#1  
Old 18-12-2015, 11:49 AM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Modded DSLR -standard naming?

There is much ambiguity in some of the references to the various mods which can be carried out to the DSLR - mainly Canons.
Is full spectrum the same as FULL spectrum?, when is a UV-IR replacement filter required? What about the clear filter replacement?
Can I work with all the filter elements removed??

The phrase "Astro modified" seems a very good choice to describe the removal of only the colour balance filter which enhances the response to the red region of the spectrum - including the Halpha.
To reduce star bloat it is usual to add back a UV-IR filter, or if you want to maintain auto focus with the standard camera lenses add a clear filter replacement.
For the mod which removes both filters (colour correction and the anti alias filters) and allows access to wavelengths from 360nm through to the NIR I think should be called a "FULL spectrum mod" Again the UV-IR or clear filters can be added back depending on the need.

What do you think?
Makes more sense?
Reply With Quote
  #2  
Old 18-12-2015, 06:36 PM
blink138's Avatar
blink138 (Pat)
Registered User

blink138 is offline
 
Join Date: Aug 2009
Location: perth w.a.
Posts: 2,275
yes i find this very confusing too ken
for instance i have the 60da, my brother has modded a 350, 450 and a 60d by purchasing whatever filter and removing one, the results are no where near the same as my 60da
he also recently acquired a "full spectrum IR" modded 5d which he wanted to try out for astro, but the jury is still out on that one, it gave a funny rendition of the LMC
pat
Reply With Quote
  #3  
Old 18-12-2015, 09:05 PM
glend (Glen)
Registered User

glend is offline
 
Join Date: Jun 2013
Location: Lake Macquarie
Posts: 7,033
I would agree with Ken that there is some confusion regarding 'naming conventions' applied to DSLR mods. Those doing mods on a commercial basis, like Gary Honis, and Yun Lee at CentralDS, and Luis Campo, and others, can describe them as they like. However, some standard conventions do shine through:
1. The removal of the Low Pass 2 filter is generally accepted as the standard spectrum mod as it will open up the camera to Ha (which was effectively blocked by that filter). For special 'astro' models like the 6Da I believe that Canon just created a special LP2 with greater Ha reach, whether it is the equivalent to a LP2 removal I can't say.
2. The Full Spectrum title can be applied to #1 above, but it is perhaps more appropriate for the combined removal of the LP1 and LP2. Some people choose to leave LP1 in place as the spectrum passed (as Ken has shown) is very close to that of a Baader UV/IR Cut filter which would likely be added back in front anyway. The anti-alysing function of LP1 may have some value in relation to smearing across the bayer matrix to prevent Miore pattern effect in sensors .
3. A complicating factor emerging now through the mono conversions (debayering) that CentralDS is offering is that the stock Canon sensor coverglass has high reflectivity and is not multicoated - something viewed as essential for serious astro work. There is a good utube video that shows the differences in reflectivity of the stock sensor coverglass and after-market Muticoated Clear replacements from companies such as Astronomik, Edmunds, and others. You can see the utube video here:

https://www.youtube.com/watch?v=pwYmmNtGTqg

There is currently no name for that mod ( I guess we can call it the Coverglass Mod) and only CentralDS seem to be offering to replace it as part of the mono mod. I should mention that it is fairly complicated as a DIY mod due to the bonding over the sensor which requires heat to be carefully applied to the sensor glass surface to enable the glass to be removed.
4. Mono mods, or debayered sensors, bring new language into the Mod conversation. CentralDS is now offering a debayering service for most Canon sensors, using a chemical stripping process to remove the microlense layer and the Bayer Matrix. The result is a sensor that is using the entire pixel population for luminance at all times, greatly increasing the apparent resolution as it is no longer constrainted by the Bayer Matrix into producing only one colour. The price you pay for increased resolution is that the microlenses are lost and thus some photons that previously would have been focused into the pixels may not be picked up now, and this may mean that sub times might have to be increased to compensate for this loss of photon focusing. A secondary effect is some slight reflectivity off the surface which may result in increased bloom effect. CentralDS is very honest about this but their test shots on the website are pretty impressive. I might add that I have had a 450D sensor debayered by CentralDS recently and am in the process of testing it now, with early results showing promise.

Re Cooling mods, I won't get into the cold finger modification process here but it can significantly reduce noise to the point where Darks may no longer be required. Testing done by Rcheshire and myself (separately) confirmed that in the case of the 450D noise ceases to be an issue when the sensor is cooled to close to 0C. At that temperature you can just use Bias/offset frames as your darks as they become near identical to sub time darks. So suffice to say there is terminology around that mod as well but it is a rather rare one at this point due to the complexities involved.

Clear as mud right?

Last edited by glend; 18-12-2015 at 10:22 PM.
Reply With Quote
  #4  
Old 19-12-2015, 07:46 AM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Hmmm,
I believe in the KISS principle...
Anyone considering the "basic" mods to the DSLR may be interested in the attached illustrations.
The first shows the bare sensor response curves with the Bayer Matrix. You will see that the response curve shows a couple of interesting things....
- Down in the UV the blue and red filters still pass a reasonable amount of light.
- Above 650nm (H alpha) the red filter has a good efficiency and the green/ blue have a resonance peak - this gives quite a high NIR response.

The camera manufacturer has rightly established that the sensor needs to be:
A. Colour balanced to suppress the high red response
B. The UV and NIR response needs to be blocked to supress "bloating" effects.

Canon have addressed this "issue" with the use of the two filter system used in their current cameras.
Filter #1 (anti-alias/ dust shake) provides the (B) blocking of the UV and NIR and Filter #2 (colour correction) provides a visually pleasing colour render (A) of "family photos" as well as additional NIR blocking.
If you then multiply the bare sensor response curve by the transmission curves of Filter #1 and/ or Filter #2 you will obtain the final actual response curve of the camera.

The "Astro Mod" would remove Filter #2
The "FULL spectrum Mod" would remove Filter #1 and Filter #2
Attached Thumbnails
Click for full-size image (Canon_450D_Spectral_Response.jpg)
63.0 KB24 views
Click for full-size image (canon filter mod.jpg)
40.4 KB24 views
Reply With Quote
  #5  
Old 19-12-2015, 09:09 AM
glend (Glen)
Registered User

glend is offline
 
Join Date: Jun 2013
Location: Lake Macquarie
Posts: 7,033
Would probably be useful to compare the spectral performance of the Baader UV/IR Cut filter as well, as that seems to be the default filter that imaging people go to after removing LP1 and LP2. Charts below.

For a comparison of the performance of the Canon 60D, 60Da, and a modded 550D, this article commissioned by Canon is a very good insight into the relativities of the three cameras. There is not much difference between the 60Da and the 550D but the edge goes to the 550D to my eyes.

http://www.astropix.com/HTML/I_ASTRO..._60D_550Da.HTM
Attached Thumbnails
Click for full-size image (canonccibaaderspectrum.jpg)
92.6 KB22 views
Click for full-size image (baader1.png)
7.1 KB22 views

Last edited by glend; 19-12-2015 at 09:30 AM.
Reply With Quote
  #6  
Old 19-12-2015, 09:40 AM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Glen,
Thanks for adding the info.
Beware, the difference between the Baader and the Canon curve doesn't represent the INCREASE in sensitivity between the two.
The final red response of the chip is determined by the red Bayer filter and the underlying "Mono" response curve of the chip.
The attached typical CCD chip curve shows the basic performance if the Bayer matrix and all filters were removed.
Attached Thumbnails
Click for full-size image (Typical CCD QE curve.jpg)
19.7 KB13 views
Reply With Quote
  #7  
Old 20-12-2015, 06:17 PM
SteveInNZ
Registered User

SteveInNZ is offline
 
Join Date: May 2008
Location: Auckland, NZ
Posts: 239
I disagree with your last statement and the response curve of a mono CCD is irrelevant. Sure, the base silicon has that response but that's not what is being modified. It's a sensor with a bayer matrix and that matrix has varying levels of transmission over the spectrum, just as your earlier graph of 450D Quantum Efficiency shows.

If you took your 450D graph with the two factory blocking filters and compared it to the response with a Baader filter mod, it would show the relative increase shown in Glen's post.

Steve.
Reply With Quote
  #8  
Old 20-12-2015, 06:30 PM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Steve,
I agree.
The Mono sensor graph is the "base"
The difference between the UV-IR graph and the Canon curve shown could be interepretated as a significant increase in the camera sensitivity.
I just wanted to point this out.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 10:58 PM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Bintel
Advertisement
Testar
Advertisement