OT: (maybe) Photo Scanning with iPhone

decent computer and EpdonV600 photoscanner.

Any cheap USB scanner will be more than adequate. I use a Canon LiDE 400 about £60.

And even a 1200dpi negative scanner is cheaper than an i- phone

I needed a setup to scan lots of my old slides, and also those of a club. I have a Canon compact camera, and I took a cardboard box and made a position for the camera, and a holder for the slide, in exactly the right place for the camera in Macro mode. The bottom of the box is clamped between the camera and the tripod base, then I point the camera at my large window, and at a piece of white perspex which sits against one of the glass panes. A light cloudy sky is the perfect backdrop. I have a USB extension cable plugged in, and after however many shots I want, I just transfer them directly to my PC. Total cost: Nil (I already had the camera, the box, the tripod and the perspex).

I am not worried about minutely correct colour rendition, and after

This might not be suitable for your requirements, but, hey, it is the d-i-y newsgroup!

Yes. I have been agreeably surprised at the quality of scanned photo prints. Epson ET-2650. A6s come back good as original on photo paper. You can tweak the colour/contrast settings as well but I guess that may be available on the iPhone set up.

For a tenner or so you can get an A4 light box / tracing board from eBay, then you don't need the perspex, window, and suitable sky.

Very handy for pre-sorting slides or negatives.

I've just resurrected a Dimage slide scanner that is no longer supported by Windows, by spending £50 or so on VueScan x64 software. A nice thing about that is that the software also works on my Canon Lide flatbed and Brother "all in one" printer/scanner.

Thanks for the input. I hadn't actually considered a dedicated scanner, as my existing scanner can't touch the quality of iphone pics of photos. Then again, maybe I've just got a crappy old scanner (Canon MP640) and things have moved on in terms of quality. I'll take a look at scanner too.

If you can scam larger photos, its better.

My canon goes down to 1200dpi, but takes forever. I scan 120 sized negs at 600dpi and that is a very good resolution for then - as good as the 1943 film I am scanning.

I will be digitising most of my 35m,mm at that res or maybe 1200 for the really good stuff.

This beast

Here is a sample of a 'half 120' sized negative at 600dpi - you can see the film grain.

Its expensive because it has a built in lightbox and will scan larger than 35mm negatives.

If photos are all you want 1200dpi usb scanner will be fine if fairly new.

I had an aged (free) canon Lide but the color was all over the place across the bed.

What I have suits me fine, and cost me nothing at all.

Not a requirement, in my case. The slides were all in cassettes or albums, and all were to be scanned.

I have a flatbed scanner which is mumble mumble years old now, and works perfectly with Linux Xsane imaging software. Also free!

Diff'rent strokes for diff'rent folks.

Thanks - The Hurricane pic looks great. The V600 looks interesting, I'm gonna take a closer look at that.

How old is that? I only ask because it appears to use USB2, so it is no surprise it is slow. Years ago (2002) I used a Perfection 2450 with Win XP. Fortunately, my computer had a Firewire port and I could use that as the 2450 also had one. Scans were fast. Many years later I used a Win7 laptop with a faster processor, but no Firewire port. I had to use USB2 to connect with the scanner, and couldn't believe how much slower it was. Of course, the software was different too, but I wonder how scanners with USB3 ports compare. Are they much faster?

There used to be a stand called Giraffe that held the phone over the picture or text for scanning. As for apps for photos, don't know I was mainly interested in accurately scanned text so we could read printed matter with the voice dream scanner app. Brian

Scanners should be sensor-limited.

The stepper "dwells" on a row of pixels for time constants in the milliseconds. The sensor has a row of pixels as wide as the bed (as in 1x8192 or 1x16384 pixels). Those are read out. Then the stepper advances to the next row.

Some more expensive scanning devices (likely not all that suitable for photographic work, more for document scanning), they use a camera and take a single picture rapidly, and they can "scan like the dickens". A Fujitsu business class machine, as tall as you are perhaps.

But traditional single-pixel array sensors need enough exposure time for the sensor to read one row, and then you're lucky if the data rate is much higher than 2MB/sec.

You can tell from the humming/whirring noises and the existence of a belt visible inside the unit, that they're based on a single pixel array.

Inside some scanners, they have functions such as color correction. There could be a small micro in there (8085 class, not a quad core ARM or anything). That might be an excuse for a slow result.

The way the scan is done, also differs with document feeders. The stepper in the bed stops moving, and it has a "parking" position. The paper is whisked through the bed area by the automatic document feeder. Some implementations do their scan at a different rate, when comparing the two methods. The "moving the sensor" scan, just reading whatever you place on the bed, might be a higher-quality run.

But generally, unless the sensor is a 2D array or is an entire camera, there's not much chance of USB2 rates being exceeded. With a 2D sensor of decent dimensions, you might be able to have the stepper do "large jumps" and overlap the chunks and weld together before transmission. I've not heard of nor read any accounts of devices using such a 2D idea. The sensor is normally 1D (a single row of pixels).

And if you look at the deployment velocity of USB3 in webcams, the scanner business, which is smaller, probably can't afford the engineering effort. The single-chip solutions of the past, came from places like National Semiconductor (bought out). There's more than one provider of scanner chips, but still, the engineering resources might not be available for "innovation". And the company making the scanning chip, is at the mercy of the company making the sensor, to make forward progress.

Paul

How do you keep control of the light levels and light purity ?

A scanner usually goes to some degree of trouble to make a stable light source, a light source with a known set of color bands, so that the scan can be color corrected.

Both CCFL scanner tubes and LED-based light sources, have "spikes" in the optical output. The scanner designer works with this light source and its quirks, to give the good results you see. If the light source is not "hot enough", you may see some problems in the appearance of the output.

The camera has some of that. The camera could use its flash unit, but I don't know if the overall effect gives as tight control as a *good* scanner.

For negative scanning, good scanners use ICE, a second scan in the infrared, for removing defects in the scan.

For scanning print paper, you likely wouldn't need that, unless the print paper was physically compromised in some way.

There are a variety of IT8 scanning targets for sale. I don't know if they make any of those for film (like Kodachrome 100 maybe). I have one target here, but don't know anything about using them. The target was provided in the scanner box.

Paul

When you hold the camera in a fixed position and do some sort of macro capture, you might benefit from applying a correction for lens effects. Barrel distortion or whatever. So your rectangular buildings don't have leaning corners.

The 1-D scanning array in a conventional $100 scanner, is immune to that. As the scanning process really is "naturally rectangular".

Paul

I only ask because it appears to use USB2, so it is no

Yeah, its USB 2, but I am not sure that that, rather than the scan process itself, is the limiting factor

another problem that I have noticed with camera "scans" is that the colour balance gets altered, Forms sent to me from black on white originals end up witn a markedly grey background. Doesn't happen with a scanner.

On 11/06/2020 08:25, charles wrote: <snip>

Many of the scanners I've used over the years give a grey background when using greyscale if the option to automatically adjust tone is not enabled (or if the tone curve is "wrong"). A but like cameras with the wrong white balance.

Your comments are confirmed by

When I've got a few spare hours I'll try it with this laptop. I'll have to get some Linux software for the 2450 (maybe here?