I recently had a suggestion to use a JPEG instead of a TIFF format for my ortho images.
?ÿ
I tried it out in a recent project and the quality of image is almost indistinguishable and the file size is 1/10 that of a TIFF.
?ÿ
Ive attached two screenshots zoomed in to a MH.
?ÿ
JPEG
?ÿ
TIFF
?ÿ
?ÿ
Am I missing something here? I cant see any reason to use a TIFF. It also seems my TIFF is shifted slightly even though they were exported using the exact same settings.
Ive used ECW's in the past because of the smaller file size but Im thinking just skip a step of converting the TIFF to an ECW and get the JPEG directly from Agisoft.
Depends.
Tiff you can use in GIS as raster data.
Jpeg not so much. Image versus data is a decision line, yes it is all data but all data is not equal.
TIFF can hold more than one page / image
JPG size depends on the compression
nzd$0.02
We always used TIFF and I could not remember why.
I asked Google.
What is the difference between TIFF and JPG?
Jpeg sets the standard for Internet images and the small file size accommodates storage. Tiff images contain all of the data of the original photograph, resulting in significantly larger files and high quality images.Oct 29, 2022
Ahhh ok, so it may make a difference if one were to try to bring it into some GIS software.
I strictly use my orthos in CAD.
With that in mind, is there any practical reason to use an image 10x the file size when the quality difference is almost indiscernible?
My understanding is that JPG is compressed using an algorithm that results in some (though not much) loss of fidelity, whereas TIFF is uncompressed. This does save considerable disk space, though when opened a JPG expands to about the same size as the equivalent TIFF, so if RAM is an issue you're in the same boat with either format.
From my point of view you can have world file exported from Agisoft, together with the ortho file and you can import/attach the TIFF to AutoCAD directly on the correct coordinates.
Someone to confirm if done it with JPG
I've always exported .tiff's from Pix4D to bring into Civil3D. Depending on the size of the flight, the .tiff usually is too large to bring into Civil so I have it tiled up and then bring those in. You can leave out ones you don't need and save space too.
Do jpeg's hold positional data as well? If so, I just learned something new. That's why I always used the TIFF.
Both TIFF and JPG files use world files to align them to specific coordinate systems when inserting into CAD or GIS applications. The world files have .tfw and .jpw extensions.
Peter is correct.
Correct me if Im wrong Peter, but the TIFF can have the location data baked into the image itself without the need of a world file but the jpeg does in fact need the tfw and jpw.
I've downloaded free topo maps from USGS that were in JPEG2000 format. I understand these can contain georeferencing information, but that's about all I know.
@jim-frame It's worth distinguishing lossless compression from lossy compression. An image that is still within the image sensor in the camera would be uncompressed. Any popular file format, including TIFF and RAW will be compressed, but many are compressed in a way that the uncompressed image can be exactly reconstructed. Formats that achieve the highest degree of compression, like JPEG, are lossy. That is, the information from the original cannot be exactly reconstructed.
Any popular file format, including TIFF and RAW will be compressed
Not necessarily. The TIFF specification allows for 3 different compression schemes, one of which is no compression. I'm wondering if most orthophotos are compressed at all, given their typically large file sizes.
@jim-frame It's worth distinguishing lossless compression from lossy compression. An image that is still within the image sensor in the camera would be uncompressed. Any popular file format, including TIFF and RAW will be compressed, but many are compressed in a way that the uncompressed image can be exactly reconstructed. Formats that achieve the highest degree of compression, like JPEG, are lossy. That is, the information from the original cannot be exactly reconstructed.
By reconstructed you mean revert to its fully uncompressed state?
What would be the circumstances that you would want to do that?
When I look at the quality difference here I’d say the trade off for file size and the fact that any image viewer can open a jpeg allowing for easier usability for clients just wanting to look at a picture, it’s a no brainer. You could create a tiff as well but I think it would rarely be used.
What would be the circumstances that you would want to do that?
In order to display the image, it has to get uncompressed in RAM. The compressed state only pertains to storage.
