Luminous Landscape Forum
Equipment & Techniques => Digital Cameras & Shooting Techniques => Topic started by: Kenneth Sky on September 18, 2006, 02:52:32 pm
-
What irony. A day after Alain Briot's article and we see the announcement of a 160 megapixel camera that would make any landscape photographer salivate. Although I suspect the price tag may make them gag.
-
Holy crap! Put some wheels on that and Shaun White can use it at the X games.
-
I'd love to borrow that thing for a week.
-
For sale: 2 parents, good condition, 2 former keepers. UK based but can be flown worldwide at your cost. Offers in the region of 35 000 euros to allow for lens and tripod.
-
Wow, a 21 250 x 7500 file. If they could use this scanning sensor in a 6x7 back for our Mamiya RZ67, it would give a 9640 x 7500 file and 72MP. And if it can scan 17cm in 0.5 seconds, it should be able to scan 7.2cm in ~0.2s (I don't know whether this follows )
...And if it costs 28000 euro for 17cm, it should only be...aw forget it.
Just thinking out loud
-
Now that I think about it, this could be a truly incredible all-in-one camera. It could be used 6x17 (1:3 & 161 MP), but of course could be cropped to 6x12 (1:2 & 112MP), 6x9 (2:3 & 84MP), 6x8 (3:4 & 75MP), 6x7 (4:5 & 72MP) and 6x6 (square & 56MP) to give all the standard formats.
I wonder about noise. Its iso range is 500-10 000. If image quality is as good as I fantasise about, thist could be IT!
Now where did I leave that spare 35K...
-
Now that I think about it, this could be a truly incredible all-in-one camera. It could be used 6x17 (1:3 & 161 MP), but of course could be cropped to 6x12 (1:2 & 112MP), 6x9 (2:3 & 84MP), 6x8 (3:4 & 75MP), 6x7 (4:5 & 72MP) and 6x6 (square & 56MP) to give all the standard formats.
I wonder about noise. Its iso range is 500-10 000. If image quality is as good as I fantasise about, thist could be IT!
Now where did I leave that spare 35K...
[a href=\"index.php?act=findpost&pid=76875\"][{POST_SNAPBACK}][/a]
OH MY GOD. That is so cool!!!!
Reading the blurb, it sounds like they have a real winner, bring on the samples and reviews. This sounds like what I really need, not just the huge files, but the whole cassette idea so it can go on a monorail for studio/architectural as well, may even be suitable for talent utilizing the smaller formats. Yay.
And it uses real lenses!
Did I say Yay already?
-
TDI? 300MB per second? Makes Better Light look positively neanderthal. Maybe practical largeformat digital for field use is gonna happen after all. Now if they could just make an insert/back that uses a MacBook for exclusive control, power and storage ... and work on the price:
http://www.roundshot.ch/ (http://www.roundshot.ch/)
-
TDI? 300MB per second? Makes Better Light look positively neanderthal. Maybe practical largeformat digital for field use is gonna happen after all. Now if they could just make an insert/back that uses a MacBook for exclusive control, power and storage ... and work on the price:
http://www.roundshot.ch/ (http://www.roundshot.ch/)
[a href=\"index.php?act=findpost&pid=76880\"][{POST_SNAPBACK}][/a]
Can't see why (yet) they couldn't subsititute any new Intel mac laptop for the mini in the package as listed. May need a separate battery/Power supply for the unit, but I'm with you on (p)review on the laptop screen with zoom and camera control. No tilt/shift but if it can take "regular" LF lenses, perhaps a Zoerk type adapter would emerge. Need more details , already started saving
-
Did someone else already make the obligatory print/share button joke?
-
At 300dpi that works out to be a 25"x71" print right out of the camera. I would expect it to uprez nicely (poor CPU) to at least double that. That might just be the digital camera for Clyde Butcher. Be sad to see a gator make off with a camera like that.
-
If the camera needs 0.5s to take a picture, how was the photo with the surfers at the Seitz-homepage done?
-
If the camera needs 0.5s to take a picture, how was the photo with the surfers at the Seitz-homepage done?
[a href=\"index.php?act=findpost&pid=76895\"][{POST_SNAPBACK}][/a]
It's 1/2 sec to complete the exposure scan, not a 1/2 sec exposure. The image at one side of the scene will be 1/2 sec delayed compared to the other side but this is not visible like it would be with other scanning backs where it takes many seconds to complete the scan.
The actual exposure would have been more like 1000th sec for each scanning line.
-
It's 1/2 sec to complete the exposure scan, not a 1/2 sec exposure. The image at one side of the scene will be 1/2 sec delayed compared to the other side but this is not visible like it would be with other scanning backs where it takes many seconds to complete the scan.
The actual exposure would have been more like 1000th sec for each scanning line.
[a href=\"index.php?act=findpost&pid=76897\"][{POST_SNAPBACK}][/a]
Yes, but assuming that each of the surfers takes up about 1/20th the width of the scene, it still took about 1/100 sec to scan that area of the scene, right?
Normally, you would need shutter speeds around 1/500th to get them real sharp.
I would be very interested in looking at a 100% version of the image, I bet that it is full of scanning artifacts.
Regards,
Bernard
-
Yes, but assuming that each of the surfers takes up about 1/20th the width of the scene, it still took about 1/100 sec to scan that area of the scene, right?
Normally, you would need shutter speeds around 1/500th to get them real sharp.
I would be very interested in looking at a 100% version of the image, I bet that it is full of scanning artifacts.
Regards,
Bernard
[a href=\"index.php?act=findpost&pid=76901\"][{POST_SNAPBACK}][/a]
True, but you have to remember that it's a line of photosites exposing in sequence so the exposure time of each individual 'strip' is actually very short. It's the 'stretching' of the moving subject that you have to watch for - remember that classic shot of the old car with the oval wheels? That was taken using an early focal plane shutter moving in a vertical direction. Thus the top of the car had moved relative to the bottom by the time the shutter exposed that part.
With scanning backs it is less the 'shutter' speed than the scanning speed that makes an image look sharp - within reason of course.
So if you are correct and it took 1/100th sec to scan over the surfer then the amount he would have moved in that time is small and the distortion is invisible (hopefully). I would expect to see a sharp image, but maybe a tiny bit of distortion.
-
So if you are correct and it took 1/100th sec to scan over the surfer then the amount he would have moved in that time is small and the distortion is invisible (hopefully). I would expect to see a sharp image, but maybe a tiny bit of distortion.
[a href=\"index.php?act=findpost&pid=76905\"][{POST_SNAPBACK}][/a]
Nick,
Wouldn't you be overlooking one phenomena?
The problem with those scanning devices is that each component of the RGB coordinates is exposed at different times for a given point in space. Meaning that it is enough for the surfer to move a tiny bit for the pixels at the edge of its body to exhibit RGB artifacts.
Indeed, let's assume that the red channel is captured first, and that we are looking at a pixel that is still inside the surfer's body at time T=0, the surfer moves enough that that same pixel is now over water at time T=T1 during which the Green channel is exposed for that pixel (the CCD moved one step) -> we have a color that is completely meaningless, since the red and green channels effectively relate to different objects due to the movement in the mean time.
All the moving objects are bound to have artifacts at their edges. This is true for the surfer, but also basically for all the waves in the scene...
Regards,
Bernard
-
Nick,
Wouldn't you be overlooking one phenomena?
The problem with those scanning devices is that each component of the RGB coordinates is exposed at different times for a given point in space. Meaning that it is enough for the surfer to move a tiny bit for the pixels at the edge of its body to exhibit RGB artifacts.
Indeed, let's assume that the red channel is captured first, and that we are looking at a pixel that is still inside the surfer's body at time T=0, the surfer moves enough that that same pixel is now over water at time T=T1 during which the Green channel is exposed for that pixel (the CCD moved one step) -> we have a color that is completely meaningless, since the red and green channels effectively relate to different objects due to the movement in the mean time.
All the moving objects are bound to have artifacts at their edges. This is true for the surfer, but also basically for all the waves in the scene...
Regards,
Bernard
[a href=\"index.php?act=findpost&pid=76906\"][{POST_SNAPBACK}][/a]
Yeah, good point, assuming that's how this particular scanning back works - maybe it's an RGB capture like the Foveon. I guess we will have to wait for samples!
Cheers
-
Yeah, good point, assuming that's how this particular scanning back works - maybe it's an RGB capture like the Foveon. I guess we will have to wait for samples!
Cheers
[a href=\"index.php?act=findpost&pid=76907\"][{POST_SNAPBACK}][/a]
Well, if the Dalsa sensor in the Zeitz has the ability to capture all three channels at once (a la foveon as you say), then I would seriously consider buying this device!
But what I described above is how current Betterlight work for instance.
Let's wait indeed.
Regards,
Bernard
-
Indeed, let's assume that the red channel is captured first, and that we are looking at a pixel that is still inside the surfer's body at time T=0, the surfer moves enough that that same pixel is now over water at time T=T1 during which the Green channel is exposed for that pixel (the CCD moved one step) -> we have a color that is completely meaningless, since the red and green channels effectively relate to different objects due to the movement in the mean time.
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=76906\")
It appears the data is aggregated over time:
[a href=\"http://vfm.dalsa.com/support/primers/HSLineScanPrimer.asp]http://vfm.dalsa.com/support/primers/HSLineScanPrimer.asp[/url]
I'm not sure if this explains why it wouldn't be a problem, but Seitz seem to think you can handhold this thing.
-
A few points, forget any kind of studio flash lighting, it's hot lights only and with the kind of lighting needed for LF style photography you are going to melt the model before you manage to finish the frame!
Computers. What kind of monster would you need to process those files, just 5 layers would grind down almost every desktop available at this point in time!
How does the price compare to a phase 1 P45? it doesn't seem that far from it or am I mistaken?
This chip in a rather more conventional medium format setup and you will have landscape photographers wetting themselves!
-
A few points, forget any kind of studio flash lighting, it's hot lights only and with the kind of lighting needed for LF style photography you are going to melt the model before you manage to finish the frame!
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=76927\")
HMI lights? Admitedely very expensive.
[a href=\"http://www.bhphotovideo.com/bnh/controller/home?O=NavBar&A=search&Q=&ci=2248]http://www.bhphotovideo.com/bnh/controller...arch&Q=&ci=2248[/url]
Computers. What kind of monster would you need to process those files, just 5 layers would grind down almost every desktop available at this point in time!
[a href=\"index.php?act=findpost&pid=76927\"][{POST_SNAPBACK}][/a]
A Mac Pro WS 8 or 16 GB Ram should be more than enough. 1GB per file is only twice more than what you get with scanned 4*5 in 16 bits and I have been dealing with those on a 2 GB Win XP without major problems.
Crucial now sells Mac Pro workstation compatible at 1150 US$ for 2*2 GB. More than RAM, storage will quickly become the problem.
A 3-4 TB Raid 5 IEEE or USB2 unit seems mandatory. Sansdigital gets my vote so far.
How does the price compare to a phase 1 P45? it doesn't seem that far from it or am I mistaken?
[a href=\"index.php?act=findpost&pid=76927\"][{POST_SNAPBACK}][/a]
It is actually a good 15% more expensive that a P45 considering the rate of the Euro vs US$. If it delivers, it is however in a different category all together.
Cheers,
Bernard
-
Digital goes nuts. I like the fact its not just another boring me too product, but something genuinly different. Imagine ther file size shot vertically then rotated. Awesome.
Quentin
-
Digital goes nuts. I like the fact its not just another boring me too product, but something genuinly different. Imagine ther file size shot vertically then rotated. Awesome.
Quentin
[a href=\"index.php?act=findpost&pid=76940\"][{POST_SNAPBACK}][/a]
And how about stitching a big bunch of vertical shots.
Eric
-
And how about stitching a big bunch of vertical shots.
Eric
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=76954\")
I think Roundshot are famous for auto-stitching vertically - in fact the second camera on the homepage does this, so we can expect a rotating monster soon - I shudder to think of the storage needed for such a supermonster. Maybe this is actually an application for which IMAX film is more appropriate
[a href=\"http://www.roundshot.ch/xml_1/internet/de/application/d438/d925/f933.cfm]http://www.roundshot.ch/xml_1/internet/de/...8/d925/f933.cfm[/url]
Edmund
-
I think Roundshot are famous for auto-stitching vertically - in fact the second camera on the homepage does this, so we can expect a rotating monster soon - I shudder to think of the storage needed for such a supermonster. Maybe this is actually an application for which IMAX film is more appropriate
[a href=\"index.php?act=findpost&pid=76997\"][{POST_SNAPBACK}][/a]
It says 470 megapixels for an 80mm lens. This would generate a 2.8GB file (48-bit RGB) but these are Bayer type pixels so two thirds of this is fudged data. RGB probably isn't the best space to edit this in, we really need Photoshop support for something with chroma-subsampling (like Kodak's YCC as implemented in PhotoCD). You've got to wonder about the market/application for such huge files/images.
-
To think that if I didn't visit the LL forums I might never have heard of Seitz or Roundshot!
I wonder how much trade-in theyll offer me for my Canon S60 P&S.
Eric
-
It says 470 megapixels for an 80mm lens. This would generate a 2.8GB file (48-bit RGB) but these are Bayer type pixels so two thirds of this is fudged data. [a href=\"index.php?act=findpost&pid=77007\"][{POST_SNAPBACK}][/a]
Stephen,
What makes you think that they are Bayer type?
My understanding was that they were true RGB, like all scanning devices until now.
Cheers,
Bernard
-
What makes you think that they are Bayer type?
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=77010\")
From the specs for the 6x17 (at the bottom):
[a href=\"http://www.roundshot.ch/xml_1/internet/de/application/d438/d925/f934.cfm]http://www.roundshot.ch/xml_1/internet/de/...8/d925/f934.cfm[/url]
"File sizes
raw (16-bit): 307 MB
uncompressed tiff (48-bit): 922 MB"
-
From the specs for the 6x17 (at the bottom):
http://www.roundshot.ch/xml_1/internet/de/...8/d925/f934.cfm (http://www.roundshot.ch/xml_1/internet/de/application/d438/d925/f934.cfm)
"File sizes
raw (16-bit): 307 MB
uncompressed tiff (48-bit): 922 MB"
[a href=\"index.php?act=findpost&pid=77012\"][{POST_SNAPBACK}][/a]
Well spotted, this is interesting indeed. I wonder how this works...
Regards,
Bernard
-
Is there a reason to assume that it will capture all colors on each photosite? Could it scan using a bayer-pattern sensor? Do scanning backs necessarily use scanning methods that capture full RGB on each photosite?
-
I'm not sure if this explains why it wouldn't be a problem, but Seitz seem to think you can handhold this thing.
[a href=\"index.php?act=findpost&pid=76911\"][{POST_SNAPBACK}][/a]
[/quote]
um...why are there finger holds on the body....?
surely if you're spending such serious bank on this thing you're not going to handhold it.....
-
um...why are there finger holds on the body....?
surely if you're spending such serious bank on this thing you're not going to handhold it.....
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=77048\")
Seitz' words, not mine:
"Thanks to ergonomical hand grips the camera is perfectly stable for handheld images."
[a href=\"http://www.roundshot.ch/xml_1/internet/de/application/d438/d925/f934.cfm]http://www.roundshot.ch/xml_1/internet/de/...8/d925/f934.cfm[/url]
I wouldn't presume to tell people what they can or can't do with it.
-
Seitz' words, not mine:
"Thanks to ergonomical hand grips the camera is perfectly stable for handheld images."
http://www.roundshot.ch/xml_1/internet/de/...8/d925/f934.cfm (http://www.roundshot.ch/xml_1/internet/de/application/d438/d925/f934.cfm)
I wouldn't presume to tell people what they can or can't do with it.
[a href=\"index.php?act=findpost&pid=77051\"][{POST_SNAPBACK}][/a]
Looking at the picture of the guy holding the camera in portrait orientation with two hands - how do you reach the shutter button (red thing poking out of the lens)? Either you have to take one of your hands off the grip or push the shutter button with your nose.
Perhaps you need two people to operate it - one to handhold the two grips and one to meter the subject, adjust shutter, aperture and then push the shutter button. Practicalities of this thing need to be worked on.
-
Well, these guys are certainly leveraging off the name of their city, Lustdorf...
-
This size is also interesting from an image stabilization point of view. I remember (can't relocate now) a gizmo that basically placed the camera in the centre of a wheel roughly the size of a bicycle wheel. A 1 or 2 degree "shake" when holding it would be minimized on the axis of the wheel.
-
Looking at the picture of the guy holding the camera in portrait orientation with two hands - how do you reach the shutter button (red thing poking out of the lens)? Either you have to take one of your hands off the grip or push the shutter button with your nose.
Perhaps you need two people to operate it - one to handhold the two grips and one to meter the subject, adjust shutter, aperture and then push the shutter button. Practicalities of this thing need to be worked on.
[a href=\"index.php?act=findpost&pid=77072\"][{POST_SNAPBACK}][/a]
Since it's a scanning back, I doubt the shutter button is used. This is just a hangover from the Technorama. I couldn't find a button anywhere on the body so I guess it's triggered by the control unit with a delay. This would also give you time to brace yourself for the shot. Provided you don't jerk it, I think their claim of being able to hand-hold it are possible. The capture of every line is very fast and any movement (of the camera) would be spread over the entire frame and probably not noticeable. It sure is an interesting beast!
-
How does a scanning back work?
Is there a small sensor of just a few pixel columns that moves across the image with the shutter curtain? (Is there a shutter curtain?) And the sensor is read-out many times during the exposure? Or is there one gigantic sensor?
I have no clue?
p
-
It is a scanner on a camera back. Works like you're scanning a sheet of paper.