When should I see the difference between JPEGs and TIFFs in film scans?

Thanks RJ... I noted in one of the posts that I had been using grads for much of the time on the west coast, but might not have been for those particular frames.

But my point in introducing that shot was more in the context that the differences visible between the different scans of the same (admittedly difficult) frame are MUCH greater than any differences I have seen so far processing a JPEG vs a TIFF from the same scan. I could find other frames that I have scanned with the same scanner, but with Silverfast SE 6 vs Vuescan, where again the scans are chalk and cheese. So, for my purposes and workflow, I am seeing no benefit in getting the TIFFs, even when "free" (I pay in significantly greater download times, and in HD and backup costs), so I'm now only getting JPEGs. ISTR that UKFL (and presumably now CFL) never even offered a TIFF option.
 
ChrisR said:
Thanks RJ... I noted in one of the posts that I had been using grads for much of the time on the west coast, but might not have been for those particular frames.

But my point in introducing that shot was more in the context that the differences visible between the different scans of the same (admittedly difficult) frame are MUCH greater than any differences I have seen so far processing a JPEG vs a TIFF from the same scan. I could find other frames that I have scanned with the same scanner, but with Silverfast SE 6 vs Vuescan, where again the scans are chalk and cheese. So, for my purposes and workflow, I am seeing no benefit in getting the TIFFs, even when "free" (I pay in significantly greater download times, and in HD and backup costs), so I'm now only getting JPEGs. ISTR that UKFL (and presumably now CFL) never even offered a TIFF option.
Click to expand...

Well, if you're like me and you don't edit your photographs much, then JPEGs will typically the way to go. For me, TIFFs ordinarily represent a huge waste of space.

That said, in an example like this that features an extended subject brightness range and no grad filter, even I would be going for a TIFF if I had the choice.
 
I thought I'd revisit this thread, given that Filmdev are now offering 16-bit TIFFs rather than JPEGs, on request. I did get TIFFs for 3 recent films, as an experiment, and the zip file was 2.4 GByte rather than the more usual 300-400 MByte; the individual TIFFs are 38 MByte rather than around 5 MByte for JPEGs, so the difference in storage is very noticeable (over 4 GByte for 3 films vs half a GByte, even if I ditch the original zip files). I have a 500 GByte hard "drive" (SSD) with about 130 GByte remaining, so I'd be in disk space trouble after about 100 more films (< 18 months!).

So the "cost" of 16-bit TIFFs is even more acute. I came to the conclusion last time that, for my usage, there was no discernible (as opposed to theoretical) advantage of 8-bit TIFFs over JPEGs. I'm inclined to that opinion again. However, for my most recent film I did ask Filmdev for TIFFs as well as JPEGs, so I can do some more tests. Last time (see posts 24 and 29) this was the test I did:

"I've done 10 cycles of importing an image into Affinity Photo, making a slight adjustment to brightness using an adjustment layer (up and down by 3% on successive rounds), and exporting (using default settings), then re-importing the last version. First for JPEGs, then for TIFFs. Then I have imported the 10th version into Aperture and compared with the original JPEG and TIFF.

"The first thing to note is file sizes. The JPEG from Filmdev was 5.6 MB; the first save from Affinity Photo was 8.3 MB, with each successive version being slightly larger until the tenth was 9.7 MB! Meanwhile, the TIFF from Filmdev was 17.7 MB; the first save from Affinity Photo was surprisingly 14.6 MB, and each successive version was either 14.5 or 14.6 MB.

"Comparing the 4 images visually, the only discernible difference was that the 10th TIFF save was slightly brighter. I'd put that down to a slight error in my brightness adjustments. Looking at 400% (!) at a few selected areas of the image, I can't really see any differences; certainly no obvious artefacts.

"Comparing the histograms, there are very slight differences in each. The differences in the first and last TIFF histogram are, if anything, slightly more obvious than for the JPEG case, but this is mostly down to a slight uptick at the bright end."

Was that a reasonable test? Can anyone think of a better test?

So far I've imported all the JPEGs into Aperture, but only one of the TIFFs (the "missed the cut" image of horses under the trees). I used the Aperture Lift/Stamp tool to apply the same adjustments to each, and there is only the slightest discernible difference in the upper mid tones of the histogram.
 
Except for the slides that got a permanent holiday at UK Film Lab you do have, of course, what is much better than a tiff or jpg, the actual slide/negative. The advantage of a lossless compressed file is that some arbitrary algorithm has not already thrown away what it thinks unimportant. Editing and resaving over the same file will affect the quality of the original file and more so where lossy compression is used. Your workflow should always aim to avoid saving over your source file. The main advantage of scans to TIFF (or PNG) is where the data is stored with greater precision using more bits per channel - so 16 bit grey (65,536 shades of grey) or 48 bit colour (281 Trillion colours) instead of 8 bit grey (only 256 shades of grey) or 24 bit colour (16 Million colours).
 
dmb said:
Your workflow should always aim to avoid saving over your source file.
Click to expand...


So much this! Harder to do with film (not sure it's possible to overwrite a slide/neg [accidentally!]) but all too easy with digital files. I avoid it by keeping a second folder in every folder which I call Originals and that file is sacrosanct - no edits allowed! Memory is now so (relatively) cheap that a few extra MB (or even GB!) of used space is worth it for the extra security of the files. Of course, an extra layer of offline backup isn't a bad idea either...
 
Nod said:
So much this! Harder to do with film (not sure it's possible to overwrite a slide/neg [accidentally!]) but all too easy with digital files.....
Click to expand...

I've done that with single frames and even whole 35mm films in camera, which is why I write date and camera on my film leader with a Sharpie before loading, also helps if I need to pull/push or throw in the bin.
 
dmb said:
Except for the slides that got a permanent holiday at UK Film Lab you do have, of course, what is much better than a tiff or jpg, the actual slide/negative. The advantage of a lossless compressed file is that some arbitrary algorithm has not already thrown away what it thinks unimportant. Editing and resaving over the same file will affect the quality of the original file and more so where lossy compression is used. Your workflow should always aim to avoid saving over your source file. The main advantage of scans to TIFF (or PNG) is where the data is stored with greater precision using more bits per channel - so 16 bit grey (65,536 shades of grey) or 48 bit colour (281 Trillion colours) instead of 8 bit grey (only 256 shades of grey) or 24 bit colour (16 Million colours).
Click to expand...

Good point about the negatives/slides being the ultimate authority, although I'll add something in reply to Nod.

I do get the point about the "arbitrary algorithm [having] already thrown away what it thinks unimportant" in relation to JPEG, though I think in posts earlier in the thread it was pointed out that just this happens when you go from raw scan data to any representation (ultimately, TIFF as well as JPEG). So the purists will be keeping their "raw" files, rather than TIFFs.

I think I showed in posts 24/29 that, for me, a large amount of edit/resave surprisingly did NOT produce a noticeable difference between JPEGs and 8-bit TIFFs! In my normal workflow, however, files go straight into Aperture (for now) and never get altered except on export (so one read, one write). Only a very small minority of files get sent to Adobe Photo and come back in, but they come as TIFFs or PSDs, and again don't get changed until final export (one read, one write as TIFF, one read as TIFF, one write as JPEG).

(There will of course be a real issue when I move from Aperture to whatever comes next, and I will probably have to double (treble?) my disk space by exporting originals as well as "final" edits... though I must say that edits are rarely final if the file is interesting, when I come back to it I usually find something I want to tweak a little bit!)

I think I understand the point in your final sentence, that doing significant edits on ANY 8-bit data (JPEG or TIFF) is more likely to produce posterisation and other effects than with 16-bit data. I think I've noticed this posterisation on occasion. Oddly the examples that stick in my mind occurred when scanning with Silverfast 6 SE Plus, inside the scanning program itself, which should have had access to the full data bit depth. But perhaps I'm misremembering that!

Anyway, I'm particularly interested in what sort of tests might reveal whether octupling my disk space consumption will give me noticeable benefits!
 
Nod said:
So much this! Harder to do with film (not sure it's possible to overwrite a slide/neg [accidentally!]) but all too easy with digital files. I avoid it by keeping a second folder in every folder which I call Originals and that file is sacrosanct - no edits allowed! Memory is now so (relatively) cheap that a few extra MB (or even GB!) of used space is worth it for the extra security of the files. Of course, an extra layer of offline backup isn't a bad idea either...
Click to expand...

When I started scanning I was dealing with negatives (and slides) that were around 50 years old. OK, maybe you don't really "overwrite" them, but very definitely they were not unblemished originals! Apart from the ubiquitous dust, some of the scratches were horrendous, and I found a few examples of reticulation as well. One negative of my oldest child as a baby had something stuck on it; attempting to remove it, I tore the negative in half! Slides, which had been used as such, were all jumbled up and took a great deal of sorting out. However, much of this was simply down to poor handling; since about 2015 all my negatives and transparencies have been properly sleeved and filed, so I can find any of them.

As noted above, the way Aperture operates the original is preserved "un-touched" (?); not sure to what extent this ill be true of its (still unchosen) successor in my workflow.

I'm beginning to realise however, that even if I stick with JPEGs (thus minimising my disk space consumption), on migration from Aperture with a set of originals and a set of edited versions, plus the old Aperture library (just in case), I'm already going to have to find a new disk space strategy. The current strategy of "everything on the main drive" plus backups in two separate places, is not going to work well. But I think that's a post for a different thread!
 
ChrisR said:
... Anyway, I'm particularly interested in what sort of tests might reveal whether octupling my disk space consumption will give me noticeable benefits!
Click to expand...

No-one had an opinion on this, so I'll ask a related question... what sort of image is most likely to show up differences between JPEG and TIFF after multiple manipulations and saves?
 
ChrisR said:
No-one had an opinion on this, so I'll ask a related question... what sort of image is most likely to show up differences between JPEG and TIFF after multiple manipulations and saves?
Click to expand...

H'mm well I'd say nearly any decent image but maybe a facial shot? But then even if the image is degraded Photoshop could probably revive it. Anyway why don't you take a tiff file convert to jpg, then alter something and save and do this for about 10- 20 times then you'll have the answer.
 
Well I wrote a paragraph of why's and wherefores, Jox tiff workflow and questioning the folly of "why" anyone would not tiff scan proof negs
In the end it didn't answer the question because I have no real World experience of the circumstances required for me to ask the same question.
I do scan to tiff
I do produce to jpgs from it
I don't edit jpgs
I don't scan everything because scans are not my end product and I can't be bothered
I have plenty of storage should I need it (which I won't)

So, my cut to the chase is, don't spoil the ship for a hapeth of tar, get more storage...:)
 
ChrisR said:
No-one had an opinion on this, so I'll ask a related question... what sort of image is most likely to show up differences between JPEG and TIFF after multiple manipulations and saves?
Click to expand...

I remember scanning some colour negs in Epson Scan one time and noticing some nasty coloured banding in the sky on some of the photos. It turned out I’d accidentally changed the scan settings to JPEG from TIFF. I re-scanned the negs as TIFFs and the banding disappeared straight away.
 
dmb said:
https://petapixel.com/2010/02/04/saving-jpeg-photos-hundreds-of-times/
Click to expand...

Thanks David. That video (when you can eventually find the original) shows an artefact within about the first second, which is interesting, as I didn't get any visible artefacts after 10 edit/save cycles. But the artefacts showed up most clearly in the continuous tone areas; the ones in the more complex areas didn't show up until much later. That might help me find a couple of alternative trial images.

Of course, I don't need to save a JPEG 500 times; once or at most twice works for my workflow!

I know some others may thing I'm mad, and yes, storage is cheap (relatively... though I can buy a lot of film for the price of another couple of Terabytes of fully backed up disk!). It's just that having everything on my laptop and backed up in 2 places does work for me at the moment; pretty much everything else would be much more expensive and AFAICS less convenient.
 
FishyFish said:
I remember scanning some colour negs in Epson Scan one time and noticing some nasty coloured banding in the sky on some of the photos. It turned out I’d accidentally changed the scan settings to JPEG from TIFF. I re-scanned the negs as TIFFs and the banding disappeared straight away.
Click to expand...

Is it possible you had the JPEG quality set too low (too much compression)? I've never seen banding in skies...
 
You must log in or register to reply here.
Back
Top