Let’s face it, digital capture systems today have far exceeded our expectations of quality, speed, cost, and size.  I am still amazed by the small camera on our smartphones that can create shockingly great images.  However, while they are amazing to have at our instant disposal, when we compare these files side by side to our professional cameras, their limitations are quickly evident.  They fall apart in almost every area of comparison.  But what are those areas of comparison, and how does a normal photographer evaluate this?  It’s clear that a working set of standards for image quality is highly beneficial to any individual or organization that needs to compare files or meet a certain criteria. 

Luckily, in the cultural heritage marketplace, a few US federal agencies established a set of specifications that allowed us to measure and compare the quality of our image output.  The Federal Agency Digital Guidelines Initiative (FADGI) provided all image creators with a complex set of parameters for digitization purposes in order to yield consistent content.  This took the difficult prospect of image quality consistency and standardized it into a simple four star rating system to compare output.   Don’t we wish we had this in other areas of photography?

Why We Conducted This Test

We have been incredibly frustrated in both the lack of information or the intentional misinformation in our market in regards to Cultural Heritage capture systems.  Nowhere online could we find any testing that illustrated ALL of the options for clients looking to achieve the highest quality capture and where today’s systems fell in the FADGI rating system. Why?  We strongly believe there are two reasons.  

First, the testing is difficult to conduct and report properly. Before the new online portal, NimbusQ∆, it took almost $6,000 of targets and profiling software to see if you are meeting current standards. This put it our of reach for most customers and retailers.

And secondly, there are financial reasons for those who specialize in this market to keep these results private. By doing so they could keep actual costs hidden and maintain inflated profit in their pricing structure. This just isn’t acceptable in today’s stricter budgets and has prompted us to work towards correcting some of this intentional misinformation. By adding some products to our offering and by utilizing the new NimbusQ∆ website, we can now proudly state that we have put in the effort to present the correct information to the cultural heritage market.

FADGI 4 and 100mp Capture

We are world renowned as the experts in medium format capture systems and we pride ourselves as the open and honest outlet in the US for photographic information.   For purposes of this test, we painstakingly tested all of the leading 100 megapixel capture systems side by side keeping as many variables equal thus allowing us to definitively determine each system’s FADGI 4 scores. Those variables that could not be equal, they are listed after each score. By following structured and detailed testing, it is now quantitatively possible to compare each 100mp capture system directly to its competition.  Our primary questions that we posed before the tests are:

Can the leading medium format manufacturer’s best products yield a FADGI 4 file shooting reflective artwork?

How does a single shot system compare to a multi shot capture system today?

In the future we will conduct other tests for each system including multiple lenses, transmissive film, and lower FADGI standards.  But for this test, FADGI 4 and the highest resulting score for each 100mp camera combination was our goal. 

Tools used in our testing standards

• Cambo RPS 255 Precision Copy Stand – Cambo is the world leader in precise copy stands. The motorized RPS allows for vibration absorption, customizable sizing and precise repeatability. Having an individually aligned and calibrated base, tabletop, column, cross arm, and camera platform means everything is consistently aligned for efficient productivity and quality results.
• Reproduction lighting – Lighting that is soft, uniform, full spectrum balanced, and satisfies the 98 CRI and CQS scores is of utmost importance.
• Golden Thread FADGI 19264 Target – In order to determine FADGI ratings it is imperative to capture the industry standard ISO compliant test chart. This Golden Thread target consists of 168 unique color patches, 5 Spatial Frequency charts, continuous greyscale bars, geometric distortion patterns, and L* scales.
• Laser alignment – CI laser aligns all planes of capture table to the capture sensor maintaining parallelism and registration.
• And finally we added an Edmunds Scientific 30×40 Resolving Power Chart, an Xrite Digital SG Color checker, and a QP Greyscale.

The Cameras, the Process, and the Scoring

The latest releases from Hasselblad and Fujifim were chosen for this test. However, in regards to Phase One, we used a sensor that has been available for more than 9 years. Why? We would have found it unfair to use the latest 150mp sensor vs the competition’s 100mp. To keep the variables as consistent as possible, 100mp sensors were chosen here and nothing more.

We then combined each of these with the same focal length lenses to keep as much consistency as possible. Since the sensors are a different size, there was discussion that we should go with the “normal” lenses of each size sensor. Instead, we chose not to change focal lengths from test to test and thus adding possible wide angle distortion characteristics. Here are the systems chosen:

• Phase One IQ3 100mp Digital Back on an XF body with a Schneider Kreuznach 80mm Blue Ring II Lens
• Hasselblad X2D 100 C Camera Body with the XCD 80mm f/1.9 Lens
• Fujifilm GFX 100s II Camera Body with the GF 80mm f/1.7 WR Lens

Each camera system was placed at a height from the testing targets that achieved a native 400 pixels per inch output for that sensor lens combination. The files were saved as a tiff and uploaded to Phase One’s NimbusQ∆ website for FADGI scoring results. For those cameras that allow 4 shot capture, they were shot in single and multi shot modes and those scores are listed here. Also, each camera was tethered and captured with its native tethering software.

Each file was then evaluated for accurate resolution, 16 separate quality categories, and scored as to whether it satisfies FADI 4 Standards. The scores are then calculated on a scale to 100, thus the quality of the resulting files can now be evaluated directly against one another.

A few other characteristics that we kept consistent were:

• White balance was created from the same location, the XRite A1 patch, for each system in their native software.
• Sharpening was turned down to half of base standard sharpening settings for each manufacturer.
• Lighting was pulled back and set to 30° angles to far edge of table top
• No custom profiles were used. Only manufacturer linear base profiles for each system.
• Exposures were based on the same 8bit numeric value on patch A1.
• All lenses were shot at f/11
• All ISO’s were shot at 100

Digital Capture Scoring Results

Phase One IQ3 100MP Digital Back

• Phase One IQ3 100mp
• Phase One XF Camera Body
• Schneider Krueznach 80mm Blue Ring II lens
• Capture One Tethering and Processing
• Single Shot Capture
• 400ppi Resolution
• 21.5 x 28.75 in. native capture

FADGI Single Shot Score = 97%

Total cost for the FADGI 4 Repro station = $46,000

Notes on Phase One workflow and results

Since there is no option for multi-shot capture, single-shot capture is the only option for Phase One digital backs. However, Phase One consistently scored at the top of our testing through each round.

We were also able to achieve scores of 98% and 99% by using Rodenstock HR lenses. However, for this test, it seemed unfair to compare $10,000 to $15,000 lenses to $3,000 to $5,000 lenses from the competitor. Since the goal was to find “out of the box” results with each manufacturer the Auto Focus 80mm Blue Ring II achieved that purpose.

The easy question to ask is “since Phase One is a partner in NimbusQ∆, do they intentionally score files higher that have Phase One Metadata?” Our gut feeling on this is no. The visual of the actual Phase One files showed superior tonal quality, color accuracy, and uniformity.

And the final question to ask, “Does Capture One Processing give an advantage to Nimbus scores?” And honestly, this question is still up for debate.

Our first system tested achieved FADGI 4 consistently for well under $50,000.

Hasselblad X2D 100c Camera

• Hasselblad X2D 100c Camera Body
• Hasselblad XCD 80mm f/1.9 Lens
• Phocus Software Tethering and Processing
• Single & Multi Shot Capture
• 400ppi Resolution
• 22 x 29 in. native capture

Single Shot = 83% ( fail )

FADGI Multi Shot Score = 91%

Total cost for the FADGI 4 Repro station = $35,000

Notes on Hasselblad workflow and results

Hasselblad historically has championed the advantages of multi shot backs and pixel shifting. Many of our clients were severely disappointed when they discontinued the Hasselblad H6D 400c multi shot digital system and gave the cultural heritage market fewer options. It sent them scrambling for expensive alternatives. And then in 2024 firmware version 4.0 gifted existing Hasselblad X2D users the ability to pixel shift again on their pre-existing systems.

Multi-Shot technology involves moving the sensor accurately by 1 or 1⁄2 a pixel at a time by utilizing the In Body Image Stabilization (IBIS) actuator, thereby enabling the camera to capture each filter in the bayer patter for each individual pixel. The quantifiable quality improvements are more accurate color data for each pixel, virtually no moire issues, and improved edge definition.

So it was not a surprise to see these results for the Hasselbald X2D 100c. The true key to satisfying FADGI 4 was the improved 4 shot capture tethered through Phocus software. With only one toggle in the Phocus interface and no changes to exposure, a 4-shot image with exactly the same pixel dimensions and framing is produced.

Needless to say, we were thrilled to see the 4 shot capture come though and save the day. While on paper and in marketing, pixel shift technology states that it can resolve edge details that single shot cannot, it was extremely satisfying to now be able to illustrate the actual advantage mathematically and directly to its competiton.

And more importantly, Hasselblad allows Capture Integration to assemble a complete FADGI 4 system for less than half the price of the leading competitor in the Cultural Heritage Market.

FujiFilm GFX 100S II Camera

• FufiFilm GFX 100s II Camera Body
• Fujinon GF 80mm f/1.7 WR Lens
• Fujifilm X Aquire Tethering and processing for 4 Shot Capture
• Capture One Tethering and Processing for Single Shot Capture
• 400ppi Resolution
• 21.75 x 29 in. Native Capture

Single Shot = 89% ( fail )

FADGI Multi Shot = Not Recommended

Total cost for the FADGI 4 Repro station = Not recommended

Notes for Fujifilm workflow and Results

And now to the most surprising results of our testing. After years and years of promoting Fujifilm GFX cameras we thought we knew this product well. The lineup of products and their broad offering have produced amazing results for such a large number of our clients. Architectural, Landscape, Commercial, and Editorial customers have excelled using the Fujifilm line of GFX camera bodies. And we ourselves have also championed the quality of the final images. None of that changes our results of this Cultural Heritage FADGI testing.

The reality is that in a controlled and rigidly strict consistent environment, FujiFilm products fail. And trust me, this is as surprising for us as it is for many of our customers to read. But if you bear with me you might have seen the same Fuji anomalies in your process that we have all just shaken off in the past.

Consistency – Those customers who have ownership of their FADGI process understand that it is important to keep all variables controlled. Many customers test for FADGI compliance daily or even multiple times a day. And this is where Fujifilm fails shockingly. While tethered to Capture One, we find a shift in exposure of up to +/- 5 points (10 point total variability) in the 8 bit 256 scale from one image to the next. This is occurring in every single body we tested (6) of the latest GFX 100S II and GFX 100 II bodies. This exposure inconsistency is occurring with both continuous and strobe capture, and with electronic shutter and mechanical shutter. And while we achieving passing scores on some of these tests, we don’t find a 10 point shift in exposure acceptable in the strict archival process.

Live View Focus – To put it bluntly, it is just does not compare to the ease of live view to its competition. It is difficult to look at all 4 corners of your flat art through live view. It takes a zoom, scroll, scroll, scroll to reach each corner. And sometimes it gives us 100% views and sometimes it just does not want to work no matter how many times you try. It is a slow frustrating process and slows down any production efficiency. And this is tethering any Fujifilm GFX body in any situation. If you own one, you know the issues.

Multi-Shot Capture – I have one word to describe their Multi Shot quality: Flawed. On one hand, the exposure consistency we have noticed is solved with 4 shot capture. These files are actually accurate in density from shot to shot within the multi-shot sequence as they should be. However, the multi-shot images require a varying amount of overexposure compared to single shot images (needing to adjust both camera exposure and/or light output value from 1/3rd to beyond 1 full stop). Secondly, assembly of the 4-shot image requires the use of Fujifilm’s Pixel Shift Combiner which has to run outside of Capture One and produces a finished DNG file that needs to be imported into Capture One manually. Furthermore, the boundary of the multi-shot frame, while not just smaller by ~1%, can include portions of the scene outside of the boundary of the corresponding single shot file. Thus changing your crop and our output file to file. And finally, each GFX camera body tested reacted a little differently, thus giving us no consistency from one repro table to the next.

Beyond the more complicated shooting workflow, the issues of geometrical distortion and scaling were like nothing we have seen before. It appears to us that the the lens correction settings for these multi-shot files are creating inaccurate scaling issues outside the center 1/3rd of the image to the corners. This creates files that are not at the same scale from center to edge. Secondly, with lens distortion uncorrected, there is massive distortion making the files completely unusable for flat art work. And finally, with our GFX 100 II, approximately every 5th image gives us a green color shift that white balancing cannot correct.

The above two images are a layered multi shot image at 50% opacity on top of a normal single shot file. The first image illustrates complete registration a the center while the second image illustrates how the multi shot is stretched at the corners creating a different scale.

Yes, we are as surprised and disappointed as you might have been to read these results. It would make sense that our commercial clients have not witnessed all of these anomalies in the field. When the sun is changing you won’t notice a 5 point shift in exposure. These problems truly only show up when tethered in a highly controlled environment. Again, while we achieve satisfactory FADGI standards on some Fujifilm files, it is disappointing to state that we cannot currently recommend any GFX camera system for Cultural Heritage workflow.

Conclusions

US FADGI 4 standards do not require a $70,000 – $150,000 capture station to satisfy their standards.

Phase One digital systems are still the leading capture device in the Cultural Heritage marketplace.

Hasselblad 4 Shot capture allows you to achieve FADGI 4 Standards at a fraction of the price to the competition.

Fujifilm as a manufacturer has some flaws in its workflow consistency that make it a poor choice for Cultural Heritage workflow.

“New knowledge is the most valuable commodity on earth. The more truth we have to work with, the richer we become.” – Kurt Vonnegut

dave@captureintegration.com

770.846.5223