gdanmitchell
Offline
Upload & Sell: Off
|
 Re: DR100% vs 200/400% | |
gyoung143 wrote:
Jack Flesher wrote:
If I edit a raw file, I have to output it in some other standard format for other software to read it, this can be tiff or psd or even jpeg and a host of others. However, in doing that, the raw file is unchanged. I can open it 5 years from now and reprocess it to anything I want. Moreover, I cannot save over that raw file; it remains the raw in its original format.
If I edit a jpeg, I can save the edited version under a different name so as not to overwrite the original. But it’s really easy to forget that step and save the edited one without changing the name, and now I have effectively “destroyed” the original and there’s no getting it back to that original state as all original pixels are different. (Yes, I understand it can theoretically be reconstructed from the exif data, but that’s an arduous process.)
Hence why raw editing is considered “non destructive” and directly editing any other standard format image can easily be destructive to the original.
You couldn't reconstruct it from Exif data as its re-compressed every time you save it. That is the primary reason why editing jpegs is 'as bad idea', its already compressed to make the jpeg, each time you save it it goes through compression again.
Gerry
I was going to say that, too, but I decided to let it go.
But when you save a jpg a whole lot of data is permanently lost and cannot be restored, even if you were to try to replicate settings used for the original capture.
Here’s an analogy that might make sense for some people.
Back in the days of cassette tapes, people used to duplicate cassettes. In fact, there were cassette player/records with two sets of heads for just this purpose. You could borrow a cassette from someone, load it into one of the cassette bays, put a black cassette into the other one, and “dub” a copy.
It worked pretty well. But the copy could never be quite as good as the original — some degradation of the information was unavoidable. The first copy was OK, though an attentive listener could hear the difference. But when someone took hat copy and duplicated IT on their machine, the degradation continued, and it wasn’t long until the quality was not good.
However, if you had access to the original reel-to-reel master recording, you could (limited only by the life-span of tape recordings, a separate topic…) make essentially unlined “dubs” from that with excellent quality since (a) the source was the original full quality original (or perhaps one generation removed) and (b) the copy was a first generation copy of the high quality original.
Think of that master tape as the rough equivalent of the raw file.
(For serious audio recording folks, studio automation moved that source tape even closer to what we get from a raw file today. Rather than doing a hard edit of the master recording, the automation data could be used to apply editing — dynamics, panning, tone controls, effects — to the output of the master.)
One of the reasons that jpg files are smaller is that the format was designed to be “lossy.” It makes some assumptions about what data can be thrown away without being noticed. This is not a significant problem for displaying images on screens or, arguably, even for transmitting files for printing. But it does mean that any alterations to the file — curves, color adjustments, noise reduction, retouching, etc. — are permanently baked into the file and simply be undone to return to the original.
|
