Jump to content
Sign in to follow this  
Alan7140

Osterley, Central Highlands, Tasmania

Recommended Posts

OgTidzP.jpg

 

Obviously a 19th Century farmhouse was once attached to these fireplaces, but beyond that I have no details.

  • Like 2
  • Applaud 1

Share this post


Link to post
Share on other sites

Reducing what was a 100MP stitched copy of the negative (10,200 x 10,200px) to 1480 x 1480px for the Web does hideous things to IQ, what with film grain being random and digital images being a regularly arranged grid, trying to replicate hundreds or even thousands of grains into one pixel even at full resolution, let alone further reducing that by nearly 7x for posting.

 

There is a mismatch there which resampling algorithms based on reducing digital images cannot reconcile, although I'm not seeing any artefacts as such on my monitor other than a general roughness of "grain" in the sky which is simply invented grain caused by the mismatch which simply doesn't appear at all in the neg. The film was Agfapan APX 25, which for all intents and purposes is grainless when printed conventionally on bromide paper, or even when printed large in inkjet without downsampling that native 100MP "scan" resolution.

 

The sky and clouds are smooth in tonal transitions on the neg, even when viewed under 15x magnification (i.e. 825 x 825 mm print), the background and foreground were out of focus as I shot this with a Pentacon Six using a CZJ 180mm lens wide open @ f/2.8, focused on the leading edge of the middle fireplace.

There is perhaps also a lack of crisp sharpness as I shot this hand-held at 1/125 sec which might be pushing my luck a bit, but with 25 ISO film this is what can be expected. Next time I'll carry a tripod, although I'm down to my last 5 rolls of APX 25, and there's no more where that came from, leaving 50 ISO Pan F as the slowest B&W film available.

Share this post


Link to post
Share on other sites

You're tempting me to try stitching to get higher resolutions of the film scans I'm working through  - but then again I could also switch from the Nikon set-up I'm still using for that to the Fuji (I just need to decide what to do about a macro lens for the Fuji - adapt the old F-mount Tamron; the Fuji 60 with extension tubes; perhaps even just extension tubes on the 18-55 or do it properly with something more expensive). 

Share this post


Link to post
Share on other sites
12 hours ago, crowecg said:

You're tempting me to try stitching to get higher resolutions of the film scans I'm working through  - but then again I could also switch from the Nikon set-up I'm still using for that to the Fuji (I just need to decide what to do about a macro lens for the Fuji - adapt the old F-mount Tamron; the Fuji 60 with extension tubes; perhaps even just extension tubes on the 18-55 or do it properly with something more expensive). 


The key to what I'm doing is the Sigma sd Quattro-H camera - there is no interpolation with adjacent pixels when shooting in B&W mode as there is with Bayer or X-Trans. That means that every pixel is indeed an individual recording of its unique part of the image, which in turn makes stitching accurate as the program can match individual pixels accurately. The lens is less of a concern, more importantly was to get a lens that had a wide enough image circle to utilise the Hartblei P6 shift->M46->Sigma SD adapter and flat field rendition which the Zeiss Jena 2,8/120 medium-format lens with 1.2:1 extension tubes supplies perfectly for these 100MP scans.

 

Using this setup with an LED panel stuck to the copy board and a Durst 1200 Laborator Neg holder on top of that means that nothing changes between the camera sensor and the neg during the exposure - one simply shifts the adapter after taking an initial centre-of-neg shot between 8mm and 11mm (as appropriate), and then it takes a further 9 exposures rotating the adapter to appropriate click-stops around its 360° circle and is perfect for 6x6 negs. Focus never changes during this, nor for subsequent negs, so after a bit of practice the procedure becomes very quick indeed.

 

The slowest part of the procedure is Sigma's notoriously slow SPP raw processor, but as everything is constant I just set the processing to batch with a custom preset and get on with something else while it trundles away in the background. It takes PTGUI approximately two minutes to add the converted TIFF images, crop, stitch and save the pano. I only invert to positive after all this is done and I'm cleaning things up in Photoshop (film reintroducing the joys of dust and micro-fibres on the negs requiring spotting).

 

All that said, for my usual copy setup I use my X-T2 and Zeiss Touit 2,8/50M Makro lens, for larger originals I take segments by shifting them around the copy-board and stitch them for a higher resolution end result. For any original A5 or smaller a single 24MP exposure is usually more than adequate, though.

 

Here's a picture of the final setup, the lens at rear on the copy-board is a Sigma 70mm macro which I use for single-frame proofing of negs - it's quicker to use as it couples with the electronics of the camera.

 

QfdCCkF.jpg

Edited by Alan7140
  • Like 2

Share this post


Link to post
Share on other sites

You make my set up look decidedly Heath-Robinson.  

 

With just 35mm film to work with I’d only be combining two shots per frame unless I go past 1:1, which would need some equipment upgrades.  I’m not yet sure if I’ll even be pushing a 1:1 macro with my Fuji setup yet.

Share this post


Link to post
Share on other sites
1 hour ago, crowecg said:

You make my set up look decidedly Heath-Robinson.  

 

With just 35mm film to work with I’d only be combining two shots per frame unless I go past 1:1, which would need some equipment upgrades.  I’m not yet sure if I’ll even be pushing a 1:1 macro with my Fuji setup yet.

 

It depends on your final use, of course, but I've found 1:1 barely adequate (or even totally inadequate) for prints of around 12x16" and above (depending on the film stock - it messes up super fine grain film like Agfapan 25 more than Tri-X due to the effect described below).

 

1:1 can sound like you're getting a true representation, but film grain can often be much smaller than a pixel, or crosses two or more pixels without "filling them up" with the tone they represent, being that film grain is irregular both in size, shape and distribution, which a sensor array is not. While the eye can't really see anything other than the overall image both microscopic methods show unmagnified, what does happen is that even a 1:1 digital copy of a film frame tends to effectively amplify the size of grain when trying to figure out how to interpret these random tiny objects smaller than a pixel or irregularly crossing two or more pixels. So the digital file looks grainier.

 

This effect often produces online comments about my photos being grainy, something accentuated when further reducing resolution for forum use, and all I can respond with is "but you should see the print...". Files that look grainy at 1:1 digitally copied look less so when stitched sections at twice that magnification are produced (so far I've peaked at around 100MP which puts out a file in the order of 1GB - I'd have to switch to a large format lens as I've reached the limits of the image circle with my medium format lenses to get a whole neg at that resolution). Fact is, though, if I printed a 16x20" print from the neg onto photographic paper, it will appear a lot less grainy than one of these 100MP stitched copies, and probably even a 150MP or more copy.

 

At least that's what I figure seems to be happening - I obviously don't have the scientific instrumentation to actually clinically confirm this theory.

Edited by Alan7140

Share this post


Link to post
Share on other sites

I’m seeing ‘grain’ effects on some of mine, even at only around 12Mpixel.  But that is with 800 speed consumer grade negative film.  

 

I’ll probably not gain too much improvement beyond 1:1 as I’ll run up against sharpness limits of my older lenses.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×
×
  • Create New...

Important Information

By visiting this website you are agreeing to our Terms of Use, Privacy Policy & Guidelines.