Please help me develop a workflow

[simato73]

I am a bit frustrated with the steep learning curve I have to overcome with PSu and would highly appreciate help to develop a workflow that suits my needs.
I am coming from Aperture, where everything worked so sleekly, and I am struggling finding similar functionality. Hopefully in the end I should find that PSu serves my DAM needs much better than Aperture, but I am not there yet.

The workflow I would like to follow is:

1) Import pictures from card. Rename files as desired using a scheme and save them in a logical (to me at least!) folder structure. I think I am managing to do this reasonably well and have saved some import profiles

2) Geotag images. I use gps4cam and a phone for this. I could use the gpx file exported from gps4cam and use it to apply coordinates upon import, but I find the camera/phone time offset is better managed externally. I like HoudahGeo for this action; this also allows to add altitude data. I still have some teething problems with this part but I may be able to sort it on my own, if now I'll call for help.

3) Cull images. Images with wrong focus or grossly wrong exposure, grimaces in portraits, unplanned for motion blur, etc should get eliminated here. I currently assign red colour for marking images for deletion, but I would like to find a quicker and more practical way to mark images for deletion. Any suggestions?
The best way I have found so far to inspect/rate single images is to simply go to full screen more with Shift + Enter. However I am at loss when comparing and picking the best among a group of similar images. I want to compare side by side each image with a reference, or current pick; if the reference is better move to the next image, and so on, until a better image is found and with one keystroke that replaces the old pick; continue until the series is complete. Is there a way to do this efficiently?

4) Remaining images are at least 1 star. 2 stars for images that I want to develop with an external RAW editor to get TIFFs. Before that, however, comes:

5) Reverse geotag. (Done with a batch action)

6) Assigning labels. I could have done this before culling and assigning start but it seems more efficient doing this way since I can avoid wasting time labelling some images and then deleting them. Overall fairly happy here, but I could improve by customizing the layout of the assign pane. If you have done it and found something that works particularly well I'd be interested in hearing from you.

7) Develop 2 star images with external RAW developer (currently Iridient for me). Resulting TIFFs inherit the 2 star rating and the labels of the originating files. When the developed files are ready (in the same folder) I verify the folder to bring them into the catalog.

8) Inspect 2 star TIFFs and decide whether any will be further edited with another external image editing software. The selected images are rated 3 star and edited (not necessarily immediately). Maybe I should give them some kind of colour tag to remind me in the future to do this action, what would you suggest?

The basic workflow ends here. Any image that is selected later for printing on paper gets 4 star rating. Those deemed worthy of hanging off a wall in a frame get 5 stars - both ratings get assigned only at a later stage, as and if necessary.

In addition to suggestions on how to do things mentioned above in bold, please feel free to chime in if you think an alternative workflow might work better.

[Mke]

- quicker and more practical way to mark images for deletion: I normally just delete them as I go...

- comparing and picking the best among a group of similar images: Select the images to compare, then click the Light Table button

- decide whether any will be further edited [and mark them]: I flag the folder icon with a color if there are outstanding actions, rather than individual files - harder to miss. (Preferences » Catalog » Folder State Definitions)

- chime in if you think an alternative workflow might work better: Here's some ideas - http://forum.idimager.com/viewtopic.php?f=57&t=23786

My only other suggestion would be to consider leaving more space at the top of your ratings - if one day you might want to do more with your photos than hang them on a wall you won't have any stars left to rate anything higher!

[fbungarz]

Hi Simone,
just a few quick comments:

1) I generally avoid importing images into PSU, which are garbage. I find FastPictureViewer an extremely efficient tool to cull images from a memory card. It is extremely fast browsing through images and with a single click you can copy the keepers to an folder on your harddisk from which you then start importing into PSu.
You can do that culling in PSu too, but I generally prefer doing it in FastPictureViewer, more efficient. I am typically fairly generous, though, even downloading images that are not quite perfect. The reason: I have learned that, unless an image is completely out-of-focus, you can often still manage to get a fairly decent results with even noisy or strongly under- or over-exposed images. And RAW editors become ever more powerful, so, who knows what kind of detail might eventually be possible to recover...
Therefore - once in PSu I generally, like you, then still mark the "worst" images to perhaps at some point to be deleted.

2) I shoot Nikon and use a foolography bluetooth GPS to geotag images already when shot. For the few images that have no geotags embedded I find PSu's GeoPanel does a good job. No idea about third party tools, sorry...

3 - 4) Rating is a bit a matter of taste, but a good rule of thumb is Peter Krogh's recommendation to rate only the very fewest, best images with 5 stars, the majority with two or three, very few with four and the worst ones with just one. If you use a color label for the ones "possibly delete", then you have an additional category for the very worst ones...

I typically develop photos only when I need them. So in my workflow that is typically the last step. The reason: why spend too much work editing files, if I won't use these images.
Also: I use DxO for batch developing large amounts of files. It generally gives good results without too many fine-tuning based on how it interprets the camera info. For the more exceptional, "valuable" photos I use Photoshop (or recently Affinity) and spend considerably more time fine-tuning.

Like Mke points out: Flagging the folders is a great way to quickly see where in your workflow you are... (you can configure what the colors mean under "Preferences - Catalog - Folder State Definitions).

5) Not sure what you mean by that, but I agree that PSu's GEO Panel is quite powerful. You can copy-paste coordinates, reverse-lookup a locality, etc.
Did you know that under "Preferences - Sync Settings - Read Settings" you can choose to have labels being created from the Geo-Panel (see the options for: Geo-location processing).

6) I do labeling much earlier, often already assigning some labels during import. Do you know that labels are not simple keywords. They are extremely powerful, in my opinion one of the best features in PSu. Right-click a label and check its properties (Details). You can automatically assign Geo-coordinates with a label and fill in one or several metadata fields via "label-mapping" and/or "apply detail profile" (expand Metadata settings).
This way, for example, you assign a single label and have it fill in who is the creator of the photo, his address, job description etc. etc.

Regarding the assignment process: you can easily create so called "label sets". See the + sign at the bottom of the panel to create a new set? Once you have a set created you can easily add a label via right-click "Label Set Membership". How you organize your labels into sets is something you yourself need to figure out, though. Difficult to advise anyone on that.

Hope that helps,
Frank

[tstoddard]

I find FastPictureViewer an extremely efficient tool to cull images from a memory card. It is extremely fast browsing through images and with a single click you can copy the keepers to an folder on your harddisk from which you then start importing into PSu.

I also use FastPictureViewer and agree with Frank about it but if Simone is coming from Aperture, he is probably running a Mac OS. FastPictureViewer runs on Windows only so it might not be an option for Simone.

[snowman1]

Geotagging - I wasn't entirely clear from your post how you are doing this, but it is easy in PSU to geotag imags from a GPX file (using the Geotag panel).

It is also easy to then geotag with reverse lookup details - again using the Geotag panel - no need to do in batch. The reverse lookup on the panel only works on a single image (bug) so the trick is to do this for multiple images is by selecting the required images and using the right-click options to do this.

Whilst the light table is the intended method to compare similar images I find it very fiddly and I will often use image basket to do this task (details and different methods were discussed in a recent thread so I won't repeat them here).

HTH

[simato73]

Thank you all for the comments.
From a quick scan there are many valuable points and suggestions, to which I will reply in better detail later.
However I have had a lot of hard work lately and now it is time to go to sleep

[simato73]

- quicker and more practical way to mark images for deletion: I normally just delete them as I go...

- comparing and picking the best among a group of similar images: Select the images to compare, then click the Light Table button

- decide whether any will be further edited [and mark them]: I flag the folder icon with a color if there are outstanding actions, rather than individual files - harder to miss. (Preferences » Catalog » Folder State Definitions)

- chime in if you think an alternative workflow might work better: Here's some ideas - http://forum.idimager.com/viewtopic.php?f=57&t=23786

My only other suggestion would be to consider leaving more space at the top of your ratings - if one day you might want to do more with your photos than hang them on a wall you won't have any stars left to rate anything higher!

Thanks Mke, good suggestions.
I have been overly cautious so far, I'll just try deleting files that are clearly not up to snuff.
Flagging the folder is definitely something that I have thought I should do. The trick is being systematic about it, otherwise things get missed out and forgotten.
I would also say that 5 stars are for those images that I consider my very best ones; probably those I would sell if photography were my business. In nearly 10 years with Aperture I have used this system and I am very comfortable with it.

[simato73]

I find FastPictureViewer an extremely efficient tool to cull images from a memory card. It is extremely fast browsing through images and with a single click you can copy the keepers to an folder on your harddisk from which you then start importing into PSu.

I also use FastPictureViewer and agree with Frank about it but if Simone is coming from Aperture, he is probably running a Mac OS. FastPictureViewer runs on Windows only so it might not be an option for Simone.

That is correct. Plus for image management (and I include inspecting/rating in it) I'd rather use only one software if possible, it makes things less complicated.

[simato73]

Hi Simone,
just a few quick comments:

1) I generally avoid importing images into PSU, which are garbage. I find FastPictureViewer an extremely efficient tool to cull images from a memory card. It is extremely fast browsing through images and with a single click you can copy the keepers to an folder on your harddisk from which you then start importing into PSu.
You can do that culling in PSu too, but I generally prefer doing it in FastPictureViewer, more efficient. I am typically fairly generous, though, even downloading images that are not quite perfect. The reason: I have learned that, unless an image is completely out-of-focus, you can often still manage to get a fairly decent results with even noisy or strongly under- or over-exposed images. And RAW editors become ever more powerful, so, who knows what kind of detail might eventually be possible to recover...
Therefore - once in PSu I generally, like you, then still mark the "worst" images to perhaps at some point to be deleted.

2) I shoot Nikon and use a foolography bluetooth GPS to geotag images already when shot. For the few images that have no geotags embedded I find PSu's GeoPanel does a good job. No idea about third party tools, sorry...

3 - 4) Rating is a bit a matter of taste, but a good rule of thumb is Peter Krogh's recommendation to rate only the very fewest, best images with 5 stars, the majority with two or three, very few with four and the worst ones with just one. If you use a color label for the ones "possibly delete", then you have an additional category for the very worst ones...

I typically develop photos only when I need them. So in my workflow that is typically the last step. The reason: why spend too much work editing files, if I won't use these images.
Also: I use DxO for batch developing large amounts of files. It generally gives good results without too many fine-tuning based on how it interprets the camera info. For the more exceptional, "valuable" photos I use Photoshop (or recently Affinity) and spend considerably more time fine-tuning.

Like Mke points out: Flagging the folders is a great way to quickly see where in your workflow you are... (you can configure what the colors mean under "Preferences - Catalog - Folder State Definitions).

5) Not sure what you mean by that, but I agree that PSu's GEO Panel is quite powerful. You can copy-paste coordinates, reverse-lookup a locality, etc.
Did you know that under "Preferences - Sync Settings - Read Settings" you can choose to have labels being created from the Geo-Panel (see the options for: Geo-location processing).

6) I do labeling much earlier, often already assigning some labels during import. Do you know that labels are not simple keywords. They are extremely powerful, in my opinion one of the best features in PSu. Right-click a label and check its properties (Details). You can automatically assign Geo-coordinates with a label and fill in one or several metadata fields via "label-mapping" and/or "apply detail profile" (expand Metadata settings).
This way, for example, you assign a single label and have it fill in who is the creator of the photo, his address, job description etc. etc.

Regarding the assignment process: you can easily create so called "label sets". See the + sign at the bottom of the panel to create a new set? Once you have a set created you can easily add a label via right-click "Label Set Membership". How you organize your labels into sets is something you yourself need to figure out, though. Difficult to advise anyone on that.

Hope that helps,
Frank

Lots of interesting points here.

I also used a bluetooth dongle when I had a Nikon DSLR (a competitor of Foolography,but it looked and worked the same) but there is nothing similar for Fuji so I have found an alternative way, which is almost as good.

I did not know what Peter Krogh recommended but it is almost exactly what I do. 5 star rated images are about 0.1% of the total in my database. 4 star are much more common, perhaps up to 5%. I'd say 3 star are about 10-20%.

I have not worked out what is best for reverse GEO lookup. However gps4cam + HoudahGeo allows me to automatically correct for the time difference between the camera and the phone, so that the correct coordinates are assigned. This is done in a clever way and I am willing to discuss this further if there is an interest. Not so happy with the locations I get from reverse lookup in general. I need to study more what are the options, but there are other more pressing issues for me to address first.

What you said about the labels is really completely new to me and potentially extremely powerful if I learn how to use it well. This is really one of the things that will make it worth the effort of switching.

[snowman1]

Simone, you do know you can correct for time differences between cam and phone using the PSU geotag panel? Once you have loaded a GPX file (by clicking on the "GPX track" button and selecting a file) the start and end times of the track will appear next to the button. Click on these to bring up the correction dialogue. Once you have set the difference (if any) click on the "GPX point" button to apply the coordinates to the images.

And as I mentioned before you can reverse lookup by either clicking on the "reverse" button (with a single image selected) or right click an image and select "operations" -> "GEO reverse lookup" (with one or more imges selected).

Don't forget to click on the apply button to save the metada that you apply using these actions.

[simato73]

Simone, you do know you can correct for time differences between cam and phone using the PSU geotag panel? Once you have loaded a GPX file (by clicking on the "GPX track" button and selecting a file) the start and end times of the track will appear next to the button. Click on these to bring up the correction dialogue. Once you have set the difference (if any) click on the "GPX point" button to apply the coordinates to the images.

And as I mentioned before you can reverse lookup by either clicking on the "reverse" button (with a single image selected) or right click an image and select "operations" -> "GEO reverse lookup" (with one or more imges selected).

Don't forget to click on the apply button to save the metada that you apply using these actions.

Sorry for the late reply.
This is indeed what I wanted to do. One could do it upon import but at least now, for control reasons, I want to do it in my own time until I am sufficiently confident the process is in hand.
Currently I am doing it externally with HoudahGeo for two reasons, both of which could easily become irrelevant.
The first is altitude data. I thought only HoudahGeo would do it but PSu seems to do it just as easily. This is good.
The second is setting the time difference between camera time and time recorded in the gpx file (which happens to be phone time).
Without gps4cam I am struggling to find an easy, accurate and objective way to set the time difference. My camera does not show seconds when I display the time, so there is an uncertainty of about 1 minute in any correction I may enter.
For shots taken while en route this does make a little, but appreciable, difference in coordinates.

Despite that I am already preferring the approach you suggest to the one I was using previously.
As mentioned in countless other threads, it is better to let only one program write metadata to files, and since I use PSu, that program should be PSu itself.

[snowman1]

Before PSU had geotagging capability I used to use Geosetter, which is free, but it is a lot easier to keep it all within PSU; and as you say it is one less "boundary" where metadata can get messed up between different software (though to be fair I don't recall that ever happening between Geosetter and PSU).

Incidentally I would think the time in the GPX file is the time from the satellite GPS signals (which is phenomenally accurate). Having said that I believe the GPX standard states the time should be recorded as UTC but my phone seems to either adjust this to timezone or to daylight saving time or both - I haven't yet got the bottom of this due to the combinations of location and season required to properly test - but this maybe something to watch out for.

If I need to edit a GPX file I typically use Routeconverter, which is free and quite simple to use, but there are many free and paid for programs of varying complexity. As I normally forget to turn the tracking app off when I finish a walk I often edit to trim the unwanted points off the end! It can also normalise the altitude from databases, cull the number of points according to various methods etc. I rarely bother with those options nowadays as the receivers seem to have got better and the accuracy seems much better than it used to (I have always used mobile phones rather than a dedicated GPS). Anyway I'm probably getting a bit off topic!

PS I and others have raised some interesting bugs and suggestions over the years in Mantis relating to geotagging in PSU, which you might find of interest.

[Mke]

The second is setting the time difference between camera time and time recorded in the gpx file (which happens to be phone time).
Without gps4cam I am struggling to find an easy, accurate and objective way to set the time difference. My camera does not show seconds when I display the time, so there is an uncertainty of about 1 minute in any correction I may enter.

Use your camera to take a photo of the time on your phone. You can then compare the time the image was taken (from the image file properties) and the time displayed in the photo to easily calculate the offset.

However you'll find that the GPS location may still be significantly out, since the GPS on a smartphone is normally only accurate to within a few metres in open country and significantly worse in some urban areas (even if you've switched it to 'high accuracy' mode). To get better, you'd need a dedicated GPS device able to support more intensive processing and with a higher quality antenna.

[simato73]

Use your camera to take a photo of the time on your phone. You can then compare the time the image was taken (from the image file properties) and the time displayed in the photo to easily calculate the offset.

However you'll find that the GPS location may still be significantly out, since the GPS on a smartphone is normally only accurate to within a few metres in open country and significantly worse in some urban areas (even if you've switched it to 'high accuracy' mode). To get better, you'd need a dedicated GPS device able to support more intensive processing and with a higher quality antenna.

I have thought about it however my camera does not display the seconds, which really means I have a time precision of 1 minute.
As I have found out, one can travel a fair bit on foot in just one minute!
In the case of my last upload my estimate of the time difference between phone and camera (from comparative analysis of locations and times across a large number of images) was about 20 seconds.

[snowman1]

Seconds will be recorded in the exif information though, so as long as the phone can be made to display seconds you can ascertain the difference. But as I've already mentioned I would expect the GPX file to contain the time as derived from the GPS signal (from the atomic clocks on board the satelites), whereas what you see on the phone's clock is simply that. What you want is the delta from the camera to the time as recorded in the GPX file not the phone's clock time. I can't off hand think of a way you could get that, though as this is the *accurate* time you can use a proxy by comparing the camera to any other *accurate* time source, but the camera not displaying seconds rather scuppers that idea. Anyway it all depends what the phone records in the GPX file.

Personally I am not bothered by inaccuracies of <1min as that will give me a location that's as near as I need it - but I can understand some may want more accuracy, especially if you wish to accurately record the position of objects (as opposed to say taking landscape pictures). As Mke says, the inherent accuracy will be limited by the receiver, the satellites in view at the time, line of sight to them, and - unless this has changed - the US do not permit domestic users to exploit the full accuracy of their system (the armed forces can exploit full accuracy and certain users such as Formula 1 have dispensation to use full accuracy). Maybe this will change when the European system is fully functional!