5

I am sending some C41 colour film to be developed. I'm not going to be requesting any prints, just scans of the negative to CD. I can choose between high resolution, high quality JPG or TIFF.

The TIFF files are more expensive and will be much larger files.

Although I'm not intending to do extensive post processing on these pictures, I may make small adjustments or bigger adjustments when I feel a picture needs it. I am used to shooting RAW on my digital camera and the extra PP latitude this brings when editing my photos in Lightroom.

I know that theoretically a TIFF file has the potential to retain more data than a lossy JPG, giving more PP latitude.

My question is does this theoretical benefit translate to a real post processing benefit when making minor/medium adjustments in Lightroom?

Would the benefit be of a similar magnitude to that of RAW over high quality JPG, or much less?

On a secondary note: I understand that there are various options when saving to TIFF (e.g. 8bit vs 16bit) though I do not fully understand what advantages these give. If I choose TIFF do I need to make sure the lab is going to use particular settings in order to get the benefit over JPG?

Regmi
  • 1,640
  • 3
  • 18
  • 35
Rich
  • 626
  • 5
  • 11
  • 2
    Additional information I would need to make a more informed decision: How much is the price difference? 10%? 50%? 300%? What is the maximum bit depth capability of the scanner(s) the lab will be using? How large is the total project? If it is particularly large, I would consider having a small sample processed both ways and then comparing the results. – Michael C Apr 09 '13 at 11:17
  • Thanks for the response

    There is an increased cost of 44% (£4.50 vs £6.50) for develop and scanning as TIFF vs develop and then scan as JPG.

    I do not know what the details of the scanning equipment, though it is a small commercial store. I have emailed them asking what bit depth they save their .TIFFs in but am yet to hear back.

    The total project is small (3 films, 96 frames) and the photographs are for personal use.

    Basically what I want to know is does someone who has got their C41 scanned in .TIFF and .JPG notice a difference when it comes to PP?

     – Rich Apr 09 '13 at 11:32
  • It would be highly dependent on the bit depth of the TIFFs, and whether the scanner can take advantage of that bit depth. For only 3 films I would just go ahead and pay for the TIFFs unless I suspected the lab couldn't exploit the advantage, in which case I would be looking for a lab that could. – Michael C Apr 09 '13 at 11:46
  • 2
    @Rich, after the job is done, please be sure not to vanish from this site like so many new user. Come at least to report the results and how pleased you are yourself with it. I find this question very interesting. – Esa Paulasto Apr 09 '13 at 12:34
  • @MichaelClark Thanks - do I need to just check that their scanner is capable of recording in 16bit, and that the .TIFFs will be saved as 16 bit? – Rich Apr 09 '13 at 12:54
  • @Esa - thankyou, I will let you know when I've got the files back. At some point I will try and do some tests to determine the answer to my quetsion, although it may not be for a while as I have a lot of exams coming up. When I do I will report back here – Rich Apr 09 '13 at 12:55
  • The compression level they are using on the JPEGs is probably the most critical part. If it is a high quality JPEG with the same color depth, there won't be any noticeable difference. If the compression is high, then there will be a night and day difference. – AJ Henderson Apr 09 '13 at 13:08
  • @AJHenderson Thanks; is the color depth the difference either 8bit or 16bit?

    If the scanner can do 16bit and the TIFF is saved as 16bit, what difference will this make (particularly in terms of PP) in comparison to a low-compression JPG?

     – Rich Apr 09 '13 at 13:54
  • 1
    @Rich - it would depend on the color gamut of the scanner and the film, but 16 bit color is able to produce orders of magnitude more distinct colors. Each additional bit doubles the number of values available for each color. Likely some of this extra headroom goes in to a more complete color gamut (smoother color changes) but some may also go in to storing an expanded color gamut with colors that are otherwise outside the range of color that the lower bit color depth can express. It's a pretty technical area that depends on a number of factors, but generally, more bits = better color. – AJ Henderson Apr 09 '13 at 15:21
  • Thanks everybody

    Just to update you - after speaking to the lab on the phone they didn't seem to know whether their scanner was 8bit or 16bit but thought it would just be 8bit, so I went with JPGs

    The chap on the phone said that he thought the 'quality' difference between the two was indistinguishable as he didn't notice any difference in prints made from either. I tried to explain that the difference might lie in post processing but he didn't seem to understand - frustrating as I am specifically using this lab as it seems to be one of the most clued up ones in the UK.

     – Rich Apr 15 '13 at 14:45
  • 1
    In the future I may try and find a place that can scan 16bitt TIFFs and then compare them to JPGs.

    My understanding from the above (@AJHenderson) is that I would mostly be gaining an advantage in colour, as opposed to detail in shadows/highlights - is this correct?

     – Rich Apr 15 '13 at 14:47
  • @Rich - a gain in color may well be a gain in detail in shadow and highlights, but it depends on the gamut. You could have a million colors between dark grey and light grey vs 256 colors between black and white. The 256 colors is going to have more detail in shadows and highlights since it is able to describe the blacks and whites. That's the difference between gamut and bit depth. Often a higher bit depth may be able to store a wider gamut though as it's fairly typical to make an 8 bit color space correspond to an established gamut like sRGB where as a 16 bit color space may be broader. – AJ Henderson Apr 15 '13 at 18:15

3 Answers3

4

If the TIFF files are only 8bit and the resolution is the same then there will be very little (unless the JPEG compression is set very high). The only difference will be slight artefacts in high frequency areas and potentially lower colour resolution if chroma sub-sampling is used on the JPEGs.

Additionally if the scan resolution itself is high compared to the resolution of the image, then there will be little difference between TIFF and JPEG as they will both contain more information than the original film.

It seems to me that they're just trying to create an artificial differentiation to increase revenues. The only time I'd consider paying half again for TIFFs would be under the following conditions:

  • The quality of the original negatives was very high
  • The TIFF files and scanner were both more than 8 bit
Matt Grum
  • 118,892
  • 5
  • 274
  • 436
  • Thanks! Did you mean either of those criteria or both at the same time?

    If the scanner and files are 16 bit, what sort of difference do you think this would make in comparison to a high quality JPG?

     – Rich Apr 09 '13 at 13:55
2

With the data you provide:

  • the files are much larger
  • more expensive
  • it's a service they chose to provide

We can assume that the TIFF format used is less compressed and higher bitrate - most likely 16bit lossless, or the service is really pushing it. And they offer this service because the bitrate of the scanner is above 8bit, so they want you to take advantage of that for a higher price. It is then similar to raw vs jpeg out of the DSLR. If you plan to do any processing like contrast, gamma, curves, highlight/shadow meddling, you'll want to get those nice tiffs to avoid posterization in the 8bit end result.

Michael Nielsen
  • 10,644
  • 35
  • 48
0

TIFF is not an image format. It is a container for varoius formats from 2-Bit CCITT to 4x16 bit or probalby even more. You can even put JPEGs into TIFF.

Therefore comparing TIFF as such with JPEG or whatever is quite rediculous.

However, when talking about TIFF from scanners or cameras (there are some or there used to be) should contain 3x8 bit RGB in most cases. So the key difference comes down to the destructive compression in standard-jpeg vs. nondestructive compression or no compression at all of the BMPs in TIFF.

My opinion: There is no advantage in saving the tiff in 16 bit (vs. 8 bit) nor is there any significant andvantage vs. jpeg saved with best quality parameters. (may that be 100% or 1.0 or what ever setting provides the lagest files and therefore least destructive compression for jpegs.)

However, there are significant advantages in the further processing. After opening a file in photoshop or whatever, the next step should be the conversion into a 16-bit format. Any interim result should be saved in a non-destructive 16-bit format too. If the final result is to be handed over or uploaded in any 8-bit (sometimes jpeg is even required) format, then it should be the last and very final step converting the file back to 8 bit.

Unfortunately my English is not good enough to explain in detail, where the advantages during the editing process are.

Hermann Klecker
  • 1,113
  • 6
  • 9
  • 1
    How can you claim there is no advantage to saving 16 bit tiff , if the scanner actually had a ADC that is over 8bit? it makes no sense to select 8 bit on purpose at first only to convert to 16bit later fort all "interim results". you've already destroyed the data. – Michael Nielsen Apr 09 '13 at 18:50
  • I know this is unpopular. However, it is fact. Did you ever acutally see a difference? Ever? – Hermann Klecker Apr 09 '13 at 20:56
  • in 8 vs, say, 10 bit data? yes. all the time. – Michael Nielsen Apr 09 '13 at 22:13
  • On a print? Which I doubt. Get away from these theroetical discussions. I am IT professional myself which includes image processing to a certain extend. I know what the data loss is. The point is what the data loss really means to a print or to its presentation on a web site. I mean that what really matters. – Hermann Klecker Apr 10 '13 at 07:32
  • yes, on numerous occasions I am knocking my head into the wall over my old images from before 2011 where I started shooting 14bit, and need to adjust the curves in jpegs where the result for facebook. Also when I was working with Nikon D200 to image peach blossoms for work, the blossoms were basically not visible in the jpegs, but the raws had nicely defined blossoms, so I exported them to 16 tiffs for further processing. – Michael Nielsen Apr 10 '13 at 11:55
  • So you did a poor job some jears ago. Quite sad. And yes, there are significant advantages using a 16 bit working colour space, as I said before. We were talking about saving the reusult of the scanning process. – Hermann Klecker Apr 10 '13 at 18:52
  • I think the point is that the original media doesn’t have as wide of a latitude as modern digital cameras, so (with contrast and blacklevel set optimilly at scan time) 8 bit is sufficient. At high resolution, you'll be recording halftoneing in space (you see the grains), not more subtle shades in one pixel. – JDługosz Nov 20 '17 at 08:54