Jump to content
UltravioletPhotography

Why no good dark Violet in digital photography ?


colinbm

Recommended Posts

One bug-bear I have with digital photography is that the sensors don't seem to give a good dark rich violet colour that Violet flowers have ?

I don't know how the emulsion films coped with Violet flowers ?

Is this a limitation of CCD's & the CFA ?

I was hoping that the Sigma Foveon cameras would do better with violet, but they don't either.

The Sigma Foveon sensors are a 3 layered sensor with the RGB photodiodes embedded in the silicon at depths that correspond to the wavelengths of blue, green & red. The blue photodiode is just below the surface at 0.2um, but it fails to do violet.

This brings us back to the colour wheel, which is a nonsense, because colour is a linear response to wavelength. The rainbow doesn't appear to have violet, but nature has violet & brown & pink too ?

So do I have to put up with nice blues to represent violet & never get a true photographic representation of a Violet flower ?

Cheers

Col

Link to comment

Violet and purple are amongst those colours difficult to reproduce and both are extremely sensitive to the colour calibration and profiling of the entire work flow. Your chosen colour space will also be important. Try a wide gamut space.

 

You may try setting the input light source for your RAW conversion software to 'linear' instead of 'daylight' or similar if the option exists (PhotoNinja supports this and probably other programs as well). Profiling your camera for UV might add more "colour fidelity" - now, that is a strange expression for colours that by definition are false :D

Link to comment

In addition to the calibration issues the Sigma doesn't perform that well for UV despite its design.

The blue photodiode is just below the surface at 0.2um, but it fails to do violet.

 

This is the junction depth and although it is sensitive to blue and UV wavelengths, it is optimized for 420nm light. For optimized 350nm you would require a junction depth around 0.01um i.e. 10nm (10nm of p+Silicon is about 70 atomic layers!!). These depths also do not account for overlying layers such as passivation which unfortunately have an additional attenuation effect on UV. Also against the odds, is the fact that UV light is more highly reflected from these structures than blue light.

 

I can tell you that I had high hopes for the SD-10 as a UV camera. So much so that I purchased one of the first available SD-Nikon F adapters (and never used it). I tested an SD-10 in 2008 by removing the internal filter and using a UV-Nikkor 105/4.5. The camera functions for UV but overall the results were quite disappointing due to noise. Further UV testing was carried out and the Foveon sensor was "deemed" a poor UV performer with respect to Nikon 2D sensor arrays. You can see some initial results using Spectralon reference standards here

 

http://nikongear.com/live/index.php?/topic/8897-uv-capabilities-of-the-sigma-sd-camera-foveon-cmos-sensor/

 

Unfortunately I can't find the followup test results from 5 years ago which put the nail in the coffin. Ideally I want a Foveon sensor with 5 pixel wells, UV, B, G, R, IR and some decent on-board processing to reduce noise issues.

Link to comment

Thanks for the useful comments Shane

Unfortunately I can not see the links in the Nikon-Gear article without paying a subscription fee :D

I will be testing the SD9 as it is the earliest model with the Foveon sensor & didn't have any micro-lenses, so that will eliminate them from depleting any UVA response.

The sensors glass cover has been said to reduce the UVA response. I will attempt to replace it with a quartz cover, if I can remove it safely ?

Cheers

Col

Link to comment
Unfortunately I can not see the links in the Nikon-Gear article without paying a subscription fee

Sorry forgot about that! I will see if I can find the images.

 

Not sure about the SD9 but the SD10 was easy - remove the camera body cap and lift out the ICF. On the Sigma SD10 I didn't replace it with anything and don't remember having any significant focus issues, although I only had one for a short time for testing.

 

I am not sure if you are confusing the cover glass with the ICF, they are in fact two separate entites. Every digital camera sensor actually has a cover glass that is typically not removable (without severe likelyhood of damage to the sensor surface) because it is glued/fused to the periphery of the sensor ceramic package. This is actually the "true" moisture/dust/scratch/bondwire protection for the sensor. Above the sensor glass lies another piece of glass, the ICF. This is "removable" and is the component often removed for enhancing the spectral range of the sensor.

Link to comment

Thanks Shane

Yes the ICF / hot mirror on the Sigma dSLR's are easy to remove & replace by the user, as they are in front of the mirror box & just behind the lens mount & is used as a dust protector.

Yes it is the, "cover glass that is typically not removable (without severe likely hood of damage to the sensor surface) because it is glued/fused to the periphery of the sensor ceramic package. This is actually the "true" moisture/dust/scratch/bond wire protection for the sensor", that I want to replace with a quartz cover.

With the Sigma SD10 & later dSLR's I am 'told' that this cover glass is a UVA cut filter & more importantly has the micro-lenses too.

This is why I want to try to replace the SD9's cover glass as it has no micro-lenses.

I have to learn what will soften the bond with the cover glass & the ceramic base ??

 

This is an interesting paper on the early Foveon technology.... http://www.foveon.com/files/CIC13_Hubel_Final.pdf

If I can get close to the UVA results in Figure 2, I will be happy :D

 

Cheers

Col

Link to comment

I have some Amici Prism Spectrograph's on my Flickr page UV Spectroscopy

These are with the Sigma SD15 dSLR without the hot mirror / ICF. The first examples are without a lens & the second examples are with an older 58mm M42 mount lens that I know passes at least to 365nm. The spectrograph's are relevant to each other as they are all centred in the frame at the same place, so the distance from the left is of use to gauge the amount of transmittance into the UVA. All the three types of lamps used are showing Mercury emission lines that give their peak references in the violet/blue range at 365nm, 405nm & 436nm.

Cheers

Col

Link to comment

Colin, the coverglass is removable (at great risk) but I have not heard of anyone removing it from a Sigma, not sure how it is attached. The problem is, it can be as thin as 0.5mm and hermetically sealed. If it fractures during removal it can impact the sensor surface and scratch it, break the surface passivation or damage the bondwires. If you search the cloudynights (astro) forum there are several postings on DIY monochrome conversions which require removal of the coverglass.

 

I have read several of the papers by Hubel of Foveon but you will probably find this Naval School postgrad thesis much more interesting, google it.

 

DETECTING NEAR-UV AND NEAR-IR

WAVELENGTHS WITH THE FOVEON IMAGE

SENSOR

by

Cheak Seck Fai

December 2004

 

Do you have any reference for Sigma microlenses on the coverglass?

Link to comment

Thanks Shane

I'll have a look at CloudyNights at the cover glass references, I'll need all the help I can get :D

Yes, Cheak Seck Fai paper is a good read too.

The only reference to the microlenses being on the cover glass of the sensor is in this recent thread on DPReview's Sigma Camera forum. I did question the remark ? Until I see it referred too with a bit more authority, I would be sceptical.

http://www.dpreview.com/forums/thread/3567089#forum-post-52448985

 

I will be in hospital for a couple of days, so if I don't appear here for a few days I will apologise now.

Cheers

Col

Link to comment
  • 3 months later...

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...