Dodt gradient contrast

Here you can discuss different microscopic techniques and illumination methods, such as Brightfield, Darkfield, Phase Contrast, DIC, Oblique illumination, etc.
Message
Author
hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Dodt gradient contrast

#1 Post by hans » Tue Nov 10, 2020 6:27 am


microbear
Posts: 8
Joined: Wed Dec 16, 2020 3:52 pm

Re: Dodt gradient contrast

#2 Post by microbear » Wed Dec 16, 2020 10:24 pm

It's not so different from oblique illumination. For many in vitro two-photon systems it's pretty standard. Some patching people buy DIC because it costs more, but with DGC there are no prisms in the way of the laser and it's said to be better for thicker slices, which is typically the case for acute brain slices. If you actually compare DIC and DGC the difference isn't as big as the price difference. I've long been wondering why people still regularly buy DIC.
I'm currently tinkering around with an attempt at DGC on my Leitz, but I've little time and the papers that Dodt published aren't that rich in optical information. Thorlabs has some interesting specifics about their DGC system, they use an offset slit and different ones for different NAs. They also run an engineered circular diffuser. I've no idea how much benefit these extras give :( It's a fairly mysterious technique.

microbear
Posts: 8
Joined: Wed Dec 16, 2020 3:52 pm

Re: Dodt gradient contrast

#3 Post by microbear » Thu Dec 17, 2020 1:54 am

I tried it. Turns out my microscope has a small slit that lets me position a thin object between lamp collimator and diffusor. I 3D-printed a quarter annulus (slightly to large diameter) and put it in. It does contrast phase objects like epithelial cells or embedded red blood cells, which are hard to see in brightfield, nicely. The strength of the effect is very similar to oblique illumination but without the intensity gradient in the image, which oblique illumination quickly produces.

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#4 Post by hans » Thu Dec 17, 2020 2:36 am

microbear wrote:
Thu Dec 17, 2020 1:54 am
Turns out my microscope has a small slit that lets me position a thin object between lamp collimator and diffusor. I 3D-printed a quarter annulus (slightly to large diameter) and put it in.
Interesting, not familiar with Leitz, what model? Would you mind taking a photo showing how the illumination system is arranged and where you put the mask?
microbear wrote:
Wed Dec 16, 2020 10:24 pm
Thorlabs has some interesting specifics about their DGC system, they use an offset slit and different ones for different NAs. They also run an engineered circular diffuser. I've no idea how much benefit these extras give :( It's a fairly mysterious technique.
I also wonder about the diffuser -- could a mask with varying optical density (like a graduated ND filter) produce equivalent effect as the sharply-defined quarter annulus combined with diffuser? Or is there something more subtle going on with the diffuser? Possibly some subtlety related to this (comparing to oblique):
Both methods use a mask to generate an illumination gradient, but in Dodt contrast, the mask occurs much earlier in the optical path.
But I don't understand the significance of the mask being earlier in the optical path if it is still conjugate with the condenser/objective apertures...

BramHuntingNematodes
Posts: 1538
Joined: Tue Jan 21, 2020 1:29 am
Location: Georgia, USA

Re: Dodt gradient contrast

#5 Post by BramHuntingNematodes » Thu Dec 17, 2020 3:40 am

Or perhaps some thing like DJs radial diffusion mask:
http://bettermicroscopy.blogspot.com/20 ... i.html?m=1
1942 Bausch and Lomb Series T Dynoptic, Custom Illumination

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#6 Post by hans » Thu Dec 17, 2020 4:21 am

BramHuntingNematodes wrote:
Thu Dec 17, 2020 3:40 am
Or perhaps some thing like DJs radial diffusion mask:
http://bettermicroscopy.blogspot.com/20 ... i.html?m=1
interesting, implies that the diffusive character of the tape is important:
Matte finish tape, however, apears to be superior to their "Satin finish" variety, which seems to exhibit less desirable diffusion characteristics, at least for this use.
But no explanation why diffusivity matters or comparison photos with, for example, a similar mask constructed using transparent neutral density film to give the same pattern of illumination in the objective rear aperture?

MichaelG.
Posts: 3976
Joined: Mon Apr 10, 2017 8:24 am
Location: North Wales

Re: Dodt gradient contrast

#7 Post by MichaelG. » Thu Dec 17, 2020 9:03 am

I’ve just found this : https://www.scientifica.uk.com/learning ... t-contrast

Not studied it yet ... my Caffeine level needs boosting

MichaelG.
Too many 'projects'

viktor j nilsson
Posts: 760
Joined: Thu Aug 23, 2018 10:12 pm
Location: Lund, Sweden

Re: Dodt gradient contrast

#8 Post by viktor j nilsson » Thu Dec 17, 2020 10:12 am

MichaelG. wrote:
Thu Dec 17, 2020 9:03 am
I’ve just found this : https://www.scientifica.uk.com/learning ... t-contrast

Not studied it yet ... my Caffeine level needs boosting

MichaelG.
That was the second link in the first post. You DO need coffee!

Chris Dee
Posts: 216
Joined: Mon Aug 12, 2019 1:02 pm

Re: Dodt gradient contrast

#9 Post by Chris Dee » Thu Dec 17, 2020 10:27 am

I needed coffee just to focus on the web page :D Looks like interesting, but like Hans I'm struggling to see the difference from a gradient NDF (or one laser printed on acetate).

MichaelG.
Posts: 3976
Joined: Mon Apr 10, 2017 8:24 am
Location: North Wales

Re: Dodt gradient contrast

#10 Post by MichaelG. » Thu Dec 17, 2020 10:32 am

viktor j nilsson wrote:
Thu Dec 17, 2020 10:12 am
MichaelG. wrote:
Thu Dec 17, 2020 9:03 am
I’ve just found this : https://www.scientifica.uk.com/learning ... t-contrast

Not studied it yet ... my Caffeine level needs boosting

MichaelG.
That was the second link in the first post. You DO need coffee!
.

My apologies :oops:

MichaelG.
.

Edit: Clever chap, Herr Dodt
https://www.researchgate.net/profile/Hans-Ulrich_Dodt
Too many 'projects'

microbear
Posts: 8
Joined: Wed Dec 16, 2020 3:52 pm

Re: Dodt gradient contrast

#11 Post by microbear » Thu Dec 17, 2020 5:13 pm

After endless search in his papers and not finding anything of value, I checked for patents. E voila, there it is, with all detail.
DE000019514358C2

At https://depatisnet.dpma.de/ the fulltext and figures are available.

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#12 Post by hans » Thu Dec 17, 2020 5:35 pm

Chris Dee wrote:
Thu Dec 17, 2020 10:27 am
I needed coffee just to focus on the web page :D Looks like interesting, but like Hans I'm struggling to see the difference from a gradient NDF (or one laser printed on acetate).
The Leica app note and Scientifica page both mention the diffuser position being adjustable axially, so perhaps sharp mask plus movable diffuser is just a way to control the strength of gradient, but for any particular setting, equivalent to a gradient NDF? However on the Thorlabs module, no indication that the position is adjustable, but they still use a diffuser. (And a pretty fancy one, too...)

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#13 Post by hans » Thu Dec 17, 2020 6:14 pm

microbear wrote:
Thu Dec 17, 2020 5:13 pm
After endless search in his papers and not finding anything of value, I checked for patents. E voila, there it is, with all detail.
DE000019514358C2
Nice find, thanks, do you read German? (I do not, although my name may give the impression otherwise...) Will try machine translation later.

User avatar
75RR
Posts: 8207
Joined: Sun Oct 12, 2014 2:34 am
Location: Estepona, Spain

Re: Dodt gradient contrast

#14 Post by 75RR » Thu Dec 17, 2020 7:08 pm

hans wrote:
Thu Dec 17, 2020 6:14 pm
Nice find, thanks, do you read German? (I do not, although my name may give the impression otherwise...) Will try machine translation later.
Do not know which translator you use. Here are two, providing both as depending on what is being translated either one or the other seems to be a bit better on the day.

https://translate.google.com/

https://www.deepl.com/translate
Zeiss Standard WL (somewhat fashion challenged) & Wild M8
Olympus E-P2 (Micro Four Thirds Camera)

microbear
Posts: 8
Joined: Wed Dec 16, 2020 3:52 pm

Re: Dodt gradient contrast

#15 Post by microbear » Thu Dec 17, 2020 8:08 pm

Yes, I'm a german scientist. I think they filed a US patent as well and google also has a translation. The condensed version of it is:
The tube is fully self contained and inserted into a part of the illumination path were the light source is focused to infinity, basically directly after the collimator of the lamp house. The tube contains to symmetrical lenses of equal focal length, with the focus in the middle. The disk is positioned in the fourier plane although in the drawing it's slightly offset from the focal point. Behind the disk is a diffuser that can be shifted forth and back to control the strength of the effect.
Now I know where to position these two things!
Thorlabs seems to use a different system, interestingly! At least based on what they show on the website, their system is no Dodt contrast. They replaced the two lenses with their engineered diffuser, if I understand it correctly. The engineered diffuser is not the Dodt diffuser but a second one before the annulus. After the annulus it seems they have a traditional diffuser.

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#16 Post by hans » Thu Dec 17, 2020 8:34 pm

Thank you for the summary.
microbear wrote:
Thu Dec 17, 2020 8:08 pm
Behind the disk is a diffuser that can be shifted forth and back to control the strength of the effect.
Does it say which direction this goes? I would guess, the effect is strongest with the diffuser close to the mask, and weakens as the distance is increased?

crb5
Posts: 101
Joined: Sat Sep 12, 2020 11:55 pm

Re: Dodt gradient contrast

#17 Post by crb5 » Thu Dec 17, 2020 9:16 pm

Thanks for the inspiring posts. I just tried a DIY Dodt method on my Amscope 120. I have been testing oblique illumination by placing a sharp edge over the light source to get axial illumination. It gives a nice 3D shadow effect across the center of the field but illumination is uneven (one side dark and one side light). By placing a diffusing disk in the filter holder below the condenser, this gradation is smoothed out. The diffusion disk was made from a 32 mm circle cut from a transparent food container which was then covered in one layer of translucent scotch tape. Having the edge stop and the diffusion disk separated by a few centimeters makes all the difference (cf. an earlier post where the scotch tape and edge stop were combined in one radial filter: #5 Post by BramHuntingNematodes » Wed Dec 16, 2020 8:40 pm http://bettermicroscopy.blogspot.com/20 ... i.html?m=1 ).

microbear
Posts: 8
Joined: Wed Dec 16, 2020 3:52 pm

Re: Dodt gradient contrast

#18 Post by microbear » Thu Dec 17, 2020 10:07 pm

What you're doing there is actually almost the same as Dodt contrast. You aperture is in a fourier plane and you use a diffuser to get a smooth gradient. Now switch to a quarter circle plate and you get the additional advantage of having a gradient in two orthogonal directions, not just one!

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#19 Post by hans » Fri Dec 18, 2020 1:46 am

75RR wrote:
Thu Dec 17, 2020 7:08 pm
Do not know which translator you use. Here are two, providing both as depending on what is being translated either one or the other seems to be a bit better on the day.
Thanks, had not tried DeepL before but it seems to be doing a better job than Google.

The last paragraph of section 2 translates:
Finally, the gradient can be created by vapor deposition of gradient reflective or absorptive layers on glass platelets, where the transmission at each point of the glass plate must imitate the effect of the combination of ring diaphragm and diffuser described above. Finally, this gradient can also be realized by liquid crystal devices (LC-SLM). This has the advantage that the gradient can be arbitrarily can be adjusted by computer. This would also allow the realization of several concentric rings, which in combination with phase-shifting components to over-resolution (Toraldo apertures).
So I guess that answers the question of whether there is anything more subtle going on with the diffuser vs. just being a convenient way to adjust the strength of the gradient.
microbear wrote:
Thu Dec 17, 2020 8:08 pm
Thorlabs seems to use a different system, interestingly! At least based on what they show on the website, their system is no Dodt contrast. They replaced the two lenses with their engineered diffuser, if I understand it correctly. The engineered diffuser is not the Dodt diffuser but a second one before the annulus. After the annulus it seems they have a traditional diffuser.
In the Thorlabs diagram it is not clear how the mask is positioned relative to an image of the light source. However it does show the mask being in a region where the field stop is collimated, which I think is not necessarily inconsistent with it being in a Fourier plane conjugate to the objective rear aperture as in Dodt contrast.

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#20 Post by hans » Fri Dec 18, 2020 6:01 am

hans wrote:
Thu Dec 17, 2020 5:35 pm
Chris Dee wrote:
Thu Dec 17, 2020 10:27 am
I needed coffee just to focus on the web page :D Looks like interesting, but like Hans I'm struggling to see the difference from a gradient NDF (or one laser printed on acetate).
The Leica app note and Scientifica page both mention the diffuser position being adjustable axially, so perhaps sharp mask plus movable diffuser is just a way to control the strength of gradient, but for any particular setting, equivalent to a gradient NDF? However on the Thorlabs module, no indication that the position is adjustable, but they still use a diffuser. (And a pretty fancy one, too...)
Speaking of needing coffee... just realized Chris was referring the "radial diffusion mask" but I responded about the diffuser in the Dodt system. There are more blog posts following the one Bram linked but no further discussion of the diffusivity of the tape that I saw, and the later posts stop calling it "radial diffusion mask", just "radial 3D mask", so maybe they changed their mind. However diffusivity issue aside it does seem like the mask would give illumination similar to Dodt, just not as smooth a gradient.

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#21 Post by hans » Fri Dec 18, 2020 9:40 pm

Chris Dee wrote:
Thu Dec 17, 2020 10:27 am
... (or one laser printed on acetate).
This would be convenient for experimentation, but it appears reputable brands like 3M and HP are no longer available or very expensive, and I mostly see generic stuff with mixed reviews that I'm not sure I would want to put through my laser printer. Anyone have recommendations for decent quality transparency film that is safe for laser printers?

BramHuntingNematodes
Posts: 1538
Joined: Tue Jan 21, 2020 1:29 am
Location: Georgia, USA

Re: Dodt gradient contrast

#22 Post by BramHuntingNematodes » Fri Dec 18, 2020 9:59 pm

Big run on that stuff to make covid masks I thought
1942 Bausch and Lomb Series T Dynoptic, Custom Illumination

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#23 Post by hans » Fri Dec 18, 2020 10:19 pm

Ah, that would explain it. I was thinking even the generic stuff seemed pretty expensive, but I don't remember how much it used to cost back when people actually used it on overhead projectors.

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#24 Post by hans » Sat Dec 19, 2020 8:33 pm

hans wrote:
Thu Dec 17, 2020 2:36 am
But I don't understand the significance of the mask being earlier in the optical path if it is still conjugate with the condenser/objective apertures...
Thinking about it a bit more, if the mask/diffuser were placed just below the condenser as is more common the diffuser would also be blurring the image of the field stop. So perhaps the motivation to add the additional pair of relay lenses and place it earlier is to avoid interfering with the imaging of the field stop?

MichaelG.
Posts: 3976
Joined: Mon Apr 10, 2017 8:24 am
Location: North Wales

Re: Dodt gradient contrast

#25 Post by MichaelG. » Tue Dec 22, 2020 10:51 am

microbear wrote:
Thu Dec 17, 2020 8:08 pm
Yes, I'm a german scientist. I think they filed a US patent as well and google also has a translation.
For convenience ... This page is in English, and includes citations, etc.
https://patents.google.com/patent/DE19514358C2/en

MichaelG.
Too many 'projects'

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#26 Post by hans » Tue Dec 22, 2020 7:25 pm

MichaelG. wrote:
Tue Dec 22, 2020 10:51 am
microbear wrote:
Thu Dec 17, 2020 8:08 pm
Yes, I'm a german scientist. I think they filed a US patent as well and google also has a translation.
For convenience ... This page is in English, and includes citations, etc.
https://patents.google.com/patent/DE19514358C2/en
Google's translation has quite a few strange phrasings. I was getting somewhat less confusing results pasting the OCRed text from Espacenet into DeepL that 75RR mentioned. I never found any US patents.

MichaelG.
Posts: 3976
Joined: Mon Apr 10, 2017 8:24 am
Location: North Wales

Re: Dodt gradient contrast

#27 Post by MichaelG. » Tue Dec 22, 2020 10:38 pm

hans wrote:
Tue Dec 22, 2020 7:25 pm
MichaelG. wrote:
Tue Dec 22, 2020 10:51 am
For convenience ... This page is in English, and includes citations, etc.
https://patents.google.com/patent/DE19514358C2/en
Google's translation has quite a few strange phrasings. I was getting somewhat less confusing results pasting the OCRed text from Espacenet into DeepL that 75RR mentioned. I never found any US patents.
.
Noted, and agreed ... I am a big fan of DeepL
The main reason I linked that page was for all the other references.

MichaelG.
Too many 'projects'

hans
Posts: 986
Joined: Thu May 28, 2020 11:10 pm
Location: Southern California

Re: Dodt gradient contrast

#28 Post by hans » Wed Dec 23, 2020 8:36 pm

In case "circular oblique", "radial diffusion", etc. sounded too dull, we now have "curved, tapered lightsaber" illumination courtesy of Google translate:
As shown in FIG. 3, the increased contrast is clarified by the fact that the object, such as brain sections, is no longer illuminated with a full light cone ( 9 ), but only with a curved, tapered lightsaber ( 10 ).

MichaelG.
Posts: 3976
Joined: Mon Apr 10, 2017 8:24 am
Location: North Wales

Re: Dodt gradient contrast

#29 Post by MichaelG. » Wed Dec 23, 2020 8:52 pm

May the Force be with you, Hans

MichaelG.
Too many 'projects'

MichaelG.
Posts: 3976
Joined: Mon Apr 10, 2017 8:24 am
Location: North Wales

Re: Dodt gradient contrast

#30 Post by MichaelG. » Wed Dec 23, 2020 9:07 pm

Curiously enough :
.
9ED07F3E-554D-4EB1-A700-4571893081B3.jpeg
9ED07F3E-554D-4EB1-A700-4571893081B3.jpeg (45.48 KiB) Viewed 9511 times
.

MichaelG.
Too many 'projects'

Post Reply