Walgreens gives me the TIFF

[kshapero]

Nowadays I have been getting scans done at Walgreen's, not very hi res, but good enough for review and uploading. Today I got a CD back and it also had 6 mg TIFF files as well as JPEG files. What can I do with TIFF? Can at least treat them as electronic negatives, much like RAW files in digital? BTW I am shooting film.

[Jason Sprenger]

TIFF is good. TIFF is a standard format that isn't manufacturer dependent like RAW and doesn't toss away picture information like JPEG.

When I use my scanner, TIFF is the format I save the files to for archiving. Any decent image editor can read TIFF.

[bmattock]

TIFF is not RAW. TIFF is not a 'lossy' compression format. However, it does not have all the data encoded in it as RAW does. Consider it better than JPG and worse than RAW.

[dmr]

Oh really! I'll have to check some of my recent CDs. All I know of are their med-res .jpg files.

[FA Limited]

Originally Posted by bmattock

TIFF is not RAW. TIFF is not a 'lossy' compression format. However, it does not have all the data encoded in it as RAW does. Consider it better than JPG and worse than RAW.

doesn't that depend on the number of bits?

[bucks11]

Maybe I'm wrong, but I thought TIFF was bigger than RAW...

Doesn't TIFF store files according to 3 individual channels, RGB, while RAW stores the same information (RGB) in 1 channel, making it a smaller filesize but essentially as accurate as TIFF...?

[bmattock]

It is not that TIFF cannot store more data than RAW, or that TIFF files are not bigger than RAW files - it is that digital cameras do internal processing on the image before storing it in either TIFF or JPG format, so data is lost. RAW is essentially unprocessed, the camera does minimal 'work' on the image before storing it.

A classic example of this is white balance. With either TIFF or JPG, white balance information is chosen by the camera, and encoded in the file. You have very limited options to change white balance in post-processing. RAW is encoded before the camera chooses white balance, so all of the potential data is present, so that in post-processing, you or your RAW post-processing software must choose the correct white balance - but there is much more data present, and much more room to maneuver.

http://photo.net/learn/raw/

http://www.ronbigelow.com/articles/raw/raw.htm

TIFF is a better choice than JPG if your camera offers it, unless file size or speed of writing to the card is a primary concern. Otherwise, RAW is technically superior to both.

[JeremyLangford]

Originally Posted by bmattock

It is not that TIFF cannot store more data than RAW, or that TIFF files are not bigger than RAW files - it is that digital cameras do internal processing on the image before storing it in either TIFF or JPG format, so data is lost. RAW is essentially unprocessed, the camera does minimal 'work' on the image before storing it.

A classic example of this is white balance. With either TIFF or JPG, white balance information is chosen by the camera, and encoded in the file. You have very limited options to change white balance in post-processing. RAW is encoded before the camera chooses white balance, so all of the potential data is present, so that in post-processing, you or your RAW post-processing software must choose the correct white balance - but there is much more data present, and much more room to maneuver.

http://photo.net/learn/raw/

http://www.ronbigelow.com/articles/raw/raw.htm

TIFF is a better choice than JPG if your camera offers it, unless file size or speed of writing to the card is a primary concern. Otherwise, RAW is technically superior to both.

But what about with scanning film? I thought that scanning to a 16 or 24 bit TIFF was as good as a digicam RAW picture.

[bmattock]

Originally Posted by JeremyLangford

But what about with scanning film? I thought that scanning to a 16 or 24 bit TIFF was as good as a digicam RAW picture.

RAW still should be superior in terms of the amount of information stored (not the amount of data stored, which is different). However, I am told that if you do Digital ICE for noise/dust reduction, this is not applied to the RAW file - it's 'post-processing' from the point of view of the scanner, so it is present in a JPG or TIFF produced by the scanner, but not in the RAW file.

Scanners frequently have their settings tweaked during the preview or prescan so that the white balance, etc, are correct or very close to it prior to scanning. If that is done correctly, then part of the advantage of saving the scanned image to RAW format would be lost, and you might like TIFF better.

[DGA]

Originally Posted by Jason Sprenger

TIFF is good. TIFF is a standard format that isn't manufacturer dependent like RAW and doesn't toss away picture information like JPEG.

When I use my scanner, TIFF is the format I save the files to for archiving. Any decent image editor can read TIFF.

TIFFs are good, as long as they are 16bit-per-channel.
8bit-per-channel TIFFs are the same as JPGs with no compression.

[bmattock]

Originally Posted by DGA

8bit-per-channel TIFFs are the same as JPGs with no compression.

TIFF files can be compressed, but they do not use the 'lossy' compression method that JPEG files do. An 8-bit TIFF would not be identical to the same file saved in JPEG because the JPG file would have thrown away data, to be mathematically interpolated (guessed at) on re-opening of the file. The 8-bit TIFF would still be superior to JPG.

[JFH]

OK, here's a related question.... The lab I've used for a while with satisfaction (C-41) does a scan they call "pro- scan" which yields jpeg files in the 1.7 to 2.0 meg range... at 2000 x 3000 dots. When I mentioned that these files seemed a bit "small," the owner said something to the effect of, "Oh, no, you don't understand.. when you open the files ( in PSE or whatever) you should then save them as TIFF's and they will be much larger with all the data."

I was a bit skeptical, but as he said, saving as TIFF's gave me files in the 4 to 7 mb range, and "expanding" the print size with my software led to TIFF files upwards of 15 to 20 mb. I haven't picked his brain any further, since he's 150 miles away and I mostly do mail order business with him. I've done at least one print of 10x20 inches (with cropping) from one of these files, and I got what I would consider good definition and detail, etc, and I'm quite happy with the results.

I don't know exactly what machinery and scanner he's using, but the little 4x6 "mini- proof" he supplies with each roll has an Agfa logo on it. Would anybody here care to take a stab at further explaining all this? Whatever it is, it seems to work well for me....

[bmattock]

Originally Posted by JFH

OK, here's a related question.... The lab I've used for a while with satisfaction (C-41) does a scan they call "pro- scan" which yields jpeg files in the 1.7 to 2.0 meg range... at 2000 x 3000 dots. When I mentioned that these files seemed a bit "small," the owner said something to the effect of, "Oh, no, you don't understand.. when you open the files ( in PSE or whatever) you should then save them as TIFF's and they will be much larger with all the data."

It is possible to store JPEG files using a lossless form of compression. This is almost never done. Generally, JPEG files are stored in a compressed format that is 'lossy', which means it applies a mathematical formula to determine which data points it can throw away during compression and 'guess' at (it is called 'interpolation') later. The degree to which it is correct or incorrect is based on how much the file is compressed.

Once this is done, it cannot be undone. Saving the file in TIFF or any other format later one does not put data back in that has been thrown away - it instead saves the data it has plus the data that was interpolated (guessed at) when the file was opened. Yes, it will be 'bigger' but that itself is meaningless.

Imagine the old days (if you're old enough) of recording an LP record to a cassette tape. It never sounded quite as good as the original. Make a copy of that tape onto another tape, and it sounded worse. Same thing if you made copies of VHS tapes one to another. JPG loses data in most normal circumstances. Once gone, it is gone for good.

Since it is theoretically possible that your lab guy could be using some form of non-lossy compression for your JPG files, I won't say he's lying to you. But I'd check it out a bit more thoroughly. The explain he's giving you is a bit on the hocus-pocus side.

[bmattock]

Originally Posted by JFH

OK, here's a related question.... The lab I've used for a while with satisfaction (C-41) does a scan they call "pro- scan" which yields jpeg files in the 1.7 to 2.0 meg range... at 2000 x 3000 dots. When I mentioned that these files seemed a bit "small," the owner said something to the effect of, "Oh, no, you don't understand.. when you open the files ( in PSE or whatever) you should then save them as TIFF's and they will be much larger with all the data."

Oh, and by the way - 2000 x 3000 is about what you get from a typical 6 megapixel camera. Yes, you should be able to print those to reasonably large size without problems, I certainly have. However, the typical frame of 35mm film is capable of giving much more detail. The question is whether you want it or not.

[JFH]

Bmattock: Yeah, that's about what I figured. I know you can't "put something back" once it's been removed. It sounded a bit hocus pocus to me too, and I've always felt that my negatives were probably at least a bit better than the scans. Only problem is that it'll cost me a bunch (in prints and so forth) to really find out. I'm not ready to buy a scanner either.
Something to keep in mind for when I go "up" to the next level in photography (whatever that is....)

[IK13]

Originally Posted by bmattock

RAW still should be superior in terms of the amount of information stored (not the amount of data stored, which is different).

This is not true if you are comparing linear gamma tiff from a scanner to a raw file from a (bayer sensor) digital camera.

Linear gamma TIFF from a scanner is as RAW as it gets.
There is no more RAW you can get from a scanner.
The RAW DNG files you get from Vuescan are rebadged tiffs (and Mr. Ed Hamrick confirmed that in an email few months back when I asked them about it).

What is the difference?

The digital camera RAW file doesn't have all the color information for every pixel. It needs to do Bayer interpolation to do in order to get RGB values for each and every pixel. Once the interpolation process (where software decides on the color values of a pixel, based on the color of the adjacent pixels etc) is complete though - you have one interpretation of the RAW data. And no matter how you save this "interpretation" - TIFF, JPEG etc. it's still only one interpretation of the RAW data (and therefore not RAW).

A scanner on the other hand collects all 3 channels (RGB) for each and every pixel at the time of the scanning (same for the Foveon sensor). No interpolation ("guessing") is necessary. So if the scanner's software doesn't apply any changes and saves the result as an uncompressed (or non lossy compressed) TIFF file, then this TIFF file is as row as it gets.
Whether or not you're getting this TIFF or an "auto-everything" image, tweaked by the software and saved as tiff is a whole other story.

[bmattock]

Originally Posted by IK13

This is not true if you are comparing linear gamma tiff from a scanner to a raw file from a (bayer sensor) digital camera.

Well, you got me there. I didn't realize we were comparing the two, but OK with me, that does sound right.

[KenD]

Originally Posted by JFH

OK, here's a related question.... The lab I've used for a while with satisfaction (C-41) does a scan they call "pro- scan" which yields jpeg files in the 1.7 to 2.0 meg range... at 2000 x 3000 dots. When I mentioned that these files seemed a bit "small," the owner said something to the effect of, "Oh, no, you don't understand.. when you open the files ( in PSE or whatever) you should then save them as TIFF's and they will be much larger with all the data."

I was a bit skeptical, but as he said, saving as TIFF's gave me files in the 4 to 7 mb range, and "expanding" the print size with my software led to TIFF files upwards of 15 to 20 mb. I haven't picked his brain any further, since he's 150 miles away and I mostly do mail order business with him. I've done at least one print of 10x20 inches (with cropping) from one of these files, and I got what I would consider good definition and detail, etc, and I'm quite happy with the results.

I don't know exactly what machinery and scanner he's using, but the little 4x6 "mini- proof" he supplies with each roll has an Agfa logo on it. Would anybody here care to take a stab at further explaining all this? Whatever it is, it seems to work well for me....

Ok, let's take a 35mm Tri-X negative and enlarge it to make an 8x10 negative. Is a contact print going to look like one from an 8x10 original camera neg? Making the new negative or the new file bigger is not going to bring back lost/ discarded detail or tonality. Sometimes size matters )