CF-39 chroma noise

[Expeditionary]

Hi Everyone,

I'm hoping the expertise here might be able to give me some guidance. I have recently acquired a Hasselblad CF-39 and I am using it on a Fuji GX680 and a Toyo VX126 currently. I am triggering it using flash sync, which seems to be working successfully so long as the shutter speed set on the back is the same or longer than the lens shutter speed on the Fuji or LF lens, e.g. no smearing or magenta casts. However, there seems to be an excess of colour noise in the shadows. I think I am seeing that this is worse when the shutter speed on the back is much longer than on the lens. The fastest shutter speed I can set on the back is 1/8 sec though, so if I am using 1/400 or 1/250 even through to 1/30 it is certainly very noticeable. When using longer shutter speeds it is easier to match them closely and the problem is less pronounced, though there is more colour noise present than I am used to seeing at base ISO with 35mm digital. Is this normal? My research suggested this sensor is the same as in the P45+ and although it is not implemented for long exposures in the same way, I still thought that the noise floor would be quite low. But it does seem like if the back is exposing without light then it produces a lot of noise.

See attached image for a example. It is at 200%, which is excessive but the easiest way to make it obvious what I am talking about. It is still clearly visible at 100% but I feared web compression might change that. At 50% viewing the issue is not as apparent, but by that point I am at lower resolution that 24MP 35mm digital, which kind of undermines the point of having a 39MP back with no AA filter.

I have also attached a second image showing what I assume are dead pixels, usually appearing as green or black spots. There seem to be a LOT of these.

Comments/suggestions would be appreciated.

Tom

[torger]

You should be able to set the back so it syncs automatically with the shutter, which works well down to say 1/125 below that there's often sync issues (magenta cast and other strange things).

The manual and user interface is very confusing on this setting, and unfortunately I don't remember now how you configure it, but you can. I'd try that first. Do watch out for color shifts, like magenta etc, because if you see that you have some sync or light leak issue and the camera is not working in the way you think it is. A couple of times I've run into that in the field when I had mounted the view camera bellows wrong so there was a tiny light leak...

If you want to maximize image quality you should be using Phocus which has finely tuned noise reduction, and hot pixel removal. I don't recommend using Lightroom. If you don't like the Phocus limited post-processing features you can just make a basic conversion to TIFF there and then process that in some other software, many Hassy users do that.

If you look at the images in their raw form without hot pixel reduction and without noise reduction there will be a fair bit of both, they're certainly not as clean as a modern CMOS. On longer exposures you will have an extreme amount of hot pixels, it's peppered with it, and it has always been the case with CCDs, but the native raw converters are good at removing them.

Hasselblad's Kodak CCD backs are more noisy than Phase One's with the same sensors, but they have on the other hand usability advantages such as the shutter sync via flash (no wakeup needed) and you don't need to wait for a black frame on longer exposures.

[Chris Gahran]

Tom,

You say "The fastest shutter speed I can set on the back is 1/8 sec..." and that's correct. When using shutter speeds faster than 1/15 sec (faster being 1/30, 1/60, 1/125 etc) set the back to the 1/8 setting. Timing issues which result in a cyan cast or smearing of the image occur when slower shutter speeds are used and the back is not set to a longer speed than the shutter. That is, if you set the lens to 1/4 sec set the back to 1/2 sec or 1 sec but a longer time than the shutter speed used.

Use Phocus for processing your images. There is built-in noise filtering in Phocus and a "Noise Filter" adjustment slider. My backs (CF-22MS) delivered much cleaner images after I moved from FlexColor to Phocus.

If you have a friendly Hasselblad dealer you might ask about having your back's CCD re-profiled. These backs cannot be repaired but you can get the IR filter replaced and if you ask very very nicely it is sometimes possible to get a new profile installed into the back for the CCD when the IR filter is replaced. A new profile would eliminate dead pixels (black spots) and pixel bursts which are very bright pixels usually green or red. Those pixel bursts look like a LED light in the image and your back does not show any of those pesky pixels.

Chris

[Expeditionary]

Thank you both for the helpful replies. I had done a brief comparison between LR (which I am accustomed to) and Phocus. Phocus didn't seem to give a cleaner image in terms of noise or dead pixels, but I will do a more careful comparison. I can't seem to find an option in Phocus for the hot pixel removal, is it fully automatic? I admit I have not read the manual for that program yet, so I suppose I should before asking questions about it here!

There may be a hint of magenta in the test images I have taken, I was shooting just before dusk so thought this was a function of light quality and auto white balance, but I will try and do another careful test with a custom white balance set and that will tell me if it is in fact sync issue causing the loss in IQ.

Unfortunately I am in New Zealand so dealers and servicing are all very distant propositions. Thanks again for the help.

Tom

[epines]

Your shadows should be much cleaner than this. I think you are experiencing a sync issue, which can happen with Hasselblad backs at some of the faster shutter speeds. I think you need to do an easy test to find out:

First, get yourself a cable release that goes directly into the back's input port, so you can trigger the back separately from (and prior to) the lens.

Shoot an exposure like you've been doing. The kind that produces the noisy shadows. At 1/400, etc.

Next, set the back to, say, 2 seconds. Trigger the back first, then, while it's exposing, trigger the lens at the same shutter speed as the prior exposure. Now compare the two images. I'm betting the second one will be cleaner. At top shutter speeds, sometimes Hasselblad backs need a sort of wake-up.

If this is the case with yours, contact Kapture Group. They make a cable release that solves the issue.

[Expeditionary]

Thanks very much Ethan, this was exactly what I suspected and to hear confirmation that this might be the problem from someone else is reassuring! I will try the test you suggest.

PS-the Solarcity images on your website made me laugh out loud. I love the concept.

Tom

[NickT]

Unfortunately I am in New Zealand so dealers and servicing are all very distant propositions. Thanks again for the help.

Tom

CR Kennedy (Greg) is your dealer here in NZ, Damon Rulach at CRK Melbourne is also very helpful.

I agree with others that the level of noise you are seeing is not right.

[Expeditionary]

Thanks Nick, I will contact Greg if I get stuck.

[Chris Gahran]

Tom,

Consider learning and using Phocus.

MFD back makers buy third party CCDs (and recently CMOS) then use electronics and software to deliver image files that confirm their reputation of producing superior images. Realize that Hasselblad has spent millions of dollars writing Phocus software to make their images the best they can squeeze out of a shared third party CCD. IMHO, you should not use generic imaging software if you use MFD backs.

Chris

[Expeditionary]

Thanks Chris,

Having spent a bit more time with Phocus now I can see that it does a better job with the RAW files. I am coming up with a workflow that works for me and learning all the Phocus hotkeys etc! Thanks for pushing me in this direction.

I am working on the sync issues with the GX680 and have the back working properly with my 4x5 camera. I have also been able to borrow a Mamiya AFDII to test it on and it exposes and syncs perfectly with that, as the i-adapter I am using is for Mamiya 645. However, I am having issues with focusing. I can't see any shims installed anywhere, but I can't get infinity focus on the AFDII and this is with MF lenses at their hard infinity stop. They still front focus and I only get something like infinity focus arriving at F16 through depth of field. This is on all the lenses I could test, which are 24mm, 50mm shift, 80mm, 120mm and 200mm. I find it really hard to believe that the i-adapter doesn't allow infinity ficus, but I can't see anything in the system which is causing this except for the sensor simply being too far back. I also have trouble with getting focus right with my 4x5 adapter and although the focus screen is very bad on that adapter (the sliding style) that could be caused by the i-adapter also. I adjusted focus on the GX680 focus adjustment screws for its focus screen, so less problem there, but again I had to do a lot of adjustment, so again it could be the i-adapter.

I was prepared to accept focusing issues and bad tolerances when using adapters on the 4x5 and GX680, but when using just the expensive and proprietary i-adapter onto a Mamiya AFDII I am really surprised to have problems. Am I being dumb? Very frustrating to have a cross-platform back and three systems to use it on and still unable to get it working satisfactorily on any system! Help!

[Expeditionary]

I'm happy to report, I was being dumb. On making a third inspection of the back without the i-adapter attached I was just able to see at the corner that there was in fact evidence of a shim just curling up. With a bit or encouragement it came away and with the i-adapter back on I now have infinity focus on the AFDII, what a relief! Thanks for the patience of people reading this thread.

Amazing that such a thin piece of metal can make such a big difference, now I really am worried about the tolerances in those other adapters...

Ok, back to fixing this sync issue with the GX680.

Thanks again for hearing me out.

Tom

[Chris Gahran]

Tom,

CF backs use shims placed between the face of the digital back and the i-Adapter. The iAdapter is attached the CF to allow mating the back to a specific camera body. A 50 micron shim was provided on every new CF back that was sold. Hasselblad at one time sold a kit (Shim kit #75020330) that had three shims of different thickness to allow shifting the focus plane to match a particular camera body.

By the way, if your CF back is used on one camera body you can possibly adjust the camera's focus screen to attain optimal viewfinder focusing.

Chris

[NickT]

On the subject of shims I have in the past successfully shimmed backs by using thin plastic (of the sort you would find covering a bound document).

Tom you are welcome to drop by the studio (assuming you are in Auckland) for a primer on Phocus.
Nick-T