Enhancement jpg with lightroom (Page 2): I like number 1 and 3...

davyboy Loc: Anoka Mn.

Ysarex wrote:

I'd say incorrect. This all hinges on what does "OK" mean and what are you doing with the photo. Any edit of a JPEG compromises quality and causes damage. It doesn't matter if it's a "one time" edit. It's a common myth that opening and re-saving a JPEG is what has to be avoided. It's true that opening and re-saving a JPEG isn't good but that's a minor concern. The real cause of damage is editing.

If you make a tone response and/or color change to a JPEG even if just "one time" you're going to cause damage that wouldn't be caused if you made those edits to data that hasn't bee JPEG encoded. Can you see that damage on your iPhone? No. Therefore the damage isn't real -- be happy.

Everybody does it of course so it must be OK. Well that sounds reasonable -- be happy.

Here's an example: In the first illustration below you see on the left the camera SOOC JPEG, in the middle the processed raw file and on the right the SOOC JPEG edited to better match the processed raw image. I deliberately reduced resolution because the damage caused by editing the JPEG disappears with the resolution drop. In other words you're going to look at it on your phone so there is no problem -- be happy.

The camera was set to auto-WB and no effort was made to adjust the JPEG settings from default. That explains the need to edit the JPEG. Auto-WB screwed up and those baptisia are not that color seen in the SOOC JPEG. So all I did using Photoshop ACR was adjust the color and contrast to better match the processed raw which has accurate color and better tone response. The edit is one time only with only that one save when I saved the illustration. The edited JPEG looks fine -- be happy.

The 2nd illustration shows an out of focus region of the image from all three files and at 100% resolution. OUCH! The edited JPEG is badly damaged. That's a one time edit and one only save of light editing of the SOOC JPEG. So don't pixel peep!!! -- be happy. This example shows the worst that can happen. That kind of damage won't show in areas that are in focus. So avoid out of focus backgrounds and don't pixel peep!! -- be happy.

NOTE: I edited the JPEG parametrically using ACR. That generally works better than doing the edit with a raster editor like PS proper. Can the damage done editing the JPEG be avoided? No. But if you can't see it's not real -- be happy.

EDIT: Worth noting that although you can't prevent the damage to the JPEG from editing there are things you can do to repair or cover-up that damage. If you find yourself doing that it might be appropriate to ask what the bleep you're doing trying to fix the screw-up you caused by trying to fix the screw-up you caused....

Joe

I'd say incorrect. This all hinges on what does &q... (show quote)

I like number 1 and 3

camerapapi Loc: Miami, Fl.

Modern JPEG files are pretty good in quality. If they need manipulation during editing I would go with the less amount of it. Remember they are 8 bits files so changes in colors and banding are always a possibility.

GrandmaG Loc: Flat Rock, MI

tenny52 wrote:

you can easily tell which one is enhanced

Sorry, but this looks a little over-done to me, although the original picture does need some adjustments.

David1943

In this particular photo....go back to LR and and simply change the white balance ...try it with "auto" first. I downloaded your photo and it certainly made the photo look 100% better !! Don't know how to attach it to this reply ...but try it !!!!

rehess Loc: South Bend, Indiana, USA

tenny52 wrote:

More question: can a better camera/lens/technique capture a similar enhanced picture?

My Pentax cameras have various settings to control how the camera 'develops' raw data to create JPEGs - "Custom Image" can be "Bright", "Natural", "Portrait", "Landscape", Vibrant", "Radiant",
"Muted", "Bleach Bypass""Reversal Film", "Monochrome", or "Cross Processing" {a random setting}. Based on what you have shown so far, you would be happier with "Bright" or "Vibrant", while you actually got something more like "Natural".

jerryc41 Loc: Catskill Mts of NY

I'm betting you have your camera set to produce neutral JPEGs. Most cameras give the user a choice of how much processing the camera should do. Neutral is nice since it allows room for the photographer to process. I shoot in raw, so processing isn't optional.

jerryc41 Loc: Catskill Mts of NY

tenny52 wrote:

you can easily tell which one is enhanced

The enhanced version looks like over-cooked HDR, which is okay if that's what you want, but it's not a natural shot.

tenny52 Loc: San Francisco

Jerry, you are right. I set my picture capture to Standard which should be VIVID instead. And most of my enhancement could be overly done since I notice I am weak at color distinction. But I am probably the only to enjoy my own captures.
The reason I shot jpg because many good Fuji reviews on jpg color rendering and a Fuji raw uses 30mb instead of 23mb per D610.

rehess Loc: South Bend, Indiana, USA

tenny52 wrote:

Jerry, you are right. I set my picture capture to Standard which should be VIVID instead. And most of my enhancement could be overly done since I notice I am weak at color distinction. But I am probably the only to enjoy my own captures.

Pleasing yourself is the important part.

JD750 Loc: SoCal

Ysarex wrote:

This is not correct. Changing the file format to other than JPEG does not remove the original JPEG encoding. Re-saving an edited JPEG is not the primary cause of degradation to the image. The degradation derives from any editing of the image tone and/or color interacting with the original JPEG compression grid -- altering the file format does nothing to prevent that as the original compression grid is saved intact in the new file format.

Joe

Joe I’m sorry but this is incorrect.

Ysarex Loc: St. Louis

JD750 wrote:

Joe I’m sorry but this is incorrect.

No, it's completely correct. Changing the format of a JPEG to something like TIFF or PSD only prevents further degradation from re-saving the JPEG which is the much lessor problem. The more serious problem when editing JPEGs is damage that results from interaction between the edit changes and the original JPEG compression grid and changing the file format has absolutely no effect on that.

Joe

rehess Loc: South Bend, Indiana, USA

Ysarex wrote:

No, it's completely correct. Changing the format of a JPEG to something like TIFF or PSD only prevents further degradation from re-saving the JPEG which is the much lessor problem. The more serious problem when editing JPEGs is damage that results from interaction between the edit changes and the original JPEG compression grid and changing the file format has absolutely no effect on that.

Joe

I'm not sure what you mean by "compression grid" here. After ten years of regularly using JPEGs, the only damage I've ever seen is the result of re-saving as JPEG, and therefore re-compressing. If I read a JPEG into PhotoShop or gimp, I can save many many times in the native format with no noticeable degradation.

Ysarex Loc: St. Louis

rehess wrote:

I'm not sure what you mean by "compression grid" here. After ten years of regularly using JPEGs, the only damage I've ever seen is the result of re-saving as JPEG, and therefore re-compressing. If I read a JPEG into PhotoShop or gimp, I can save many many times in the native format with no noticeable degradation.

Compression grid is how JPEG works it's magic. JPEG's job is to create redundancy in the data to allow compression. To do that the JPEG algorithm lays a grid (most commonly 8 pixels square) over the image. Photo data is dense and the odds are that in any one grid cell there will be 64 unique pixels. JPEG's job is to alter pixels so that when it's finished there are for example 32 unique pixels in the grid cell -- at very high compression rates maybe only 24 unique pixels, etc. It is brilliant and works beautifully. However once the job is done there's no going back and that compression grid is now part of the image -- think of it as embedded in the image.

JPEG degradation due to editing then occurs from two sources. 1. Re-saving and re-compressing can cause additional loss but it is the lesser of the two causes of degradation. 2. Changes to the tone and color of the image will interact with the original compression grid. For example you're trying to alter the color and density of the sky which is a gradient. The alteration you perform should preserve the original gradient but instead it runs into a stack of 6 pixels that are all the same. Originally they weren't but they are now. The edit then causes the compression grid that was previously hidden to start to become visible. This is the greater of the two causes of degradation and there's nothing to do to avoid it -- the original compression can't be undone.

Here's an example where you can see both types of damage.

What I did:

I selected an old camera JPEG from a 10 megapixel camera (Sony R1). The lower res of this test JPEG makes it easier to see what's happening. I selected a JPEG that needs the sky, a gradient, corrected. Changes made in the sand for example wouldn't show.

I opened the camera JPEG (copy), started a count and re-saved it re-compressing it and overwriting it. I closed it. I reopened it and I repeated 7 times total adding to the count each time.

I opened the camera JPEG (another copy) and immediately converted it to 16 bit and saved it as a TIFF file. I then opened and edited that TIFF file in ACR (parametric edit) increasing contrast and adjusting the sky color to blue.

Below you see first the original camera JPEG. Then an illustration which is a PNG file and so there is no additional lossy compression being applied to any of the three images there. The right image was compressed only once by the camera. The middle image is the camera original JPEG and the left image is the 7 times re-save and re-compress.

The 7 times re-save and re-compress did do additional damage but it's slight -- difficult to see the seam in the sky between it and the original. The real damage shows in the edited version that changed the tone and color of the sky. The mottled appearance you see there is very literally the original compression grid becoming visible as a result of the editing changes. Converting to a 16 bit TIFF before editing was of no value to prevent this 2nd (and more severe) type of JPEG degradation. Again note that the only lossy compression applied to that (right side) image was the original camera compression.

We have a modern advantage that I noted in my earlier post in this thread. Most of us now use 16, 24, 36 and even higher megapixel cameras. The same JPEG degradation due to editing occurs for the same reasons but it's much less visible at the higher resolutions we now use. In other words our butts are covered by megapixels.

Joe

EDIT: I popped over to 500px and found a breathtaking sunset that was originally a camera JPEG and got some pretty heavy editing. The JPEG degradation was visible on my screen at viewing res but not horrible. I copied it and enlarged a section of the sky to 400%, saved it as a PNG and posted it below. I present it here because rehess was curious about the compression grid. Look real close and you can actually see the JPEG 8x8 pixel grid cells in the sky. The heavy editing is literally causing the grid to lift up out of the photo and become visible -- we're not supposed to be able to see it -- that's the point.

(Download)

rehess Loc: South Bend, Indiana, USA

Ysarex wrote:

Compression grid is how JPEG works it's magic. JPEG's job is to create redundancy in the data to allow compression. To do that the JPEG algorithm lays a grid (most commonly 8 pixels square) over the image. Photo data is dense and the odds are that in any one grid cell there will be 64 unique pixels. JPEG's job is to alter pixels so that when it's finished there are for example 32 unique pixels in the grid cell -- at very high compression rates maybe only 24 unique pixels, etc. It is brilliant and works beautifully. However once the job is done there's no going back and that compression grid is now part of the image -- think of it as embedded in the image.

JPEG degradation due to editing then occurs from two sources. 1. Re-saving and re-compressing can cause additional loss but it is the lesser of the two causes of degradation. 2. Changes to the tone and color of the image will interact with the original compression grid. For example you're trying to alter the color and density of the sky which is a gradient. The alteration you perform should preserve the original gradient but instead it runs into a stack of 6 pixels that are all the same. Originally they weren't but they are now. The edit then causes the compression grid that was previously hidden to start to become visible. This is the greater of the two causes of degradation and there's nothing to do to avoid it -- the original compression can't be undone.

Here's an example where you can see both types of damage.

What I did:

I selected an old camera JPEG from a 10 megapixel camera (Sony R1). The lower res of this test JPEG makes it easier to see what's happening. I selected a JPEG that needs the sky, a gradient, corrected. Changes made in the sand for example wouldn't show.

I opened the camera JPEG (copy), started a count and re-saved it re-compressing it and overwriting it. I closed it. I reopened it and I repeated 7 times total adding to the count each time.

I opened the camera JPEG (another copy) and immediately converted it to 16 bit and saved it as a TIFF file. I then opened and edited that TIFF file in ACR (parametric edit) increasing contrast and adjusting the sky color to blue.

Below you see first the original camera JPEG. Then an illustration which is a PNG file and so there is no additional lossy compression being applied to any of the three images there. The right image was compressed only once by the camera. The middle image is the camera original JPEG and the left image is the 7 times re-save and re-compress.

The 7 times re-save and re-compress did do additional damage but it's slight -- difficult to see the seam in the sky between it and the original. The real damage shows in the edited version that changed the tone and color of the sky. The mottled appearance you see there is very literally the original compression grid becoming visible as a result of the editing changes. Converting to a 16 bit TIFF before editing was of no value to prevent this 2nd (and more severe) type of JPEG degradation. Again note that the only lossy compression applied to that (right side) image was the original camera compression.

We have a modern advantage that I noted in my earlier post in this thread. Most of us now use 16, 24, 36 and even higher megapixel cameras. The same JPEG degradation due to editing occurs for the same reasons but it's much less visible at the higher resolutions we now use. In other words our butts are covered by megapixels.

Joe

EDIT: I popped over to 500px and found a breathtaking sunset that was originally a camera JPEG and got some pretty heavy editing. The JPEG degradation was visible on my screen at viewing res but not horrible. I copied it and enlarged a section of the sky to 400%, saved it as a PNG and posted it below. I present it here because rehess was curious about the compression grid. Look real close and you can actually see the JPEG 8x8 pixel grid cells in the sky. The heavy editing is literally causing the grid to lift up out of the photo and become visible -- we're not supposed to be able to see it -- that's the point.

Compression grid is how JPEG works it's magic. JPE... (show quote)

If that is all it is, then I'm not worried. If a file began as a JPEG, then it was created by a camera. I might decide to plumb it, to crop, and to adjust levels; with my KP I don't even need to apply sharpening. It is usually done in last than five minutes and always in one write.

selmslie Loc: Fernandina Beach, FL, USA

rehess wrote:

If that is all it is, then I'm not worried. If a file began as a JPEG, then it was created by a camera. I might decide to plumb it, to crop, and to adjust levels; with my KP I don't even need to apply sharpening. It is usually done in last than five minutes and always in one write.

Simply leveling an image is enough to introduce degradation. The grid that was created by your camera gets rotated slightly and laid over a new grid. It's bound to cause problems but you may not be able to detect them unless you are looking very closely at a smooth tone like a clear sky.

Adjusting brightness might not be a problem if it merely brightens everything by the same amount but changing contrast or applying a gradient can also be detrimental.

If your habit is to do the least amount of editing possible then you will do the least harm.