|
Post by michelb on Feb 7, 2019 22:52:01 GMT
Another piece of the puzzle: Is Photos correctly color managed? I have searched and found a lot of posts in the MS forums and other sources. There seems to be a consensus that it is partly, but far from perfect. The most definitive one : The guru behind the user name (unknown member) is... the Digital dog, Andrew Rodney.
So, Photos is sort of color managed with its own interpretation of what a goog picture should look like. You have to disable that automatic color 'enhancement' in Photos.
What Sepiana and I have seen is that with a Prophoto jpeg file properly tagged, we don't see the usual huge loss in colors which happens when the Prophoto file is untagged, and the software assumes its in sRGB (that's a recommendation of the competent authorities). That does not mean the result is perfect, but that was rather a surprise since we expected Photos not to be color managed.
By the way, Bayley, my LaCie 324 is not a 'wide gamut' display, but it is calibrated with my Datacolor Spyder puck. That has absolutely no bearing to the present discussion.
|
|
|
Post by michelb on Feb 7, 2019 21:39:22 GMT
I am currently testing (playing...) with my new smartphone (Samsung S8) which has the ability to output superb DNG raw files.
MichelB, I saw your reply in another thread and thought I would start a new thread to ask my question.
I have a Samsung S8, but I didn't know it could take raw files, and I don't have a clue how to output DNG raw files. How did you do this if you don't mind me asking? I have been connecting to my computer and copying the photos over from my phone. I assumed they were jpeg files? (...I am constantly amazed at the quality of the photos.....)
I was also wondering if the new Topaz software which converts jpeg to raw files would work well on smartphone photos.
Pat The raw option is also available on the S7. When you get the first screen after opening the camera, you have to swipe to the right until you get an array of icons in the bottom of the display. In the middle of the top row, there is a 'shutter' icon called 'Pro'. That enables storing both a jpeg and a DNG file. Warning: this is not 'sticky', you have to think about it before starting your shooting session... I have read interesting posts both in the Adobe feedback forum and at Dpreview in the retouching section. I don't know much about Topaz, but some find this standalone app interesting. What is criticized is the misleading advertisement about the program converting jpeg to raw (nonsense). What it does is about the same as when you open a jpeg in Camera Raw in PSE. They claim to use AI for various corrections, but I don't see myself any advantage compared with PSE. The 'Auto' setting and the ability to batch edit jpegs in ACR. The Topaz module is the same price as the full PSE...
|
|
|
Post by michelb on Feb 7, 2019 7:23:36 GMT
What do you use for the Color Space? Thank you for asking the same question I posted earlier. So the first thing I would suggest is check in your Windows 10 settings what is your screen's default colour space and secondly confirm whether it is actually embedded into your exported file. Bailey, No, not the same question.
Sepiana question is most probably on the right track.
Here is why. You asked two different questions. 1 - Windows 10 colour space settings. Good question, but totally different. That won't probably help much, except confirm that the screen is not calibrated. 2 - The question about "embedding" the display profile is not the same as the question of which profile is used by LR to convert (recalculate all the pixels values from the internal ProPhoto space to another one). Generally (and for good reasons) LR users don't select the display profile, but one of the three other standard modes. In those choices, a conversion takes place, the resulting file is tagged with the colour mode and no embedding takes place. If you choose display profile (I have already noted that it's your debatable choice) LR has to tag the file and embed the profile without pixel conversion. Only with that display profile choice there is question about embedding or not. So, your question comes after Sepiana's question of how LR is set. If your belief that Fauxtoto has chosen display profile, new questions do arise: - does LR automatically embed the display profile? - is the embedded profile kept in the various steps in his workflow? - is Photos able to take an embedded profile into account?
Our common experience with Sepiana is that Photos recognizes the ProPhoto colour space. That's a significant difference with the enlightening discussion she linked to. Now, we are waiting for more details by Fauxtoto. We also can try to reproduce the same behaviour with the different possible LR options.
|
|
|
Post by michelb on Feb 6, 2019 21:02:12 GMT
I am sorry to make things still more complicated... I am currently testing (playing...) with my new smartphone (Samsung S8) which has the ability to output superb DNG raw files. In my tests, I also tested the advantages of Photoshop or LR to work in the wide color gamut of ProPhoto compared with aRGB and sRGB. The DNG was saved from Photoshop with the Prophoto profile.
Elements recognizes the ProPhoto mode and displays it correctly, as expected.
But also does Windows Photos! (Other browsers and even FastStone viewer can't). So, my guess is that Photos has been updated since the 2016 link given by Sepiana.
|
|
|
Post by michelb on Jan 21, 2019 15:04:27 GMT
Hi Michel,
Thanks.
My album has over 2000 photos. Dragging from the bottom to the top takes about 2 minutes - certainly not immediate. What am I doing wrong?
My Plan B is to split the album if there is not a quicker way to move photos in such a large album.
Robert
Robert, With 2000 photos, you are right, scrolling to drag images will be slow even if you choose small thumbnails. My album are rarely over 300 items (for books or slideshows). For such a big number of photos, you must have different sorting criteria apart from the available ones by filename, date or import batch. Very often, one of those criteria will be the main one; very often, you can build your album by importing smaller batches in your required order. Suppose all files are in custom order inside each batch, you simply have to import all batch in your preferred order.
|
|
|
Post by michelb on Jan 21, 2019 14:32:14 GMT
Just two images to make my purpose still clearer. The first one is jpeg in aRGB. The second one is the result of the comparison from difference blend mode and auto level.
So, Bailey, while it is obvious that the result will be seen differently on each monitor and with everyone's personal vision, the resulting file pixels will be strictly identical since it is strictly calculated from the numbers.
The process is identical to comparing two aRGB files, the second having been converted first to sRGB, then back to aRGB. All pixels within the sRGB range are not altered. The second conversion can't restore exactly the out of gamut colors. You don't see any change on screen when converting any sRGB file to aRGB. When we paste an sRGB file over an aRGB file the conversion is implicit.
My purpose is still to reveal the difference. Anyone can get the same difference "by the numers". And anyone with Elements can experiment to find which colors in which images may be "out of gamut". You may find that your monitor is better than you think, that some subjects are worth a wider and truer color range.
|
|
|
Post by michelb on Jan 20, 2019 21:51:28 GMT
Elements supports two color modes, aRGB for print, sRGB for display. It can also support other modes like prophoto when opening from Lightroom or Photoshop, but when you download from your camera those two modes are the one you can choose. With raw files, the conversion from raw is set from your choices in the menu Edit >> color settings: aRGB if you optimize for print, sRGB for display. Even if they are acquainted with color management and understand the advantage of a wider color gamut like aRGB, most Elements users don't have professional printers or displays which can show them the difference. I may add that we don't have all the ability to distinguish between very close hues. There are tests which show that I have lost some ability in getting older. So, my purpose is to use a simple Elements procedure to reveal the differences even if my display or printer can't show them.
To begin, the advantage of wider color range modes like aRGB or ProPhotoRGB is to better render some hues. And that is dependent on the colors of an image. If the colors in a given image are all included in the sRGB color space, nothing will be lost. So, it's useful to know which colors are hard to render in sRGB because they are 'out of range'. Generally those colors will lose saturation.
The second thing to consider is that neither your printer nor your display are exactly sRGB or aRGB. They may be inferior for some hues, but the good thing is that they often are better in other ones. If you can reveal the differences calculated by Elements, you'll be able to check if your display and printer can do better than sRGB and enable you to use an aRGB workflow. You also will be able to check if your favorite scenes are critical for color range (flowers...)
The idea is simple, it's to use the difference blending mode in layers. - first set your Edit >> Color settings to always optimize for print (aRGB) for raw files, and your camera settings to aRGB if you shoot jpeg. - Open the file and check that it is in aRGB mode - File >> duplicate - Menu Image >> Convert color profile (to sRGB) (Now you can toggle between the two modes in the history panel, and maybe notice differences in colors.)
- Select all and copy (Ctrl A then Ctrl C) - choose the original aRGB image and paste the sRGB layer. - set the sRGB layer mode to difference - stamp visible to copy the result to a new layer (Shift + Ctrl + Alt + E) - Do an auto level to increase the contrast. Now, you can see the real differences even if you did not notice them before. Elements can't show out of gamut regions easily like its pro companions, but it can be done.
My hardware is not too basic, LaCie 324 and HP C7280 (6 inks) and Spyder calibration. I was pleasantly surprised that the differences revealed by the above procedure are generally visible both in the display and the printer.
|
|
|
Post by michelb on Jan 20, 2019 15:33:22 GMT
Hi all, I have a large number of photos in an album, and as I add new photos they appear at the end of the album. When I want to place them somewhere else in the album I click and drag. This can take up to 2 or 3 minutes, or more, especially if the new location is near the top of the album. Is there a quicker way to do this? There doesn’t seem to be an equivalent of a Cut and Paste. Any suggestions, please? Thanks Robert The only way is drag and drop. I don't understand your issue since dragging (a single photo or a batch of highlighted ones) is always immediate.
|
|
|
Post by michelb on Jan 20, 2019 15:30:06 GMT
You're just like an old dog with a bone, aren't you? Nope, I am just asking a legitimate question the answer to which I am not sure of but interested in. If you think this question is illegitimate, not useful or potentially helpful to anyone, then since you are an administrator I don't understand why you simply do not delete this thread. It's a legitimate question which can have two kinds of answers in my opinion. The first one is that I am convinced that many printing services operators don't really understand the isue themselves. Their background is printing technology, not photography and their goal is to make their job easy by receiving files of decent size compliant with the kind of technology available to them. It's the same for web hosting or publishing sites. They should concentrate on defining pixel sizes but we see too many requirements which make no sense. The second one is history. Many of the misunderstanding and meaningless rules are coming from old practices which may not be totally false, but which are totally useless today. That is true for printing technology, but that is also true for the software used in the printing industry. The best explanation I have seen has been in the French Adobe forum, where specialists of other softwares like Indesign, Illustrator, Acrobat explained how older sofwares (now extinct) did require totally useless ppi measurements. What was necessary 20 years ago for the softwares to comply with the hardware of that time is now obsolete and a huge diffculty for teachers.
If you consider the printing options for books or scrapbooking, I think that we have a wide choice of solutions. What is more important: optimizing your choice for 300 or 360 ppi or on the other hand, choosing between various paper qualities, glossy or matte, inkjet or silver printing? Taking into account the necessity of cropping for full bleed images?
|
|
|
Post by michelb on Jan 17, 2019 8:11:22 GMT
Wrong. They are not based on any printer (even if my printer is a HP and all of my external printing services require or recommend the mythic 300 ppi). They are based on common values which will be used by most beginners, whether you like it or not. It's already difficult to be at ease with the two parameters: dimensions in pixels or in in/cm when you are using metric and non-metric data.
With beginners: The pre-requisite to master the relation between pixels and inches, and to evaluate how they can view and judge the differences in different situations by themselves. Once that is clear, the viewing distance factor must absolutely be taken into account if they want to grasp the idea of resolution. My examples should bring them to think about why common usage would make them choose the same ppi values for 4 x6 inches, A4, 12" x 12 " and poster size when commonly viewed at the same distance. Common use, but coherent?
With experienced users: You may not agree, but my firm belief is that what you are doing with photography should be judged by the personal visual effect. Even if therory tells you that more pixels, wider bit-depth, color depth are better, you'll find where your own limits are. Just be curious about how other users answer this question: "If my poster is viewed at two meters instead of one, how do you calculate the ppi to get the same visual resolution (which is an angle, not a pixels/length ratio)?
Then, you have to take into accound a lot of other factors like display or printer technology, old rules based on outdated softwares for mysterious reasons, Web sites with incoherent requirements, and data efficiency (why keep and transfer a lot of totally useless pixels?)
For instance the last efficiency factor has to be taken into account in your two asumptions: - ppi is irrelevant for display - process multiple files manages perfectly a batch of portrait and lanscape images.
Generally true, but not optimal, and process multiple files is limited.
About your explanation about default printer ppi: I challenge you to make users see the difference. True and interesting in theory, even if that does not take into account the various printing technologies.
|
|
|
Post by michelb on Jan 16, 2019 22:25:32 GMT
The current challenge for us is now to dispell wrong ideas as you are doing. That is exactly why I pointed out that Sepiana's claim that batch processing portrait and landscape images separately will give better results is not true. I posted why it is not true and asked her why she believes it is as anyone is entitled to. I believe I won't get an answer because her claim is clearly not true. Just try this one:
The dimensions in decimal inches are not accepted, but that's not the issue, you can use pixels instead or cm.
The conflict is in the 'constrain' proportions and the input of width and height. If you don't constrain, the image is distorted. If you constrain, you won't get the respective dimensions. So, the process multiple file process can't solve the original question, which is not totally defined and can be understood in two ways. - is it for max pixels dimensions for displaying (requires the correct orientation and a cropping in width).
- or for printing, where orientation can be ignored and the vertical pictures can be rotated to fill to fit in the maximal printing dimensions?
For viewing you can define height, for instance for a slideshow. The result will be that you'll have to downscale and lose resolution. Or you choose width, which will upscale (more data to store and transfer) and you let the displaying software adjust height. The process multiple file can do either way with both landscape and portrait files, it's up to you to choose W or H.
If I print a batch of 10 x 15 cm files in both orientations, it's critical that I get the most of my resolution, which means rotating the vertical ones. That's automatic when I print on my home printer with the organizer, but I am not sure that will be the case with external services. That may explain the old advice to work separately with portrait and landscape.
For more flexibility in batch printing, I also use FastStone resizer.
|
|
|
Post by michelb on Jan 16, 2019 20:52:34 GMT
I don't teach photo editing to beginners, but I think that I would ask them to discover by themselves what 'resolution' means for their own vision. That requires to get acquainted with: - pixels - dimensions in cm/in - viewing distance.
And taking the trouble to train do the necessary basic calculations.
Then to check where is the optimum resolution for their own vision (no loss of visible detail, no details that you can't see - for instance reading very smal fonts).
Different use cases:
- I am viewing on a 24" monitor, 1920 pixels wide at 60 cm. I can get nearer or farther. I can harly read the PSE user interface at 2 meters. ca 2,5 Mpix
- I can view on a 19" laptop, 1600 pixels at 30 centimeters. Not quite 2 Mpix
- ... on a tablet - on my Samsung A310 (2016), 8cm wide, 1280 x 720 pixels at 20 cm viewing distance ca 1 Mpix
or in print: - 4" x 6" (10 x 15 cm) in at common printing resolution (300 ppi or 120 ppcm) - 1200 x 1800 pixels - ca 2 Mpix. Viewed at any distance from 20 cm to 60 cm.
- A4 format (21 x 29,5 cm) at 300 ppi - 2500 x 3500 pixels - 9 Mpix - Photobook 30 x 30 cm (scrapbook format) at 300 ppi 3600 x 3600 pixels - 13 Mpix. - Poster from the same 9 Mpix photo in 80 x 60 cm, ca 100 ppi, looks very good at viewing distance > 1 meter.
To compare, my cameras range from 8 to 12 Mpix, my first Fuji in 1999 had 1.3 Mpix!
(but this is not the moment to compare the resolution in prints, displays and... sensors).
I am afraid beginners won't catch the meaning of printing and display resolution until they can judge by themselves and they are at ease with the above calculations.
(Note: it's a bit late, please correct me if some calculations are wrong... )
|
|
|
Post by michelb on Jan 16, 2019 17:42:57 GMT
Bailey, We are drifting far away from the purpose of the original post. The danger is that very useful information (such as the issue is points vs pixels, for me) will be buried and not useful for beginners or expert users. I do understand the purpose of cats4jan and I think that such purpose would be best served in a new post. We all have a different background of technical and scientific data, little or advanced knowledge, we have different purposes and we are deeply influenced by the softwares we have used in the past and litterature we have read and studied from books or forums, yesterday or 15 years ago. In our community, everybody is trying hard to be useful: - to help beginners understand the basics
- to help beginners as well as experts to solve problems - to introduce new practical or scientific knowledge - to dispell a lot of misconceptions or outdated ideas - to discuss and share our own ideas and our history (and the history of the question, since a lot of the digital concepts were set even before modern printing.)
The current challenge for us is now to dispell wrong ideas as you are doing. Since those misconceptions are deeply entrenched in today's digital culture, we'll have to debate with so-called experts as well to help beginners to start on good tracks. There are a few themes along with this one (cropping, resizing, resampling) which can't be explained and discussed in a single post or even a single discussion. The main issue is that we are using the same words with different meanings, in different contexts and for different purposes.
My suggestion would be to start a new discussion about the word "resolution". I have too much to say about it myself, so I am sure that there are both expert / experienced users and teachers here to deal with all sides of the resolution issues. Ready for the challenge?
|
|
|
Post by michelb on Jan 14, 2019 10:53:05 GMT
Sorry I have no better idea to solve this issue.
|
|
|
Post by michelb on Jan 12, 2019 21:03:26 GMT
This is an old faq by John R Ellis.
There is a situation where the impossibility to select raw files only in version sets may be a big problem. Suppose I want to recover disk space by keeping only the 'top of version set' and remove the large raw file (32.2 MB for my Fuji RAF files). That makes sense for a lot of pictures I'll never need to reconvert and re-edit. It's easy to "Flatten version sets" when you select collapsed version sets in your browsing space, but when your selection is the result of a search, you are shown both the best match files and the non matching ones as described in the faq. If your files are in stacks and not in version sets, you can use the menu View >> collapse all stacks for the highlighted files. There is no similar menu for version sets which are the general rule for raw files.
The workaround which seems to work with recent PSE versions is the following. - you create an album such as 'Versions to flatten'. - You launch the search and select all - You move the album name over the selected files. - The album now only shows collapsed version sets. Now you can select all the version sets of the album, right click and choose to flatten the version sets.
Flattening is rather slow, which may explain why it is not offered for version sets like in stacks.
|
|