Pixel Shift Scanning 2
Вставка
- Опубліковано 28 лип 2024
- A redo of the video because my previous version was not well done.
Philip Bloom's Video on The difference between sharpness and detail
• The difference between...
The Gear
Camera Bodies - kit.co/Patrick.Doyle/camera-b...
Nikon Lenses - kit.co/Patrick.Doyle/nikon-le...
Photo Accessories - kit.co/Patrick.Doyle/photo-ac...
You don't mention it, but to me it looks like the color of the pixel shift scans are a lot better
Thank you for the incredible amount of work that you put into this video. I appreciate it a lot. Makes me wanna scan even more with a newer camera that offers pixelshift. This video confirmed my choice of buying a new camera since my 35mm scanner broke down xD.
Love your camera scanning mount, I built one similar a couple of yrs back, also out of aluminum extrusions. Regarding the lower apparent noise in the pixel shift scans that you observed, downsampling does increase your signal-to-noise ratio (SNR) because digital noise follows a normal distribution across the sensor, and gets averaged out during downsampling (or binning). IIRC, you get roughly a 2x decrease in noise floor for a 4x reduction in resolution. The noise/grain problem is also worse with negatives in part because gradients get stretched out during inversion. The other major source of noise in camera scanning is heat from the sensor itself, which is hard to eliminate unless you use an astrophotography camera with active cooling like I'm attempting to do.
As you so rightly said, the diminishing returns you get (in some circumstances) with pixel shift may make the whole process less worthwhile.
Perhaps attempting to correct the flaws inherent in single image scans, especially with smaller formats, may be an ok compromise. I think I just said the same thing twice... in the most verbose way possible.
Love your channel, thanks for the great work. 📸
I'm not sure the application of equal amounts of sharpness to the one-shot version and its pixel-shift counterpart is relevant in regards to the 35mm film images. In most of the "Unsharpened" slides - the exception being the seemingly out of focus candid of the woman with the soft drink - the one-shot versions don't seem to need any additional sharpening; the pixel-shift versions do appear to require more. It might be helpful as far as the 35mm images are concerned in the "Sharpened" slides to only increase the sharpness of the pixel-shift captures to the same level of apparent sharpness of the one-shot scans in the "Unsharpened" slides. The difference in detail may seem less in the examples made with lower ISO films or larger media sizes. However, it is difficult to determine because the zoomed/cropped portions of some of the pictures are not of the sharpest part of the image. For example, a portion of the wall supporting the woman with the drink is actually the sharpest part of the image. Later on in the medium format pictures, the focused part in the cherry picture is a green leaf to the lower left, not the section with the bulk of blossoms.
I appreciate you taking the time to make this video. Having just inherited thousands of 35mm and 110 slides and negs from my family, I am doing the same experiment, for archival purposes. Just bought an A7r4 specifically for this, a laowa 2:1 magnifying macro lens, & did my first couple 35mm slide scans. I'm at first very amazed at the detail that half a century old slides manage to contain, but not so sure I'm going to keep this expensive & complex rig for this task. Perhaps the clunky, slow plustek scanner is fine.
Eldrich scannin'. Thanks for doing this, I was looking for many answers you did provide.
I agree with @ikorin - based on my monitor, the pixel shift images appear to have slightly more grain and less detail. Perhaps this is a UA-cam video issue and the images you see are different. Regardless, thank you for this. The pixel shift feature has been something I've wondered about when scanning film. Thanks!
regarding the eagle photo comparison: on the video I prefer the photo on the left and I do not see more details on the right one. Anyway thank you for the video. I think the right conclusion is to switch to large formats when details retention is priority
My conclusion as well, but if you do not see the difference between an eagle and a seagull, your powers of observation become suspect. My personal evaluation of "pixel shift" is that it is a lot of technology looking for a reason to exist, and got lost along the way.
I’m already half mad. Sounds like this would be right up my alley 😂
6:40 the picture on the left has more color contrast. Next time just take 4 images and stack them on top of eachother. It will be less noise and much clearer and better
Hi there! I absolutely love your channel! I currently scan my negs/slides with a Sony a7R IV (using pixel shift) and a Voigtlander 110mm f/2.5 APO-Lanthar Macro (for medium format & 4x5) and a Printing-Nikkor 105mm f/2.8 "A" version (for 35mm). I’m very close to getting the GFX 100S II, hoping that its lowered native 80 ISO would improve my scans.
I’m currently using the same Albinar copy stand that I think I saw in your earlier setup (and it’s a pretty terrible copy stand to be sure). That said, I did want to ask you, if I may: Where did you purchase the parts to make your incredibly well-built, ultra-sturdy copy stand? Are these parts I can get off, say, Amazon? Additionally, I wanted to ask, in your opinion, if the Rodenstock 2x enlarging lens that you use in this video is the best lens I can get for scanning medium format & 4x5 negs/slides. I’d love to hear your take on the matter.
Anyway, congrats to you on you channel and all the great info you put out. Truly terrific and informative. I’m glad that I’m not the only one who geeks out on things like this. 🤓
All the frame parts are 2020 aluminium extrusion cut to length and bolted together using standard 2020 plates and fasteners. I bought all the parts from amazon. The film carriers are all negative supply but I think you could bolt the valoi one to it and make it a fair bit smaller as well.
Hello Patrick. Thanks for doing this - it's so good that you put this video together.
I've some issue with sony a7riv pixel shift and negative lab pro. When I do 4 image pixel shift on colour negative scans - it throws off the ability for negative lab and adobe light room to do colour adjust and white balance. Image comes out all colour shifted.
I've no issue with the same camera and negative lab pro with a single, non pixel shifted image, from the same scan, camera and light setup. What do you think it could be?
Try converting the pixel shifted image to a tiff and then using negative lab pro.
@@SprocketHoles thanks. I did and didn't work. For some reason using the sony app was the cause. I used a different converter, different app, and it works now.
I wonder how many situations the shift mechanism has before it fails (akin to shutter life) and can it be fixed.