The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Fotodiox challenges medium format resolution

David Klepacki

New member
Well no, because for 1/20th of the price of the MF system, the Rhinocam cannot do what the MF system can do - which is of course to create an image on a medium format size sensor with a single exposure.
The issue is how "best" to capture an image with resolution of 140+ megapixels. Basically, there are three ways to achieve this:

1. Use a single chip sensor and stitch multiple shots (e.g., Rhinocam)
2. Use a multiple chip sensor and stitch with less shots (e.g., MFDB)
3. Use a single chip sensor with a single exposure (e.g., Sietz: overview)

If single exposure is the most important criteria in getting there, then the Seitz looks to be the best tool for the job. Or, if cost is the most important criteria in getting there, then the Rhinocam (or similar stitching kit) appears to offer the overall lowest cost.

And, none of these solutions will be perfect and without problems.
 

torger

Active member
We also have motorised panorama heads, or manual panorama heads. I'd say for typical stitching applications a manual panorama head like nodal ninja with a DSLR is the "best" alternative. Quick to work with, forgiving when the right software is used, very high quality and fairly low cost.

To me stitching with an MFDB is not exactly the best use of money. One big reason for doing MF for landscape/architecture is to be able to capture all the quality you need in one shot, it's not only simpler but to me also a more pleasing way to work - if possible I prefer if my fine art prints is one single exposure and not something merged in the computer. But sometimes you need wider angle than the focal length you happen to have.
 

Stefan Steib

Active member
The most interesting part of this device is that it seems such technique has now reached mainstream. It uses a camera that is made in the millions, uses widely available (leftover) lenses from former highend MF that can be had cheaply now and does so at a stunning low price of 499 $.

A better and larger image for a small price for many , that alone is remarkable. And it has started some ideas in my head already...... :)

regards
Stefan
 

gerald.d

Well-known member
The issue is how "best" to capture an image with resolution of 140+ megapixels. Basically, there are three ways to achieve this:

1. Use a single chip sensor and stitch multiple shots (e.g., Rhinocam)
2. Use a multiple chip sensor and stitch with less shots (e.g., MFDB)
3. Use a single chip sensor with a single exposure (e.g., Sietz: overview)

If single exposure is the most important criteria in getting there, then the Seitz looks to be the best tool for the job. Or, if cost is the most important criteria in getting there, then the Rhinocam (or similar stitching kit) appears to offer the overall lowest cost.

And, none of these solutions will be perfect and without problems.
The Seitz 6x17 is a scanning back. It doesn't really do a "single exposure" in the typical sense. Because it's a scanning back, it actually introduces a whole bunch of problems that you'd never get from shooting and stitching multiple exposures. Also, it limits you to its pre-determined aspect ratio. Not saying it's a bad aspect ratio of course, but it is a limitation.

I doubt there are many people out there who have as much experience of stitching across multiple formats and approaches as I have. I've done 60 gigapixel images created from 4,500 individual photos with a Canon DSLR and a 300mm lens using a VR Drive 2, down to a two shot shifted stitch on a MFDB using an Alpa Max, and pretty much everything in between - including shooting, scanning, and stitching 5x7 large format film (nice idea in theory, PITA in practice).

I remain to be convinced that the Rhinocam approach would be any better than shooting around the nodal point with a lens that is actually designed for the NEX sensor. If you've got a constrained budget and already have a NEX, you don't need to spend a dime on the Rhinocam and MF lenses, just use what you have already.

That's not to say the Rhinocam is a bad idea - I think it's a great bit of kit. It would be a load of fun to shoot with, and I don't doubt for one moment that people will get some excellent results out of it. But it simply doesn't make sense to me if your main criteria is cost-driven. You don't need to use one of these in order to deliver the results you can get from it.

Speaking from considerable experience, there is absolutely no question whatsoever in my mind that to "best" capture images in the 140MP+ range, there is only one approach. You shoot with an IQ180 using Rodenstock glass on an ALPA FPS under the control of a Seitz VR Drive 2. Pricey, yes. But if you truly want the best, then that's the way to do it.

Regards,

Gerald.
 

Stefan Steib

Active member
Hi Gerald

that is all true for images that don´t have a foreground and no parallaxerror.
The closer you go the more valuable the Rhino will be. And not to forget, productshots, Furniture, industry...... a lot more than pano´s involved here.
Not to forget the wideangles used by Rhino will get a better image angle than on any other 4,5x6 solution as the image area is larger than any existing chip.

regards
Stefan
 

gerald.d

Well-known member
Hi Gerald

that is all true for images that don´t have a foreground and no parallaxerror.
The closer you go the more valuable the Rhino will be. And not to forget, productshots, Furniture, industry...... a lot more than pano´s involved here.
Not to forget the wideangles used by Rhino will get a better image angle than on any other 4,5x6 solution as the image area is larger than any existing chip.

regards
Stefan
Hiya -

You don't get any parallax errors if you spin the camera around the nodal point of the lens.

The wide angles used by the Rhino will not give you a better image angle than on "any other" 645 system. If you're shifting, then you're limited by the image circle of the lens - regardless of what sensor and shift mechanism you're utilising. If you're panning, then you have no limitation - you can shoot and render images that can mimic any wide-angle focal length for a real lens, and well beyond that into the realm of lenses that physically don't exist.

Again - to stress - I'm not saying the Rhinocam is a bad system, and I don't doubt that many people will find a use for it, but it's important to look beyond the marketing hype and understand which (very limited) scenarios it would actually provide a real benefit.

Regards,

Gerald.
 

Stefan Steib

Active member
Hi Gerald

??? The format of the Rhinocam with an 8 Stitch is 62,4x46,8mm. With a 4/40mm Distagon this will result in an image angle of 88,5 degr. diagonal.
If you use the same lens on an IQ180 53,7 x40,4mm you will get 80,1 degrees.

And parallaxerror: In theory you could do a pano on a tabletop Gerald. But have you ever done this ? Good luck ! I don´t doubt you have a lot of experience with architecture but you will go completely mad before this will come to work. There is : A) Focus changing B) Perspective changing and C) falling lines you will have to correct otherwise in Photoshop.
After some hours of postprocessing when you "may" have gotten one image that is usable you will long be done with a flatstitch.........

Regards
Stefan

PS.: of course you could do panning with several flatstiches also Gerald............... :)
 

Ben Rubinstein

Active member
Not quite sure this is new in any way? Stitching, including nodal point stitching or the use of a t/s lens for flat frame stitching has been around for many years. So have adaptors for putting your DSLR's on a stitching adaptor on the back of your 4X5 from the cheap ebay adaptors to the Camera Fusion adaptor that I used to own. Why the hubbub now? I've been matching the resolution of MF for many years using a 12 megapixel 5D and a variety of lenses. That isn't really the point though. Yes the requirement for MFDB's are only a small percentage of the photographic industry and that has been the same for many years. However the percentage using 60-80 megapixel cameras, in the main, will not be in any way impressed with giving them up for stitching, if they had then they wouldn't have bought them in the first place methinks.
 

Ben Rubinstein

Active member
And parallaxerror: In theory you could do a pano on a tabletop Gerald. But have you ever done this ? Good luck ! I don´t doubt you have a lot of experience with architecture but you will go completely mad before this will come to work. There is : A) Focus changing B) Perspective changing and C) falling lines you will have to correct otherwise in Photoshop.
After some hours of postprocessing when you "may" have gotten one image that is usable you will long be done with a flatstitch.........
I have done a lot of stitching architecture, both indoors and out Stefan, with modern software it is extremely easy. Sorry. You don't need to even correct for verticals when shooting, it's a simple click or two in the stitching software, faster than doing it in camera or in PS from my experience. What you are talking about above is about 5 years out of date.
 

Stefan Steib

Active member
Hi Beni

try a tabletop maybe 50x60cm and packshots or a closeup (maybe a watch) with panning. That was what I was writing. And then tell us your findings.

Have fun.

regards
Stefan
 

gerald.d

Well-known member
Hi Stefan -

Hi Gerald

??? The format of the Rhinocam with an 8 Stitch is 62,4x46,8mm. With a 4/40mm Distagon this will result in an image angle of 88,5 degr. diagonal.
If you use the same lens on an IQ180 53,7 x40,4mm you will get 80,1 degrees.
As I said in my earlier post - "If you're shifting, then you're limited by the image circle of the lens - regardless of what sensor and shift mechanism you're utilising."

You can of course also shift the IQ180 on the 4/40mm, so your point is not relevant.

And parallaxerror: In theory you could do a pano on a tabletop Gerald. But have you ever done this ?
Yes.
PS.: of course you could do panning with several flatstiches also Gerald............... :)
And I've done this too - just last week in fact.

All you're doing here is reinforcing the very point that I was making - there are only a few limited scenarios where the ability to shift a NEX sensor around the image circle from a MF lens will actually produce a better result than pan-stitching using lenses designed specifically for the sensor.

IF you want to do panos of close-up objects and need to keep a flat focal plane, and IF you can get hold of a MF lens that is going to work well with the 4 micron pixel pitch of the NEX 7 sensor, and IF you need the extra resolution, and IF you don't have the budget for "proper" MF kit, then yes. This would be a solution worth investigating.


Kind regards,

Gerald.
 

yaya

Active member
Can the NEX do the stitching in-camera? Can it be used tethered or do you have to shoot those 8 frames, download them from the card and then stitch them? Can you do lens corrections (CA, vignette etc.) on the stitched image or do you do it on the separate smaller images? All in all IMO it's a nice gadget that expands the capabilities of small cameras. It is also great (for us) that they mention the Credo 80 as the MF reference (but do not show any images...)

Yair
 

gerald.d

Well-known member
Yair -

The images would not be stitched in-camera (there is no way the NEX could know that it was being used for shift-stitching). You'd use either Photoshop or one of the pano stitching programs (Autopano, PTGui, etc) to do the stitching. These programs have very advanced lens correction algorithms.

Regards,

Gerald.
 

Stefan Steib

Active member
Hi Gerald

"All you're doing here is reinforcing the very point that I was making - there are only a few limited scenarios where the ability to shift a NEX sensor around the image circle from a MF lens will actually produce a better result than pan-stitching using lenses designed specifically for the sensor."

there is about ONE central condition that applies for probably 99 % of the people who will buy this device from Fotodiox: they do not have around 50-60k $ to spend for such a setup.

They will get a 140Mpix file (or more if they pan) for around 3k$.
I´d call this pretty essential.

Nobody says it will be better than the Large solution. I say it will be good enough for maybe 99 % of the people.

Regards
Stefan
 

Stefan Steib

Active member
Hi Yair

A Nex 5R(500€) or a Nex 6(600€) can do Wlan out of the box. so you just shoot into your laptop´s hotfolder and then hit go/or do a droplet when the 8 images are there.

Regards
Stefan
 

yaya

Active member
Hi Yair

A Nex 5R(500€) or a Nex 6(600€) can do Wlan out of the box. so you just shoot into your laptop´s hotfolder and then hit go/or do a droplet when the 8 images are there.

Regards
Stefan
Do they transfer the RAW files or JPEG images?
 

gerald.d

Well-known member
Stefan - this isn't about comparing with a MF set-up.

This is about comparing with using a NEX and simply nodal stitching. Which, for the vast majority of usage cases, I believe would provide better results than using MF glass and shift-stitching.

Regards,

Gerald.
 
Top