Back to


Warning: This post is somewhat technical, kind of lengthy, and it has an example of early recycling near the end.

A former professor of mine recently called with some technical questions about the Conservation

Department’s new x-ray equipment. Instead of responding to him via e-mail I thought I would provide a kind of open response here. And why not, right? It just might be interesting to you and perhaps also to my colleagues in the conservation world. Besides, in my mind, the IMA’s blog seems to be the perfect place to discuss x-radiography considering the previous non sequiturs of Cheesecake and Redbuds.


I’m a little worried, though, because I doubt many of the folks upstairs realize that we have the capacity to produced x-rays down here in the basement and this might make them a bit nervous. I think it’s fair to say that most people have a fear of x-rays for good reason, because they have the potential to change humans on the atomic level, and that’s unnatural at best; but, trust me, we’ve taken a lot of safety precautions to make sure that all of the x-rays that we produce stay in the room that they are produced in. The room is an enclosed space that is lined all the way around with an 1/8 of an inch of lead. We use a Geiger counter to confirm that this room is successfully containing the energy.

Now I’m not going to take the time to fully explain what an x-ray is or how an x-ray tube actually produces x-rays. If you’re interested, go here to learn what x-rays are or go here for a brief discussion of how they are produced (the links will introduce you to such things as anode, cathode, colliding electrons, KV, MA, and so on). I will say that an x-ray is a form of electromagnetic radiation that carries a relatively large amount of energy and momentum. Depending on an x-ray tube’s capabilities, x-rays can penetrate through various objects – from canvas paintings to this covered bronze vessel with handle (60.24, called a “You”) from the Western Zhou Dynasty (c 900 BCE) and even through large pieces of stone. Our “you” (pronounced: yo) is a traditional type of Chinese ritual vessels; you can go here to find out more about the other types of Chinese ritual vessels.

In the radiograph above of this Chinese vessel (notice I removed the lid), you can “see through” it because the x-rays have penetrated all the way through the copper alloy from which this vessel is made (it’s about 3/8 of an inch thick). You should be able to see that this vessel was actually broken into many pieces and then the fragments were re-assembled at some point (Jim Robinson, the IMA’s Jane Weldon Myers Curator of Asian Art, believes this was likely re-assembled shortly after it was recovered from burial. Side note: you can go here to see a video of Jim and the director talking about the new Maya Lin installation, Above and Below.)

X-ray technology has been used frequently in the field of conservation for many years. In fact the IMA purchased its first x-ray tube more than 30 years ago; but last year that tube finally broke down. We were all pretty bummed about that until, out of the blue, the IMA was given a gift from an anonymous donor that allowed us to purchase a bigger and better x-ray tube and shift to a Computed Radiography system. I really have no idea who this donor is, or these donors are, but I’d like to say thanks. Seriously, thanks a lot! I know that the whole conservation department has been energized by this new piece of equipment.

So, the old system that we had produced radiographs just fine but we had to use x-ray film to capture the images – if you want to get a sense of what this film was, think of large piece of black & white film (14” x 17”) and you’re in the right ballpark. Though you can make high quality and clear images using film, the process is time consuming, messy, and, in a way, a thing of the past. Plus to view film radiographs they need to be lit from the back (or they can be scanned as a digital file, but you then loose some of the resolution). There are plenty of examples of analog technology that already are or soon will be digital.

As I mentioned, the digital technology we now use to capture images is called computed radiography (CR). While CR requires the same kind of tube to generate x-rays, one of the main differences between CR and film is that instead of shooting on film, we now shoot onto reusable phosphor imaging plates that are scanned and then viewed on our computer. So, instead of developing film in a series of chemical baths, we simply stick this “plates” in a scanner (here’s a spec sheet on our imaging plates and here’s a spec sheet on our scanner). If you don’t mind hearing high-pitched noises, you can even watch one of these plates go through the scanner at this link (and, no, I don’t need to be reminded how geeky it is that I made a youtuber of this process; clearly I realize I’m a bit geeky, but oh well). Needless to say, this process is much easier and faster than developing film by hand.

With this new system we now have the capacity to do things we’ve never done before – plus we’re kind of proud to be among the first institutions to use CR in a museum conservation department. CR has been used widely in the medical and industrial fields for some time now but it’s relatively new to the museum world. When I say that we can do new things with this system this is because with our new x-ray tube we can produce x-rays of higher energy. Our old tube could only produce x-rays up to 110KV but now our new one can go to 200KV. Here’s the spec sheet on our new tube.

With our old tube I didn’t have the capacity to generate strong enough x-rays to penetrate through the Chinese vessel pictured above. Previously to do something like this we would have to collaborate with nearby industries to use their x-ray equipment. For example, in 1981 the IMA worked with Detroit Diesel-Allison to produce radiographs and other technical analysis of this vessel and others like it for the out-of-print catalogue Beauty and Tranquility: The Eli Lilly Collection of Chinese Art. Of course collaboration is always a good thing, and we’re always looking for new partners with which to collaborate, but in this case it’s much faster to be able to do this work in our own lab.

Since this object was radiographed in 1981 I have an excellent film-based example to which I can compare the new digital file that I created with the CR system. This is important because one of the main questions about the difference between film-based radiography and CR revolves around the idea of the quality of the image, its resolution. That is to say, can the CR system make an image that is of the same quality as what used to be made with film? The other part of this question is if CR can produce high-quality images when higher KVs are used.

I think so, but it may be a bit difficult for me to prove it here because, of course, I can’t really show you the film-based radiograph over the internet and the monitor I used to view CR files is a high-resolution monitor (here’s a spec sheet on our new monitor). Clearly, the best way to do this kind of comparison would be with me down in the department’s analytical room. Since we can’t do a side-by-side comparison, I’ll explain to you an example of what our CR system can do.

Jim Robinson mentioned to me one day that this vessel had an interesting example of early recycling. I’ll let Yutaka Mino and Jim explain what they observed about bottom of the You in their catalogue back in 1981:

“The exterior of the base has a grid design of raised lines, and the chaplets in the bottom are randomly placed. Two chaplets are of particular note. One is under part of the inscription, showing that the characters were not cast with the vessel. The other interesting chaplet has a fine spiral pattern of leiwen on it. This means that an old, possible broken or unsuccessfully cast vessel was salvaged and used in the manufacture of this object. The practice of recycling old bronzes is manifested in other bronzes.”

I think you can see these chaplets clearly in the radiographs I created (here’s a spec sheet on the software we use). Have a look: can you find the three chaplets in this image? In case you’re wondering, chaplets are used in the casting process to act as a kind of spacer between the inner and outer molds. In this image the chaplets appear kind of like islands in the rest of the copper alloy. These chaplets were inserted into the mold before the liquid metal was poured in.

Look closer, can you see the chaplet with the spiral pattern on it? Can you also see the grid design of raised lines that Yutaka and Jim were talking about? Pretty cool, huh?

So, finally, I think that CR produces images that are of equal or higher resolution than film radiography, and it’s faster and easier to use. Plus, the images that are created are easier to share, and even potentially be blogged about. If you’ve read this whole post, thanks and congratulations; I really didn’t mean for it to be this long, it just sort of happened.

If you have questions, post them in the comments and I’ll answer them the best I can.

More Info:
You can check out more images that I put up on my account.

The digital radiographs I produced were shot at these settings.

Radiograph from the side 170 KV 3 MAs 1.7 minutes
Radiograph top 125 KV 3 MAs 1.4 minutes
Radiograph of lid only 110 KV 3 MAs 1.1 Minutes

Further Reading:
Here’s an interesting article from JAIC:
“Technical Examination of the Classical Bronze Horse from the Metropolitan Museum of Art”

From the California State University Northridge:
“Piece Mold, Lost Wax & Composite Casting Techniques of the Chinese Bronze Age”

Filed under: Conservation

15 Responses to “X-Radiographic”

  • avatar
    Ed Says:

    I was wondering, since I know that ordinary sunlight can fade materials over time, is there any concern when using x-radiography that the process will cause significant damage to artworks? Are there certain materials or objects of a certain age that you wouldn’t use this technique on?

  • avatar
    JP Brown Says:

    > is there any concern when using x-radiography that the process will cause significant damage to artworks?

    There are measurable alterations in some properties of some objects as a result of exposure. One such property is the thermoluminescence date of ceramics — x-ray exposure will make the TL date appear older than is actually the case. This can also be an issue with cast bronzes where dating any clay remaining from the casting mantle may be the only scientific dating method available.

  • avatar

    Thanks for this helpful introduction and the invitation to the conservation field to start a broader, collaborative conversation about computed radiography.

    As many are following the rapid developments in computed radiography, it is worth noting that a parallel conversation is underway in the field to address the fate of existing archives of x-ray films that are housed in a large number of museums throughout the US and abroad. Efforts are being focused on identifying the most suitable digitization technology and methodologies, along with the necessary agreed-upon set of metadata to insure long-term preservation and access.

    More information on this developing initiative can be found at

  • avatar
    Richard Says:

    Thanks, JP Brown, for answering Ed’s good question. I agree with you. I’ll add that if and when I expose any artwork to x-rays, I make note of the dosage that the piece received in the event that is useful for future generations.

    And thanks, Alison, for your good mention of the x-ray film archiving initiative. That’s a great project, one that I’m going to continue to watch develop. I understand the project to be aimed at investigating the “feasibility of establishing a centralized digitization service.” I wonder, though, if there is interest in making these radiographic images available to the public, once digitized.

    Wouldn’t it be cool to have radiographic images associated with the on-line collection database!

    In an act of blatant thievery, I’m going to post three links that are from Alison’s link:

    Understanding Scanners: A presentation by Robert Simcoe, Harvard College Observatory

    National Gallery of Art Guidelines for X-Radiograph Film Preservation, by Constance McCabe, Elizabeth Walmsley, and Sarah S. Wagner

    Cold Storage Options: Costs and Implementation Issues, by Sarah S. Wagner

  • avatar
    Karen T. Says:

    One of the nice things about traditional x-radiography is having a 1:1 film to compare with the art object. Are there output options with your setup that you feel are on par with traditional film? I’m approaching this from the standpoint of paintings — in our museum we regularly view a painting in the studio on the easel with an oversized lightbox adjacent for the x-ray. Many conservators feel a printout (usually limited to 8×10 or so) doesn’t offer quite the same level of detail and information. Any thoughts?

  • avatar


    Thanks for “stealing” those links! The more people who know about this good stuff, the better. And apologies for taking your thread off-track, since computed radiography is ripe for an active and focused discussion of its own. But to briefly answer your question:

    > I wonder, though, if there is interest in making these
    > radiographic images available to the public, once digitized.

    Yes, in fact that is one of the long-term goals that is under discussion for x-ray archives (the final bullet point on the agenda).

    ARTstor is one possible venue that could provide access to these federated collections, given its teaching and research mission. The side-by-side viewing tools available through the site would allow for pairing visible light images (where available) and x-rays for comparative study across numerous institutions.

    Of course, whether scanned or born-digital, individual institutions are free to present their content however they choose. Similar side-by-side views could certainly be achieved in local online collection databases. It is more often an issue of resources and priorities than stated interest.

    The x-ray archive agenda is one part of the Mellon Foundation’s broader initiative exploring the challenges, complexities, and opportunities of Conservation Documentation in Digital Form. The topic of public access continues to figure prominently in all of these discussions (transcripts, documents, publications, etc. all on

  • avatar
    Christina Says:

    I would like to respond to Karen T.’s comment on how computed radiograph compares to film when examining paintings. You pose a very good question in regards to the 1:1 comparison of the painting to tradition film. Printing out an image of the computed radiograph to simulate the same practical comparison in one option. One limiting factor would be the printer’s features (not only size limitations, as you mentioned, but image resolution as well). We purchased a large format photo-quality pigment printer so that we could print out bigger, more practical images of our x-radiographs. There are settings to print 1:1, so with the larger papers, it’s possible to create a printout that is essentially the same size as traditional film. The resolution quality of the image can be quite good, depending on the particular printer used. The second option would be to set up the computer with the CR review software in a space that allows you to also have your object close by. With this option, you can review the x-radiograph on the high-resolution monitor in a 1:1 comparison, or magnified for greater detail. I recently used both methods to examine a heavily overpainted area of a painting to determine whether there were losses to the original paint beneath the overpaint.

  • avatar
    Angeliki Says:

    I work on the industrial field but since you are among the first tou use the CR system I could not help myself from asking you..
    Do you know which is the lifetime of an imaging plate?
    I am worried that it could be overstimated.

    Moreover I would like to know of how you deal with the problem of dead pixels. What is the limit of screen-pixel drop-off? Which are the equivalements of requirements of film type and denisty. I mean that in film radiography we have classes of films is there any similar standard with classes of pixels?

    Kind regards

  • avatar
    Mark Mc Says:

    In response to Angeliki’s questions:

    “Do you know which is the lifetime of an imaging plate?
    I am worried that it could be overstimated.”

    The physical plate will wear out before the chemistry of the plate stops holding x-ray images. If proper care is taken with the CR plates, you can expect many thousand exposures on them. Typically, scratches, dirt, and rough handling will bring about the end of a CR plate’s life.

    In regards to “dead pixels”, this really is not an issue with CR imaging technology. The CR plates emit light when stimulated by the scanner’s laser. This light is integrated to a digital signal by the scanners photomultiplier tube. There are no pixels on the plate, as is one continous surface. This is in contrast to flat panel imaging detectors, which have a finite number of elements which collect the x-ray energy (we will call them “pixels”). If a pixel on a flat panel is bad, it can typically be calibrated out by taking the values of surrounding good pixels.

  • avatar
    Richard Says:

    Thanks, Mark Mc, for your comment.

    I hope this answers your question, Angeliki.

    I asked Mark Mc from GE to respond to this comment because I thought they would be better positioned to provide an answer.

    I can only tell you anecdotal information as we’ve only had the system for about a year.

    The one thing I’ve noticed with the plates is that you have to take care not to leave dust on them or on the scanner itself (the plates move across it during the scanning process). I have seen very thin lines in the image when I haven’t taken the time to dust the scanner before I scan.


  • avatar
    Angeliki Says:

    Thanks so much for your immediate reply,
    it was really helpful but what about the athe equivalements of requirements of film type and denisty. The imaging plates correspond to what class of film? D4? d2
    Thanks again, keep up the good job with this blog.


  • avatar

    Hey, Angeliki —

    I’m sorry to have taken so long to respond. Thanks for your comment and kind words.

    I think it’s hard to say what film the Phosphor Imaging Plates compare to. Here’s a link to some info about the plates:

    I’ve been working with the system now for a long enough to know that it takes a little while to learn how to take a very good image. It’s kind of a new way of thinking; maybe it’s just me, but I haven’t tried to think of correlations between the film and phosphor, I just try and make it work on the phosphor.

    I hope that makes sense.


  • avatar

    I’m interested in how the Rhythm software works. I understand it cannot stitch seperate scans together. How do you get around this. Are there any other quirks of this system.

  • avatar

    Hi, Stewart —

    We’ve been told by GE that they are working on the stitching function and that it is coming “soon,” but we haven’t been given any firm date on when that is gonna happen.

    We have exported the DICOM files as a TIFF and then used Photoshop to stitch them together. This can get a bit unwieldy, though, if you don’t watch your file size. Can you imagine stitching together 10 20mb images? Yikes.

    The other issue we haven’t fully resolved is the archiving process, but this could likely be somewhat easily resolved if we were to put in the necessary time to figure it all out. I know we made some major headway on this recently, but I haven’t had a chance to get back to the software and really see what it can do.

    Thanks for the comment!


  • Trackbacks