Question about Digital Photo Professional

[Manfred M]

To disagree with the majority; I often work from jpegs only, and frankly for anything being posted on the web or viewed from the computer. It is certainly easier to get artifacts when one edits a jpeg, but if all one does is a bit of trimming and some other minor tweaks on a high quality jpeg, generally one cannot see any quality loss.

1. There is a bit of a myth regarding quality loss of resaving a jpeg (yes I have done this to verify this). The compression and data loss occurs on the first iteration, after that, there is no change in image quality (and with the tools I was using, the image size was identical); I ran a good 10 or 12 re-saves. From a software standpoint this does make sense; the biggest data loss will be from going from 12 or 14-bit data down to 8-bit. Throw the same image at the compression algorithm; it will compress the same way over and over again.

Compress to a lower quality level, you can start seeing these compression artifacts - something gets thrown away to create smaller file size.

2. Most computer screens you view the image on are at best 8-bit natively. The common TN displays are natively 6-bit and use dithering (cycling adjacent pixels rapidly) to emulate 8-bit. Higher end IPS screens are 8-bit natively and use dithering to emulate 10-bit colours.

3. The human eye can distinguish almost 10-million individual colour shades. 8-bit images are 16-million colours (assuming that your screen can reproduce them). Pro colour printers are actually limited to reproducing a few hundred thousand distinct colours (remember, these printers use a limited number of cartridges (8 on my Epson 3880) and vary dot size to create colours).

4. Any image you see in your web browser is going to be 8-bit at best.

So, why do I use RAW files for my more serious work? Simple; the more data one has, the smoother the editing program can create intermediate colour values, and this virtually eliminates creating artifacts while manipulating colours. As RAW files do not have colour balance (or gamma settings) "baked" in, these values can be changed too, which is really quite useful as well.

[bnnrcn]

Manfred what I understand from your post is that 'if you are a good shooter you don't need PP very much and you can make your minor adjustments even with a jpg format'.Thank you

To disagree with the majority; I often work from jpegs only, and frankly for anything being posted on the web or viewed from the computer. It is certainly easier to get artifacts when one edits a jpeg, but if all one does is a bit of trimming and some other minor tweaks on a high quality jpeg, generally one cannot see any quality loss.

1. There is a bit of a myth regarding quality loss of resaving a jpeg (yes I have done this to verify this). The compression and data loss occurs on the first iteration, after that, there is no change in image quality (and with the tools I was using, the image size was identical); I ran a good 10 or 12 re-saves. From a software standpoint this does make sense; the biggest data loss will be from going from 12 or 14-bit data down to 8-bit. Throw the same image at the compression algorithm; it will compress the same way over and over again.

Compress to a lower quality level, you can start seeing these compression artifacts - something gets thrown away to create smaller file size.

2. Most computer screens you view the image on are at best 8-bit natively. The common TN displays are natively 6-bit and use dithering (cycling adjacent pixels rapidly) to emulate 8-bit. Higher end IPS screens are 8-bit natively and use dithering to emulate 10-bit colours.

3. The human eye can distinguish almost 10-million individual colour shades. 8-bit images are 16-million colours (assuming that your screen can reproduce them). Pro colour printers are actually limited to reproducing a few hundred thousand distinct colours (remember, these printers use a limited number of cartridges (8 on my Epson 3880) and vary dot size to create colours).

4. Any image you see in your web browser is going to be 8-bit at best.

So, why do I use RAW files for my more serious work? Simple; the more data one has, the smoother the editing program can create intermediate colour values, and this virtually eliminates creating artifacts while manipulating colours. As RAW files do not have colour balance (or gamma settings) "baked" in, these values can be changed too, which is really quite useful as well.

[dje]

Hi Binnur

If you open a RAW file in DPP for editing, it will come up with the RAW tab selected in the Tool Palette. You can make adjustments here and then select the RGB tab and the adjustments made in RAW will carry over. You can then make separate (additional) adjustments in the RGB tab. All of these adjustments are saved when you Save As ...

If you open up a jpeg for editing in DPP, only the RGB tab is available for editing. Even if you open a previously edited image, the settings in the RGB tab will initially be the default values. This part of the software is not a parametric editor.

Hope this makes sense.

Dave

[bnnrcn]

Thank you Dave it really makes sense.When I edit a raw image in RGB tab and then if I convert it to a jpg image,the edits made in the RGB part is not shown.That's why I have to edit the jpg format again after converting from raw format.It is good that you you confirmed how DPP worked because I was a bit confused when I converted from raw to jpg and saw that some of my edit wasn't there

Hi Binnur

If you open a RAW file in DPP for editing, it will come up with the RAW tab selected in the Tool Palette. You can make adjustments here and then select the RGB tab and the adjustments made in RAW will carry over. You can then make separate (additional) adjustments in the RGB tab. All of these adjustments are saved when you Save As ...

If you open up a jpeg for editing in DPP, only the RGB tab is available for editing. Even if you open a previously edited image, the settings in the RGB tab will initially be the default values. This part of the software is not a parametric editor.

Hope this makes sense.

Dave

[ajohnw]

As some one who also works on jpgs I will add to Manfred's comment. It's all down to the information needed to capture the shot. For information read stops / exposure values. Raw has more but has to be manipulated to less stops to display it on a PC screen. The camera also does the same thing when it produces the camera jpg. Often providing highlights aren't severely clipped all that will need to be done to a jpg is to brighten up dark areas a little if needed. Then the usual processing steps can be used. One of the more interesting facts is that to some extent it's rather difficult to get any more stops out of a raw file than what the camera will put in a jpg. Many cameras basically map 9 stops or even more into a jpg and if that is sufficient to capture the scene the chances are that it can successfully be processed without going near a raw file. Where all of them are used some areas of the shot will be too dark - if the high lights aren't clipped. Depending on the degree they can be brightened up. Interesting thing is that similar problems will also occur working from raw. In a nut shell say a raw file has 12 bit colour information it needs to be squeezed in some way to get it into the 8bit colour space of sRGB. Squeezing results in a loss of contrast some where. A camera jpg usually squeezes more at the dark end than at the bright end. The user has the choice from raw.

Best thing to do really is set the camera for both and select best quality jpg's and try it.

One of the reasons for people having mixed views on the use of jpg's is down to some not noting the changes that have been made in cameras either stop wise into jpg's or the quality of the jpg's they produce. I have one that can put over 10 stops into a jpg. Not used it yet. Be interesting to see what can be done with it. That one seems to squeeze all of the tones.

There is no point on commenting about jpg's being lossy. People continue to assume that they are noticeably so.

Where I find raw most useful is in situation where I can't obtain a shot that looks natural from a jpg so I develop 2 exposures from the same raw file and blend them using HDR software. Also in other high dynamic range situations that need a good capture of both the dark and bright ends. One way of doing that is to develop raw in a way that gives good contrast to both the dark and bright ends and then restore the mid tones manually. Another way is to get the mid tones right and leave space for contrast improvement at both the dark and bright ends and then restore those manually. This is where those curves I suggested you should leave alone for a while come in. Many don't bother and one side effect of that you will see is rather unbelievable clouds and other odd things where these have been bought up in an entirely different way. There are many other options in this area as well.

Raw can be advantageous because it can be used to produces deeper colour image formats. Tiff's etc. When contrast etc is stretched in a jpg bands can appear because of the precision it has. A deeper colour space image will allow the hidden precision to be slowly bought in. Being honest though I have never experienced this on a camera jpg. The usual limit on those is colour artefacts in area that have been brightened too much. Where I have seen it is on work produced by others from raw. it can crop up far too soon. I've no idea why this happens.

John
-

[Manfred M]

Manfred what I understand from your post is that 'if you are a good shooter you don't need PP very much and you can make your minor adjustments even with a jpg format'.Thank you

No Binnur, I'm not saying that at all.

What I am saying is that depending on one's needs, working from a jpeg (with minor edits) might be good enough.

As as example, the images I post to Facebook are often "snapshots", which don't deserve or need the quality of an image that goes through PP. I would never post an image to a serious photography website without at least a bit of PP work.

[jcuknz]

Binnur I would go further and say you are not a good photographer if you do not accept and use the tools that have been devised for post processing when they are needed bearing in mind that merely downloading from the camera is a form of PP.

PP is a part of photography and has been from its earliest days ... it is just the form and way of it that has changed.

If your DPP editor has been devised this way you should be able to adjust it so that saving a jpg incurs little or no compression, I have my PSP organised this way except when I want to make a compressed file which is a deliberate action on my part rather than an automatic one by the programme for me.