10 min subs versus 20 min subs - empirical results...

[Darth Takahashi]

Sorry the images are great and I would have loved to have taken them myself. I have found the article again and it is attached below.

Lets take this further by PM or start a new thread because its very interesting info...

Aza in my experience going beyond 10mins with a DSLR is a waste of time and that's with both my 1DMIII and Kiss 2X (450D). I can post some 20min subs of B33 & flame which are very disappointing . I kept going long and longer 3-5-10 & 20 mins and the noise kept pace!

Neil.

DSLRvsCCD_API.pdf

[Psychobilly]

The technique I use to compare the noise performance of a camera is to ....

Starting at the highest ISO setting...with the Viewfinder and lens capped take darks...

Take 1,2,4,8,16 min subs leaving a time equal to the sub length between subs...

examine the subs for "noise"

Drop to the next ISO setting and repeat the sequence

this until your at the lowest ISO setting you want to test

Next pick a setting where there is noise and take a series of subs of 3 leaving the same gap between subs ands the sub length

Examine the subs and look for changes in the noise

Next take groups of 3 subs leaving gaps of 1/2 , 1/4 , 1/8 and 1/16 of the sub length between subs again look for differences between subs in the group...

What your looking for is any growing "glow" in the images or other noticeable changes

What you will eventually end up with is some idea what the max sub Length you could hope to use is at each ISO setting and what time interval you need to leave between subs....

My Nikon D200 needs almost the same time as the sub length between subs

My Canon 350D needs at least half the sub length between subs

My Canon 1000D needs only a few secs between subs...

Just "play" with the camera and see what happens...

Apologies TJ for the drift... answering a question in thread....

Peter...

[Aza]

Cheers... I have read through it, and to be honest I dont understand most of it

But what I can make of the graphs he shows there, it seems the best result would be 1 minute subs at ISO 200.

Now this is not the case at all for me....

[Aza]

Thanks Peter, I shall have to have a play around with your advice there

[Tim]

Aza, take a 10 and 20 min on the same night. The stretch them to the same degree in PS. Might help???

The DSS website has a list of dslrs and the recommended iso speed to use with them.

It is possible though that individual cameras behave differently, that is true with CCD cams.

[Tim]

Keep the posts coming in the thread btw, its what its for

Lets not confuse amp glow and noise though, its a different problem.

[Psychobilly]

Keep the posts coming in the thread btw, its what its for

Lets not confuse amp glow and noise though, its a different problem.

Yes its a "different" problem but is still "detrimental" to the wanted signal and could therefore be descibed as "noise" and with some DSLR's is probably one of the biggest limiting factors...

Peter....

[ngc2403]

I have not had time to read through the report but it looks like the images (graphs) have not been bias subtracted, and normalized counts is never a good way to go.

anyway i ope to have a wee read through it soon.

ally

[Darth Takahashi]

I think the best advise is as follows;

1. If you have a cooled camera go as long as you can with your sub's

2. If you are a patient man experiment as Peter suggests to find the optium settings.

3. If you are just a pratical man wanting to image start with 3 min subs and build up to 5min then 10mins

However, be warned going over 10min's with a DSLR might not bring you what you wanted.

As for the article the problem is the SPREAD of the normal distribution for the DSLR's tested. The cooled camera's maintain nice tight distributions where noise is well controlled. Obviously thermal noise is the issue and the reason that the HYPER mod's exist.

Neil.