Camera Labs

I've been trying to work out how to incorporate DXO into my workflow.

Here's what I've settled on.

1) Download photos via Photoshop Elements.
2) Tag, rank and keyword photos in Elements browser.
3) Load up DXO
4) Create a new project
5) Add photos that I like to the project.
6) Do edits. This is actually rather quick since I find DXO's defaults to be very sensible.
7) Batch convert the project into the same folder as source images.
Back in Photoshop Elements, a popup appears saying that new images have been detected. Add them to the library.
9) The processed photos should appear right beside the original RAW files. Select the RAW + converted and hit CTRL-ALT-S to stack them.

Aside from the working in DXO and then manually stacking the output, this is almost the same as my current JPEG workflow.

_________________
PhilipGoh.com

Still playing with it...

Tonight I tried putting through some Sony A350 raws and compared them against camera jpeg. High ISO noise reduction was turned off.

Again I'm too lazy to put crops up right now...

ISO100 - very little difference in it. Faint chroma noise in shadow areas in camera jpeg if you look for it.
ISO200-400 - DxO takes a lead here. Camera jpeg starts to show chroma noise all over, particularly in shadow areas. DxO output is sharper and has much lower chroma noise, although some can be faintly seen in shadow areas.
ISO800-1600 - in camera jpeg chroma noise is clearly visible everywhere except the highest exposed areas. DxO output is still fine for the most part, with low frequency chroma noise being noticeable in shadow areas.
ISO3200 - camera jpeg is awash with chroma noise. DxO output is also significantly affected by low frequency chroma noise but maintains more sharpness and contrast generally.

Overall verdict: For the A350 it gives better results at both low and high ISO. To my tastes I'd say it gives about one stop advantage generally for sharpness, and maybe a bit more for noise control. It makes ISO1600 quite usable, with further noise reduction processing to remove the lower frequency component it leaves, but even ISO3200 is a bit too far for anything but smallest images.

Anyone know of software than can do batch lossless crops? For example, I just want to feed it a bunch of identical sized jpegs and want to crop the same region from all. I'm getting too lazy to do it manually for each of these as I normally do...

Another time, I have to repeat this for the 50D...

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

A batch cropper would be very useful indeed

I've downloaded some of the Canon 50D RAW files from Imaging Resource, and have played with them in Bibble, RT, ACR and DXO. Would be nice if there was an easy way to crop them as I'm rather lazy to go through some 20+ images and manually do multiple crops.

What I've noticed, while DXO may smudge the fine details at high ISOs it does make high ISOs usable for small prints. ISO 12800 on the 50D is notorious for chroma noise banding and hot pixels. None of these are visible in DXO(!). Admittedly, it takes about 2 minutes to process one file as I imagine it's doing some really severe interpolation.

_________________
PhilipGoh.com

Detail loss is inevitable at high ISO, so even with the best software in the world you'll never get full resolution ISO100 like results.

I suppose there's two questions on the high ISO output:
1: is it "good enough"?
2: is it better than something else?

#1 depends on your needs, but #2 is easier as a comparison.

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

Of course, but if you look at the ISO 12800 results from the in camera JPEG, ACR, Bibble and any other product you'll see lots of brightly coloured "stuck" pixels. You'll also see some extremely annoying colour banding.

DXO gets rid of these, which I'm thoroughly impressed by. It's still a little too smudgy for regular use, but it's good to know that ISO 12800 is "sort of usable" now in emergencies.

_________________
PhilipGoh.com

Finally getting around to trying DxO on the 50D raw output. Here's some quick test shot resizes to compare the light and bigger picture variations. Here I'm using DxO defaults. There's practically no difference between ISO100 and ISO3200, but you get the low frequency noise appearing at ISO6400 and getting pretty bad at ISO12800.


ISO100


ISO3200


ISO6400


ISO12800

100% crops to follow later.

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

Hi popo,

The differences looks interesting but I haven't followed this thread in detail, I'm afraid - sorry. Are the left hand versions DPP and the right ones DxO or are they both DxO but the left hand ones without tweaks?

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Doh! I'm really not awake today am I?

Left image of the pair is camera jpeg. Right image is DxO. I never figured out how to use DPP effectively.

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

Have the 100% crops!

Again, left set is from camera jpeg, right is DxO defaults. Shot by 50D with 50mm f/1.8 at f/5.6. Due to interactions with shooting in portrait mode, the camera jpeg crops had to be rotated manually in photobucket which doesn't seem lossless.


ISO100


ISO200


ISO400


ISO800


ISO1600


ISO3200


ISO6400


ISO12800

Also I noticed this artefact being generated by both, but more noticeable in DxO output. On the box, the word "decaf" is only in red and white, no dark spots.



Overall, this test doesn't seem as convincing as the earlier landscape test shot. The DxO output is sharper than jpeg in this case, but there doesn't seem to be any more detail. There isn't much in it at the standard ISO range up to ISO3200. In the extended range ISO6400 and ISO12800 DxO does control the bright spot noise, but doesn't seem to able to reign in the low frequency colour noise.

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

Thanks for the crops popo. One thing I notice is that the DXO images appear to have better handled the greenish colour cast introduced by your lighting. Still it doesn't add enough of a benefit over shooting straight JPEG for me to put up with the 1 minute processing time for each image.

_________________
PhilipGoh.com

Too lazy to post more pics, but I've changed my mind about DxO yet again!

Today I went to the park, and while it was dark and rainy, was shooting mostly at ISO1600 with some underexposure. I find this the optimal point as higher ISO seems to raise the noise more than brightening from 1600.

Anyway, for a change, I shot only RAW. No camera jpeg. I got home and put it through my DxO defaults which I found to give a good balance without manual tweaking. As the first few were coming out, they really sucked. It looks like DxO was trying too hard to recover detail, and it was introducing significant colour noise in what was a very grey scene.

I finally got off my bottom at this point and fiddled about with DPP for comparison. I left it on defaults, raw conversion only and put the raws through that. Looks much nicer, closer to what I'd expect from camera jpeg. Interestingly I found DPP to be slow running. Not DxO slow, but nowhere near ACR fast. Is that normal?

While undoubtedly raw can allow more flexibility in both level and detail adjustment, for the effort required I still think jpeg is "good enough" most of the time.

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

DPP is horribly slow. Unlike DXO where it's the batch operations that take a long time, DPP seems to apply things like lens correction and noise reduction on the fly. This is really slow so while the final conversion may not be as slow as DXO, I'm guessing that you've already paid for the time.



Unless you're pixel peeping, JPEGs from your camera are sufficient for most things. What I've noticed is that the vocal photographers shoot RAW while the ones who just get the job done shoot JPEG.

_________________
PhilipGoh.com

My trial's running out and I'm still mulling over buying a license. I don't use it much, but I can see how the lens corrections and lighting tools could be really useful...

*sigh* Software lust.

_________________
PhilipGoh.com

Get a 7D. I don't feel the need to buy anything at this moment

I've gone off DxO. Strangely I found I like its low ISO output compared to jpeg more than likewise at high ISO, which was its original attraction for me. Kinda decided to get to know DPP better as that might be good enough when I do decide to shoot raw, even if it is unlikely to ever support 3rd party lenses.

That reminds me, I need to update the lens correction data in the 50D to add the 15-85 some time...

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

You're not helping me contain my gear lust >.<

In any case, I'm waiting for the 60D.

_________________
PhilipGoh.com