Raw to tiff??

[domromer]

For the past few days I've been organizing my pics in lightroom. I'm still trying to figure out my workflow. I've been shooting a ton or raws lately and can't decide whether to save them as raws. Then only save a tiff if I end up tweaking them, then finally make a jpeg out of the tiff if I need one.

So today I started to convert raws to tiffs. My raw files are around 6-8megs each. Where the converted tiffs are around 20 megs. How is this? I though a raw was an unprocessed jpeg. Why are the tiffs so much bigger?

 

[Alex_B]

they are 16 bits, which already makes some difference as RAW data is either 12 or 14 bits usually.

also, depending on the RAW format there is lossless compression of data.

 

[domromer]

they are 16 bits, which already makes some difference as RAW data is either 12 or 14 bits usually.

also, depending on the RAW format there is lossless compression of data.

That being said, why do more people seems to convert raws to tiffs for storage rather than save the raws?

 

[Alex_B]

That being said, why do more people seems to convert raws to tiffs for storage rather than save the raws?

- because there is a standard for TIFF, but not for RAW. All RAW formats are totally different. If you dig your archives in 10 years time, you might be happy to have the TIFFs

nevertheless, one should archive the RAW data as well, since you might want to reprocess them with different settings one day.

 

[Big Mike]

As far as I know...you don't or can't save RAW files. A RAW file is just the data from the camera (with an embedded JPEG so that you can see it). So once you open/convert the RAW file...it must be saved as something (JPEG, TIFF, PSD etc).

Now with Lightroom...it's a different story. Lightroom is a non-destructive workflow, so when you open a RAW file and make adjustments...it doesn't actually make changes to the image. What it does is to save the changes/adjustments that you made.

I'm not sure if Lightroom embeds this info into the file or if it saves a separate file but the point is that anything you do in Lightroom, won't damage the integrity of the image because you aren't actually making changes to the file.

When you do save (output) the image...then lightroom applies the changes to the image. So, as I understand it, you don't need to save the file in an image format until you have a reason to do so. If you plan to work on the image again later, just keep it in Lightroom (save the changes) and then when you open it again, you will only be editing the changes.

As for why TIFF files are so big (PSD files as well)....I'm not sure. I think it's because they are actual image files, which need to have a lot more information...RAW files are just the gibberish that comes out of the camera.

 

[kittymaguire]

In Lightroom you have two choices when you import images, the first is to keep the raw format, any changes made to the image will then be stored in a separate file, the 2nd is too import as dng which is Adobe's digital negative format. Changes made to the image by Lightroom will be written to the dng file but in such a way that Lightroom will be able to revert to the original as shot image.

The tiff is bigger because it is an agreed standard were as the raw format varies between cameras which means that the camera manufacturers can take short cuts when writing the image data.

If you are using Lightroom to manage your work flow then don't keep to store your images as tiffs. You can export the tiff or jpg only when you need to use them.

 

[EOS_JD]

One thing that has not been mentioned here is that RAWs are actually Compressed files. it has nothing to do with tif being an agreed standard or that they are actual image files... they both just hold digital information. So this is the reason file sizes are smaller than for the tifs you export. All manufacturers RAW images are different sizes and I think this is down to how each compresses their files.

Also your converter will normally create an uncompressed (24-bit) TIFF files, so they'll be larger. A 24 bit RGB file has three 8 bit colour channels - 3 channels x 8 bits = 24bit image.

Do you export your tifs as compressed files? You normally have the option of LZW compression which is a lossless way of compressing the tif image and reducing your file size.

What bit depth do you export as? 8bit or 16bit? By exporting a 16 bit image, this will create a 48-bit TIFF(16 x 3 channels - RGB) file that retains the full dynamic range of the raw file.

These files will be double the size of the 24 bit files.

 

[Alex_B]

One thing that has not been mentioned here is that RAWs are actually Compressed files.

Has been mentioned, first reply

 

[EOS_JD]

That being said, why do more people seems to convert raws to tiffs for storage rather than save the raws?

Also to add.... Everyone should keep their RAW files. To delete them is crazy. I don't believe many do though.

I keep the RAWs and convert to jpg. Do my edits then save only once. It's rare that I need a tif unless I need to save and come back to a file. JPGs generally work fine for me though - Even with a little loss (which is never noticed).

I've got to the stage that unless I edit in photoshop, I don't save the jpg or tif and keep only the RAWs. Lightroom is used more now than photoshop in my workflow.

 

[EOS_JD]

Has been mentioned, first reply

Sorry Alex. Must have read it differently in my eagerness to respond!

 

[Alex_B]

np

 

[Helen B]

I don't think this has been mentioned in detail before in this thread, though Big Mike has touched on it, so here goes: Raw files from a normal Bayer-pattern sensor have only one channel per pixel. Each pixel has only luminance data - the colour data are added and interpolated during the conversion.

The Raw files can be uncompressed, losslessly compressed or slightly lossily compressed depending on the camera and the options available.

Best,
Helen

 

[Alex_B]

I don't think this has been mentioned in detail before in this thread, though Big Mike has touched on it, so here goes: Raw files from a normal Bayer-pattern sensor have only one channel per pixel. Each pixel has only luminance data - the colour data are added and interpolated during the conversion.

good point!