How different are the P45+ and P30+ as regards long exposures and noise levels?

[GrahamWelland]

The back will do it automatically. It means that when shooting long exposures that you need to figure that each shot will take 2x as long, once for the original scene capture and then the same amount of time again while you patiently wait for your MFDB to do the dark frame and processing. Obviously when it's doing this you can't shoot another image until it has finished.

Some camera systems allow you to override this and not capture the dark frame and do noise reduction later in post processing. Phase backs, at least the ones I have, don't allow you to do this.

 

[sonar]

Thanks Graham.

Is there a difference in quality doing it in the back vs. doing it in post? Presumably the back is applying algorithms that clone the data around 'noisy' pixels the same as the computer software would do, so I am wondering which would do a better job, if there even is a difference.

I shoot a lot of [upwards of] 10-minute shots in the daytime using a 10-stop ND filter, using various medium format film stocks. I have noticed that even the drum scans that come back from the lab have quite a lot of grain. I am hoping that getting into a digital back will yield more impressive results. Does anyone here have any advice/experience in this particular issue with shooting long exposure on film vs. digital? Perhaps this should be a new topic of discussion, but I figured I would ask!

 

[sonar]

Here's a 30 second, ISO 50, P45+ capture:

Jack, this image is beautiful, thanks for sharing. Where is this location?

 

[Shashin]

Sonar, are you compensating for reciprocity law failure with your film? If not, you are underexposing and contrast increases by the loss of shadow density. This could give you more grain.

Digital does not suffer from reciprocity both with the exposure time nor change in contrast. But noise will build up. Some things you can do when you shoot, use ISO effectively and dark frame subtraction, some can be done in post, stacking several shots to reduce random noise. Since digital is more sensitive to IR, the quality of the ND filter is important for long exposure daylight work--if it does not cut IR, you will end up with a color headache.

 

[sonar]

I generally expose about 20% longer than what the light meter tells me at a 10-stop increase, but I don't have a formula for compensating for reciprocity. Is there a general rule of thumb I should know about?

Interesting what you say about IR showing up in the digital. This could explain something that has been perplexing me for some time. When I shoot long exposure in the daytime with a 5DMK2 I often get this distinct line in the lower 1/3 of the frame where the color values are significantly different. I can't explain it, it doesn't seem to matter what brand ND filters I am using (Lee, B+W, etc). I never considered that it could be infrared so I've not tried using a UV filter to see if that corrects the issue. I figured it was a limitation with the Canon itself. I have two 5D's and the issue occurs with both. I have not noticed this with film yet, but I have noticed that with slide film I get a strange redness at times over the entire image- which could be the same issue just expressed differently in the film.

Here are two of the most apparent examples I got with the 5D earlier this year. The first image I stacked many ND filters on the lens to experiment bringing out this effect, and the second image just had one ND.

Here is an image shot on film where the same phenomenon might be occurring, but is not as apparent (notice the lower rock cropping has red in it):

Do you think the use of a UV filter would correct this issue?

 

[dougpeterson]

Hi Sonar,

I wanted to pitch in one WHY a dark frame is done.

When dealing with mathematical reduction of noise in an image (or music, or any other electronic recording) the type of noise is just as important as the amount. Specifically, if the software knows what the type of noise is in an image it can much more effectively reduce the noise in that image without significant loss of detail.

For instance software has a much harder time removing "white noise" from an audio recording than removing a single constant tone (e.g. a buzzing noise).

The "dark frame" is the digital back taking a measurement of both the amount, and the type, of noise being caused by the long exposure. By taking a 1 minute image and then a 1 minute blackframe the software is given a lot of additional information to work with and can do a very good job of providing you a final clean image.

Notably for a Phase One digital back this extra information is only read by Capture One (due to extensive cooperation, co-development, and mutual interests of the software and hardware teams at Phase One).

There remains no digital back in the world in the same league as a P45+ with current firmware for anything in the many minutes and above category.

If you'd like to work with a Phase One partner with extensive long exposure experience we'd be honored to help; we frequently work with photographers on evaluation rentals which can be counted towards purchase .

Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870 *| *Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off

 

[GrahamWelland]

Thanks Graham.

Is there a difference in quality doing it in the back vs. doing it in post? Presumably the back is applying algorithms that clone the data around 'noisy' pixels the same as the computer software would do, so I am wondering which would do a better job, if there even is a difference.

This is actually a very interesting question!

I think that it partially depends on what you are shooting as to whether the in-back dark frame subtraction and noise reduction is 'better' than post processing techniques such as the image stacking or noise analysis & subtraction.

I'm no expert in this area but I do know that there are two types of noise that need to handled with long exposures - pattern noise and random noise.

Pattern noise is best handled by dark frame subtraction, either in camera or it can be handled in specialized image processing software applications.

Random noise can be tackled by shooting multiple images and then averaging them either in a tool like photoshop or better through a specialized image stacking application. Another technique is to shoot multiple short images and combine these in a stack which will in effect sum together the exposures - the benefit for noise reduction is that each shorter image may have less noise than the equivalent longer exposure image and so you'll combine these into a cleaner image.

In in an ideal world, you'd get the very best results from a combination of dark frame subtraction & image stacking in one of the astrophotography applications since these can utilize algorithms that are more sophisticated than the in-camera processing of the raw file. However, at some point you have to be practical and that means with a MFDB like the Phase One P series, you are going to have the dark frame subtraction done in camera for you anyway. You don't have a choice. Then you have to decide whether you want to go through the hassle of processing a boat load of individual images to produce the equivalent long exposure. If you only have the ability to shoot at most up to 15-30s this may be your only option (other than film).

I bought a P25+ specifically for long exposure work because it works well for my kind of shots of several minutes. I'd rather stack images from my IQ160 but the available software options to do this were somewhat limited and the ones I looked at wouldn't handle 60mp 16bit files well at all. If I'd hit the lottery then I'd have bought a P45+ because that's the king of long exposures with the higher resolution and longer image handling. If you ever want to use a technical camera you might want to avoid the P30+.

One down side of using the P+ backs though is that you don't have the option to turn off the dark frame subtraction - not a problem on a single shot so much but if you want to shoot multiple images of a scene it would be nicer to be able to shoot ONE dark frame image and then multiple shots later at the same exposure settings. If you're shooting a 10 minute image then that saves you 10 minutes per exposure. Oh well, it is what it is.

p.s. I'd take Doug up on his offer to have a chat ...

 

[sonar]

If you ever want to use a technical camera you might want to avoid the P30+...

Graham, curious to know what you mean by this. I am seriously considering buying one of these right now, for both fashion work (which pays the bills) and long exposure work (which serves my soul).

Is the P30+ the only back that doesn't have the option to turn off dark frame subtraction?

 

[GrahamWelland]

The micro lenses on the P30+ make it unsuitable for use with technical cameras (Arca, Alpa, Cambo etc) where you may have lens movements. You'll get horrible lens brightness and color casts on the this back.

If you intend to shoot using a DSLR body you'll have no problems at all. It's not a bad back in any way, it's just not suited to technical cameras with movements. For everything else it has a great reputation.

You might find this helpful: http://forum.getdpi.com/forum/showthread.php?t=15269

 

[dougpeterson]

I generally expose about 20% longer than what the light meter tells me at a 10-stop increase, but I don't have a formula for compensating for reciprocity. Is there a general rule of thumb I should know about?

Interesting what you say about IR showing up in the digital. This could explain something that has been perplexing me for some time. When I shoot long exposure in the daytime with a 5DMK2 I often get this distinct line in the lower 1/3 of the frame where the color values are significantly different. I can't explain it, it doesn't seem to matter what brand ND filters I am using (Lee, B+W, etc). I never considered that it could be infrared so I've not tried using a UV filter to see if that corrects the issue. I figured it was a limitation with the Canon itself. I have two 5D's and the issue occurs with both. I have not noticed this with film yet, but I have noticed that with slide film I get a strange redness at times over the entire image- which could be the same issue just expressed differently in the film.

Here are two of the most apparent examples I got with the 5D earlier this year. The first image I stacked many ND filters on the lens to experiment bringing out this effect, and the second image just had one ND.

Here is an image shot on film where the same phenomenon might be occurring, but is not as apparent (notice the lower rock cropping has red in it):

Do you think the use of a UV filter would correct this issue?

That looks like viewfinder light leak. A very small amount of light leaks through the viewfinder and will cause issues during extremely long exposures, especially when there is a bright source of light behind the camera (e.g. sunlight). Just place a piece of black tape over the viewfinder.

P.S. I do not claim expertise on film. However you can find emulsion-specific reciprocity failure (which includes both color and density shifts) on the film box itself, or, failing that, you can google "reciprocity [name of film]". It varies a lot between types of film (e-6, color neg, black and white) and a little between specific emulsions of that type of film.

Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870 *| *Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off

 

[jlm]

any reason one couldn't create a catalog of dark frame shots of different exposure times and simply use them much as a LCC correction during processing? almost the same question as graham's

 

[sonar]

That looks like viewfinder light leak. A very small amount of light leaks through the viewfinder and will cause issues during extremely long exposures, especially when there is a bright source of light behind the camera (e.g. sunlight). Just place a piece of black tape over the viewfinder.

Doug, do you mean in the digital shots or the film shot? That's a very good theory that I haven't considered. From now on I'll bring a little piece of gaff with me when I'm shooting these, just in case.