4 image viewers and what they display on my screen!

[soundbyte]

Had a task yesterday to get a photo of a Little Wattlebird ready for possible posting and during the process I decided to see what differences if any I could see when the image was loaded into 4 well known image viewers, Faststone Image Viewer 7.5, Gimp 2.10.22, XnViewMP .98 and IrFanView 4.56.
All screenshots done with Greenshot and saved as jpg.

Image that was processed with XnViewMP, Topaz Denoise and Irfanview and saved as a jpg at quality 100%.

You can only see EXIF info for this image if you are logged in.

Screenshot results from the 4 image viewers.

Faststone Image Viewer 7.5

You can only see EXIF info for this image if you are logged in.

Gimp 2.10.22

You can only see EXIF info for this image if you are logged in.

XnViewMP .98

You can only see EXIF info for this image if you are logged in.

Irfanview 4.56

You can only see EXIF info for this image if you are logged in.

An interesting exercise, apples aren't apples when it comes to what image viewers display on my screen it would seem, YMMV.

 

[JP Onstwedder]

That’s amazing that a jpg would display so differently! If it was a raw image I would expect results like this. I can add a few image viewers to the comparison; is the top image full size?

 

[soundbyte]

I thought the image would be somewhat similar whichever viewer I used but as it was not I thought I should share the results.
The image I used for display is a 4064 X 2882 pixels X 72dpi X 7,296,145 Bytes, the one posted above is a 50% resize done after watermarking.
The complete bird image posted above could be downloaded and used as control sample for anyone to use I guess.
It has been changed from what I posted by the web site software, it is now 2000 X 1419 pixels X 96dpi X 452,419 Bytes.

Nothing better than doing your own comparisons with your own equipment and choosing the viewer that suits you best.

Worth a try to add more simple comparisons if you wish.

 

[JP Onstwedder]

I looked at four applications on the Mac - Capture One 20, Affinity Photo 1.8, GrowlyPhoto and the built-in Preview. No differences in how they display JPGs, or rather, just very minor differences in colour. So maybe it's a Windows thing?

 

[Woodpecker]

Had a task yesterday to get a photo of a Little Wattlebird ready for possible posting and during the process I decided to see what differences if any I could see when the image was loaded into 4 well known image viewers, Faststone Image Viewer 7.5, Gimp 2.10.22, XnViewMP .98 and IrFanView 4.56.
All screenshots done with Greenshot and saved as jpg.

Really interesting thought !

I have been using several raw processing packages in the past (View NX2, Capture NX2, vIew NX-i, Capture NX-D, Capture One) and am now using DxO PhotoLab 4.
On the viewer side used IrfanView and then moved to XNViewMP because of its much better organizing and meta data management functions as well as the multi-platform availability (beside photos I have to juggle big quantities of graphics that I manage by using key words).
When looking at processed photos I was more concerned about differences between what I see in the raw processor and the viewer rather then between dfferent viewers, so I will definitely have a closer look with my equipment, which currently is a DELL Precision M6800 with a 4GB AMD Graphics and connected to either two DELL UP2716D or two DELL U2413 - depending on the location where I am working -, in both cases connected via miniDP through a docking station.

With my setup I could not see visible differences between JPG's using the programs I mentioned above. Differences between the two work places were minimal, considering that I use two pairs of differetn monitor models that are both calibrated, but used under differetn lighting conditions.

Something I can confirm that color management under Windows can make life a bit complicated sometimes . It is not so bad as long as you use applications that allow settings for color management. After getting these settings right jumping back and forward between different software packages isn't really an issue. Problems might occur if you use Windows' internal things like the photo viewer or using JPG's of your photos as wallpaper or background slide show. E.g. iIf you save JPG's with integrated color profile and use these JPG's for this you might have severy, ugly color bending caused by this integrated profile, because Windows can'T cope with it.
The solution is to store these JPG's separately without an integrated color profile. This can be done easily with batch cvonversion e.g. with XnViewMP.

Another thing I had to learn is that as a default Windows is importing JPG's for display with 80% quality setting - officially in order to save loading time - which IMHO is a joke with the hardware we have today. This can be changed by an additional entry in the registry that has to be added manually. Interestingly this not only shows an effect for wapp papers and the screen saver slide shows, but also for the photo viewing itself.

Coming back to the display quality of viewers there is one thing I haven't tested recently, but I remember discovering when dealing with my business graphics exported as JPG's in the past. Some of the viewers offer different options for the algorithms to be used for resizing pictures including resampling them. However, although they are mostly found somewhere among the picture processing preferences, they might have an impact on the way pictures are displayed as well ... but as I said, this was just an observation - no scientific test behind it ☺ . It is worthwhile playing around with the display and processing options in the program setup a bit to get rid of this kind of deviations.

Another source of trouble is saving JPG's from different programs. Many of them have options for JPG export like the quality setting (in the end defining the compression level). Unfortunately a quality setting of 85% is not doing the same in all the programs which also can lead to this kind of deviations.

 

[dabhand16]

I don't know the Greenshot software but I'd rather have seen jpgs direct from the apps.

 

[soundbyte]

Thanks everyone for your thoughts and comments.

I don't know the Greenshot software but I'd rather have seen jpgs direct from the apps.

The viewing of one file with various viewers which showed in my opinion some quite large differences, then capturing those differences from my screen was what I was commenting on, that is all.
Consistency of screen capture was important in attempting to show the differences, one file for display in each viewer and one capture option (Greenshot - https://github.com/greenshot/greenshot) for all the viewers.
Saving from the viewers, then viewing the results is another matter entirely, adding more steps in the process that would most likely influence the results in many other ways as mentioned by @Woodpecker.

It was giving a heads up that there may be problems with viewers and that one should check any viewer/s they may be using for any such display differences which could influence decisions when Post Processing(PP).

Most image viewers have ways to change the settings to make the displaying, saving, converting files etc; more consistent and I highly recommend finding how to do so, to hopefully improve your results.

As always I find it best to check what PP software is available for my systems (Windows, Linux Mint, MAC uses different screen), try them out to see what I prefer overall, then stick to my selections.
I regularly calibrate the screen, check for any updates for my chosen PP applications (updates do not always go as planned though!), as well as revisit the settings after updating any software, they may have been reset during the update installation.

 

[dabhand16]

Consistency of screen capture was important in attempting to show the differences, one file for display in each viewer and one capture option (Greenshot - https://github.com/greenshot/greenshot) for all the viewers.

Here's the thing though. If someone is using XYZ viewer and has been for years and is very happy with it, then downloads the ABC viewer to make the comparison you suggest - and finds differences - which one is correct? If you tried to get a reference image you would need to match so much that I doubt it could be done, and how do you know the reference image is correct?

I'm wondering if the effect you have is moire. Does your camera have an anti-aliasing filter? If not moire can be an issue.

 

[bleirer]

Assuming the original was an out of camera jpeg? The camera could have been set for either srgb or Adobe rgb and it would be embedded in the jpeg. The internet speaks srgb, so unless we view the download Adobe rgb will look weird. Color managed application like Photoshop has rules that the user can set for what to do with incoming color spaces. The best bet for the web is to convert to and imbed srgb upon exporting to the web, then it is all good. Viewers I'm guessing don't alter the colorspace or maybe they are not color managed at all, so that would be worth checking,

Raw images don't have a colorspace no matter what is set in the camera until one is applied by the raw converter. Lightroom uses a version of prophoto no choice, but you can export in prophoto, srgb, or adobe rgb. For the web pick srgb and imbed the profile. Photoshop is the ultimate color manager.

 

[soundbyte]

I am talking image viewer/s here not image processing and/or organising images and/or data management etc, they are in and of themselves other subjects entirely!

@bleirer The image is the same jpg with no metadata, only the image viewer changed, each viewer will/may have different displays for whatever reason/s.
Can you change the settings in the viewers? Maybe, but bog standard installs of 4 image viewers produced these results on my system, what about yours?

@dabhand16 Agreed it is a moire pattern that is seen, worse in some, better in others.
If moire was a problem inherent in the jpg image it should have been apparent with all 4 viewers?

So many different methods of display, conversion etc, it comes down to what you personally decide as being the most usable/best/correct(?) for you.

For me IrFanView is the best overall image viewer I have tried, followed by XnViewMP (I use this for its PP options), for you it will/may well be some other image viewer.

As I said above "Nothing better than doing your own comparisons with your own equipment and choosing the viewer that suits you best. ".
You may find nothing, that is OK, but you never know until you try some different image viewers?

@JP Onstwedder, @Woodpecker, thanks for posting your results and conclusions.

 

[dabhand16]

I am talking image viewer/s here not image processing and/or organising images and/or data management etc, they are in and of themselves other subjects entirely!

Not disagreeing with most of what you say, (in your quote and elsewhere) but Gimp is a full-blown powerful editing application, not just a viewer, so out of your original images the Gimp one is the one I'd trust. Apart from the AI features that PS now has it can hold its own against it - and I've seem the proof with Gimp edited images that are awesome.

I've been using Faststone Image Viewer for around 15 or more years, but I only use it as a viewer. I occasionally use the built-in magnifier but never go to the magnification you did in post #1. If I want to examine an image more closely I do it in Photoshop and usually using the tiff rather than the jpg that has already been degraded in its' production. Maybe a comparison with tiff files would have been better? So much information is discarded when a jpg is produced and it is the way that apps do this which is causing the differences you highlight. So as I said, the only one I'd trust from you first post is Gimp. Faststone I mage viewer has some editing capabilities including two things that Photoshop Elements does not have (or didn't last time I looked) - cloning and curves but I don't consider it as anything other than a viewer.

 

[bleirer]

I am talking image viewer/s here not image processing and/or organising images and/or data management etc, they are in and of themselves other subjects entirely!

@bleirer The image is the same jpg with no metadata, only the image viewer changed, each viewer will/may have different displays for whatever reason/s.
Can you change the settings in the viewers? Maybe, but bog standard installs of 4 image viewers produced these results on my system, what about yours?

@dabhand16 Agreed it is a moire pattern that is seen, worse in some, better in others.
If moire was a problem inherent in the jpg image it should have been apparent with all 4 viewers?

So many different methods of display, conversion etc, it comes down to what you personally decide as being the most usable/best/correct(?) for you.

For me IrFanView is the best overall image viewer I have tried, followed by XnViewMP (I use this for its PP options), for you it will/may well be some other image viewer.

As I said above "Nothing better than doing your own comparisons with your own equipment and choosing the viewer that suits you best. ".
You may find nothing, that is OK, but you never know until you try some different image viewers?

@JP Onstwedder, @Woodpecker, thanks for posting your results and conclusions.

I think it is as simple as some viewers are color managed and some are not. I believe Faststone and Irfanview have settings for color management and I'm sure others as well. GIMP would be similar to Photoshop I described above. But it is the wild west to view an image in a non color managed program.

Was the jpeg out of camera? If so would you know the colorspace? If you saved the image did you imbed the colorspace? What colorspace? All things that could be factors.

 

[dabhand16]

I think it is as simple as some viewers are color managed and some are not. I believe Faststone and Irfanview have settings for color management and I'm sure others as well. GIMP would be similar to Photoshop I described above. But it is the wild west to view an image in a non color managed program.

Was the jpeg out of camera? If so would you know the colorspace? If you saved the image did you imbed the colorspace? What colorspace? All things that could be factors.

I don't think this is a colour management issue - although I agree that using an application that can read ICC profiles is desirable. Colour management only affects the colours - not how the app renders the image.

 

[Andrea]

I think it is as simple as some viewers are color managed and some are not. I believe Faststone and Irfanview have settings for color management and I'm sure others as well. GIMP would be similar to Photoshop I described above. But it is the wild west to view an image in a non color managed program.

Was the jpeg out of camera? If so would you know the colorspace? If you saved the image did you imbed the colorspace? What colorspace? All things that could be factors.

I fully agree, you must be shure that all the used viewers are able to manage the color space embedded in the jpeg, or verify if they use this color space or assign by default different colour space.

 

[soundbyte]

Thank you everyone for your comments and thoughts.

I see very little in difference between the colours in the five images posted on my system screen, the screen display is calibrated regularly.

If you discern any colour differences between these images would you please identify such colour differences and an explanation of where these differences are in the images?
What part of any colour space error/s could cause the varied (moire?) patterns seen in the four magnified images?

Is this a problem with calibrated screens and if so what do you use to identify any colour errors? How do you know if there are any?

Is there a way to identify the actual colour space that is displayed?
If the system says that for instance you are using AdobeRGB colour space, how do you check?

Can you see any colour space errors in your image viewer display?
How do you identify and resolve these colour space errors?
What about on a non calibrated screen?
How much does it matter?

I could see differences with other quality aspects of the image (other than colour) displayed by the four image viewers and that got me thinking of whether what we see is consistent across viewers, obviously it's not, but exactly what causes these differences is not agreed upon....yet?

I am sure there are other image viewers that also show differences with a simple jpg display, but I have not tried any others than those named.

I was hoping that a few more readers would conduct their own comparisons and post the results.
Saying that one image viewer over another should be trusted jut because it is popular or has a lot of features etc... does not guarantee that is the case.

As to what causes these differences may become more evident with more data.
I suspect it is more to do with the decoding algorithms than anything else, how do we choose the best?

SO many questions, I would like to learn more.

 

[dabhand16]

SO many questions, I would like to learn more.

I think at the end of the day if you are happy with your images, don't beat yourself up trying to figure out why different software works differently. Use what gives you what you like and be happy.

I know a guy who wants to delve into the minutia of pretty much everything and is always reading or watching YouTube videos to the extent he does not actually DO anything. He wants to master every aspect of Photoshop before he starts editing. Lots of knowledge, but no idea. He has a Nikon D750 but his pictures are point and shoot standard.

 

[soundbyte]

@dabhand16 I am totally satisfied with my software selections, I note others are not.
Really only trying to get people to try things out for themselves as I see a lot of dissatisfied people on forums that might get something from the exercise.

Not interested in knowing everything but some knowledge comes in handy at times.