possible if you change the lens or body type - null flexibility; its no leightweight either; last but not least you´ve to pay via internet (UK company) before and be patient with them (it took 4 weeks and some reminding mails) - it seemed they dont like to answer your mails after you payed...

2) stitching software: Easypanos Panoweaver was good for use with the sigma 8mm, did its job automatically and quick - IF IT DID! If not there was no way to help and it was (may be not the actual version 6) limited on some preset body/lens-combination to work with. PTmac from kekus (seems to be discontinued in development) with panotools is very flexible but if you use it pure there is a lot work to do because usually you have to manually set all stitching points between the pics. It is very exact and works reproducable but - although there is a batchversion aviable - it is a slow workflow. I also tested PTGui and some other knowen stitching software but now I am very glad to use Autopano pro and giga. It is nearly as flexible as panotools, accurate, quick and its usability value (perfect GUI) is high AND it has a batch cue!

3) pano/tour software: Easypanos Tourweaver is a sophisticated tool, at the moment it is - i think so - the only software that puts a complete GUI-driven assembling tool on your hands. It is good if you can live with its schematas because it quickly delivers a complete tour with mainly and detail features a tour needs, BUT: It is expensive; it has major restrictions in where to put which information, in sizeing the components and many other details because it follows a concept you cannot change; it was VERY errornous (very often it crashed and all work was gone and even destroyed sometimes) and - belief me - you suffer if you only want to change text in a textfield after the project is published and uploaded.

There is no software out there with a development history like the "big" ones (Photoshop, etc.), so: seen (bought and used) fpp , fpp with flashificator , panosalado , cubic connector and some other packages - if you want to give your creativity a chance, want to control the quality, want to deliver individual tour projects - take krpano, learn to use its scripting technique and help youself with actionscript is the best choice now!

KRPANO is GREAT!
btw: i think krpano is low in price (even if you additionaly buy the google-plugin) - we all should buy the brandfree-version for to ensure Klaus will earn enough money to continue developing (this is of major interest!!)

Great thread with excellent info :-)

I'm using a canon 1000D with a sigma 8mm fisheye and panosaurus head. It took a while to get the head perfectly aligned so now the 1000D is *never* unscrewed from the pano head's base plate. With the added Canon battery grip this makes for a pretty big chunk of gear.

I tried many different shooting patterns based on many forum searches across the wonderful interweb. I used 8 around; six around and one up; four around 10-degrees down and four 10 degrees up. Now I just use four around 5-degrees up. I get coverage of the zenith (sometimes just a pinhole to patch), and just a little patch on the nadir which I need to fix because of the tripod anyway. With HDR that makes 12 shots. With my old standard lens I was shooting 140+ shots per pano !

I stitch in PTgui which now usually gives me "very good" result without any manual point alignments. I always shoot in HDR and find that PTgui does a good job with true HDR. IN fact I find it is much faster than going to Photomatix. Sometimes I can mess with shots in there for ages and still not have something I like.

I too fell for the old "forgot to take the ring off" trick on the Sigma ! Luckily I was shotting 4 down and 4 up when I did that, so still got cover.

As for focus, I started on auto, went to manual, and am now trying auto again. I'm a bit confused there too, but auto gave me excellent results last time so maybe I'll just stay with that...

Usually after outputing the main panorama (5000px wide for the web) I'll take the image into Lightroom and tweak it up a bit. But it can be dangerous fiddling too much.The images are already HDR and so can begin to look overcooked. When that's done I convert the pano to cube faces so I can easily patch the nadir. I don't like photoshop too much, and stopped upgrading my copy long ago, so now use Fireworks which I know extremely well.

I then drop the six faces into a multires which creates the images folder, an html page and xml file. All that I then drop into my standard KRpano "starter folder". It has all the elements and a good xml base, but of course there are hotspots and viewpoints to figure out and lots of litte things that always seem to need adjusting.

Of all the stages KRpano is by FAR the most time-consuming. I wish there was a faster way to do all the hotspots, viewpoints and linkages. I find it very tedious and dull. But I did check out all the main competing software and KRpano just kept winning in all the categoried that were important to me. And it has already added new features I very much appreciate, so I'm very happy.

The main things that still bother me:
- Still undecided about focus settings on the Sigma
- Still seeing "stuttering" on a range of computers when panning around my shots
- It is very slow work to customise controls, add tour maps etc in kr pano

I always export 5000px for the web.

Example:
http://www.multimediaaction.com/citypark.html

How much RAM do you have ? What are the PC specs ?