Should I be using RAW?

[SBrooks1]

I've got the Canon 10D and I shoot all my images using the best image quality JPEG (least compression). I shoot a lot of weddings, bridals and family portraits/group shots. I see all of you talking about using RAW and it intrigues me that I'm not doing this. Should I be using RAW instead? If so, why?

Thanks!
Scott

[Wildman]

I use Canon RAW mode all the time, primarily because of the capability to modify white balance after shooting the picture. The file size is somewhat larger and an extra processing step is needed. These are really not much of a problem and have saved a couple of pictures for me.

[Roger_Cavanagh]

Scott,

If your objective is the best possible image quality with the flexibility and margin for error that Wildman mentions, then shoot raw. What this costs you is possibly a more complicated workflow.

Shooting JPG's cost you quality because of the data compression and because of the use of 8-bits instead of the 12 that the camera captures - 256 values per colour vs 4096.

What you gain with JPG is speed, reduced storage and possibly a simpler workflow. If you are careful about correct exposure and WB, then most of the time no-one will be able to tell the difference.

In difficult conditions, shooting raw will give you much more latitude to successfully edit your pictures.

Personally, I shoot raw because it seems daft not to capture the best quality picture possible. I have developed a workflow that takes some time, but is mostly automatic, so it's not my time, but my PC's that is being wasted.

I think you should experiment with raw and see whether it suits your modus operandi, and you can see the benefits for you.

Regards,

[AJSJones]

With the 10D you don't have to choose one OR the other ahead of time - you can get both the best jpeg and the raw saved for each shot (in the preferences for raw+jpeg) If the jpeg is good enough, and quite often it is not a whole lot worse than the processed raw, you saved time and effort, if you don'y like the jpeg, process the raw. Reviewing the best quality jpeg is close to instant while seeing the processed raw takes a while (even if you're doing something else while it does so). The only downside is that you get a smaller number of shots per card than raw + the lowest jpeg

Andy
(wishing my D60 could do this!)

[Amazer]

I also shoot everything at max size and min compression on my S45, and often wonder whether the difference between this and RAW is worth the hassle.

Rog, would you be so kind as to give us an easy-to-follow workflow method for processing RAW.

I've read about it many times, but get confused at some stage or other as it seems to depend on what software one uses. Also, I never seem to understand at what stage you have to save to TIFF (which always crops up in the workflow) - is it the first thing to do, or do you do it after you have sharpened, cropped, adjusted, etc. And is this step (saving to TIFF) absolutely necessary? If you save directly to JPEG from RAW, is the resulting image better, worse or the same as a shot which has been captured as JPEG in-camera?

Am I right in assuming that it does not make sense to save to a lossless format except from RAW? In other words, does the situation ever arise when you need to save to TIFF (lossless) from JPEG (lossy)? What is the difference between linear and normal conversion?

Because of all the above doubts, I have always shied away from RAW. Please help me, and others like me.

I have: BreezeBrowser/PhotoElements2.0

Thanks for listening,
Ian

[gcogger]

Amazer wrote:
Rog, would you be so kind as to give us an easy-to-follow workflow method for processing RAW.

I'll let someone else talk about workflows, since I've not settled on mine yet!


I've read about it many times, but get confused at some stage or other as it seems to depend on what software one uses. Also, I never seem to understand at what stage you have to save to TIFF (which always crops up in the workflow) - is it the first thing to do, or do you do it after you have sharpened, cropped, adjusted, etc. And is this step (saving to TIFF) absolutely necessary? If you save directly to JPEG from RAW, is the resulting image better, worse or the same as a shot which has been captured as JPEG in-camera?

Am I right in assuming that it does not make sense to save to a lossless format except from RAW? In other words, does the situation ever arise when you need to save to TIFF (lossless) from JPEG (lossy)? What is the difference between linear and normal conversion?

Ian



Saving to JPEG, whether you start with a JPEG or RAW file, should only ever be done if you are absolutely sure that you are never going to make any more changes to the file. This is because every time you save as a JPEG the quality degrades - even if the file was a JPEG to start with.
In other words, nearly all the time you should save the file as a TIFF (or other uncompressed format) whether you start with a JPEG or RAW. Even if you end up with a JPEG, keep the original file from the camera, as you never know when you might want to come back to the image!

[Roger_Cavanagh]

Amazer wrote:
I also shoot everything at max size and min compression on my S45, and often wonder whether the difference between this and RAW is worth the hassle.

Rog, would you be so kind as to give us an easy-to-follow workflow method for processing RAW.

Ian,

I'll have a go, but since I'm getting a 10D and looking at using Adobe Camera Raw my old workflow, which used linear TIFF is subject to change. I did write up this workflow: http://www.rogercavanagh.com/helpinfo/06_workflow.htm. It's not entirely appropriate to you, but you may get something useful from it.

I'm also using a different toolset from you.

OK, caveats out of the way I'll try to deal with the points you raise. I'm not going to discuss the relative pros and cons for shooting raw: the assumption is we have shot raw what's the best way to handle it.

I'll deal with some of the specific points you raise and suggest an overall workflow at the end of the post.

Also, I never seem to understand at what stage you have to save to TIFF (which always crops up in the workflow) - is it the first thing to do, or do you do it after you have sharpened, cropped, adjusted, etc.

Conversion to TIFF comes at an early stage. Raw format cannot be processed directly. BB is a good tool for this conversion.



And is this step (saving to TIFF) absolutely necessary? If you save directly to JPEG from RAW, is the resulting image better, worse or the same as a shot which has been captured as JPEG in-camera?

Conversion to TIFF or another lossless format such as PNG or JPEG 2000 is necessary. There is not much point is taking the trouble to shoot raw, and then converting to lossy JPG.

I don't know whether conversion to JPG gives the same result as saving JPG in-camera. I would guess not necessarily because you can adjust additional things in conversion - most notably, JPG quality.

Am I right in assuming that it does not make sense to save to a lossless format except from RAW? In other words, does the situation ever arise when you need to save to TIFF (lossless) from JPEG (lossy)?

No you are not right. Even if you shoot JPG, when you load pictures to your editor - Photoshop Elements in your case, you should save the edited image in a lossless format. The PE native format would be acceptable. I guess it's PSD same as Photoshop. The reason is that every time you save and edit a JPG, compression routines are applied and you lose even more data. The only time I ever save a JPG is for use on the web or in my local PC photo album. This is only done after I have completed all edits. If editing has been complex, I will save a PSD version that I can go back to.


What is the difference between linear and normal conversion?

The linear conversion is the closest we can get to the "raw" image data. It is the picture data without being corrected for saturation, contrast, sharpness and most obviously gamma/brightness, which is why linear TIFF's look dark. Users of linear TIFF's like them because they give the cleanest, most unprocessed data so we can use the power of Photoshop to turn our pictures in great images. :) If you've been hanging around here long enough, I expect you'll have seen references to LinearSharpen.

Non-linear conversion does apply adjustments for brightness, etc., so the picture is immediately usable. We may wish to do further edits, but you can see what you've got a picture of.

I have: BreezeBrowser/PhotoElements2.0

PE is not capable of editing 16-bit images. You do, therefore, lose some of the benefit of shooting raw. Raw images are 12-bit, which gives 4096 values per colour - 4096 red x 4096 green x 4096 blue different colours. 8-bit has only 256 values per colour.

OK, using the software you have, here's an outline of a workflow.

- Download the images from the camera

I would get Chris Breeze' Downloader program to do this. The current version 1.5 is free; he's working on another version for which he will charge.

- If the pictures are important, make a copy of the files.

- Use BB to review the pictures and delete anything that's no good or you don't want.

- Output Option 1: the pictures don't really need special editing, so use BB to convert and process to final output.

If you open the "Convert raw image" window in BB, you will see you can specify settings for contrast, saturation, etc. To some extent, these are a matter of taste and the best settings will probably vary with subject: scenery can look good with saturated colours, people not always so good. :)

BB also has post-processing options so you can make adjustments and sharpen, and even change the size. If you are converting a batch of files, then auto-levels usually works OK. Personally, I wouldn't clip as much as the default 0.5%. Saturation agains can be tuned to taste.

Sharpness settings will need to be set to suit the S45. I've never used one of these, so I don't know what settings are sensible to suggest, but a little experimentation will tell what's reasonable. I would be inclined to use a small radius 0.5 - 0.7 and increase the amount to increase sharpness - too much and you will see artefacts such as haloes around edges and noise appearing.

Since these are going to be "final" images, you could choose TIFF or JPG (to save some space) with JPG you need to choose a quality setting. This will vary depending what you want the image for - medium is OK for web or emails, but you will want higher quality for printing.

Once you've set everything to your requirements, click "convert all" and go and have a beer. :)

Option 2: Images will be edited in PE

In this case I would turn off or set to lowest possible settings for sharpness, contrast, etc. You can then adjust these in PE with much more control. Output format should be 8-bit TIFF. Unless you get hold of a profile for using linear TIFF, then conversion method will have to be normal (or composite if you are trying to recover detail from an over-exposed image).

I would set raw conversion preferences to embed the sRGB profile. Start conversion and go and have another beer.

At the end of all this you end up with a bunch of TIFF files that you will edit with PE. In Photoshop, it is possible to create actions to batch process all the images without manual intervention. You can't do this in PE, so it's one at a time.

Make your edits, and save the PSD file.

Obviously, edits will be custom to each picture, but I would adjust levels, saturation, and finish with sharpening. I've never used PE, so I'm not sure what's available compared to PS.

And go on to do whatever you what for print, etc.

Some people will save the converted TIFF's. I don't this is necessary. They will take up a lot of space, and you can always reconvert, if you need it again.


Does that make sense? Does it help?

Regards,

[Amazer]

Hi Roger,

Just seen your reply to my post. You rock man!!

Thanks ever so much for taking the trouble to go into so much detail. You were very very helpful. The RAW workflow is now making much more sense to me, although there are still a couple of points which are not quite clear - aren't there always? :) But thanks to this forum and people like you, they are getting fewer and fewer.

But I will make a "dummy run" this evening at home (I am writing from work right now :p) and will write down any steps which I am not quite happy with.

I will then abuse some more of your time and patience and post another message here. If you don't mind.

Thanks again.

Ian


ps: Hope you enjoy your new 10D - lucky you!!

[Amazer]

Hi gcogger, and thank you for your reply.

gcogger wrote:

This is because every time you save as a JPEG the quality degrades - even if the file was a JPEG to start with.

Referring to your comment on quality degradation when saving, does this mean that if I shoot JPEG in-camera and transfer (edit copy, edit paste) the files from my Card Reader to my hard drive I am already losing quality? Or does this happen only if I use "save as" to transfer between drives or folders, or to rename. This seems pretty drastic to me. Why should a file lose data simply because you save it to another folder? Excuse my ignorance! :)

Regards,
Ian

[Roger_Cavanagh]

Amazer wrote:
Hi gcogger, and thank you for your reply.

gcogger wrote:

This is because every time you save as a JPEG the quality degrades - even if the file was a JPEG to start with.

Referring to your comment on quality degradation when saving, does this mean that if I shoot JPEG in-camera and transfer (edit copy, edit paste) the files from my Card Reader to my hard drive I am already losing quality? Or does this happen only if I use "save as" to transfer between drives or folders, or to rename. This seems pretty drastic to me. Why should a file lose data simply because you save it to another folder?

Ian, no you don't lose anything just copying the files from CF to camera or using Explorer to move them around your PC. Only when you load the image to Photoshop, make one or more edits and then save an updated version of the file. It is at that point that PS applied JPG compression and a little more picture data goes down the cyber drainpipe. :)

Cheers,

[Amazer]

Hi there Roger,

I have gone through your first (long and very much appreciated) reply to my post. I have also printed it for easy reference!

I wish to clarify a few points and am using your post as a prompter. Could you please reply or comment my observations/conclusions. If a YES/NO or TRUE/FALSE is enough, I am happy with that; but feel free to expand if you have the time and inclination. I have numbered my comments for easy reference.

Here goes.


Roger_Cavanagh wrote:

Conversion to TIFF comes at an early stage. Raw format cannot be processed directly. BB is a good tool for this conversion.

Conversion to TIFF or another lossless format such as PNG or JPEG 2000 is necessary. There is not much point is taking the trouble to shoot raw, and then converting to lossy JPG.

I don't know whether conversion to JPG gives the same result as saving JPG in-camera. I would guess not necessarily because you can adjust additional things in conversion - most notably, JPG quality.

1A- If you shoot in JPEG, the image cannot be bettered much, whichever software one uses later, because JPEG does not react favourably to a lot of afterwork. The most one should attempt is a bit of sharpening, colour correction and cropping.

1B- If you absolutely must do a lot of editing, save the out-of-camera JPEG immediately to lossless TIFF, do your thing and save again to TIFF when happy. Use this as your stock image for that particular shot. Do not save back to JPEG as this will lose you some quality. Always keep a copy of the original file, in whatever format you shot it, just in case. If you need your photo for the web or e-mail, save as JPEG from the stock TIFF, then post to the web. This will give a better result than the original shot, even if you did not apply any correction.

1C- In any case, if you intend to attempt a high-quality print from your work, do not use JPEG. Use RAW.

Even if you shoot JPG, when you load pictures to your editor - Photoshop Elements in your case, you should save the edited image in a lossless format. The PE native format would be acceptable. I guess it's PSD same as Photoshop. The reason is that every time you save and edit a JPG, compression routines are applied and you lose even more data. The only time I ever save a JPG is for use on the web or in my local PC photo album. This is only done after I have completed all edits. If editing has been complex, I will save a PSD version that I can go back to.

2A- The native format of your editing software (if it uses one) is always the preferred lossless format to save to. This has nothing to do with the quality of the final image, it's just that the data will be handled faster. If you decide to print, even at a commercial printer, the file format is not important as long as the quality and required resolution are there.

The linear conversion is the closest we can get to the "raw" image data. It is the picture data without being corrected for saturation, contrast, sharpness and most obviously gamma/brightness, which is why linear TIFF's look dark. Users of linear TIFF's like them because they give the cleanest, most unprocessed data so we can use the power of Photoshop to turn our pictures in great images. :) If you've been hanging around here long enough, I expect you'll have seen references to LinearSharpen.
Non-linear conversion does apply adjustments for brightness, etc., so the picture is immediately usable. We may wish to do further edits, but you can see what you've got a picture of.

2B- In BreezeBrowser, always choose linear conversion for the best results. Useless to use 16-bit for post-processing in PE2.0, even though it is available in BB, so use 8-bit. Un-check "Enable Post Processing" box as we will be doing this in PE.

PE is not capable of editing 16-bit images. You do, therefore, lose some of the benefit of shooting raw. Raw images are 12-bit, which gives 4096 values per colour - 4096 red x 4096 green x 4096 blue different colours. 8-bit has only 256 values per colour.

2C- The only way to get maximum benefit from RAW is using 16-bit. Forget it if you don't have Photoshop proper. There is no way around this.

Option 2: Images will be edited in PE

I would set raw conversion preferences to embed the sRGB profile.

2D- Help ??? No idea what you're talking about Rog.

At the end of all this you end up with a bunch of TIFF files that you will edit with PE. In Photoshop, it is possible to create actions to batch process all the images without manual intervention. You can't do this in PE, so it's one at a time.

Make your edits, and save the PSD file.

[b]2E- I'm a little lost here too. When you mention "batch processing," are you referring to the procedure of exporting the images from BB to Photoshop?

Obviously, edits will be custom to each picture, but I would adjust levels, saturation, and finish with sharpening. I've never used PE, so I'm not sure what's available compared to PS.

2F- Yes, PE can do all the above no problem.

That's it. I'm done. Hope I did not utter too much nonsense :)

Best regards and thanks.

Ian.

[Roger_Cavanagh]

Let's hope the quoted quotes don't get too confusing. :)

Amazer wrote:
Hi there Roger,

I have gone through your first (long and very much appreciated) reply to my post. I have also printed it for easy reference!

I wish to clarify a few points and am using your post as a prompter. Could you please reply or comment my observations/conclusions. If a YES/NO or TRUE/FALSE is enough, I am happy with that; but feel free to expand if you have the time and inclination. I have numbered my comments for easy reference.

Here goes.


Roger_Cavanagh wrote:

Conversion to TIFF comes at an early stage. Raw format cannot be processed directly. BB is a good tool for this conversion.

Conversion to TIFF or another lossless format such as PNG or JPEG 2000 is necessary. There is not much point is taking the trouble to shoot raw, and then converting to lossy JPG.

I don't know whether conversion to JPG gives the same result as saving JPG in-camera. I would guess not necessarily because you can adjust additional things in conversion - most notably, JPG quality.

1A- If you shoot in JPEG, the image cannot be bettered much, whichever software one uses later, because JPEG does not react favourably to a lot of afterwork. The most one should attempt is a bit of sharpening, colour correction and cropping.

Your statement is too black and white, :) images that start out life as JPG can be edited fine. Don't forget you're not editing a JPG image, you are editing a Photoshop image. Raw format will give more margin for error in difficult situations. For instance, a lot of the pictures I took at Petrohue Falls in Chile needed a lot of adjustment because I had exposed for the highlights and details in the under-exposed area needed bringing out during editing. (Examples here: http://www.rogercavanagh.com/galleries/petrohue/index.htm).

1B- If you absolutely must do a lot of editing, save the out-of-camera JPEG immediately to lossless TIFF, do your thing and save again to TIFF when happy. Use this as your stock image for that particular shot. Do not save back to JPEG as this will lose you some quality. Always keep a copy of the original file, in whatever format you shot it, just in case. If you need your photo for the web or e-mail, save as JPEG from the stock TIFF, then post to the web. This will give a better result than the original shot, even if you did not apply any correction.

I would make this more positive. If you have made any edits to an in-camera JPG, always save it to a lossless format. Otherwise, I agree.

1C- In any case, if you intend to attempt a high-quality print from your work, do not use JPEG. Use RAW.

Plenty of people do make good prints from JPG's.

Even if you shoot JPG, when you load pictures to your editor - Photoshop Elements in your case, you should save the edited image in a lossless format. The PE native format would be acceptable. I guess it's PSD same as Photoshop. The reason is that every time you save and edit a JPG, compression routines are applied and you lose even more data. The only time I ever save a JPG is for use on the web or in my local PC photo album. This is only done after I have completed all edits. If editing has been complex, I will save a PSD version that I can go back to.

2A- The native format of your editing software (if it uses one) is always the preferred lossless format to save to. This has nothing to do with the quality of the final image, it's just that the data will be handled faster. If you decide to print, even at a commercial printer, the file format is not important as long as the quality and required resolution are there.

By and large, yes, but some commercial printers will only accept certain formats.

The linear conversion is the closest we can get to the "raw" image data. It is the picture data without being corrected for saturation, contrast, sharpness and most obviously gamma/brightness, which is why linear TIFF's look dark. Users of linear TIFF's like them because they give the cleanest, most unprocessed data so we can use the power of Photoshop to turn our pictures in great images. :) If you've been hanging around here long enough, I expect you'll have seen references to LinearSharpen.
Non-linear conversion does apply adjustments for brightness, etc., so the picture is immediately usable. We may wish to do further edits, but you can see what you've got a picture of.

2B- In BreezeBrowser, always choose linear conversion for the best results. Useless to use 16-bit for post-processing in PE2.0, even though it is available in BB, so use 8-bit. Un-check "Enable Post Processing" box as we will be doing this in PE.

My one experiment with LinearSharpen and 8-bit linear TIFF proved that it doesn't work well. The aggressive processing in LS caused awful posterisation. I have not tried using 8-bit linear TIFF with other options such as Fred Miranda's Linear Pro Batch or the About Digicam profiles (some of which are bundled with BB).

PE is not capable of editing 16-bit images. You do, therefore, lose some of the benefit of shooting raw. Raw images are 12-bit, which gives 4096 values per colour - 4096 red x 4096 green x 4096 blue different colours. 8-bit has only 256 values per colour.

2C- The only way to get maximum benefit from RAW is using 16-bit. Forget it if you don't have Photoshop proper. There is no way around this.

Probably true, but there are other editors that can handle 16-bit images that don't cost as much as PS.

Option 2: Images will be edited in PE

I would set raw conversion preferences to embed the sRGB profile.

2D- Help ??? No idea what you're talking about Rog.

This takes us into the realm of colour management, which boils down to making sure that the colour we see that is captured by our camera, displayed on screen and printed all match as closely as possible. See this page http://www.rogercavanagh.com/helpinfo/10_colourmanagement.htm on my site for links to some introductory material.

At the end of all this you end up with a bunch of TIFF files that you will edit with PE. In Photoshop, it is possible to create actions to batch process all the images without manual intervention. You can't do this in PE, so it's one at a time.

Make your edits, and save the PSD file.

[b]2E- I'm a little lost here too. When you mention "batch processing," are you referring to the procedure of exporting the images from BB to Photoshop?

No, by batch processing I mean a sequence of edits that are applied to a bunch of images without intervention once the process has been started. In Photoshop terms, the sequence of edits is called an "action". PS allows me to aim that action at multiple files and come back when processing is finished.

BB does batch conversion of multiple raw files when you choose "convert all" from the raw conversion window and batch editing, if you enable post -processing (per Option 1).

I want to finish by making it clear that the case for raw over in-camera JPG is not cut and dried. Raw does give you more margin for error and more opportunity to produce a good image in difficult shooting conditions. I like to use a golfing analogy here: playing with a 500 dollar graphite-shafted driver doesn't guarantee that every drive I hit will be better than that of the guy using the 50 buck special. Once in a while I hit one perfectly and outdrive him by 50 yards whatever he does. If we both hit one off the toe, my technology will usually get me a better result than his, but a lot of the time we're gonna be not to far apart.

In other words, using raw doesn't get you a better result every time. I am not saying that. What I am saying is that rather than worry about when raw is appropriate and when it it is not, I'll use raw all the time. I have a workflow that does this for me without extra effort, though it took a good while for me to work it out.

Feel free to ask supplementaries - I have started to confuse myself by getting over-complicated. :)

Regards,

[AJSJones]

Roger_Cavanagh wrote:


In other words, using raw doesn't get you a better result every time. I am not saying that. What I am saying is that rather than worry about when raw is appropriate and when it it is not, I'll use raw all the time. I have a workflow that does this for me without extra effort, though it took a good while for me to work it out.




Thanks Roger - your posts are always helpful! Also glad to see the discussion of when raw helps - I agree having the raw is always a good idea, but the in-camera jpeg is often sufficient (and eliminates the need for the raw processing) and the 10D has the advantage over the D60 of allowing you to have both the raw and the best jpeg so you can decide post facto

Andy

[Roger_Cavanagh]

AJSJones wrote:
Thanks Roger - your posts are always helpful! Also glad to see the discussion of when raw helps - I agree having the raw is always a good idea, but the in-camera jpeg is often sufficient (and eliminates the need for the raw processing) and the 10D has the advantage over the D60 of allowing you to have both the raw and the best jpeg so you can decide post facto

Andy,

Well, tomorrow possibly, or the next day, I shall have my 10D. :) Then I can find out whether blowing an extra 2mb per image for the super-duper embedded JPG is worth it. Personally, I am little peeved that there isn't an option not to have a JPG at all.

Regards,

[AJSJones]

Roger - the more choices the better, so I agree there should be a No jpeg option. If you have a good raw workflow, especially if it can be batched, and want the most from each and every image (you know you want to process each raw) then the jpegs don't help much. With full-size jpegs I can quickly decide which raws deserve the time and effort and save time on the ones that don't. My position is clearly tainted/influenced by the fact that I have 2 1GB microdrives and a 340MB so I rarely run into storage problems, and can therefore "afford" the extra couple of MBs per file. Otherwise I'd also be wanting a no jpeg option!

Looking forward to your impressions of the 10D

Andy

[Amazer]

Hi again Rog (and all),

After going through all your replies (thanks) and visiting a couple of websites like Philip Ganderton's and yours Rog, for example I feel much more comfortable now with RAW, and, seeing it is available on my S45, I have decided, at least for the time being, of taking your advice and shooting RAW as a default.

Although it is true that most of the time one is hard-pressed to see the difference between a converted RAW and an in-camera best JPEG, the standardisation and discipline that shooting RAW imposes makes for more consistency. At least that's how I see it.

One thing that I am still unsure of concerns the resizing that one invariably needs to do depending on whether one is printing or posting to the web, or both.

I haven't decided yet whether it would be better to completely finalise your PSD or TIF in its original size (including sharpening), then re-sharpen the resized file, or, do not do any sharpening at all until after resizing. And if this was the case, would you resize your PSD/TIF, sharpen and convert to JPEG or would you sharpen your JPEG.

Ian

[AJSJones]

For printing, sharpening is the last step before going to the "Print..." operation (and some will not save the resulting sharpened file unless they know they'll be printing at that size frequently). Personally, I feel that some modest USM while in 16bit immediately after the CRW -> TIF conversion is not a bad thing, but save any 8bit sharpening till the end.

The details of sharpening (i.e what is optimum) depend on the size and resolution of the print, so there isn't a "one setting fits all" even for a single image.

A little USM is usually in order after downsizing for the web.

Never sharpen a jpeg.

Andy

[Roger_Cavanagh]

Ian,

I think Andy's comment that one size doesn't fit all is entirely appropriate. One of the difficulties with digital image processing is that it is hard to decide when enough is enough.

To answer your question specifically, I (almost all the time) finalise my images at full-size, then downsize for web use - currently, my preferred size is 720 pixels on the long side. If the image is not yet in 8-bit then I'll convert to 8-bit sRGB and apply a little sharpening. For my old D30, I used to use one of the actions from Fred Miranda's CS Pro.

Downsizing tends to take away a little "pop" from an image, and a touch of USM brings it back.

Regards,

[Blast]

Folks,

Just want to say thank-you for this whole thread. I have gained, I hope, at least perspective. This wont mean much to most, but for me, I am a sponge when i am sitting in a duck blind with two or more old veterans talking about "working Mallards" with a duck-call. I am an old world class duck caller and ALWAYS learn when I can hush long enough to HEAR what they are saying. I love to listen when I can.

Thanks again for the "pearls" you folks hand out. You will never know all the people you help as with this medium, you never know who is just sitting around listening.

Blast- Harry