[Ernie C.]

Not Pentax but Hasselblad. I want to share with you a very interesting project by the Rijksmuseum. The Night Watch in 717 gigapixels: Ultra high resolution photo

---

Last edited by Ernie C.; 01-09-2022 at 01:43 AM.

[beholder3]

Not Pentax but Hasselblad. I want to share with you a very interesting project by the Rijksmuseum. The Night Watch in 717 gigapixels: Ultra high resolution photo

Two comments:

1) Especially with true art (regardless of oil paining or photography) the pixelcount is completely irrelevant. It either impresses at VGA-resolution or not and "impressing" is not related to any technical aspects. So from the creative/artistic/skill/value side this is worthless.
It probably is good for theft protection, art history, studying painting technique etc.


2) If the museum is smart they turn this image file into 10 NFTs and sell them. I am pretty confident that will generate lots of money for the museum which is always good for the arts.

[BigMackCam]

1) Especially with true art (regardless of oil paining or photography) the pixelcount is completely irrelevant. It either impresses at VGA-resolution or not and "impressing" is not related to any technical aspects. So from the creative/artistic/skill/value side this is worthless.
It probably is good for theft protection, art history, studying painting technique etc.

It's a fine art reproduction photograph, so clearly there's no creative or artistic element involved, nor should there be... but there's definitely skill in setting up and aligning the equipment for each of the thousands of shots, ensuring even lighting, determining optimum exposure for the entire image, ensuring uniform sharpness across each frame, stitching them, reviewing the results and making any necessary corrections to colour and tone curves etc. The amount of work from planning to final result must have been huge...

[AfterPentax Mark II]

Two comments:

1) Especially with true art (regardless of oil paining or photography) the pixelcount is completely irrelevant. It either impresses at VGA-resolution or not and "impressing" is not related to any technical aspects. So from the creative/artistic/skill/value side this is worthless.
It probably is good for theft protection, art history, studying painting technique etc.


2) If the museum is smart they turn this image file into 10 NFTs and sell them. I am pretty confident that will generate lots of money for the museum which is always good for the arts.

The object for the Rijksmuseum is to share this painting of Rembrandt with as many people as possible, they do not want to sell it, but their aim is that you can watch it, even if you are a million miles away, for free. It is not always about making money.

[Sandy Hancock]

Remember dial-up internet? This is really impressive.

[BigMackCam]

A 5.6 terabyte file... That's mind-boggling. Once you've created the stitched image, what can you actually do with it? I guess you'd need a custom app to load sections of the file in sequence and render each separately. I'll bet that takes a while!!

[biz-engineer]

"This is the largest and most detailed photo ever taken of a work of art. It is 717 gigapixels, or 717,000,000,000 pixels, in size."
That's only a 3.5 meters print at 600PPI or ~7 meters at 300PPI. But at 600PPI or even 300PPI , printers can't print that wide. It'll have to be down-sampled to be printed at 300 or 600PPI. We don't do this everyday, usually we up-sample for prints because our cameras are too cheap to print 300PPI straight out of our camera's sensor native resolution. Photo lab doesn't accept file over 1GBytes, I'll pass on that one, gotta have to down-sample that file and cook eggs on the back of my notebook while processing.

"The distance between two pixels is 5 micrometres (0.005 millimetre), which means that one pixel is smaller than a human red blood cell."
And CPU transistors are much much smaller than that, 5 micrometers was the standard 30 years ago. Artists need to get a little refresh about computer tech. And also , the human blood cell isn't the reference anymore, since 2019 the most popular small size reference is the corona virus.

"The team used a 100-megapixel Hasselblad H6D 400 MS-camera to make 8439 individual photos measuring 5.5cm x 4.1cm."
I didn't know one man isn't enough to handle an Hasselblad H6D 400 MS camera. I guess the photographer is very skinny, not enough muscles, should register to the gym (after covid is over of course).

"Artificial intelligence was used to stitch these smaller photographs together to form the final large image, with a total file size of 5.6 terabytes."
I know that AI sounds very high tech, but stitching software doesn't use AI at all. Image stitching is done via image registration algorithms (SIFT etc), lots of RAM and CPU power, although for such image size the stitching is done chunk by chunk because it's not possible to store all files in memory at once.

---------- Post added 09-01-22 at 14:59 ----------

Once you've created the stitched image, what can you actually do with it?

You can use that file for computer/memory tests. Advantage is you don't need to duplicate 100 smaller files to perform the test, only one file is enough to overwhelm the computer immediately (fortunately computers don't have emotion, no panic attack).

---------- Post added 09-01-22 at 15:03 ----------

1) Especially with true art (regardless of oil paining or photography) the pixelcount is completely irrelevant.

One pixel is art, and such art agrees with everyone because it hands over the total freedom of imagination to the viewer. It's a concept.

---

Last edited by biz-engineer; 01-09-2022 at 07:05 AM.

[zapp]

Knowing the guys who did it, I really like the work and all the projects around it. This is a science project not an art project - this should be obvious.

[Digitalis]

Didn't your mother tell you pixel peeping is a habit that will lead to blindness?

[luftfluss]

2) If the museum is smart they turn this image file into 10 NFTs and sell them. I am pretty confident that will generate lots of money for the museum which is always good for the arts.

Does the museum own this painting, or is it on loan from another museum or individual? That would, of course, impact the ability to create NFTs.

[AggieDad]

If you want to duplicate this process in some small way, try and take a panorama macro shot.

Instead of a coin filling your sensor, close in until it takes 4 or 5 overlapping shots to shoot the complete coin. That will give you some idea of what has been done here. It will also give you one helluva photo of a coin.

[photoptimist]

The NFT idea would be brilliant. I'd bet more than a few wealthy art lovers would jump at the chance to own one of a very limited number (e.g., 10) of nonfungible tokens created by the famous Rijksmuseum of this famous Rembrandt painting. These limited numbers of NFTs could even be auctioned in serial number order -- NFT #1 would likely be more coveted than NFT #10.

NFTs do not preclude public release of all the data. Anyone could still "have" a copy of the painting (assuming they have the bandwidth and storage for it). However, those copies of the data would only be fungible copies and not worth more than the cost of copying the data. In contrast, the NFTs of the paintings cannot be copied in any meaningful sense because an NFT is a statement of ownership of the thing, not a copy of the thing itself. With NFTs, only a limited number of people would be the acknowledged as owners of these tokens.

What makes NFTs more valuable than the underlying data is the immutable record of ownership on a blockchain somewhere. An NFT is like a deed to a piece of real estate. That crypto-data creates undeniable truth about the ownership and provenance of the token (i.e., you can trace the ownership back to the Rijksmuseum itself). Within the limits of the security of the blockchain and the owner's data systems, no one can steal or counterfeit an NFT. Thus, a person can only boast that they own an NFT of this 717 GPix scan of the Rijksmuseum's Rembrandt if they can prove they own it and that proof sits on the blockchain for allthe world to see.

[BigMackCam]

The NFT idea would be brilliant. I'd bet more than a few wealthy art lovers would jump at the chance to own one of a very limited number (e.g., 10) of nonfungible tokens created by the famous Rijksmuseum of this famous Rembrandt painting. These limited numbers of NFTs could even be auctioned in serial number order -- NFT #1 would likely be more coveted than NFT #10.

NFTs do not preclude public release of all the data. Anyone could still "have" a copy of the painting (assuming they have the bandwidth and storage for it). However, those copies of the data would only be fungible copies and not worth more than the cost of copying the data. In contrast, the NFTs of the paintings cannot be copied in any meaningful sense because an NFT is a statement of ownership of the thing, not a copy of the thing itself. With NFTs, only a limited number of people would be the acknowledged as owners of these tokens.

What makes NFTs more valuable than the underlying data is the immutable record of ownership on a blockchain somewhere. An NFT is like a deed to a piece of real estate. That crypto-data creates undeniable truth about the ownership and provenance of the token (i.e., you can trace the ownership back to the Rijksmuseum itself). Within the limits of the security of the blockchain and the owner's data systems, no one can steal or counterfeit an NFT. Thus, a person can only boast that they own an NFT of this 717 GPix scan of the Rijksmuseum's Rembrandt if they can prove they own it and that proof sits on the blockchain for allthe world to see.

Sorry to the OP for going a bit off-topic here, but...

Is there a secondary market for NFTs? I assume there must be... otherwise the NFT has no value beyond the initial sale...

[texandrews]

These techniques have been around for a while. The Japanese have used them for smaller works in temples that are light sensitive, and they aren't that uncommon in the West. The only significance I can discern here is the size of the painting---it's possible no one has yet attempted to try this with such a large work. Otherwise the whole story seemed like click-bait to me.

[JohnMc]

Otherwise the whole story seemed like click-bait to me.

Size matters to/for some, otherwise yes the original NFT, The Click