[butangmucat]

So I am spending my summer back home in China, and I have brought some film and my ME Super with me. I am in the process of finding a good photo lab. I have just had my first roll of Portra 800 returned from one of the candidates. That lab used automatic C-41 minilab and Fuji SP-3000 scanner, and returned TIFF files. However, the scans of my Portra 800 seems to be off in color. I am posting the originals and the best-effort adjustments I have made in DxO for some opinion.

Original:


Adjusted:


Original:


Adjusted:


I am also looking into another lab, and have sent them a roll of Provia 100F waiting for results. One of their advantages is their choice of scanners (obviously some of them are more expensive to use). However I do not know any of them, and I also need some opinions:
- Fuji SP-3000 (cheapest, 3000x2000 JPEG or 4600x3050 BMP)
- Noritsu QSS3233 (same price as Fuji, 3000x2000 or 5000x3350 TIFF)
- Plustek OpticFilm 120 (decently expensive, unknown format, probably TIFF)
- Hasselblad X5 (pretty damn expensive, 3000x2000 or 4400x3000, FFF raw format for negatives or TIFF for slides)

Sincerely

[pathdoc]

Maybe the train really was that spotlessly, brilliantly white, but in the second photo your efforts to correct it have produced an over-processed feel that I do not like. While it's nice to have a C41 mini lab that gives you large TIFF files to play with, IMO you should have left well enough alone. What I'm discovering about scanning even B&W film is that any post work you do on the digitised image of a film negative becomes very obvious very quickly, and the results are often not pleasant.

[Theov39]

I am not familiar with those particular scanners as I use a discontinued Nikon IV but regardless of which scanner you choose, a lot of it will come down to the skill of the operator and the particular scanning software used.

In my limited experience, feed type scanners (where the film is fed into a slot as in the Nikon series of scanners) generally give better results then flat bed scanners but then flat bed scanners are much quicker to use so it's a trade off.

Also, each scanner, like film and sensors, renders images differently so it is not so much a question of which is the 'best' scanner but of the 'look' you prefer. I suggest you ask them to make a scan on each and see which one you like best.

This article discuses different scanners, including the Plustek OpticFilm 120 and the Hasselblad X5, and their options which might be useful to you:

https://www.bhphotovideo.com/explora/photography/buying-guide/film-scanners?...BID=6867&SID=f

[edmundrt]

The slightly warm tone of the lab scans looks about right for Portra 800 to me. Fuji Superia 800 will give cooler colour rendition if you need an 800ISO film.

[LesDMess]

All Kodak Portra films are intended to be less contrasty - closer to flat. However, since there are no standards in scanning and with post processing at our disposal, results can vary greatly. I say go with which ever one is more pleasing to you.

[TonyW]

I believe as already state the Portra film is very likely to be warm as it seems to be designed for what Kodak claims to be 'beautifully natural skin tones' - whatever that is supposed to mean. Therefore not unreasonable to expect this warmth to transfer to white/off white objects.

Most labs will aim for a pleasing colour (to them) in most cases if there is not a reference such as a colour chart. Even then it is not always desirable aesthetically to reproduce the chart as if it was in a standard lighting condition with all colours reproduced as such.

I also agree that I do not care for the adjusted rendition particularly in the second shot as the trees have a flourescent look to them (viewing on a calibrated monitor covering most of Adobe RGB). Still that is a personal preference and if printed may look quite different

I feel that the lab has provided you with a pretty good starting point for your own correction preferences and TIFF a bonus. Your choice of editing software will also influence the look of starting point images in many cases as will having a colour managed workflow including a calibrated and profiled monitor. Without such you are really making guesses and what looks good on your system may not be such on another's either calibrated or not. Scanners like cameras, monitors and printers can be calibrated and profiled in this case with an IT 8 target, however whether the lab you choose undertakes such profiling may not be clear.

I am not that familiar with the scanners you mention but would suggest:

1. Should be the Hasselblad X5 (Flextight?) as raw from negs. TIFF ok for slides. A pro piece of kit
Test report drum scanner Hasselblad Flextight X5: Image quality, scan speed, measured values, image samples, comparison with Nikon Super Coolscan 9000 ED

2. Fuji and Noritsu. These designed pretty much for mini labs

3. Plustek OpticFilm 120. Designed and priced for the home/amateur user. Quality OK but not in the same league as the now discontinued Nikon or Minolta scanners.
Plustek OpticFilm 120 film scanner: Detailed review about the medium format scanner and 35mm format scanner

Note: The result in any case will also depend on the skill of the operator

[Alex645]

If you really wish to shop labs and scan quality, Iʻd recommend that you send to those labs the same film type (preferably the film youʻre mostly likely to use) and the same images. That can be problematic if you donʻt have two cameras with the same two lenses, but even around your home or apartment, you could go through a routine where you shoot some interiors under tungsten or fluorescent lighting, same with flash, go outside and shoot in the shade, then in daylight, pick a high contrast scene, then shoot a bowl of fruit or a bouquet of flowers, and last close ups with skin tone. Then rewind your film, and before the daylight shifts, shoot the same in reverse sequence.

Now you can truly compare the results from different labs. The lab with multiple scanners should be able to tell you what are the pros and cons of getting your work scanned from one or the other.

In general, I concur with what others have noted. To me, the main distinction of each in order from best quality to least:
- Hasselblad X5 (pricey and slow, but the best optical scans of the bunch)
- Noritsu QSS3233 (During the minilab era, Noritsu was the best vs. Fujifilm, Kodak, Agfa, etc. Not cheap, but very fast, and thus less labor cost passed onto the customer.)
- Fuji SP-3000
- Plustek OpticFilm 120 (agree with TonyW. More of a prosumer model; Nikon and Minolta clone but disappointingly not an equal.)

The biggest factor, as has been stated by others, is the operator; the person running the scans. Sloppy work on the best scanner will look worse than TLC using the worst scanner. Meaning, on any given day, the same lab could give you different results depending on the operator.

One of my first jobs was managing a One Hour Lab. Although we had state-of-the-art equipment, the person behind the machine could produce junk or gems. If you get great results, take the next step and ask the lab who did the work; you want to send them a thank you card. Develop a rapport with their manager or that employee. They will know you care and know you know itʻs their work.

[EssJayEff]

I recommend becoming familiar with the publication Technical Guidelines for Digitizing Cultural Heritage Materials
Creation of Raster Image Files produced by the U. S. Federal Agencies Digital Guidelines Initiative (FADGI).

[John Poirier]

I'd say that the colour balance of the original scans is quite decent. I'll discuss the third and fourth scans from the top in the original post.

Yes, the white of the train is not pure. It may, however, be accurate. Note that adjacent to the train there is a large area of warm-tone tile. It is quite possible that the illumination from the skylight to the left of the train was bouncing off that tile and imparting some warmth to the colour of the left side of the train. With a subtle effect such as that, given the way our minds make assumptions about colours, it is entirely possible for a viewer to perceive that something is white when in fact it is slightly off.

Nothing wrong with correcting that colour cast in post processing. It is unlikely that even a very competent scanner operator in a mass production situation would make that sort of minor adjustment as it is better done by the photographer to suit his or her own tastes.

I agree with others that the green bushes in the processed version of the image are excessive. In Photoshop, I would correct this using the hue/saturation function. Select the green in the foliage and reduce saturation.

[butangmucat]

Got my Flextight X5 scans back today. Looks descent. The original scan was four uber-large TIFFs containing multiple frames. I was able to cut them out with Affinity Photo and so some minor color adjustment.

Wuhan Station by Sijie Bu, 於 Flickr

Wuhan Station by Sijie Bu, 於 Flickr

Wuhan Station by Sijie Bu, 於 Flickr