do modified dslrs need ir cut filter?

[alacant]

Modified 700d.

After over a year doing this stuff I've only just realised that the red seems way over. Aligning the peaks in dss wrecks the image so I'd rather not do that. I was thinking an ir cut somewhere in the train. Any ideas?

Cheers, TIA and clear skies

[david_taurus83]

Are you not using an LP filter? I have a clip filter in my 600D and it's got UV/IR cut but still picks up Ha emission.

[spillage]

From what I can remember with my memory I am sure that when you modify a dslr you are only doing this to remove some of the blocking ir/uv filter to allow more Ha in. As you cannot just remove some of it you have to remove it all but you really need to but back a filter that stops the back wavelength coming through and allow more of the wanted stuff.

So to cut a long story short I am sure you need to a uv/ir filter to the train. I used to use a baader uv/ir filter that screwed into the end of the nose piece. This meant that I could use a clip in Ha filter at the same time.

Cheers

Mark.

[Merlin66]

The remaining anti-alias/dust shake filter does a good job of UV-IR blocking.

IMO an additional UV-IR is not required.

 

[Guest Tuomo]

 

18 minutes ago, Merlin66 said:

The remaining anti-alias/dust shake filter does a good job of UV-IR blocking.

IMO an additional UV-IR is not required.

 

Then again, if you have removed everything in front of the filter you will need IR/UV filter.

[Marci]

Aye - all depends on the modification. My EOS1000D has had everything removed and _really_ needs a separate IR/UV filter adding when I’m not shooting narrowband...
@alacant - What specifically was removed from your camera when it was modified?

[LightBucket]

1 hour ago, Merlin66 said:

The remaining anti-alias/dust shake filter does a good job of UV-IR blocking.

IMO an additional UV-IR is not required.

 

This comment is correct, removing just one filter, is the “normal” Astro mod, removing both is the “full spectrum” mod, the later requires an IR cut filter to be added to the i aging train...

Aslo don’t do any processing whatsoever in DSS as it’s rubbish at it, but perfect for stacking...

Just use it to stack your image, then take it straight to the processing programme of your choice for anything else...don’t even bother trying to align the peaks, just make sure you have the box checked that says “save changes but to not embed them” and then use the autosave file...

[alacant]

1 hour ago, Marci said:

What specifically was removed from your camera when it was modified?

The hot mirror.

[LightBucket]

Just now, alacant said:

The hot mirror.

Yes, that is the “normal” Astro mod, single filter removal...

[alacant]

34 minutes ago, LightBucket said:

that is the “normal” Astro mod

So no, I don't need an extra ir cut filter?

Cheers

[Owmuchonomy]

No you don’t. Don’t do any processing in DSS but use a proprietary piece of software. I use PS but there are others. Have you watched Rosen’s YouTube tutorials on processing? They are excellent.

[tooth_dr]

Certainly that’s what mine look like in DSS. Although sometimes I just tick align channels during stacking and it comes out ok. Otherwise I follow the instructions in Steve’s book about aligning the peaks. 

[alacant]

2 hours ago, tooth_dr said:

that’s what mine look like in DSS

Hi. Thanks for the confirmation. It seems a pity to have the red off scale and having to have such low green and blue. I wonder if an astro osc does this too? I thought it may be the light pollution with grade 3 to 4 sky here, but I've tried at a darker place too; the same. 

[Stub Mandrel]

34 minutes ago, alacant said:

Hi. Thanks for the confirmation. It seems a pity to have the red off scale and having to have such low green and blue. I wonder if an astro osc does this too? I thought it may be the light pollution with grade 3 to 4 sky here, but I've tried at a darker place too; the same. 

But that's the whole point of the mod - letting in a lot more red

 

[michael8554]

Hi alacant

If you look at your R,G and B "skyscrapers" in the DSS graph you'll see that G is highest, B is lower, R is lowest - but not by much.

So actually red isn't the largest signal.

The image appears to be red because R has a pedestal on it - slide the leftmost R adjustment over until R lines up with the others, job done.

Michael

[alacant]

7 minutes ago, Stub Mandrel said:

letting in a lot more red

Yeah, it works well but e.g. last night's m101: At 8 minutes, ISO800 looking at the APT histogram, red moves off scale whilst green and blue are barely 1/3 the way across. I find that the further right the red is, the less detail (contrast maybe?) the stack has.

[alacant]

9 minutes ago, michael8554 said:

red isn't the largest signal

Ah ok. Thanks. So what's this rule about having to keep the histogram left of centre? When I look at the APT rgb histogram, it looks the same as the dss output.

10 minutes ago, michael8554 said:

slide the leftmost R adjustment over until R lines up with the others

If I do that then take it into e.g. StarTools, it makes it look like when you've enlarged an image too much and  you can't stretch it nearly as much. Same in GIMP:(

[michael8554]

Hi alacant

18 minutes ago, alacant said:

Ah ok. Thanks. So what's this rule about having to keep the histogram left of centre? When I look at the APT rgb histogram, it looks the same as the dss output.

If I do that then take it into e.g. StarTools, it makes it look like when you've enlarged an image too much and  you can't stretch it nearly as much. Same in GIMP:(

The APT histogram is of just one image, the DSS  image is of a stack, very different as a result.

That's the whole point of stacking, to bring up the image without bringing up the noise as much.

Your DSS image looks the same as my DSS results, matching the R signal to the others in DSS has never given me a result like you describe.

If you've got StarTools, doesn't that equalise the levels for you?

Michael

[alacant]

27 minutes ago, michael8554 said:

StarTools, doesn't that equalise the levels for you

Hi. Yes, or I can do it in GIMP. In fact the APT histogram for the individual light frames are in the same position as the dss stack. I thought I had to keep them left of centre. What I see after a certain exposure time is that the red peak reaches far right leaving the b and g well to the left of centre. 

I'm curious as to whether a camera designed for astro would show me the same distance between peaks. I'm considering an osc to replace my dslr.

[Stub Mandrel]

1 hour ago, michael8554 said:

Hi alacant

If you look at your R,G and B "skyscrapers" in the DSS graph you'll see that G is highest, B is lower, R is lowest - but not by much.

So actually red isn't the largest signal.

The image appears to be red because R has a pedestal on it - slide the leftmost R adjustment over until R lines up with the others, job done.

Michael

Not quite, the DSS histos are normalised to have the same areas. Although the red signal is on a pedestal, it's largely one of LP.

 

56 minutes ago, alacant said:

Ah ok. Thanks. So what's this rule about having to keep the histogram left of centre? When I look at the APT rgb histogram, it looks the same as the dss output.

If I do that then take it into e.g. StarTools, it makes it look like when you've enlarged an image too much and  you can't stretch it nearly as much. Same in GIMP:(

You can't just align the peaks in DSS because you are also applying the luminance adjustments. You need to align the peaks and then either edit the luminance curve or place the three peaks (using linked settings) at a good place on it.

[Stub Mandrel]

The above is a modded 450D's response curve. You can see how the modified red sensors are pulling in photons from across a much wider range of wavelengths, so a lot more signal but also a lot more sky background too.

Here's an OSC camera for comparison (a ZWO), the red isn't as flat topped, the blue is lower and the green is broader. All I can be sure of is their histos would look different from the same source. The truth is every camera I have seen a curve for is different! In fact, I haven't really seen two curves for the same brand of camera that look the same.

[alacant]

3 minutes ago, Stub Mandrel said:

Here's an OSC camera for comparison

Thanks. A nice way to visualise it.

 

14 minutes ago, Stub Mandrel said:

You can't just align the peaks in DSS because you are also applying the luminance adjustments.

Another good one. I didn't know the curve represented luminance. Getting there...

[LightBucket]

14 hours ago, alacant said:

So no, I don't need an extra ir cut filter?

Cheers

No you don’t...but also see my post above about not doing any processing in DSS as it’s pretty useless at it...

[Dragon_Astro]

What sort of telescope are you using?

I was under the impression it was refractors that benefited from the IR/uv filter for a shaper image, and newtonians were fine without one?

[Wirral man]

i have just removed my filter from a nikon d5300 and yes everything is red lol so i have just ordered some 50x50mm borofloat uv/ir glass which allows Ha wavelengths to pass through this glass i will cut myself to refit in the camera plus will be able to auto focus again!!