PDA

View Full Version : Adobe Photoshop RAW converter plug-in released


mskad
18th of February 2003 (Tue), 22:24
Works with PS 7 (but not 6) and with Elements 2.0.

Have a look here (http://www.adobe.com/store/products/plugIn.jhtml?id=catPlugins_Adobe_CameraRaw).

Mskad.

pigasus
19th of February 2003 (Wed), 04:18
I've just had a look, but it doesn't tell me which cameras/RAW files are supported. The link to further info doesn't work. Any knowledge on this front?

pigasus
19th of February 2003 (Wed), 04:42
Found the answer to my question - and a fairly detailed review: http://www.outbackphoto.com/artofraw/raw_04/essay.html

Now - will someone with a D60 buy this plugin and compare it to Pekka's D60 LinearSharpen? :)

Roger_Cavanagh
19th of February 2003 (Wed), 05:46
Doesn't seem to be available in the UK yet. The US site will only sell to US and Canada. Anyone know any different?

Regards,

wcapald
19th of February 2003 (Wed), 07:16
It needs 7.01 I think as well

RAW files supported

The Photoshop Camera Raw plug-in supports many
digital cameras that can produce raw image files.
Canon
EOS-1D
EOS-1Ds
EOS-D30
EOS-D60
PowerShot 600
PowerShot A5
PowerShot A50
PowerShot G1
PowerShot G2
PowerShot G3
PowerShot S30
PowerShot S40
PowerShot S45
PowerShot Pro70
PowerShot Pro90 IS
Fujifilm
FinePix S2 Pro
Minolta
DiMAGE 5
DiMAGE 7
DiMAGE 7i
DiMAGE 7Hi
Nikon
D1
D1H
D1X
D100
Coolpix 5000
(with firmware version 1.7)
Coolpix 5700
Olympus
E-10
E-20N
C-5050 Zoom

Wayne

pigasus
19th of February 2003 (Wed), 07:37
From the early reports I've seen this looks like a winner. And if it's as good as they say, then as new cameras come out we won't have to arrange for Pekka to own them so that we can get a new LinearSharpen action to suit. :D

Assuming, of course, that Adobe adds them promptly to the plug-in.

pigasus
19th of February 2003 (Wed), 10:13
Roger_Cavanagh wrote:
Doesn't seem to be available in the UK yet. The US site will only sell to US and Canada. Anyone know any different?

Try here Roger:

http://www.transaction-one.com/gs/html/a/adobe/shopping001/en/start.jsp

Cheers.

mrbobco
19th of February 2003 (Wed), 12:47
here's the info on the supported cameras...

i'm buying it now...well let all know what transpires :)

adobe raw plugin...supported camera info (http://www.adobe.com/products/photoshop/cameraraw.html)

bob

rrb01
19th of February 2003 (Wed), 19:59
Well, I bought it but the download isn't working. Adobe support says they've had problems all day with it, and maybe tomorrow.

T a z
19th of February 2003 (Wed), 21:30
I haven't as yet purchased it (although I probably soon will) however I don't see how it could possibly offer the level of sharpening quality that Pekka's LS actions do.

As I understand it, the reason the LS actions produce such stunning sharpening quality is because all the sharpening is happening on the linear raw data (correct me if I'm mistaken on this, Pekka). Apparently, sharpening applied to the data in it's linear raw state does far less damage to the image than does other forms of sharpening, and thus delivers a better quality image.

Now...Adobe *MAY* be doing some sharpening to the linear data...but considering how quickly it converts the image, I would think any sharpening algorithms being done are no where near the level of complexity that are being perfomed by LS. Pekka's actions involve a huge amount of data manipulation by the computer's processor.

But as I said, I haven't yet purchased the Adobe product, so all the above may be balderdash speculation on my part. I would HOPE that it would sharpen as well, but I guess only our experimentation will tell, yes?

Dans_D60
19th of February 2003 (Wed), 22:15
I think Adobe has a winner. Very flexible and the idea of directly working in one application has value.

White Balance supports all camera settings, direct image read, and color temperature.
http://www.pettusphoto.com/Adobe001.jpg

http://www.pettusphoto.com/Adobe003.jpg

Image shows instant results with any tonal, sharpness, or filtering
http://www.pettusphoto.com/Adobe002.jpg


Dan
http://www.pettusphoto.com

mebow
20th of February 2003 (Thu), 00:24
Bought it !! Great match to PS7/1Ds. Mike

T a z
22nd of February 2003 (Sat), 00:04
Well...I took the plunge too. However, after some very preliminary tests, it appears that I was right in my assumptions. And that is, LS (in my case, LS342 as I'm using a D30) does indeed offer much better sharpening than I'm able to obtain with ARC and any other multitude of sharpening actions and products. Over all, it delivers a cleaner image.

Now as to the color fidelity...still too early to say. It'll take a little more investigation. But then, I'm sure Roger Cavanagh (should he decide to buy it...and I'd wager he might) will do FAR more extensive testing than I could ever dream of doing, and give us the real scoop. Right Roger? :)

pigasus
22nd of February 2003 (Sat), 05:01
T a z wrote:
Well...I took the plunge too. However, after some very preliminary tests, it appears that I was right in my assumptions. And that is, LS (in my case, LS342 as I'm using a D30) does indeed offer much better sharpening than I'm able to obtain with ARC and any other multitude of sharpening actions and products. Over all, it delivers a cleaner image. :)

I've been considering the same problem - and I think you're dead right about alternative sharpening routines. I haven't been able to find anything to apply to an image following conversion that compares with LS342.

Except possibly the sharpening steps from LS342. :)

Specifically, I converted an image using Roger's LinearSharpenMenu using no sharpening. (Note - this not the convert only action, as that action includes an USM step and excludes anti aliasing.) Then I used the sharpening steps from LS342 - create a mask and apply selection, sharpen with 1st USM, deselect mask, sharpen with 2nd USM. Results are pretty close. Very slightly more sharpened when done after conversion.

I haven't tried this on an image converted with Adobe Raw. Perhaps Roger?

Another thing I wonder about is whether I will miss the anti aliasing which Pekka's action includes if I move to Adobe Raw.

T a z
22nd of February 2003 (Sat), 08:57
pigasus wrote:
Specifically, I converted an image using Roger's LinearSharpenMenu using no sharpening. (Note - this not the convert only action, as that action includes an USM step and excludes anti aliasing.) Then I used the sharpening steps from LS342 - create a mask and apply selection, sharpen with 1st USM, deselect mask, sharpen with 2nd USM. Results are pretty close. Very slightly more sharpened when done after conversion.


I know you've done some work on LS342, Sally, (specifically, I believe, we have you to thank for the Higher Gamma and Greater Saturation options in Roger's LS Menu?) so I'm encouraged to hear that you've been able to separate out some of the sharpening routines.

I hadn't, as yet, tried to do that as I didn't think the results would be good unless you started with the initial linear file. Hearing that you've had passable results using an already converted image makes me curious as to how ARC might be incorporated here. I'll see if I can tear into it a bit and report back. Hopefully, we might also get our Mad Scientist of LS (i.e., Roger) :) to give it a go.

Cheers,

-Taz

DWard
22nd of February 2003 (Sat), 08:57
I have included text from the review on Luminous Landscape that briefly describes the function of each control in ACR.

Note that Sharpening, smoothing and other capabilities can be used or not used as one prefers during conversion.

I have found that I prefer to leave sharpening at Zero and do it at the end of the post processing flow as before. (D30 RAW files.)

What I like are the abilities for adjusting exposure, WB contrast etc. while getting feedback in histogram and preview windows.

(Luminous Landscape Text follows)

Temperature:

This is the colour temperature of the White Balance setting. The value initially displayed is that set by the camera (usually "Auto" if one knows that Camera Raw is going to be used). You can now set the desired White Balance in one of four ways;

— by adjusting the slider so that the image appears the way one prefers

— by setting the numeric value to a specific value

— by clicking on a "neutral" toned subject (using the eye dropper) within the preview image

— by choosing a standard value from the White Balance selector at the bottom left of the window

Tint:

This is the "second axis" used in setting colour balance and allows you to fine tune it. If you've used the one click method with the eye dropper this typically needn't be used.

Exposure:

This sets the "White Point", and for the first time in any program that I've ever seen allows you do so with a visual hint. You can hold down the Option key while moving the slider and the screen will show the exact point at which areas in the image begin to burn out.

Shadows:

This performs the same function but for the "Black Point". It can either be set by eye or with the aid of the visual clue produced when the Option key is depressed.

Brightness:

This control adjusts the "Gamma" or center point of the brightness range.

Contrast:

What is says — adjustments the dynamic range of the image.

Saturation:

Again, what it says, from mono to blow your mind Velvia+

Sharpness:

I suppose that this was added for those people that want the output of Camera Raw to produce essentially a final image. I don't use it, preferring to sharpen my files just before printing.

Smoothness:

This is a noise reduction setting. I have not used it much as I find that there are preferable ways of doing noise reduction in Photoshop.

Arrayed along the bottom of the window are a number of fairly self explanatory controls. Space allows you to set the colour space that the image will be converted to when brought into the main Photoshop program; Size permits you to change the size of the file; Moiré Filter removes moiré patterns that might exists in your file; Depth lets you set whether this file will be brought into Photoshop in 8 bit or 16 bit mode, and finally Resolution lets you set the default resolution which the file will adopt once in Photoshop.

There is also a histogram that can be overlayed on the image, and it is dynamically updated as you make alterations.

In addition to what you can see, there is what you can't see. Camera Raw features an always-on hot pixel removal capability that comes into play on long exposure images. Normally it has no effect but can make quite a difference on nighttime exposures.

pigasus
22nd of February 2003 (Sat), 09:35
T a z wrote:
I know you've done some work on LS342, Sally, (specifically, I believe, we have you to thank for the Higher Gamma and Greater Saturation options in Roger's LS Menu?)

The gamma adjustment was my work, the saturation adjustment was originally suggested by Pekka.

so I'm encouraged to hear that you've been able to separate out some of the sharpening routines.

I hadn't, as yet, tried to do that as I didn't think the results would be good unless you started with the initial linear file. Hearing that you've had passable results using an already converted image makes me curious as to how ARC might be incorporated here. I'll see if I can tear into it a bit and report back. Hopefully, we might also get our Mad Scientist of LS (i.e., Roger) :) to give it a go.

I've continued to play with LS342 sharpening after the fact. It's better than everything else I've tried. But using it as it was intended definitely gives the best results. Unless Adobe Raw gives me something very special as compared with LS342, I'll stick with LS for the D30.

With all the new Canon DSLRs due to come out in the next year or 2 it's a safe bet that Pekka won't be able to give us a version of LS for each one. When the day comes that I replace my D30 (I'm waiting for an EOS 3 digital) then Adobe Raw will probably be a blessing for me.

T a z
22nd of February 2003 (Sat), 10:18
Sally...

If you have the time, could you perhaps detail exactly what steps you are taking to use the sharpening actions alone? I tried following your above post for use on an ARC converted image, but I'm still not sure how you've managed to separate things out (I'm using LS Menu too, btw).

Also...it sounds like we're of like minds when it comes to replacing the D30. I know I will eventually, but like you, I'm addicted to the stunning sharpening that LS342 renders. In fact, LS has been one of the reasons I haven't upgraded. This is why I had been anxiously waiting for ARC. But I must say, as good as it is, I'm afraid it will be very hard to "settle" for the sharpening that other routines and stand alone products offer. While good, they're just not what you get with LS.

-Taz

pigasus
22nd of February 2003 (Sat), 10:54
T a z wrote:
Sally...

If you have the time, could you perhaps detail exactly what steps you are taking to use the sharpening actions alone? I tried following your above post for use on an ARC converted image, but I'm still not sure how you've managed to separate things out (I'm using LS Menu too, btw).

Taz - try this: Make a snapshot of the converted but unsharpened image and call it: Preproc. Then find the sharpening action you want to apply - ie Sharpness 1 or Sharpness 2 ... etc. In the action, untick the step: Play action "Common 100-200...". Then play the sharpening action. Voila.

I don't have Adobe Raw yet, but Roger (bless his cotton socks) converted one of my images with it for me. I tried applying LS sharpening to it and it wasn't anything like as good as applying it to a LS Sharpness 0 image. The Adobe Raw image was done with 0 sharpness and smoothing, by the way. The difference might be because the image I got from Roger had been jpeg'ed. Or maybe not. Be interesting to see what your tests show.

By the way, I also ran a LinearSharpen Convert Only on the raw image - excluding the USM step. This is the same as a Sharpness 0 image without the anti aliasing. Then I sharpened it with LS. Terrible! The Adobe Raw conversion + LS sharpening looked more like this result than like the Sharpness 0 image + LS sharpening.

I'm thinking LinearSharpen for the D30 still rules. :)

Roger_Cavanagh
22nd of February 2003 (Sat), 10:55
we might also get our Mad Scientist of LS (i.e., Roger) to give it a go.

I thought my ears were burning. :)

I have indeed bought ACR and have been playing with very much over the last two days. I do like it, but most conclusions I drawn are still tentative. The only thing I decided for certain is that the smoothing sucks. If you crank the slider high enough to remove background noise, the detail in the disappears. Pekka's HQ Noise Reduction from LS D60 does a better job - much slower, of course. :)

I've been looking at colour, which, by and large, I like. Reports of problem reds are no surprise: they've always been a problem with the D30. I've just gone through my archives and pulled out some very red images to examine.

Workflow is another area I've been thinking about. You can use ACR in a batch action, but not interactively to adjust conversion settings. Anyone who uses George Smith's PSSC might be interested in a new version (may not be an official release yet) that include an Open Next File command, which works handily for interactively opening and editing raw files with ACR.

I have not tackled the sharpening issue as Taz and Sally have done, so nothing decided beyond the fact that ACR sharpening ain't bad and will probably do for batch runs. "Special" images will always get custom processing. ACR is so much faster than LSM that re-runs are less of an issue.

One thing to check is whether the built-in ACR upsizing will work to reduce the jaggies as well as the stair interpolation in LSM and LS D60.

My favourite thing so far with ACR is the real-time adjustments to exposure (and the histogram) especially with the Alt+slider showing clipped highlights.

Ain't this fun?

T a z
22nd of February 2003 (Sat), 11:13
Roger...good to know that you are deep into your laboratory testing things out!

Also good to hear about the upcoming new feature's of George's excellent PS Slide Click Pro. I don't know how I ever did without it.

My early impressions with color from the Adobe product are also favorable. And the speed...oh, the speed! Even with my relatively "old" P3 933MHz PC I'm amazed at how quickly things happen.

Thanks for chiming in here, Roger.

-Taz

T a z
23rd of February 2003 (Sun), 00:36
Sally...

Somehow I missed your post earlier regarding how to break out the sharpening routines. Thank you very much for the step by step instructions.

However, I must say, I've got somewhat of a mixed bag of results. I'm surprised to report that on some images, the "Combo" approach (i.e., ACR at 0 sharpness combined with LS) provided better sharpening than the full LS routine! Although, looking (very) closely, the need for the anti-alaising filter was apparent. But yet on other images...usually low contrast images...there was no way I could out do the sharpening of the full LS model. Suffice it to say, that at this point I'm somewhat amazed at what the Combo approach can do for some images. However, I feel I need to do some more experimentation before making any sort of pronouncement.

One thing that IS starting to emerge is just how well ACR handles color. For example...skin tones, while better with Roger's addition of the Magenta Fix, were still somewhat hard for me to nail down. With ACR, I'm finding it much easier to do.

And, oh yes, being able to change WB on the fly so quickly is starting to become VERY addictive!

I'll continue to play around with the sharpening routines and report back. Thanks again for showing me how to "disect" the action.

-Taz

pigasus
23rd of February 2003 (Sun), 03:59
Some more thoughts on this. I woke up this morning and realised that I had neglected to look at one consideration: bit depth. (I really must get out more. :) ) The image that Roger converted and sent to me was in 8-bit. It's clear this makes a dramatic difference in applying Pekka's noise reduction technique. It may well affect the sharpening routines also.

Taz, did you keep your unsharpened ACR files in 16 bit before applying LS sharpening? Your report on your trial to date is encouraging. Have you found a specific set of ACR settings that might work as a better default starting point for the D30? Keep up the good work!

Roger_Cavanagh
23rd of February 2003 (Sun), 05:42
I haven't made a decision yet about the best colour, so I'm posting these pictures for their curiosity value. All these crops are unsharpened:

The ACR Version

http://www.pixelpixel.org/images/linked/cdp/acr/020729-0087A.jpg

Fred Miranda's Linear Pro (converted with BB)

http://www.pixelpixel.org/images/linked/cdp/acr/020729-0087LP.jpg

LSM Version (converted with BB)

http://www.pixelpixel.org/images/linked/cdp/acr/020729-0087LS.jpg

Non-linear TIFF using TrueColor Non-linear Profile:

http://www.pixelpixel.org/images/linked/cdp/acr/020729-0087TCn.jpg

Anyone got a pillar box outside their front door?

macondo
23rd of February 2003 (Sun), 07:36
One question here Roger, regarding your reds.Are you considering that the Adebr Raw is using the Adbe09 color space for their handling of the files?

I ask this because the D30 and D60 ( up to my knowledge) can be said to be shot in a color space resembling or approaching sRGB and it has become almost standard to run the worflow assinning sRGB and then converting to your favorite color flavour.
My point is that if you assign Adobe98 while opening an untagged image from those cameras, the differences in Reds are huge , compared to results obtained in sRGB. Reds have been my nightmare in this sense, and I still favor the" sRGB look ", and then making adjustments in PS later.

Your tests show the same drama with the reds.

If the Adobe Raw is assignning Adobe 98, then there is something here to ponder. What do you make of this??

Thanks.

Roger_Cavanagh
23rd of February 2003 (Sun), 08:25
macondo wrote:
One question here Roger, regarding your reds.Are you considering that the Adebr Raw is using the Adbe09 color space for their handling of the files?

I ask this because the D30 and D60 ( up to my knowledge) can be said to be shot in a color space resembling or approaching sRGB and it has become almost standard to run the worflow assinning sRGB and then converting to your favorite color flavour.
My point is that if you assign Adobe98 while opening an untagged image from those cameras, the differences in Reds are huge , compared to results obtained in sRGB. Reds have been my nightmare in this sense, and I still favor the" sRGB look ", and then making adjustments in PS later.

Your tests show the same drama with the reds.

If the Adobe Raw is assignning Adobe 98, then there is something here to ponder. What do you make of this??

Thanks.

Maconda,

I have ACR set to convert to ProPhoto RGB - I adopted ProPhoto as my preferred working space a few weeks ago. ACR only gives you that plus 3 other options: Adobe RGB, sRGB and ColorMatch. I then cropped, downsize and converted to 8-bit sRGB for the pictures I posted in my last thread. The non-linear version was assigned the TrueColor profile, then converted to ProPhoto.

So I'm happy that I didn't do something silly. I'm also pretty sure that Thomas Knoll who wrote ACR (and was the original author of Photoshop) wouldn't get confused about colour spaces. :D

Assigning sRGB to D30 images is OK, if you are shooting JPG, but the colour gamut for raw is much wider than sRGB so you'd be throwing away some colour data by starting in sRGB.

Also the benefit of converting from a small space (sRGB) to a wider open is debatable. It doesn't buy you any extra colour.

I think the differences in reds just demonstrate that profiling and colour management ain't an exact science. :) For more on "dodgy reds" see my next post.

Regards,

pigasus
23rd of February 2003 (Sun), 08:38
Roger,
About those pillar boxes. Working from memory, I think the ACR and the LSM versions look closest. The non-linear TIFF is awful and I don't think Fred has ever been to the UK. :D

Roger_Cavanagh
23rd of February 2003 (Sun), 08:46
I dug this picture out of my archives - pretty easy to see why. Here's an assortment of treatments:

The ACR version with no adjustments:

http://www.pixelpixel.org/images/linked/cdp/acr/020704-0034A.jpg

The ACR version with -30 saturation:

http://www.pixelpixel.org/images/linked/cdp/acr/020704-0034A-30.jpg

The ACR version with no adjustments then the magenta fix(double strength):

http://www.pixelpixel.org/images/linked/cdp/acr/020704-0034Amag.jpg

The LS version:

http://www.pixelpixel.org/images/linked/cdp/acr/020704-0034LS.jpg

The LS version with magenta fix:

http://www.pixelpixel.org/images/linked/cdp/acr/020704-0034LSmag.jpg

In fairness, the conversion to 8-bit sRGB makes the first picture look worse than it does in PS in 16-bit ProPhoto.

It's a while since we had these roses - one from a bouquet I bought my wife last anniversary :) - so it's guess which version is more like the real thing. It's obviously not the first, and it's not the last either.

Regards,

T a z
23rd of February 2003 (Sun), 11:36
pigasus wrote:
Taz, did you keep your unsharpened ACR files in 16 bit before applying LS sharpening?

Sally...yes everything was kept in 16bit. I really hadn't thought about it, but you're right...it would seem that one should feed the LS sharpening routines a 16bit file. As far as optimal settings are concerned, I still need to do some more experimenting. I'll keep you posted.

Roger...as far as your Rose image is concerned, how did you have the WB set in ACR? I assume, "As Shot"? If so, did you use the same "As Shot" settting in BB or Yarc Plus when you converted? I imagine you did, but even so, I'm wondering if we would still be comparing apples to apples (or roses to roses, as the case may be :) ). In other words...given the additional parameter of "Tint" in ACR and who knows what else, I'm wondering if we can directly compare the "As Shot" modes (or any of the other modes for that matter) between software. It would *seem* we could, but then we really don't know what's going on under the hood in ACR, and BB and YP are using the Canon SDK which ACR is not. Just a thought...

-Taz

Roger_Cavanagh
23rd of February 2003 (Sun), 12:13
T a z wrote:
Roger...as far as your Rose image is concerned, how did you have the WB set in ACR? I assume, "As Shot"? If so, did you use the same "As Shot" settting in BB or Yarc Plus when you converted?


Puhleeze. :)

I imagine you did, but even so, I'm wondering if we would still be comparing apples to apples (or roses to roses, as the case may be :) ). In other words...given the additional parameter of "Tint" in ACR and who knows what else, I'm wondering if we can directly compare the "As Shot" modes (or any of the other modes for that matter) between software. It would *seem* we could, but then we really don't know what's going on under the hood in ACR, and BB and YP are using the Canon SDK which ACR is not. Just a thought...

Well, there are more things that can be tweaked in ACR, but I don't see why we shouldn't compare the as-shot-out-of-the-box results between different converters. I used BB for the rose, but YP wouldn't have given a signficantly different result. It just seems that ACR can't deal very well with the reds-verging-on-magenta. It seems to be something that we D30 owners have to live with. Ideally, we can work out custom settings that can be saved as an ACR parameter file.

Was just reading the thread at Rob G's you've been contributing to (and the Colorsync one someone pointed to). There are some teddies being thrown out of the pram, aren't there? :)

Have you been able to make any sense of the suggestion that working space should be set to a camera profile?

Cheers,

T a z
23rd of February 2003 (Sun), 13:25
Roger_Cavanagh wrote:
T a z wrote:
Roger...as far as your Rose image is concerned, how did you have the WB set in ACR? I assume, "As Shot"? If so, did you use the same "As Shot" settting in BB or Yarc Plus when you converted?

Puhleeze. :)

Was just reading the thread at Rob G's you've been contributing to (and the Colorsync one someone pointed to). There are some teddies being thrown out of the pram, aren't there? :)

Have you been able to make any sense of the suggestion that working space should be set to a camera profile?


Ahh...silly me...I must have taken leave of my senses when I asked the above question! What was I thinking? Of course, our resident master of experimentation would have had things set correctly in his laboratory :)!

Seriously...regarding comparisons of WB interpretations between BB or YP and ACR, isn't it possible that because BB and YP use the Canon SDK and ACR does not that WB renderings may be slightly different? As a non-programmer, I haven't a clue how this type of software engineering is done, so I'm really just wondering aloud. It would *seem* we would be comparing apples to apples at least with the As Shot setting, but I just thought I'd throw the question out there none the less.

And no...I really haven't been able to definitively determine if it would be possible to use a custom profile with ACR. It seems that Andrew is on the verge of saying it can be done, but obviously since he's trying to make a point about not needing custom profiles in ACR, I doubt he'll provide much more info. But as I said over there...so far, I am happy with what I've seen out of ACR for the D30 (in spite of Ian Lyon's criticism). But then, I haven't used it on a large number of images as yet either. Also haven't used it in Batch mode.

Still playing with the sharpening, and I think Sally may be on to something with the "Combo" approach. Have you had a chance as yet to experiment with that? I'd be interested to hear what you think.

Thanks,

-Taz

Roger_Cavanagh
23rd of February 2003 (Sun), 15:13
T a z wrote:

Still playing with the sharpening, and I think Sally may be on to something with the "Combo" approach. Have you had a chance as yet to experiment with that? I'd be interested to hear what you think.

Taz,

If you mean by the combo approach, conveted with ACR unsharpened, then applying an LS sharpening (LSS) procedure. I hadn't, but since you asked did a few quick test runs. I picked an image with a lot of texture. I ran LSM for all sharpness levels and picked 4 as the sharpest I could get away with for my comaprison standard.

LSS worked OK, but I couldn't apply such high levels without getting much more serious artefacts.

I am currently working on some sharpening in ACR plus LSS at a low level. This looks quite promising.

WRT to sharpening in ACR, I've pretty much decided that a level of 75+ (i.e., aggressive) followed by an application of HQ Noise Reduction from LSD60 gives a good result.

Regards,

Dale
25th of February 2003 (Tue), 06:32
Roger,

Will you be updating LSM to incorporate some of ARC's applications?

Roger_Cavanagh
25th of February 2003 (Tue), 13:15
Dale wrote:
Roger,

Will you be updating LSM to incorporate some of ARC's applications?

Yo, Dale,

As of now, I'd have to say "prolly", but it would be an alternative rather than an update 'cos the linear sharpen bit wouldn't actually be there. :)

I'm still trying to get a good handle on a sensible workflow with ACR, so any suggestions or requests will be gratefully considered.

Regards,

twwilliams
25th of February 2003 (Tue), 19:10
I still haven't convinced myself of this yet, but it seems like I'm getting better quality (especially in areas such as our German Shepherd Dog's fur where I see a cross-hatch pattern for fine detail) by bringing in the file from ACR at the maximum resolution it supports and then downsampling.

I have tried a variety of sharpening options, both doing sharpening within ACR and doing it afterwards in Photoshop.

I'm finding the best results with the max file size in ACR, 0 sharpening, and smoothing between 25 and 35, although I think that may depend on the ISO of the image. I haven't verified that one yet.

I have been getting my best results with a two-pass sharpening (low percentage and high radius, then high percentage and low radius), both faded to luminosity. When I did that on a file brought in at max resolution, then downsampled and sharpened again, I got results that were as good as with LinearSharpenMenu, at least on the two files that I experimented with (I had so many permutations that I didn't have time for more).

Going back to the cross-hatch noise pattern, I also tried Yarc with no noise reduction, with ARF1, and with ARF2. ARF had almost no effect on the cross-hatch pattern, and the quality of a linear TIFF with ARF2 and LinearSharpenMenu (in terms of sharpness and noise) was no different than what I got with the technique in ACR.

On the other hand, the color in the ACR version was definitely much better. I have never been able to get color I am truly happy with out of my D30, at least until now. Most of that is due, I'm sure, to the quality of the white balancing tool in ACR.

On yet another hand, bright reds -- especially bright reds in the sunlight -- come out very strangely in ACR: they're completely clipped, even though they come out that way in linear TIFFs processed with LinearSharpenMenu.

How was that for a long and rambling post?

-- Tommy

Sporto
25th of February 2003 (Tue), 23:20
Is this a free plug in? If so could you please tell me what the link is?

Sporto

reittila
26th of February 2003 (Wed), 01:04
Sporto wrote:
Is this a free plug in? If so could you please tell me what the link is?

Sporto

Sporto: No it's not a free plugin. It's Adobe's plugin ;)
You can find it here :
http://www.adobe.com/products/photoshop/cameraraw.html
Or if you live outside U.S. you can buy it from here :
http://www.transaction-one.com/gs/html/a/adobe/shopping001/en/start.jsp