unity gain and iso...

[Krawler14]

Was hoping I could get some clarification on this forum about this... I recently asked else where what iso I should be using and was told to find out what the unity gain on my camera was (I have a rebel t3 (1100d)

 

So I did a bunch of digging... And found sites that basically say unity gain is a garbage stat and it isn't really relevant to your preferred iso setting... I also find sites stating the opposite and you should really only be using the ISO thats closest to your cameras unity gain. I also could not find any information on the unity gain of my camera... Only other canon models.

 

Can anyone here offer advice on this? I'm a beginner and have always used iso 1600, I see other people using this setting but also see a lot of people using only 800. Any feedback on the subject would be appreciated! I want to make sure I'm getting the best subs I can!

[dts350z]

I have a bunch of bookmarks on this. Here's one that seems pretty exhaustive:

 

 

http://www.cloudynights.com/page/articles/cat/articles/astrophotography/sub-exposures-in-dslrs-r1543

[s3igell]

After reading through that Article, I fail to find where it even Mentions let alone Discusses the concept of Unity Gain.  Nor does it discuss ISO in terms of an Image Quality issue, rather only that one should lower the ISO if the Stars are in danger of being Over-Exposed.

 

In addition, it is rather dated in terms of both the Canon 20D and the software - and specifies 2006 as "current".

[dts350z]

Well I took it as "look at the histogram", and make sure you've got that gap between the origin and the skyglow mountain. What Iso you use to do that at is a balance between the number of exposures you want to manage and how long you can successfully guide for (and not mentioned but the overhead of dithering, mirror settle, etc.).

 

That is similar to other "don't stress the iso" posts I've seen.

 

If the read noise was not a significant factor in 2006, it can only be less so (or equal) with more modern Canon cameras.

 

For me, imaging in a orange/yellow area with a CLS filter @ f6 to f8 on my astromodified t3i I take 480 second subs @iso 1600. 

 

I plate solve with 10 second exposures @ iso 12800. If I want to sanity check the exposure via histogram, prior to taking a sub, I do a preview of 60 seconds @ iso 12800 (equivalent to 8 min @ iso 1600).

[dts350z]

Other bookmarked articles:

 

http://photo.stackexchange.com/questions/40188/longer-exposure-lower-iso-or-shorter-exposure-higher-iso-what-gives-better

 

http://questions.astrobin.com/question/1309/longer-exposure-time-or-higher-iso-which-is-the-best-for-a-dslr/

 

http://www.cloudynights.com/page/articles/cat/articles/astrophotography/finding-the-optimal-sub-frame-exposure-r1571

[smithn00]

I believe unity is around 300 but you only have 200 and 400 available. I plan to use 200.

[dts350z]

Re Unity gain:

 

http://www.clarkvision.com/articles/digital.sensor.performance.summary/#unity_gain

 

and yes it's hard to find this type of in depth data on newer cameras but read his comments on Unity gain.

[s3igell]

I believe unity is around 300 but you only have 200 and 400 available. I plan to use 200.

For which DSLR ??

 

Until you get to the very New 60 and 7DmkII and possibly the latest 5D, Canon Sensors have had Unity Gains between 800 and 1600.

 

(Unity Gain is a bit different for every combo of Sensor and its accompanying ADC and Processor...)

[s3igell]

Well I took it as "look at the histogram", and make sure you've got that gap between the origin and the skyglow mountain. What Iso you use to do that at is a balance between the number of exposures you want to manage and how long you can successfully guide for (and not mentioned but the overhead of dithering, mirror settle, etc.).

 

That is similar to other "don't stress the iso" posts I've seen.

 

If the read noise was not a significant factor in 2006, it can only be less so (or equal) with more modern Canon cameras.

 

For me, imaging in a orange/yellow area with a CLS filter @ f6 to f8 on my astromodified t3i I take 480 second subs @iso 1600. 

 

I plate solve with 10 second exposures @ iso 12800. If I want to sanity check the exposure via histogram, prior to taking a sub, I do a preview of 60 seconds @ iso 12800 (equivalent to 8 min @ iso 1600).

Basing your Exposure on a combo of your SkyGlow and the Histogram is a Good Practice.  However, many folks will take this to Extremes such that one does "Damage" to the Image by ending up with ridiculously High or Low ISO.

 

Yes, your Mount and AutoGuiding equipment does provide Limits to your potential Exposures.  And one must always Respect those Limits.  But most folks live In-the-Middle, where advice about "Reasonable ISO" - whether you wish to subscribe to the explanation of Unity Gain or Not - is rather well Applicable.

 

Read Noise was Definitely an Issue in 2006.  But as most users were limited to performing Stacking manually in Photoshop, it was often "hidden" by other Issues.  Now that we have the so much more capable and accurate Stacking capabilities of IP and PI and DSS, we are able to Stretch to the point that even Modern Sensors will show Identifiable (if not Intrusive) Read Noise.

 

Interestingly, 1600 ISO is essentially Unity Gain for your T3i...

[dts350z]

There's a 2012, updated in 2015 article on clarkvision on iso and exposure:

 

http://www.clarkvision.com/articles/iso/index.html

[smithn00]

I quote from a DSLR photometry course I recently took:

 

I measured my Canon 1100D camera’s gain as 3.41 electron/ADU at ISO 100, 1.66 electron/ADU at ISO 200, 0.84 electron/ADU at ISO 400 and 0.40 electron/ADU at ISO 800. So at ISO 100 it takes 3.41 electrons to generate 1 ADU. For maximum sensitivity we want to detect every electron read from the potential well, i.e. a gain of 1. For my 1100D this is between ISO 200 and ISO 400, but intermediate ISO values are not available on this camera.

Higher ISO result in the ADU value incrementing in steps bigger than 1 so dynamic range is lost. For instance at ISO 800 my camera’s gain is 0.40 electron/ADU so for each electron read out the ADU value increases by 2.5, and it’s even worse at higher ISO settings. This is the reason I recommend using ISO of 400 or less, perhaps up to 800 if dynamic range is not an issue for a particular project (e.g. exoplanet transit detection).

[dts350z]

But Roger Clark comes to a different conclusion about where the maximum dynamic range is, in the last link I posted (for a different camera).

 

How does one measure a camera's gain? Actually, I guess the dynamic range at each iso, would be most instructive.

 

Some measurement instructions here: http://www.astrosurf.com/buil/20d/20dvs10d.htm

[dts350z]

I found the dynamic range chart I was looking for (600D/t3i) here:

 

http://www.dxomark.com/Cameras/Canon/EOS-600D---Measurements

 

Click on dynamic range and mouse over the orange dots to see the data. Also take a look at the SNR and Tonal range charts.

 

The 1100D/t3 is here:

 

http://www.dxomark.com/Cameras/Canon/EOS-1100D---Measurements

 

My conclusion from all of this would be to shoot at iso 800 for the t3, and 1600 for the t3i (highest iso in the dxomark green{ish} zones on the charts).

[anofeles]

When we right-click on an image, a information about it appears (L, R, G, B, mean, median...) would it be possible to add the measure of the ADUs in a zone or certain point? It would facilitate the calculation of time exposition...
Thanks in advance and pardon for my english.

Regards.

[admin]

When we right-click on an image, a information about it appears (L, R, G, B, mean, median...) would it be possible to add the measure of the ADUs in a zone or certain point? It would facilitate the calculation of time exposition...

Thanks in advance and pardon for my english.

Regards.

 

That would be a nice feature, you should make a feature request in the Feature Suggestion Box.  This is where I go when I have a few hours to spare

 

Regards,