Holtermann Collection Digitisation Project - Technical

The panoramas emerge

Fri, 04 Sep 2009 12:11:22 +1100

Apart from their well documented house front, store front and goldfield photography, Merlin and Bayliss they were also known for their landscape panoramas.
To capture these 180° views a number of glass plate photographs (usually three or four) had to be taken in succession and the prints trimmed and overlapped to form the illusion of one long, uninterrupted landscape view.

Although most of the panoramas were taken using mammoth plates so that they could be printed in a large exhibition size, we can also find some examples of these panoramas in the smaller quarter plate negatives.

We recently digitised the below four quarter plates and - using digital photo-stitching technology - were able to recreate the full panorama of Gulgong NSW taken from Church Hill in the 1870s.

Below we have an example of what the panorama would have looked like using the traditional print layering method of panorama construction - note that the joins are clearly visible and the areas of vignetting around the edges interrupt the flow of the landscape.

And here is our newly constructed panorama created using the digital intelligence of Photoshop CS4's Photomerge facility along with some manual retouching, dodging and burning - the result is far smoother and gives a clear impression of the view from Church Hill over 100 years ago.

Please take the time to view an enlarged version of the above panorama - it is well worth it!

This digitised panorama has the ability to be enlarged as a high quality photographic print - with no loss of image quality - to five metres in length. With such exciting results from negatives measuring only 3.25x4.25 inches [83x108mm] we look forward with great anticipation to seeing results from the mammoth plate panoramas, ranging in size from 10x12 inches [25x30cm] to 1.6x0.9 metres.

Collodion is alive and well!

Sat, 22 Aug 2009 10:36:26 +1100

I just came across this Youtube video submitted by modern day exponent of the collodion process, Quinn Jacobson (http://www.studioq.com/). Seems there is quite a vibrant collodion subculture in Europe and the US. There are lost of fascinating videos on Youtube about collodion processes and artists.

It gives you a real appreciation of the skills involved and the difficulties faced by Merlin and Bayliss (and the many other photographers in the field using wet plates - the American Civil War photgraphers spring to mind here). I can just picture them cramped in the darkroom wagon, trying to keep the dust out of everything, while they floated the emulsion onto the glass plates.

Resolving the image

Fri, 07 Aug 2009 12:22:35 +1100

As discussed in last week's post we have recently made important decisions on the Holtermann Collection digitisation standards. Although the library already has an established set of digitisation guidelines new considerations needed to be made for such an important and high quality collection of glass plate negatives along with the introduction of our new glass plate scanner.

We did not take this process lightly and so began by researching the digitisation practices and standards of our fellow cultural institutions, most notably the National Information Standards Organisation of America, National Government Archives of America and the National Library of Australia. We also looked at glass plate digitisation case studies, such as the Wellcome Library's John Thompson collection, to get a feel for the workflows and technical specifications used by other world class institutions.

Based on this gathered information we determined to perform a series of test scans in three different pixel dimensions (approximately 5 000, 7500 and 10000 pixels in length) to see how much detail the Holtermann negatives really hold and how well we could capture it. We performed these tests using the scanner's true optical resolution to ensure that no pixel interpolation occurred. This means the actual pixel dimension is closer to 11000 pixels and can vary slightly because the glass plates are not all exactly the same size. Below is an image we selected for test scanning due to it's high level of detail and sharpness showing the detail sections of the tests.

Once we saw the truly impressive amount of information stored in these negatives we were determined to save archival master files to our highest test scan resolution of approximately 10000 pixels on the longest edge creating a 450MB, uncompressed, 16 bit, Adobe 1998 RGB TIF file from each plate. Including our 8 bit, Gray Gamma 2.2 sub-master file derivatives that's nearly 3 terabytes of total storage space needed in total.