Lightroom version and other preferences

I have just discovered your software and it is the closest to what I have been looking for for some time. However, there are still some problems that prevent it being ideal for me, some of which may be bugs and some just personal preferences, so I will initialy put them all together in General discussion.
My requirement is basicaly some software that will run on a low performance notebook that not only enables me to reject images that are of no use, and also to produce an image of sufficient quality to identify what I have photographed underwater and show them to other divers at the time
1) Any pictures opened in FastRawViewer then in Lightroom appear as Process 2010 and if Lightroom 4 or 5 is being used they need converting to Process version 2012
2) Auto exposure processing seems to move the white point to the maximum and I then need to manualy adjust it to set the EV0 to the centre of the histogram to produce a result nearer to what I hope the end result will be. Setting the curve to vivid makes this even worse (though once the EV0 is adjusted this is the better curve). Much of my photography is underwater and images tend to be flat and an Auto Levels option would probably give the best aproximation to what I am aiming for, for rapid sorting.
3) I use an Olympus E520 with the white balance set to 4600 but my pictures appear in RFV when displayed 'As Shot' as having a value of 4209K/0

1) New BETA of FRV (expected in several days) will allow to choose Adobe Process version for newly created XMP files (process version will not change if already specified in XMP file)

2) By default, Auto-exposure sets 1% of pixels to saturation. You may set this to 0% (for example) in Preferences->Image Display->Autoexposure sets... box.

Also, it may be useful for you to set 'Keep manual exposure correction for next file' setting  (in same Preferences section) to on. This setting is very useful if you work with image sequences shot with same exposure

3) Temperature/Tint are calculated using FRV's built-in color profile. It may differ from camera internal profile. Anyway, we'll investigate E520 case. Could you please provide some samples to work with (you may, for example, put files on Dropbox and send link via Contact page)

Also, for 2)-nd question.
 Adobe products (LR, ACR) apply 'Hidden exposure correction' when processing RAW files (to compensate for the difference between camera lightmeter calibration and Adobe's implementation of ISO standard). Obviously, the value for this correction is different for different cameras (and, worse, different ISO setting). More, it is not always what the photographer wants; and on certain occasions it makes highlights to appear clipped.

To match FRV rendering with Adobe's you need to tune 'Adobe/FRV exposure difference' setting in Preferences-XMP section. Default value (1.0EV) works well with modern cameras (with lightmeter gray value about 4EV from saturation). For older cameras with less 'highlights headroom' you may need to decrease this value.

More on Adobe correction here: http://photographylife.com/adobes-silent-exposure-compensation

I have put some pictures at http://www.richardy.co.uk/FRV/ for you. The original 'Crab' 140823_103042_E-520.ORF file (not one of my best pictures but it illustrates the point) which was taken at 4600K (being reported as 4360K/2) and the way it appears with different settings in FRV and Lightroom. It would be nice if I could set FRV so that the Auto exposure correction produced a result similar to the result from the Auto button in Lightroom.
I have also added an ORF of the garden taken at the 'Cloudy' setting which should be 6000K but is being reported as 5499K/-1
As for the 'hidden exposure correction', unless I have got totaly confused this makes no difference to what is produced by FRV but changes the way the image is displayed in Lightroom so that it can be adjusted to match what is seen in FRV. I was hoping for the reverse, that is to change what is seen in FRV to match what is seen in Lightroom, without changing the appearance in Lightroom.
Thanks again for your fast responce :-)

The color temperature difference is because different color profile is used by Camera (4600K), Lightroom/CameraRaw (4350K) and FRV (4209K). We'll try to do something with it (with this camera only) before next beta.

And, yes, 'exposure difference' changes the crs:Exposure value FRV writes in XMP file. Adobe products uses unknown (to us) hidden exposure correction, so FRV's 'exposure correction' (without any hidden additions) should be translated to Adobe's value. So, you need to try-evaluate-try. You can pass image to Photoshop or Lightroom by single click (FRV Menu: File->Run-> up to three external programs; Use Preferences-External programs if FRV have not found your Photoshop and/or Lightroom at first run).



 

I don't want to change the Lightroom appearance to match FRV, I would like to make the FRV auto exposure result match the Lightroom Auto, as that is far better, so changing the  crs:Exposure value will not help, I think it is the contrast curve options that would need changing, and while the Lightroom algorithm is no doubt secret, something along the lines of auto levels would probably come adequately close for a quick visual assessment.

If you need to exactly match Lightroom rendering, you need to use Adobe programs only (with version match). Unfortunately, Lightroom algorithms are not documented (not surprising for non open-source software) and there is no correct way to duplicate them in non-Adobe software.

Also, there is no way to precisely match in-camera JPEGs (many photographers want it, not Lightroom match) for the same reason: camera algorithms are unknown.

FRV's exposure adjustment method is simple, intuitive, and linear in photograpic terms. For example the 4-stop underexposed image with +4EV exposure correction will look very similar with normally exposed image (with different noise, of course). No color shifts that we'll see in Adobe's products.  No hidden exposure corrections.

FRV is mostly technical tool, made for image inspection without any glamour added. It made to be predictable and repeatable. There is no goal to match any automatic image enhancement tools, such as autolevels: if image is low contrast, it should be clearly visible immediately.

Of course, some kind of contrast increasing may be useful if you need to show RAW image to someone immediately after shooting (e.g. to client or model). For now you may change tone curve to more contrast one. Autolevels may be useful, we considering adding it to FRV, but definitely not in first versions.

I fully understand that FRV is for inspection, not producing a finished product, and I don't want an exact match to Lightroom, just something approaching the contract levels available within the auto button. For many subjects FRV it is fine as it is. However, underwater images are a big proportion of my pictures and tend to be 'flat' so I will certainly add my vote for the autolevels option to appear, even if not till later versions. Vivid is the best at the moment but a step further would be very welcome.

We'll consider to add 'underwater tone curve' (high contrast) to 3rd public beta. No way to catch 2nd beta, that mostly done, added to TODO for 3rd beta.

Music to my ears. It is so nice to have developers that listen :-)

Dear Richard,

Are you using any filters on (or behind) your lens for underwater photography? If not, you may consider using some, depending on the colour of water, bluish or greenesh. That helps with contrast and colour saturation.

One of the main reasons for using RAW is that, having all the data, you can effectively apply filtration during the processing. Filters are a big help if only taking JPGs but the filtration required varies constantly with depth. However, at the depths I often work at there is pretty much no red light left so filters are no help as you can't add what is not there. I get round this by using 2 strobes (powerful flash guns) to add the red back. Theoretically if I set my camera the the colour temperature of the strobes there should be no problems but in practice it is not that easy as colour is very subjective, particularly when recording pictures in an environment that is different from our usual one. It is like taking pictures in tungsten light. When we sit in a room with tungsten light everything looks fine, but take a picture with a camera that does not adjust to colour temperature like our eyes do and everything looks very yellow. The difference underwater is that we never see it in daylight, so we just adjust the images to what we imagine it would look like in daylight, and once achieved it looks natural, even though it is actually artificial.

Dear Richard,

> having all the data, you can effectively apply filtration during the processing

Well, what filters do is they help expose certain channels higher. Say, the red channel is underexposed by 4 stops without filter. In raw processing it will be multiplied by 16 to get proper white balance, or one can put a 2-stop red filter and the multiplication will be 4x times. In my experience it makes for a visible difference in shadow noise, contrast, and acutance.

Having all the data is the goal indeed. With the filter we have more data in weak channels, that is why the benefit.

PS. Found a link, what they are saying coincides with my experience: http://www.magic-filters.com/

Sounds like I could be talking to a diver :-) I can't disagree with what you are saying, but for me, in most instances it does not apply as you are talking far more about taking available light pictures. These are the sort of wide angle shots of large objects (be it wrecks or fish) that are taken in crystal clear water. Last weekend we had good viz for Wales, 4 to 5 metres, The south coast this year is even worse, and in these conditions macro is the name of the game and the subjects are within the range of the strobes so the different channels are not that far apart in exposure requirements. If you are interested you can see some of my pictures at http://www.richardy.co.uk/ The Rapid Bay and Poor Knights Islands might possibly have benefited from filters, but I don't think the difference even there would have been that great.

Dear Richard,

Lovely shots you have. I do dive, but now it is only occasionally, when I need to shoot some wreckage or ruins.

Have you tried B/W Movie curve? It is quite contrasty, mimiced after (once famous) APX 25 film. The image on the display will remain in full colour, of course. This curve seems to help a little, but we are thinking of more drastic measures for the case. Based on non-filtered underwater shots, including the one you uploaded, the best approach may be not through the curve.

Thanks. I had not tried the B/W Movie curve. It never occurred to me that if would work in colour! It is a little better, but I am glad to hear you are considering other measures. Most of my diving is with SeaSearch. There is a bit about it on my page in the left column of the diving gallery section. This is basically a voluntary organisation that surveys areas and records the type of habitats and the life that is there. Photography is a great help to identification but, for example, last weekend I came back with over 300 pictures and while away I only have a rather slow notebook so software to help me rapidly sort the shots and show them to other divers at the time, while being limited to a slow computer is just what I need. Even better is that having done something with them while away, I won't have to do it all again when I get back to Lightroom. Keywording, when you get round to it, will be another fantastic help on top of all that :-)
If you want any more typical underwater RAW files just let me know. I will be happy to help your development in any way I can

I'm sorry, I don't seem to be able to edit my last reply. The 140823_103042_E-520.ORF file was taken at 4600K and is reported as 4209K/0 by FRV, not 4360K/2 as I put in my reply, that was after some editing of the image.

Dear Richard,

The difference between 4350K and 4209K is 8 mired, while the lightest light colour correction technical filter is 20 mired. It is only 12 mired between 4600K and 4350K. In real world such difference in mired constitues a negligible tint, but the descrepancy in readings is certainly something that jumps out. One just need to realize that the tints are linear in mired (1000000/ColourTemperature), not in colour temperature readings and ignore these small differences. Also, we will be working to bring the calculations closer, however the big question is - closer to camera. or to ACR/Lr? Well, maybe both...

The main reason I would like a more consistent figure is that, having adjusted an image's colour balance to suit my taste, I would like to know what to set the camera to, to replicate that result without needing post processing, so reducing the number of adjustments needing to be applied during the initial sorting process. This would be easy enough using the same light source on land just using trial and error, but adjustments underwater are not so easy, so the ability to make an adjustment based on what previous shots needed would be great. For that reason closer to the camera would get my vote, if I had to choose just one option.

Add new comment