What's in a Web Image ?
 

Never fails to amaze me of the detail you can get out of a good telephoto lens when shooting from a distance ,cropping and processing.
The following three pics show:
Pic 1 un-cropped and un-processed raw file (just converted and resized)
Pic 2 Cropped but un-processed
Pic 3 The finished article.

Of course, the finished shot is only good for the web as it is almost 100% crop so would not print more than 6 x 4 at the best, but pefectly acceptable as a web image IMO.

It looks good enough to be worth trying to print a bit larger than that?

I am pleased you have posted the series, as I often wonder if this rather strange process is what is meant to happen to an image or whether I was doing something wrong and creating unecessary work.

The only comment is whether you are converting slightly early in the process? The result says 'no' and it is what I now do for older 350D .CR2 images*. However it does mean that the last phase does not have access to the full range of stuff in the RAW file doesn't it?

Not sure what you use for conversion, but Duncan sent over some Capture 1 screens (and I looked at one or two macThings) and they seemed to have a lot of fine tuning gadgets if one can get round the technology behind it. Fortunately no longer a problem now I have gone Nikon and use NX2 where there is no need to convert at all except for web jpgs.

*some grovelling apologies :o due in this area, I think you did try to point out how limited DPP was before I came to the same conclusion empirically

Hi Chris, generally I do very little processing in raw apart from tweaking the exposure and cropping - I then convert to a 16bit TIFF and send to CR2.

A big advantage of shooting a RAW file is that you can convert to 16-bit TIFF file before the main editing starts. TIFF files are larger than JPEG files, but they retain the full quality of the image. They can be compressed or uncompressed, but the compression scheme is lossless, meaning that although the file gets a little smaller, no information is lost. It is only when I have finished editing that I convert to a 8 bit file which allows to be saved as a jpeg.

Re the printing size, the image is only 800 pixels which means that printing at say 200 dpi would only give a 4" image. For a 6" image you would have to print at 133 dpi. If you printed at 10" you would be at 80dpi which would be cr*p quality IMO.

edit Where a lot of people go wrong IMO is that they convert the raw to a jpeg (8bit) and then do more editing with this lossy format.

I appreciate that a tiff is lossless on successive saves, but does it retain everything there was in the RAW? Working on the RAW(nef) the way I do, I can and sometimes do go back to the exposure compensation and colour temp if I have not got what I want through non RAW editing
Agreed, no good. A problem with my even older files from E4500 and FZ7, but on the old favourites that I still hope to improve, I do the 1st save from original in tif; on the FZ7 which had the option to shoot in tif or jpg (but not RAW), I couldn't see any difference, though the possibility of sunsequent lossless operation is there

Yes I believe it does Chris. I never alter my raw file so it will always remain as a digital negative to re process if required. You can also work on a uncompressed tiff in DPP if you want so you could bring it back from CS2 (though I have never had a need to do this myself).

I do almost all my PP in Lightroom, working on the RAW file. Lightroom doesnt actually alter the RAW at all, just creates an editing sidecar file with the changes in, which you see rendered on what is in effect a preview image. Then when you wish, you can create a converted image as a TIFF or JPEG (or DNG, though I dont quite see the point!) in 16 or 8 bit for further work. I must admit, since using Lightroom, I very rarely need Photoshop - only for stuff involving local corrections or layers, which I hardly ever do. (This is why I havent got to grips with CS3 yet!)
To answer your original point Roy, with good glass you can get astonishing detail, even at 100% crop, although as you say, the printing options decrease.

I also have Lightroom Gordon (as well as Raw Therapee) and tried it for my raw processing for several months before reverting back to DPP and CS2. Just did not feel at ease with the interface but I guess that was down to my reluctance to change from doing the bulk of my processing in CS2 which I know pretty well. Although Lightroom has vastly superior editing tools to DPP I am convinced that DPP does a better job of converting the raw than ACR.

I had a 2006 version of DPP - cant remember which that was - when I first went digital. I found it really cumbersome to use, and wasnt that happy with the quality of files it generated either. At the time, I went for Capture 1 v3.5, which I found to be much easier to use and more flexible, as well as given better IQ. Lightroom takes a bit of getting used to ( I spent about 4 weeks in front of the computer learning the programme while recovering from a knee ligament repair), but after some initial struggles, I quickly got the hang of it, and it means I can find my images so much more easily without haveing to catalogue them in another application. It gives very good print quality too, so I can go from import, through cataloguing, development and printing in one programme.

Roy - Chris is right to question this because this is definitely NOT the case. ANY edit in PS modifies pixel values and therefore destroys information. I appreciate that if you work with layers that is not destructive editing and you can return to the original, but when you flatten the image for web or print it you are then taking the modified values. In contrast editing in RAW does not. You should do as much editing in RAW as possible and then switch to PS for edits that can not be carried out in your RAW processor eg selective edits, althhough I believe that LR in V2 now has some selective processing capability.

By way of an example here are three version of an image that I took the lake district a few weeks ago. This is deliberately underexposed as part of a bracketed series but it serves to show the difference between processing in RAW and PS. Obviously with correctly exposed images, the effects are less exederated, but will still be there.

Image 1 is the as shot RAW converted to 16bit with no adjustments and then resized and converted to JPG
Image 2 is the above image with a curves adjustment added in PS.
Image 3 is the RAW image with exposure, fill light & black level adjustments in LR and then converted to 16bit and the resized and converted to JPG.

In each case I have superimposed the histogram in the corner. Note in images 1 & 3 the histogram is continuous, were as in image 2 there are large gaps - denoting the complete absence of pixels with that value, the net result being loss of tonal range & shadow detail in particular, which is visible in the shadows on the hills.

To lean how to get to grips with the LR/PS RAW converter interface try either of these books: -

Real World Camera RAW by Bruce Fraser
Photoshop Lightroom by Martin Evening


Its worth the effort :D

I assume you mean CS2. Does CS2 (or 3) allow use of layers on 16bit tiff? If not it seems one of the major advantages of CS, ie use of layers, is gone.

PS I only have PSE4 and not inclined to upgrade for the little I use it for....like converting CR2, :mad: grrrr, as DPP batch is the one casualty of mac OS10.5, waiting with partly bated breath for reply from Canon.