DXO Optics Pro 9 Noise Removal (a quick test)

If you shoot with a wide variety of camera bodies and lenses but want to shoot in RAW, there are a few options available to smooth out the work needed to process your shots.

_MG_0115_DxO_s

The original scene. The enlargement is from the centre left.

Converting to DNG format, then opening in Photoshop is an option, but correcting 3rd party lens distortion on each individual shot is laborious. Alternatively you could switch between the RAW converters provided by the camera manufacturer, but they won’t correct 3rd party lenses either. This is where DXO Optics excels. It can load and process most camera/lens formats without any fuss – a real time saver.

In addition to lots of advanced image processing options (including integration with Filmpack 4), it offers a new noise reduction called PRIME (Probabilistic Raw IMage Enhancement), which takes a few minutes to process an image. As it looked like it was doing a lot of work it seemed worth a quick test.

_MG_0115_DxO_nonoise_s

An enlargement from the centre left. No noise reduction – and pretty grainy. Good enough for a small print but quite ugly.

I don’t often shoot above ISO 800, but with a slow wide-angle (a Sigma 10-20mm F4-5.6), no IS and no tripod in a dark interior, higher ISOs are needed. This was taken on a Canon 60D at 6400 ISO – an insane sensitivity for an old film shooter – the nearest film I can remember which would come close was Kodak’s Professional T-Max P3200, but the results would only be useable if you really wanted a very grainy look.

_MG_0115_DxO_standard_s

Standard noise reduction setting. This is very good but there’s still a fair degree of visible noise (look at the pillar on the left).

The standard noise reduction offered by DXO is better than most, but it can’t work miracles as the image above demonstrates. It’s fairly quick to process an 18Mp image however.

_MG_0115_DxO_prime_s

Processed using DXO’s PRIME noise reduction. This takes a few minutes to complete processing on a basic spec quad core I5 PC.

Personally I’d say this is excellent – better than anything else I’ve tried. There’s inevitably a tiny loss of detail (check the detail in those gold finials), but it’s more than worth it for the improvement in noise over the standard processing. It would be better to keep a tripod in the car of course, but in an emergency it’s good to know it’s possible to shoot at high ISOs in an emergency and still get useable results.

Hope you find this useful – thanks for looking.

p.s. I’m not connected with DXO in any way – just using their software.

Advertisements

Olympus Dramatic Tone meets a Lensbaby Plastic Lens

After messing about with the ‘Dramatic Tone’ on the EPL5 using a conventional lens, I resurrected the idea of using a Lensbaby. Rather than using the Sweet 35, the beautifully soft, single element plastic lens was given a chance to show what it could do (remember to hover your mouse pointer over the shots to get a proper level of contrast).

This is something like what I’d hoped for – that fence is surprisingly in focus!

I’d hoped that the interaction between the Dramatic Tone filter and the vague blurriness of the plastic lens would produce something a bit different. All shots taken in RAW using the f4 aperture disk, the lens was mounted in a ‘Composer’, the results post processed in DXO Optics 9 and Filmpack 3.

This is a difficult lens to focus (there’s no autofocus here!), as it never really looks sharp even using focus magnify. The best approach is just not to¬† worry about it – just get close enough to give some sort of idea what the subject is! If you’re going to have a try at this, take lots of shots and expect lots of failures. When it all works though it’s worth the effort….

The Dramatic Tone seems to automatically extend the contrast of what are very low contrast images, which saves a bit of post-processing.The ‘dark glow’ around the branches and rooftop is very nice.

Simple, bold compositions work best – any complexity just ends up as a mushy mess, so keep it simple.

The only slight niggle I have is that the this is a 50mm lens, making it a 100mm equivalent on the EPL5 – and using a moderate telephoto for every shot isn’t ideal. A 0.42x wide angle converter is available but what would be perfect would be something around the 12mm mark, which would mean a 24mm equivalent.

This looks like it’s been layered – but it hasn’t. The odd texture behind the railing seems to be lots of out of very soft focus highlights crossing each other.

So quite a successful experiment which yields some interesting results. The overall softness of the images interacts nicely with the strange Dramatic Tone effect, producing images which are very different from those of the glass lenses. I’ll use this more often!

All taken for the book cover market – hope you like them and thanks for looking!

A Jupiter 85mm f2 on an Olympus EPL5

The popularity of the review of this lovely lens back in February on a Canon 60D’s APSC sensor made me wonder how well it fares on a smaller Micro Four Thirds camera, as some readers are obviously interested (all three of you!). The 2x ‘crop factor’ makes this a 190mm equivalent, so getting into mid telephoto territory. All shot on an unseasonably warm and bright winter’s day, here are the results.

_DSC1480

The ‘brute’ on the EPL5.. The adaptors are a Micro Four Thirds to Canon EF then to M42 thread mount (the lens mount of the Helios). You may notice that the lens isn’t quite lined up – due to the cheapness of the EF to M42 adaptor!

The Olympus EPL5 is a very nicely built 16Mp mirrorless interchangeable lens camera, and it usually has an ultralight Lensbaby Sweet 35 ‘welded’ to it’s front. Surprisingly the weight of the all metal bodied Helios feels quite reassuring on the small camera body, and doesn’t seem overbalanced, but it’s close. The use of the EPL5’s (optionally attached) grip helps handling a lot.

In terms of appearance, it looks exactly like it was made on heavy machinery in a russian industrial complex (it was – probably in Krasnogorsk – ‘Made in the USSR’ is stamped on the underside), with basic aperture and focus markings and an intermittently milled focus ring. Though a bit of a brute I really like it, and the results justify it’s use. There’s obviously no autofocus, and centre weighted metering should be used, along with RAW as exposures can be a little random.

OLYMPUS DIGITAL CAMERA

That fantastic bokeh reappears on the little Oly! This is quite encouraging. As with the Canon 60D the depth of field close up at f2 is razor thin.

On the Canon 60D (below) the lens produced some odd ‘swirly’ bokeh, and a orange colour cast. On the Micro Four Thirds sensor the results are more conventional without any colour shift, but still very good.

_MG_8462s

On the 60D and an APSC sensor

Generally the results weren’t as spectacularly good on the EPL5 as on the 60D. Then towards the end of the test my brain engaged and the reason is obvious:-

On a 60D (1.6 ‘crop factor’) this is a 135mm equivalent, on the EPL5 (2x ‘crop factor’) it’s closer to 200mm. This means to frame the same shot I’m standing further away with the Oly, and as depth of field increases the further away a lens is focussed (or conversely narrows the closer a lens is focussed) the extreme depth of field effects on the 60D just weren’t being allowed to happen on the EPL5 for the same shot.

OLYMPUS DIGITAL CAMERA

Though they don’t have the same ‘erased away’ appearance as the 60D shots, this is still pretty good – much better than a kit lens.

This isn’t a macro lens as such, but with a minimal focus distance of 80cm and a 200mm equiv focal length you can get quite close :-

OLYMPUS DIGITAL CAMERA

Close up (a few feet), and the colours are great.

All in all then, a pretty good lens for the EPL5, even though getting the special ‘look’ of those images on the 60D is more difficult. It’s certainly better than a kit lens on the EPL5 for achieving some decent bokeh – more or less everything is in focus at most apertures with the 14-42 kit lens.

Thanks for looking – hope you find this useful!