Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: uaiomex on July 03, 2014, 03:57:44 pm
-
I got this job in which I'm selling 20 prints for decoration at a hotel. These images are part of my stock. I need some to be double the size and a few a little more. The files are mainly from Canon 5D and 5D2. A few are stitched files done with TS glass.
Years ago, I had GF but lost it when I switched to Mac. I've been experimenting upsizing in PS in 10%'s while applying a very tiny (ten times too) amount of USM. They look sharper than without USM but display a little artifacts at 100%. Most possibly they won't show in print but I'm afraid they do.
I was thinking that maybe it was time for me to spend a little money to get a state of art application of this kind.
Final work will be printed up to 41 inches on Epson EhMat
Can someone help with suggestions, ideas and comments?
TIA
Eduardo
-
One thought process, I have gotten away from the incremental process, for me takes too long and I don't see much gain.
Software to consider would be Blow up, Perfect Resize (former GF) and Photozoom Pro. I would lead with the later and use the Spline 64 as the uprez algorithm and look at Lanzcos. My method is to get the image sharp before the uprez, then use Focus Magic after the uprez, if needed.
I print everything now from LR, and for Epson you need to use 360 or 720ppi for the best results. For this size work, I would stay with 360ppi.
You can also simply try allowing LR to uprez the image for you if you are within 180 to 240ppi, and apply some output sharpening. LR does a very good job getting out the final image and often times, I can't see much difference between a 240dpi shot uprez'd to 360 with one of the software tools I have referenced and printed in LR at 360ppi and using LR to print a file from 240dpi, as 360ppi output with respective output sharpening.
This works for me and my customers but I am very sure you will get a lot of other ideas.
Paul
-
Thank you Paul. Terrific so far.
Eduardo
Paul
-
You might want to try using software you own (?) like Photoshop/ACR or Lightroom:
http://www.digitalphotopro.com/technique/software-technique/the-art-of-the-up-res.html?start=1#.U7X52VYZovg
-
Photoshop CC new Image Size is sooo much better
-
Photoshop CC new Image Size is sooo much better
Better than what? Lightroom? Can you scan and post a jpg of a comparison for two prints from the same photo using CC versus LR?
-
CC does have a bit of an advantage with the process where you can select it to "retain details". The advantages are slight to me eyes, and on larger prints like the OP needs, probably hard to see any differences. Jeff Schewe mentioned in a different post he had noticed a possible improvement with the new CC algorithim, but it didn't seem enough for him to change his workflow yet.
To me one of the hardest things to do, is print and scan back to show on the web. A lot of details can be lost back in the scan and more than likely more sharpening needs to done to the scan, just do to the nature of scanning.
Displaying crops of the final uprez'd images at 100% might better, but even then it's pretty hard to see the differences.
You have to go back to the OP's original need, hotel work. Most of the folks looking at these will never be pixel peeping IMO. They will see them in passing. Any of the current software tool sets out there can do the job.
Paul
-
I think you are right.
-
... I've been experimenting upsizing in PS in 10%'s while applying a very tiny (ten times too) amount of USM. They look sharper than without USM but display a little artifacts at 100%. ...
I am not sure that upsizing in PS by small increments is a good idea as it may actually add more artefacts than a one step process. I think that Adobe Chris Cox has stated such and if memory serves Bicubic Smoother rec. with the one step process to get the best out of the image.
Having tried Perfect Resize with a few images vs PS CS5 and CS6 I came to the conclusion that there was not a huge difference between applications and although the PS versions slightly softer (brought back by modest sharpening) these were my preference. Of course original image content and quality must play a significant role so it may be that certain images will respond better in one application than another.
IMO worth while downloading the developers demo's and actually having a play with your images and comparing the results against your versions PS or LR - you may be pleasantly surprised that you do not have to incurr added expenditure. On the other hand... :)
-
Having tried Perfect Resize with a few images vs PS CS5 and CS6 I came to the conclusion that there was not a huge difference between applications and although the PS versions slightly softer (brought back by modest sharpening) these were my preference.
Exactly what I've observed as well. Over the years, I've tested a number of these 3rd party products. Output has always been to a print as viewing the differences on-screen is not indicative of what one would see output on the print. Last year I did a Webinar on this and again found Perfect Resize with default settings (it has far too many IMHO) didn't produce output as good as using Lightroom or Photoshop with proper capture sharpening (default in the case of LR). That was more important to the output quality than any of the various products or techniques including Step Interpolation. Output sharpening was equally applied to all images. LR and Photoshop were faster to process the data by far, LR at the time was a tad better than PS all things being equal. My friend Mac Holbert (formally of Nash Editions) did similar tests and concluded that if one has to take a tiny original, say 640x480 and make huge enlargements, these 3rd party products did produce better results for him but anything short of 200-300X, Photoshop did the job better, faster, cheaper.
-
My friend Mac Holbert (formally of Nash Editions) did similar tests and concluded that if one has to take a tiny original, say 640x480 and make huge enlargements, these 3rd party products did produce better results for him but anything short of 200-300X, Photoshop did the job better, faster, cheaper.
Hi Andrew,
A 200-300% enlargement is peanuts (like upsampling from 360 PPI to 720PPI without real details). No wonder you see little difference, especially when you do not properly sharpen after upsampling. Try producing decent magnifications of 4x to 12x, like discussed in this thread (http://www.luminous-landscape.com/forum/index.php?topic=62609.msg505337#msg505337) and this thread (http://www.luminous-landscape.com/forum/index.php?topic=72738.msg578386#msg578386). And those 10.5x upsampled image crops are not even post resize output sharpened ...
It also explains why people prefer the output from Qimage rather than PS or even LR. Much sharper output even in print on matte media. Quality upsampling followed by quality output sharpening makes a difference, an increasingly huge difference with increasing upsampling percentages.
Cheers,
Bart
-
Another shameless plug for "Qimage Ultimate."
Might pay to install Windows via Parallels since it is only Windows based. Apple seems to screw things up with their Colortune idea and locking out certain print driver and management aspects. Apple, for me, is a very poor printer's OS choice - short of going into QTR or something exotic that isn't Windows supported as well.
Sure beats Adobe on sharpness and noise, imho. You can fine tune the sharpness within it too, or even print RAW straight with newer version. Can't stand to print via any other software now as it manages the print aspect so much better. Plus, it does new stuff like unclog nozzles, etc. that printer people need.
One can "Save (image) to a File" rather than print after they apply its sharpening algorithms and then print via other methods and compare the sharpness and noise reduction too.
Seems it is a serious labor of love for the owner/programmer. Least he addresses issues with it faster than Adobe can too who seems to lag behind him.
SG
-
I've been enlarging images from a Canon 1Ds I/II/III to tractor-trailer rig coverings for about ten years. I've used Photoshop, Genuine Fractals and Blow Up. At the pixel level there was a visual difference. At reproduction size, viewed from 15', the differences in enlargement algorithms were not significant. With the new algorithms in PS, I have stopped using 3rd party software for any enlargements. I have not tried PhotoZoom Pro, though.
(http://www.signaturegraphicsinc.com/wp-content/gallery/trailer-wraps/johnsonville-tastyville-trailer-angle.jpg)
-
Another shameless plug for "Qimage Ultimate."
Sure beats Adobe on sharpness and noise, imho. You can fine tune the sharpness within it too, or even print RAW straight with newer version. Can't stand to print via any other software now as it manages the print aspect so much better. Plus, it does new stuff like unclog nozzles, etc. that printer people need.
I have not looked at Qimage in several years, always liked their upgrez options, but reall photozoom pro has most of the same ones. I am also curious, how is Qimage checking or clearing nozzles? I can see it running a pattern, but for it to do a clean it would have to interface to the utility level of the print driver at least with Epson. Not doubting your word, just interested.
As I stated earlier in my first post, LR and CC work great for sizes up to 36 x 72 at 360 ppi, which is about the largest print I make personally. If you to billboard size or semi-truck, all bets are off as now you are talking so large, that no one is going to walk up and pixel peep. Here is one of the areas where "blow up" works very well and "perfect resize". As both of these software tools were meant really for very large prints originally.
As LR and CC have gotten so much better, I tend to stay with them, as it's so nice to just let LR do the final uprez and output sharpen at 360ppi.
Lots of good info in this thread to consider.
Paul
-
Thanks guys. Very helpful information.
Since the upsizing won't go more than 400%, I'm thinking on using PS bicubic smoother and perhaps buy Focus Magic for final sharpening. Its $69.00 price seems appropriate. Gf or PhotoZoomPro are mucho dinero, imo.
Eduardo
-
I've been enlarging images from a Canon 1Ds I/II/III to tractor-trailer rig coverings for about ten years. I've used Photoshop, Genuine Fractals and Blow Up. At the pixel level there was a visual difference. At reproduction size, viewed from 15', the differences in enlargement algorithms were not significant.
Hi Chris,
Viewing distance is indeed a great equalizer. However, even then it would be wasteful to not optimize output sharpening for those larger viewing distances. Human vision has a contrast sensitivity function that peaks around 8 cycles/degree viewing angle, and drops off at higher spatial frequencies. Not optimizing for that by sharpening the upsampling blur, and boosting the small to medium detail contrast would leave viewing quality on the table, unused. Topaz Labs Detail is perfect for addressing both issues (without adding halos), but also earlier in the workflow, for Creative 'sharpening'.
I have not tried PhotoZoom Pro, though.
It's somewhat similar to Blow-up in adding vector-like edge detail that exceed native resolution, but with less rounding of sharp edges than BU. Also recent versions of 'Perfect Resize' allow to add 'real' resolution to edges.
Cheers,
Bart
-
A 200-300% enlargement is peanuts (like upsampling from 360 PPI to 720PPI without real details). No wonder you see little difference, especially when you do not properly sharpen after upsampling.
Yup, but it's rare I'll ever need to upsize much past that amount. Real data is more desirable! ;D
-
..... I am also curious, how is Qimage checking or clearing nozzles? I can see it running a pattern, but for it to do a clean it would have to interface to the utility level of the print driver at least with Epson. Not doubting your word, just interested. ....
Paul, here is the video on the newest Qimage version regarding the "Unclog feature": https://www.youtube.com/watch?v=UcmmyYroPOQ (https://www.youtube.com/watch?v=UcmmyYroPOQ)
I've used it to unclog a PK and a C ink in an Epson recently. It was off only 3-4 days before it dried up maybe 3 nozzles of those two and I got a bad nozzle print off the printer's menu. Nice that it doesn't blow all the ink, compared to the Epson software or driver that use all inks.
Now it's running well again, but I just need to make sure I run the thing every 3 days maybe since it is hot and very dry here with the A/C running 16 out of 24 hours. Gotta love them $600 utility bills! NOT!!!
SG
-
I looked at a good part of the video and indeed it is a neat feature for those using printers that require all the channels to clean a printhead. Of course in the Epson x900 series printers, and I believe several others though not sure, one cleans the channels in pairs, and a regular cleaning cycle for a pair uses VERY little ink. So the usefulness of this feature depends on the printer one is using. For those of us on OSX, of course QImage is not available without installing Windows as a VM under Parallels or some such. While workable, I must say having been on OSX for close to four years now after using Windows for the previous fifteen, I don't think there's much value-added going that route and there are potential headaches.
The real solution to nozzle clogs in Epson printers isn't any of this, however. It is to keep the printer in a sufficiently humidified room and to print at least several real photos with it every two to three days.
But this is all kind of OT.
-
Mark and all,
As the subject has been discussed Ad nauseam it still seems to never answer all the questions.
Take this workflow for example. To control file size we resample all of our large files for canvas prints in CC or Perfect resize to 180 ppi.
Then uprez in Lightroom to 360 for printing. We save the the smaller 180 ppi print file to our print folder and it is ready for future uprezzing/printing.
A 40x60 print at 180 ppi is 221mb the same file resampled in CC to 360 ppi is a 884mb file,the difference in file size is huge.
We have done this dozens of ways but just do not see differences that would warrant resampling to 360 pip in CC or PR and then be stuck with all those extremely large files.
Missing anything?
-
Mark and all,
As the subject has been discussed Ad nauseam it still seems to never answer all the questions.
Take this one for example. To control file size we resample all of our large files for canvas prints in CC or Perfect resize to 180 ppi.
Then uprez in Lightroom to 360 for printing. We save the the smaller 180 ppi print file to our print folder and it is ready for future printing.
A 40x60 print at 180 ppi is 221mb and at 360 ppi it is an 884mb file,the difference in file size is huge.
We have done this dozens of ways but just do not see differences that would warrant resampling to 360 ppi and then be stuck with all those extremely large files.
Are we missing anything?
Dan, I don't think so, but perhaps worth adding - if I am not mistaken about this, that if you are printing from LR, any uprezzing you order-up in the Print Module by specifying the output PPI for printing, is done on the fly and does not change the size of the master file from what you started with. By this understanding, you eat your cake and have it.
-
Mark and all,
As the subject has been discussed Ad nauseam it still seems to never answer all the questions.
Take this workflow for example. To control file size we resample all of our large files for canvas prints in CC or Perfect resize to 180 ppi.
Hi Dan,
Why 180 PPI? Do you only have a single input pixel size that, at 180 PPI, produces the correct output dimensions (in inches)?
Also, do you optimize sharpening for this resampled resolution? If you always resample up then that may be not the best practice, because any artifacts created by resampling to 180PP and sharpening at that size will be magnified when you then upsample again. It seems that it may be better to do a single upsample to outputsize, regardless of the original file size in pixels. It will even reduce your storage requirements for these intermediates, and save time.
Then uprez in Lightroom to 360 for printing. We save the the smaller 180 ppi print file to our print folder and it is ready for future uprezzing/printing.
I stiil have difficulty in seeing the benefit of the intermediate resizing. By stopping mid-way, you do not use the full potential of Perfect Resize. LR is also full well capable of upsampling from native/original pixel dimensions to final pixel dimensions, in a single go, but PR will do it even better in a single operation.
A 40x60 print at 180 ppi is 221mb the same file resampled in CC to 360 ppi is a 884mb file,the difference in file size is huge.
Is this by any chance a compromise to still allow printing from LR without choking it with very large input files? That would make some sense, but at the expense of potential output quality.
We have done this dozens of ways but just do not see differences that would warrant resampling to 360 pip in CC or PR and then be stuck with all those extremely large files.
Why save the resampled file? All you need is a good quality upsampling solution that can be used on-demand, which then also allows to add an amount of output sharpening (from a preset if simplification of the output sharpening process is that important). That is exactly what Topaz Labs Detail allows to do, because it does not target fixed pixel detail dimensions, but rather a 'class' of detail and therefore independent of original pixel dimensions but rather actual detail at the actual output image dimensions.
Cheers,
Bart
-
We have done this dozens of ways but just do not see differences that would warrant resampling to 360 pip in CC or PR and then be stuck with all those extremely large files.
Missing anything?
Not as far as I'm concerned as I've seen the same results. YMMV of course and the image data, printer and so forth must also play a role. Some of us don't see any differences outside of viewing the image with a loupe, other's report they do see differences with their naked eyes (and I assume at viewing distance). Bottom line, test this for yourself, adapt your workflow accordingly.
-
Hi Dan,
Why 180 PPI? Do you only have a single input pixel size that, at 180 PPI, produces the correct output dimensions (in inches)?
Also, do you optimize sharpening for this resampled resolution? If you always resample up then that may be not the best practice, because any artifacts created by resampling to 180PP and sharpening at that size will be magnified when you then upsample again. It seems that it may be better to do a single upsample to outputsize, regardless of the original file size in pixels. It will even reduce your storage requirements for these intermediates, and save time.
I stiil have difficulty in seeing the benefit of the intermediate resizing. By stopping mid-way, you do not use the full potential of Perfect Resize. LR is also full well capable of upsampling from native/original pixel dimensions to final pixel dimensions, in a single go, but PR will do it even better in a single operation.
Is this by any chance a compromise to still allow printing from LR without choking it with very large input files? That would make some sense, but at the expense of potential output quality.
Why save the resampled file? All you need is a good quality upsampling solution that can be used on-demand, which then also allows to add an amount of output sharpening (from a preset if simplification of the output sharpening process is that important). That is exactly what Topaz Labs Detail allows to do, because it does not target fixed pixel detail dimensions, but rather a 'class' of detail and therefore independent of original pixel dimensions but rather actual detail at the actual output image dimensions.
Cheers,
Bart
A fair amount of my larger canvas prints are reprints of same size work for artisans so saving the file like I do equates to just hitting the print one button.
In the past I almost always resampled to 180 ppi in PR or CC and printed from Lightroom with no further uprezz with excellent results.
I have seen more issues on these very large prints when resampling straight to 360 ppi in CC or PR when the resample was 400% and larger.
A 180 ppi resample usually kept me in the 200% range. Then send the file to Lightroom for printing and let it handle the 360 uprez. Smaller file,much quicker spool and at least for us the same results or better as doing it straight to 360.
Which begs the question I guess I started with. For anyone else does a file resampled straight to 360 in PR or CC with no further uprezz in Lightroom print any differently then a resampled file to 180 in PR or CC and then uprezzed to 360 in Lightroom for the final print?
For me no different and I have several bonuses as noted.
-
Hello Dan
My process for about the last year is to output a raw conversion to 300dpi. Canon Nikon Phase or Fuji. Then when I print go to LR and use it to get to the final output size at 360 ppi.
I agree this does make for some big files easily approaching 1GB in 16 bit especially with multi-image stitched panoramas.
Paul
-
Using Raw Developer from Iridient has the best upsample output to 200%-250% at 240 PPI from your Canon Raw files. Use RL Deconvolution sharpen at 50% of the default setting and save out as 16 bit Tiff.
Michael Creedon
-
Thanks everybody. I upsized in PS, bicubic smoother, one time. And "refocused" with Focus Magic. The prints look good if not great to me. I decided for this after reading your responses and for the inexpensive cost of FM.
Eduardo
-
Thanks everybody. I upsized in PS, bicubic smoother, one time. And "refocused" with Focus Magic. The prints look good if not great to me. I decided for this after reading your responses and for the inexpensive cost of FM.
Eduardo
With all the time, effort and repeated cost of ink and paper you put into your printing, don't you think it's a bit dumb to worry about saving a few bucks on the sharpening software? I decided long ago, based on the numbers, that software is the LEAST expensive component of the whole digital image processing chain and any additional cost of software that really makes an observable difference in print quality is well worth it. That money disappears in the decimal places, but gives you a better chance for prints to look "great", not just "good".
-
Actually, Focus Magic, as a purchase is not "dumb", it's an excellent tool. The Topaz Solution (I believe theirs is called In Focus) has more adjustments, but to the results are the same.
I use Focus Magic, in my workflow before I print from LR, depending on the image. I have started using less Capture Sharpening in both C1 and LR and use FM now or Topaz. Then still use the output sharpening in LR.
Deconvolution Sharpening is pretty powerful from what I have seen, and this is used by both FM and Topaz.
Bart's articles and posts steered me towards Focus Magic, and it's well worth the investment.
Paul
-
I wasn't referring so much to the product as to the idea of using it because it's cheaper. IF it's better AND cheaper, that's a win-win. But is it really? The testing I've done with deconvolution sharpening is that unless it is used VERY carefully, it produces much less satisfactory output than I get with PK2/LR.
-
I wasn't referring so much to the product as to the idea of using it because it's cheaper. IF it's better AND cheaper, that's a win-win. But is it really? The testing I've done with deconvolution sharpening is that unless it is used VERY carefully, it produces much less satisfactory output than I get with PK2/LR.
Hi Mark,
I suggest you give FocusMagic a try, if you haven't already.
I have no commercial involvement with them, I'm just a someone who bought it after trying out lots of different sharpening methods and products, and reading on-line reviews/comparisons with images examples. But they use a pretty simple to use interface, not much one can do wrong there, it's what's going on under the hood that sets it apart from other offerings. That doesn't mean it always produces the best results, but the differences with much more involved alternatives is very small in the majority of cases in general photography.
A simple trick to avoid using the 'wrong' parameters with FocusMagic is the following.
- Set the amount control to its maximum, 300%.
- Start with the 'Blur width' setting at zero
- Now gradually increase the Blur width size by 1 at a time, while watching the small preview window
- there is usually a setting where, all of a sudden, the preview will start showing fatter contours/edges instead of sharper ones
- at that point just go back one Blur width value
- then reset the Amount control to a more 'reasonable' value, which depends on the image use, usually larger radii (such as after upsampling) can tolerate higher settings of 150+, smaller radii do best at 100 or less
One can play with various types of image source settings, but the default "Digital Camera" one is usually fine.
Deconvolution, a mathematical concept, is the most common way of actually restoring resolution, many other methods only adjust edge contrast which suggests higher resolution. There is no reason why the two cannot be used together for even higher (real and suggested) resolution.
That's why I also recommend having a look at 'Topaz Labs Detail'. It offers an amazing amount of control for the rendering of detail, which is not the same as resolution. Detail has to do with material surface structure and spatial dimensions. That is also very useful to improve the image quality for different viewing distances. 'Detail' offers a very creative way of interaction with the image, due to the real-time updates of the preview. The adjustments are not pixel dimension related but relative image size related, small medium and large detail, and produces no halos, so they should work well at various output sizes.
Cheers,
Bart
-
Just reading all of this tells me I should stay with LR sharpening. It's super easy and the results are superb.
-
Just reading all of this tells me I should stay with LR sharpening. It's super easy and the results are superb.
No Mark,
Good enough is not the same as superb (http://www.merriam-webster.com/dictionary/superb)!
Besides, out of camera JPEGs are easy, good enough for most. ;)
Cheers,
Bart
-
No Mark,
Good enough is not the same as superb (http://www.merriam-webster.com/dictionary/superb)!
Besides, out of camera JPEGs are easy, good enough for most. ;)
Cheers,
Bart
You didn't hear me being satisfied with "good enough" :-)
OK, for clarity, let's get down to the "objective function": it's partly subjective: I aim for a natural looking sharpness - not exaggerated sharpness; I don't want the sharpening to look like sharpening, or sharpening for its own sake. I don't want to see any more real or especially pseudo detail in most of my photographs than I see "in the real world" for comparable distance viewing, unless I were to have a special purpose that requires a different kind of treatment. Do you ever make a point of consciously looking at scenes or objects "x", "y" or "z" meters away specifically for the amount of textural and edge detail you can perceive at those distances? I do this periodically as a "reality check" on the kind of sharpening I do; then when I look at a photo I've printed, if it doesn't trigger me to think about whether it's "sharp enough" I'm done. 99.5% of the time LR does this for me. (The other 0.5% is probably operator error or poor focusing of the camera!)
-
You didn't hear me being satisfied with "good enough" :-)
OK, for clarity, let's get down to the "objective function": it's partly subjective: I aim for a natural looking sharpness - not exaggerated sharpness; I don't want the sharpening to look like sharpening, or sharpening for its own sake. I don't want to see any more real or especially pseudo detail in most of my photographs than I see "in the real world" for comparable distance viewing, unless I were to have a special purpose that requires a different kind of treatment.
Hi Mark,
That's actually a pretty decent description of what Deconvolution (e.g. FocusMagic) attempts to achieve. Restore what got lost in the earlier process, but not exaggerate. It's what saved the first images from the Hubble Space Telescope from a design error.
Do you ever make a point of consciously looking at scenes or objects "x", "y" or "z" meters away specifically for the amount of textural and edge detail you can perceive at those distances?
Yes, that's the great thing about e.g. Topaz Detail. That's why I mention it, because it is not only good for Creative 'sharpening', but also for output sharpening where we want to pre-compensate for losses that have yet to occur due to the output process (e.g. ink diffusion), and maybe something additional if the viewing distance is not too variable.
Viewing distance can alter our impression of detail quite a bit, as nicely demonstrated in this example (http://www.123opticalillusions.com/pages/albert-einstein-marilyn-monroe.php). That's why I recommend using tools that bring such improvements within the reach of us mortals, albeit a bit more subtle, we wouldn't want to freak out people at specific viewing conditions. ;)
I do this periodically as a "reality check" on the kind of sharpening I do; then when I look at a photo I've printed, if it doesn't trigger me to think about whether it's "sharp enough" I'm done. 99.5% of the time LR does this for me. (The other 0.5% is probably operator error or poor focusing of the camera!)
Lightroom, as nice and powerful as it is, does not allow the kind of extreme control that a more involved workflow can offer. As such it will not be able to exceed the potential of that more involved approach. It may come close (enough) for some scenarios, people can try for themselves (after being handed some guidance!) as to what fits their requirements best. Thank goodness we are allowed to make our own choices, but it would be nice if its an informed choice.
Cheers,
Bart
-
Everyone has their own style of printing for sure. I agree finding a balance between "reality sharpness" and "digital sharpness" can be hard to do especially if you print the files.
I tend to flow more to the more detailed shot, it may be considered less natural I agree but it's a style that I prefer and it works for me.
This is why I found that less Capture Sharpening during the raw conversion and adding a bit of deconvolution sharpening works better, especially on larger prints, 20 x 30 and up.
It's hard to say just how sharp the real world is, all depends on your eyesight I guess. Mine is less than perfect for sure!
Paul
-
Hi Mark,
Viewing distance can alter our impression of detail quite a bit, as nicely demonstrated in this example (http://www.123opticalillusions.com/pages/albert-einstein-marilyn-monroe.php). That's why I recommend using tools that bring such improvements within the reach of us mortals, albeit a bit more subtle, we wouldn't want to freak out people at specific viewing conditions. ;)
Lightroom, as nice and powerful as it is, does not allow the kind of extreme control that a more involved workflow can offer. As such it will not be able to exceed the potential of that more involved approach. It may come close (enough) for some scenarios, people can try for themselves (after being handed some guidance!) as to what fits their requirements best. Thank goodness we are allowed to make our own choices, but it would be nice if its an informed choice.
Cheers,
Bart
Bart,
That example is priceless. Thanks.
As for the worth of various approaches, the LR sharpening workflow is an adaptation (by the same people) of the Sharpening Workflow originally developed by Bruce Fraser and his associates. In PK2 it is a 3 stage process, in LR 2 stage. The output sharpening in LR is an automated process - once the application knows the output resolution and the sharpening is active, the strength and character of the sharpening is done automatically on the fly; likewise in PK, there isn't scope for creativity with output sharpening - it is formula-wise resolution, size and paper dependent. Capture Sharpening is where most of the control occurs in LR, and between the four controls offered there including masking, I find it meets the overwhelming bulk of my sharpening needs very well. It is also very easy to master and occurs right within the raw workflow pipeline so no need to convert the image into a much larger PSD, yet another file with layers etc. By the time Adobe got to the degree of proficiency of LR4, I was satisfied that I could go from camera to print without leaving LR for almost all my editing and printing. Whenever I leave LR and go to PS, it's mainly for some masking or skew correction that LR is still not as good for. Otherwise the quality and convenience of a total LR workflow including sharpening, once you really know the application, is brilliant. So I think part of the choice of a sharpening solution also depends on one's overall approach to the editing workflow.
-
That example is priceless. Thanks.
You're welcome Mark. I knew you could appreciate it.
As for the worth of various approaches, the LR sharpening workflow is an adaptation (by the same people) of the Sharpening Workflow originally developed by Bruce Fraser and his associates. In PK2 it is a 3 stage process, in LR 2 stage. The output sharpening in LR is an automated process - once the application knows the output resolution and the sharpening is active, the strength and character of the sharpening is done automatically on the fly; likewise in PK, there isn't scope for creativity with output sharpening - it is formula-wise resolution, size and paper dependent. Capture Sharpening is where most of the control occurs in LR, and between the four controls offered there including masking, I find it meets the overwhelming bulk of my sharpening needs very well.
I agree that it probably is good enough for most people, and it is not really that bad either. I never said such a thing. But there are even better solutions available, especially for those who need to do exceptional things, like (2.5 - 3 metre high) wall display sizes in trade-show booths. The same tools can be used for more modest challenges, at optimum quality without compromises (even at nose length viewing distances by photographers).
It is also very easy to master and occurs right within the raw workflow pipeline so no need to convert the image into a much larger PSD, yet another file with layers etc. By the time Adobe got to the degree of proficiency of LR4, I was satisfied that I could go from camera to print without leaving LR for almost all my editing and printing. Whenever I leave LR and go to PS, it's mainly for some masking or skew correction that LR is still not as good for. Otherwise the quality and convenience of a total LR workflow including sharpening, once you really know the application, is brilliant. So I think part of the choice of a sharpening solution also depends on one's overall approach to the editing workflow.
Fully agree, as long as people don't say its quality is as good as or even better than some of the more involved workflow alternatives that are discussed, e.g. here (http://www.luminous-landscape.com/forum/index.php?topic=91500.msg744978#msg744978). Close is close, but only better is better.
Cheers,
Bart
-
Hi Paul:
I rarely print these days as I've been on a sabbatical for 3 years. As I said, some of the images were mainly from "old" 5D files and some from an Epson 4870. (It is amazing what these scanners can do with 120 chromes!)
I actually used a printing service for this job in lieu of using my Epson 7880. In the end, only 10 pictures were ordered. I didn't have any matte paper, only a roll of 24" Epson Luster. This job screamed for matte paper. Some were bigger than than 24" wide. So I decided to send them to a print service which I trust as if it was me pushing the Print button.
This job was (as usual) for "yesterday". I had to take a quick decision and according to most responses to my question, it seemed to me that getting Focus Magic was the way to go. After looking at my clients face, I have no doubt Focus Magic was a really smart decision. They are not in the art or graphic biz but they are very wealthy people surrounded always by art and beautiful things. They own a few boutique hotels.
However, in time I'll investigate some more on the suggestions proposed here.
Thanks guys!
Best
Eduardo
With all the time, effort and repeated cost of ink and paper you put into your printing, don't you think it's a bit dumb to worry about saving a few bucks on the sharpening software? I decided long ago, based on the numbers, that software is the LEAST expensive component of the whole digital image processing chain and any additional cost of software that really makes an observable difference in print quality is well worth it. That money disappears in the decimal places, but gives you a better chance for prints to look "great", not just "good".
-
Eduardo.
Glad it worked out for you. I followed Bart's info on Focus magic, both his posts and articles and tried it. I don't use it as much for capture sharpening, but I have found that by backing down my capture sharpening in LR or C1, then adding the FM step, I get better results. This combined with the Topaz Clarity tool have really made for changes in my workflow. (another of Bart's suggestions).
I still use LR for all printing at 360ppi and just let LR do the final upgrez/sharpening steps or downrez depending on the file and print.
Paul
-
[...] I have found that by backing down my capture sharpening in LR or C1, then adding the FM step, I get better results.[...]
Hi Paul,
I agree, if one uses sharpening in LR or Capture One at all, do not do more that the bare minimum, Capture sharpening, not a combination of Capture and Creative sharpening. The user interface of both LR and C1, but the same applies to almost all Raw converters, is not all that helpful for accomplishing that crucial task. Human Interface Design innovation is long overdue in that area.
I generally disable C1 sharpening on Export of the actual image, there's toggle in the output recipe. I leave it on in the presets for display purposes, to identify correctly focused images and the effects of other adjustments easier.
This combined with the Topaz Clarity tool have really made for changes in my workflow. (another of Bart's suggestions).
You're welcome, Topaz Clarity can really lift a veil of drabness from most images. Topaz Detail is another a gem for Creative sharpening and modification of surface structure, and for giving output another boost when upsampling, especially when using a significant magnification.
I still use LR for all printing at 360ppi and just let LR do the final upgrez/sharpening steps or downrez depending on the file and print.
Although it adds some inconvenience in the workflow, there is still some improvement to be had if you export a 'print master' from LR and 'manually' Output sharpen that. Both FocusMagic and Topaz Detail can help with that.
Cheers,
Bart
-
Where does Qimage fit in here? I have been letting Qimage upres the image and then sharpen on the way to the printer. How does it compare to the other methods? Can I do better in LR or what?
Mike
-
Where does Qimage fit in here? I have been letting Qimage upres the image and then sharpen on the way to the printer. How does it compare to the other methods? Can I do better in LR or what?
Hi Mike,
Qimage also produces excellent results, especially because it resamples all the way up to 600/720PPI before Output sharpening is applied on the result of good upsampling algorithms. But Qimage is only available on Windows (unless one uses Parallels or similar on a MAC).
Qimage also offers a lot of other extras, so it's definitely a good choice to use for producing consistently high quality output. It can remember lots of earlier settings for re-use at a later date, which makes it very efficient also for repurposing the same image (different sizes and output media).
Cheers,
Bart
-
Hi,
QImage is a great product and LR is a great product. My guess is that it does not really matter. The way I see it, both QImage and Lightroom prepare the image for printing. I am pretty sure that output sharpening in both tools is based on a lot of real world testing.
Best regards
Erik
Where does Qimage fit in here? I have been letting Qimage upres the image and then sharpen on the way to the printer. How does it compare to the other methods? Can I do better in LR or what?
Mike
-
Qimage has something called DFS-Deep Focus Sharpening. Here is a you tube video for it
http://www.youtube.com/watch?v=AiVoXcB1uzk
all this for $69.00
-
Qimage has something called DFS-Deep Focus Sharpening. Here is a you tube video for it
http://www.youtube.com/watch?v=AiVoXcB1uzk
all this for $69.00
Not sure about the point of the video. I see USM, set too high with an ugly halo yes. We see no sharpening as well, good. We see a better result for sure with DFS. But USM was just not set correctly (or the product doesn't do this well). Is it impossible to use USM, produce sufficient sharpening (for what BTW, capture or output?) without halo's? Was that USM setting a default?
My experience with sharpening for output, it can look pretty ratty on-screen and print better than the opposite. So when the video tells us "you'll see those halo's", that may indeed be possible and it may appear better on the print.
DFS looks better on-screen, I'd like to see the prints! And I'd like to see USM settings that don't hose the image if possible.
-
Not sure about the point of the video. I see USM, set too high with an ugly halo yes. We see no sharpening as well, good. We see a better result for sure with DFS. But USM was just not set correctly (or the product doesn't do this well). Is it impossible to use USM, produce sufficient sharpening (for what BTW, capture or output?) without halo's? Was that USM setting a default?
Hi Andrew,
I agree that the comparison is not how I would have done it (but rather with a much smaller USM radius), but as a long time Qimage user I can confirm from personal experience that the DFS is a rather unique halo free, edge enhancing, sharpening method. As such it is excellent for Output sharpening especially at 600/720 PPI settings which can actually resolve the edge detail that Qimage produces.
My experience with sharpening for output, it can look pretty ratty on-screen and print better than the opposite. So when the video tells us "you'll see those halo's", that may indeed be possible and it may appear better on the print.
DFS looks better on-screen, I'd like to see the prints! And I'd like to see USM settings that don't hose the image if possible.
Correct, but there is not much ratty about DFS to begin with, although real detail may start to be non-existing at high magnifications. What's more, it is applied automatically with printed output which Qimage scales to 600/720 PPI on-the fly during printing (no need for large intermediate files to allow sharpening at that size), although the amount is user selectable. One can instead choose the USM version of smart sharpening, but that also uses a much smaller radius (variable with magnification), however, since DFS there is little reason to choose USM.
So again, the demo is perhaps not the best to demonstrate output sharpening, but it does work very well. I don't know how Mike Chaney did it, but it is a very clever innovation.
Cheers,
Bart
-
Hi Paul,
I agree, if one uses sharpening in LR or Capture One at all, do not do more that the bare minimum, Capture sharpening, not a combination of Capture and Creative sharpening. The user interface of both LR and C1, but the same applies to almost all Raw converters, is not all that helpful for accomplishing that crucial task. Human Interface Design innovation is long overdue in that area.
I generally disable C1 sharpening on Export of the actual image, there's toggle in the output recipe. I leave it on in the presets for display purposes, to identify correctly focused images and the effects of other adjustments easier.
You're welcome, Topaz Clarity can really lift a veil of drabness from most images. Topaz Detail is another a gem for Creative sharpening and modification of surface structure, and for giving output another boost when upsampling, especially when using a significant magnification.
Although it adds some inconvenience in the workflow, there is still some improvement to be had if you export a 'print master' from LR and 'manually' Output sharpen that. Both FocusMagic and Topaz Detail can help with that.
Cheers,
Bart
Thanks, Bart. My workflow has been to use FM for capture sharpening upon export from LR to PS, and then
PK for output sharpening to print from PS. Are you saying it makes more sense to use light capture sharpening in
LR and then use FM for output sharpening for print from PS?
-
Thanks, Bart. My workflow has been to use FM for capture sharpening upon export from LR to PS, and then
PK for output sharpening to print from PS. Are you saying it makes more sense to use light capture sharpening in
LR and then use FM for output sharpening for print from PS?
Hi Howard,
No, FM is much better at Capture sharpening, so I would in that case only Capture sharpen upon Export from LR and use FM then. You can then use your favorite sharpening tools for Output sharpening. You can use the adjustment brushes in LR for only applying local Creative adjustments based on the settings of the Detail panel, but not use the Detail panel for overall sharpening. I assume (haven't tried myself) that you can disable the Sharpening panel but still use its settings for adjustment brushes (but use with care because of the shifted to later Capture sharpening).
In PS you have the possibility to also do other things, like make a sharpening layer with settings based on the center of the image, and one for the more blurry corners. Then blend them with a radial gradient mask.
FM is capable of sharpening the upsampled output (if available memory allows) very nicely. Should you run into memory issues, you can divide the image in selections (using guides to make exact selections) and sharpen those individually in PS.
Cheers,
Bart
-
Thanks, Bart. One disadvantage to my workflow of using FM for capture sharpening after the raw file that I have worked on in LR is exported to PS is that the use of Clarity in LR before the capture sharpening step is not ideal. It would be great if FM could work directly on the raw data in LR as a plug-in, but I guess Adobe won't allow it.
-
Thanks, Bart. One disadvantage to my workflow of using FM for capture sharpening after the raw file that I have worked on in LR is exported to PS is that the use of Clarity in LR before the capture sharpening step is not ideal. It would be great if FM could work directly on the raw data in LR as a plug-in, but I guess Adobe won't allow it.
Hi Howard,
I agree it's not ideal having to do these things outside of LR. It's time that Adobe got their sharpening act together ...
Cheers,
Bart
-
Hi Howard,
I agree it's not ideal having to do these things outside of LR. It's time that Adobe got their sharpening act together ...
Cheers,
Bart
From my rather extensive experience using Lightroom I don't think this is a matter of whether they have their act together. It works very well when used well. Of course everything can always be better, but since I don't have a PhD in the mathematics of digital imaging it would be well beyond me to advise on whether the mathematical properties of working with raw data differ from those working with rendered pixel-based images, posing different kinds of computational issues. And perhaps the kind of algorithms being used in FM are not compatible with a raw converter working on raw data. I wouldn't know - just suggesting that the issues may be more complex than meets the eye.
-
I use Qimage and I find it extremely useful. I also have Perfect Resize, and I use it when I'm printing on canvas.
-
I use Qimage and I find it extremely useful. I also have Perfect Resize, and I use it when I'm printing on canvas.
Hi,
I agree that they both allow to produce very high output quality by using advanced resampling algorithms, and by adding a high quality (adjustable) Output sharpening after that resampling. It becomes clear when one compares output quality with that from other programs, as many people that started using those programs (and thus can compare output quality between workflows) report.
Cheers,
Bart
-
Hi,
I agree that they both allow to produce very high output quality by using advanced resampling algorithms, and by adding a high quality (adjustable) Output sharpening after that resampling. It becomes clear when one compares output quality with that from other programs, as many people that started using those programs (and thus can compare output quality between workflows) report.
Cheers,
Bart
I print almost exclusively with Qimage mainly at 13" x 19". Simply my go to even though I have LR 5 and PS CS6 available.
-
I print almost exclusively with Qimage mainly at 13" x 19". Simply my go to even though I have LR 5 and PS CS6 available.
I do not think much can be gained in the resampling (up + down) + output sharpening phase by using other applications than Qimage Ultimate. Its nice DFS sharpening can be used in Qimage's image editing phase too but QU lacks a proper deconvolution sharpening as a first step in image editing / raw conversion. I use very few of its image editing functions anyway but am hooked on its printing functions. The customers like the output too. For a fast RAW conversion + printing on well exposed images I go in automatic mode with QU and it can be better/faster than with other RAW convertors but only then. Beyond that the image editing functions like the curves tool are too odd to get used to. Sometimes I think that Mike should abandon that image editing part totally but the crop function + the rotate function or rewrite that image editing part in total. For me it is just the RAW conversion in its auto mode, the crop function and then that wealth of printing functions it already has. Enough able RAW converters with integrated image editing exist already. From RawTherapee to Photoshop. There is even a kind of mental separation between good/elaborate image editing and making the physical output that makes me think the two tasks should be split.
--
Met vriendelijke groet, Ernst
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
April 2014, 600+ inkjet media white spectral plots.