>>2692450 I picked up an epson v550 on sale recently and that's been working pretty well for me. I feel like I can get sharper results with dslr scanning, but it's a pain in the ass to set up, and manually tweak in photoshop to remove the color cast.
>>2692580 >>2692597 And have some 120 as well. Sorry, this isn't actually full res, the files are normally 9000x6000 ish. I had the wrong box checked on hugin for this roll. But I already had to use 70%jpeg to fit it on 4chan, so the smaller size is probably better.
But the real big deal about using a camera to scan is the quality of files. You can make ridiculous colour and contrast adjustments whilst still keeping relatively smooth tonal gradations, and without exciting digital noise.
And this is with one of the cheapest, worst DSLRs on the market, with a really strong AA filter and low DR and terrible noise and canon banding and only 18MP.
Oh yeah, and you scan a whole roll in like 5 minutes.
What I don't understand is why flatbedders and drumscanners et al can't get with the program. Those are expensive machines, they can almost certainly afford a camera and lens. If they're shooting MF film, they clearly care about their IQ. Don't they want better scans?
>the war is over, you lost >welcome your new rabal overlords
>>2692450 currently using a Coolscan 8000 and looking to upgrade to an Imacon so I can scan my 4x5 shit.
DSLRs or flat beds look fine if you are only posting to web but don't bother if you are printing
>>2692580 you do realize that your camera costs less than a single drum scan
No scanner can make up for resolution that is not there, if you are shooting kodak gold 800, my coolscan will out resolve it. What you get with higher end scanners is a better Dmax and Dmin, or better colour and gradation.
None of it really matters when it is posted to web as a JPEG and browsers do whatever they want with the file.
What you scan with only really matters when it comes to what you print with and what you print on and how large you print.
If you ever get a chance to stand in front of a Jeff Wall, you will know why drum scanning is still around.
>>2692664 Monitors have better dMax - dMin and gamut than prints do.
Scanning with a digital camera is a lot better resolving than with a flat bed.
Flextights have a lot of distortion and image warping problems, you don't normally notice this with most things - it is because of the "virtual drum" design (worst design ever) - scan a bit of film twice in a row without touching anything and you'll see the warping across the image change between the two shots drastically. It does provide good IQ though otherwise.
4x5 is limited to 2000dpi on Flextight scanners.
Make sure you get some kind of warranty, they can go bad, and need boards replacing, which are extremely expensive - despite just being a surface mounted 486 cpu and SDRAM.
Found the gif I made - one reason I wont buy one for myself, Imacons are not drum scanners by any means, the "virtual drum" is a marketing gimmick, and is actually a negative against the scanner.
Flextight's are just line scanners like flatbeds, they just happen to have a rather mediocre lens system in them (Magnagon 75mm f/8 enlarging lens iirc) that can resolve about 125-130 lp/mm decently off of film (not that hard for many enlarging lenses which would go even higher), as opposed to absolute shit in a flatbed scanner that can resolve a just barely/soft/blurry 25 lp/mm.
If you were technically minded enough you could build you own from off the shelf parts and line sensor and micro controller and processor, and cheap enlarging lens (or camera lens) inexpensively if you really wanted to.
>>2692729 Yeah buddy, I bet the minilab put that banding in the sky for you too.>>2692664 >>>2692580 >you do realize that your camera costs less than a single drum scan >>2692580 >>>2692569 >>inb4 some drumscanning dinosaur posts full res 135 scans that are better than mine >oh wait, that never happens... oh wow it still hasn't happened... >>2692664 >if you are shooting kodak gold 800, my coolscan will out resolve it. The plastic meniscus lenses in the disposables they load that shit into nearly out resolve it too. Put up or shut up m8.
>>2692885 There you go OP, you have your answer. 3 great DSLR scans, one screenshot of someone scanning single frames at a time on their potatotek who claims that the banding goes away in the final product, and one google images photo of a gallery exhibition of optically printed large format cibachrome transparencies posted by someone who claims they want to buy an imacon. >dslr scans btfo...
crop into the image enough so it's just your shot (no borders)
set your white balance on something that's as close to neutral grey as you can find
adjust each RGB channels curves so you top and tail the histogram for each colour
invert your main curve to invert the negative
adjust the levels of each RGB channel again paying close attention to colour shifts in the highlights/shadows
if you upload a raw I'll see what I can get and screen shot my settings. The pic attached is one I just did using JUST the workflow above and no other changes to the image except brightened by 0.3 and saturation at -10. I would usually then go onto my normal (inverted) lightroom digital workflow.
>>2692963 Gitzo Explorer tripod Arca-Swiss p0 head EvaCool "Hahn Super Dry" esky Godox Ving V680(?) Li-Ion flash, w/ bundled wireless trigger. (The trigger's normally on the scanning camera of course, but I put it on the GR to take this shot.) Piece of glass. Electrical tape on top the width of a film strip. A busted open film canister, so you can hold down the film and also wipe dust off it with the velvet strip. On the underside of the glass are a few layers of diffusion material. I have used cut up translucent plastic document wallets. They are taped so they're bent, and sit an inch or so back from the glass. You do this to keep them out of your depth of field, otherwise their own dust and scratches might get in your shots.
>>2693143 I've added alfoil around it to shape the light and make sure the illumination is even. That's probably to most important thing to get right in this whole system. I shot tethered, and shifted the flash heads angle, zoom, etc, as well as applying foil to mask off my diffusion panel, and tested it. Just open the last shot, jack up the contrast, and if a gradient appears, make an adjustment. I shoot at f8 and keep my flash around 1/16th most of the time. With the lithium flash, recycle is never a problem, even at 1/1, but there's no point roasting it unnecessarily. f8 minimises the visibility of sensor dust as well. I adjust my focus and framing using live view, and by racking the camera up or down on the long arca-style tripod plate I use and shifting the glass around. I line up the edge of the frame with the paper tab taped over the film channel. That makes it easy to just pull the film strip to the next frame, without having to line everything up in live view again. The slight natural curvature of the film towards the emulsion side props the film off the glass a fraction, eliminating the newtons rings you get when shooting through a glass sandwich to flatten the negs. Film not being flat enough is rarely a problem with Ilford or Rollei films. Kodak and Fuji curl a little more, but holding them down either side with the paper flap and the film canister works fine. I control exposure using the wireless trigger to adjust flash intensity. With the flash at full power, you can blast through any density of negative. That's about it.
>>2693152 To do 120, I back the camera away from the glass a little. I still use the taped guide, but I use the full width of it. Because I shoot 6x9, and I want to get good detail and big files, I stitch 4 shots. This is where you'll see the biggest payoff from making sure you've got even illumination from your diffusion panel. By framing to just inside the taped guidelines, I find I've got a perfect amount of overlap between the 4 shots. Making sure your glass surface is well dusted, your film is dry and your hands are clean is important, because you have to handle the film a lot more. Also, 120 can curl like a bitch. To process B&W, I use the Canon DPP raw converter, turn off all sharpening, desaturate, and adjust the Contrast, Highlights and Shadows controls to give me the flattest neg. Export as JPG, batch invert in GIMP. Edit as desired. For colour, I'll invert the tone curve as best I can in the RAW converter, and adjust the WB to almost right, then set it back to a negative. I will have set the in camera WB to custom, using a blank shot of the film leader, but you almost always need an additional shift. I apply that to the whole roll, as well as turning off all sharpening. I then tend to keep more contrast in the exported JPGs, because colour negs are lower in contrast to start with. Then batch invert in GIMP, edit individually. I don't use auto adjustments, or catch all presets for colour balance, because I don't want my photos to look like shit, but if you don't care, knock yourself out. If you've shot whole rolls in the same light, then you might be able to make a curve that you can apply to everything. and that looks right.
I'm currently working through a couple thousand old family photos, this is working pretty well-- $40 light table from amazon, lens hood, spacer made from balsa wood (popsickle sticks would work too), Remote trigger. Can do 200 photos in an hour with this setup-- No tape, no guides, the weight of the camera holds the film flat. Only downside is you can't batch-crop because they're not always in the same spot.
>>2693187 Yeah dude, this is fucked, for a number of reasons. Long exposures eating your battery, roasting your sensor (less of a problem on a modern mirrorless than on a rabal, but still) and increasing likelihood of unsharp shots. As you said, having to reposition for every shot. Fuck that noise, find a way to feed strips through. Also, you can pretty much see the unevenness of the illumination you're getting, as well as the *potential* limited spectrum of light from the LEDs (although I'm pretty dubious about how big a deal that is), as well as the inner workings of the device being too close to your plane of focus, and thus being in your shots.
I'm actually really happy with it, but it's the only one I've used. There's flicker that anything faster than 1/120 or so will catch, but I'm shooting at .6 seconds and it's working great.
>>2693211 You might be right, I haven't done much of this before and I pretty much pulled it out of my ass. That said, I'm doing .6 sec exposures which are definitely slow, but not battery-eating slow. It lasts longer than my motivation on any given evening. The camera is an NEX-5R which goes pretty cheap on KEH and has had the shit beat out of it anyway, so if it's damaging the sensor I'm not super concerned. Using an IR remote, camera shake hasn't been a problem at all.
I think a feeding system would be more fiddly than this method. Repositioning isn't bad at all, you settle into a move-shoot rhythm that only needs a couple seconds per shot.
Light evenness and color don't seem to be a problem, but I'm not that good at the post processing side of it yet so it may be that I just don't know any better.
I have an Epson V500 which I have hated since day one.
Pic related, my new scanner setup. Going to build a desktop rig with 2x4's to mount the bellow rail to and create a lightbox using my studio flashes. This setup will cost ~$3k so I wouldn't recommend it to anyone unless they already have the expensive parts.
>>2693389 The Alfa Romeo logo is oof also. GP3 isn't the sharpest film around, but it sure does look nice, anyway you can see the grain pattern is better resolved on the Flextight. Not that you could come close to this enlargement size using a single shot from a digicam, you'd need to enlarge more and stitch.
>>2693401 I've got a photo of taking one apart before, they're literally powered by a 486 (and I mean the 949 not older ones, which is also the X5 model iirc) and SDRAM, line sensor like a flatbed, and generic enlarger lens, though it does autofocus and adjusts enlarger head height for different enlargement sizes to match film size to full width of line sensor.
>>2693397 >>2693394 Agree with everything you said. FYI the Ferrari is a 9 shot stitch, from PanF+ in rodinal, which isn't a particularly fine grain combination at all. But the DSLR crop is from an area around a third of the size, and I'd say the only reason the grain isn't as crisp is the AA filtered sensor on the Canon. Scaled and sharpened, there's fuck all in it. Using a new high MP camera to scan with, it would be no contest.
I heard it was actually a bad combo, that that specific combo didn't play that well. Most films can be made sharper too with simple developer modification works in both solvent and non-solvent equally well (tried both Xtol and Rodinal)
>>2693411 And it's ultimately a pretty simple device.
Your dslr scan bullshit is so annoying. Just let people digitalize their negatives the way they want to. You can get a plustek for around 100 bucks and have a decent and uncomplicated scanning rig. Quality is fine.
>>2692450 I'm actually working on a design for a new scanner that should be commercially available in a few months, via kickstarter.
Attached is a crop from 120 film (mamiya 6, 50mm), the crop is about 5mm on the longer size. A close up is in next post.
It's kind of like the DSLR-approach; providing similar but slightly better results. The main advantage of this scanner is that it's super easy to use, is fast (about 1 sec/35mm frame for the quality of the crop shown here) , loads similarly to the Pakon scanner, and can do up to 120 or 4x5, and also should have outstanding results.
Would anyone here be interested in one of these? Probably would cost around $1200 USD.
>>2693445 I don't think you have your priorities straight.
most people who continue to shoot film, and are willing to spend $1200 on a scanner, would consider image quality (tonality, colour reproduction, range, detail, and sharpness to name a few qualities) to be an essential part of the photographic experience. this is especially true of people who shoot medium or larger formats.
from the sounds of it, your scanner is aiming for so-so quality (DSLR scan grade) but high speed. poor quality is in direct conflict with one of the main priorities of the film enthusiast. speed would be great (current scanners are slow slow slow), but given how slow the film workflow is anyways, why sacrifice quality to save a little time after the photo has already been taken? sure, time would be valuable to a commercial film photographer, but then again those shooters also require the highest quality.
to be fair, you have not yet produced any specifications for your device, so we cannot really comment on it. however, looking at your image in >>2693445 and assuming that >>2693446 is a 1:1 crop, then it looks like you're aiming for 250 points/mm? this seems rather unnecessarily high, and beyond the resolution of nearly all emulsions. going back to the samples, I see crushed blacks and blown out whites. the grain is also soft and muddy, with visible banding in the enlarged sample.
I'm happy to spend on photography, and have over $8k in camera gear alone. however, the gear that I buy must meet my needs, and to be honest, the current product you're developing doesn't excite me.
>>2693835 see >>2693171 >Set custom wb in camera on blank frame >In raw conversion, invert the tone curve so you can see most of the image as a positive >find the WB adjustment tool that lets you pick a spot on a colour wheel, and pick the spot that makes the colours look most natural >put your tone curve back to normal, export as jpg, invert properly in photoshop or gimp >Use curves adjustments to get your colours right. Usually take away some blue, add some red to the highlights, give green a subtle s-curve, increase global contrast a lot.
>buy super expensive imacon deluxe scanner >tfw no money for any film to scan
Have an ancient coolscan 4000, but that eats only 135 film, so the DSLR has to do for dealing with larger formats. Which made me think, I guess I should try feeding the DSLR scans to Vuescan and see if that would produce something slightly less terrible as a starting point?
>>2693944 Hey bruh I am on your side in regards to DSLR scanning.
I am >>2693367 and in the process of designing my rig. I do agree with other though, you need to post better examples and not really shit snaps. Your other thread has some good stuff though. I'm saying you're not going to convince anyone that DSLRs are superior with dark, low detail, and plain shitty snaps.
>>2693957 The orange mask is a mask, its called the integral mask, it is thinnest where there is plenty of exposure/density, and thickest where there is no exposure. It's self masking according to exposure, it isn't a flat orange colour.
>>2693958 This is a gay as fuck technical thread, I think snapshits are entirely fit for purpose. I won't be posting them anywhere else, so why not? Also, for the purpose of essentially bragging about sharpness and proving your colour reproduction isn't entirely fucked, it's better to find a snapshit that's on fresh film, plenty of dof, no motion blur, and plenty of exposure, and doesn't have people in it. So that reduces my pool of eligible shots substantially. I like to shoot trash film, you can't get very good colour out of bad film a lot of the time. Here is some fucked Portra 400; it looked pretty crap so I crushed the blacks, it's not going to convince anyone that it's the last word in scanning quality.
>>2693975 And by fucked I mean several year expired. And probably should have been shot at 200. I found that whenever I put the contrast where I wanted it, it was too contrasty, and when I put the skin tone where I wanted it, the jumper was over saturated, or the rocks were too blue. etc. It's a shot on a very cloudy dark day, of a contrasty shot with several competing colours, on daylight film. Almost any way you edit it gives some idiot ammo to attack your method, so why post it? Even though that's the very nature of shooting film, you can't just say that to people, they'll say their potatotek gives them perfect colour every time. It's better to make a staightforward example based on sharp shots on good film in easy light.
Here is SOOC Inverted, my actual exported raw inverted, and an alternative edit, with a little more blue and lower contrast.
>>2693988 I meant your comment didn't make any sense. Flat as fuck highlight boosted in post? Didn't use enough light? I flattenned the highlights in post, you drooling retard. There's still contrast in the specular highlights on the raindrops for fucks sake. Think about it, that means that the darkest possible parts of the negative had enough light coming through them to register above 0,0,0 in the processed jpg, and yet none of the shadows are clipped either. But you know, probably not enough light...
>>2694023 That looks like LF to me. What lense? And it looks more like noise than grain to me. Shadows on slide don't turn to rainbow speckles like negs do. This is 6x9 ancient Velvia (not Velvia 50), I've posted it before.
One of the things I always loved about colour neg was it's insane highlight dynamic range, even with the flash head next to the wall, hitting a reflective white tile, it still retains detail on the neg itself.
Behind the scenes shot from several years ago on 400H, Flextight 949 scan (just now).
>>2694068 >>2694081 I used to blame my lack of ability to push my negative scans without colors and banding going crazy to the poor range of the negatives, before I realized it's just my shitty V500 that can't make a file that can handle a push of more than about a stop.
>>2694429 I've posted this a dozen times too, but this is what I mean about the flexibility of DSLR files. The negative itself is practically clear where her face is, if you hold it up to the light you need to shine another light through it from the side to be able to see the trace of the image.
>>2694651 >Can you please adopt a trip or something. And then you'd "filter" me, to preserve the harmony of your experience here, cupcake? This is four channel, mate. The whole point is to post anonymously.
All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
This is a 4chan archive - all of the content originated from them. If you need IP information for a Poster - you need to contact them. This website shows only archived content.
If a post contains personal/copyrighted/illegal content you can contact me at firstname.lastname@example.org with that post and thread number and it will be removed as soon as possible.