Large Format lens with HasseCVF39 high Luminance Noise ISO 50

[neil]

I have been using a HasseCVF39 digital back with 200 series and 500 series cameras for the last few months and have obtained some exceptional results. Definitely the CVF39 back has brought the V series into the digital age.

This month I started to use the digital back on a large format camera. All seemed to be going well but occasionally I got a dud image. After a bit of experimentation I managed to repeat the set of circumstances that produced the dud image. I am wondering if anyone on this forum has come across a similar problem. Would appreciate some advice on how I can narrow down or solve this problem. It could be user error ( am I doing something wrong ? ) could be hardware error or even a software/firmware problem?

The problem only occurs when photographing a very high dynamic range scene.
Using large format lens connected to the back with a sync cable.
The highlights are blown by 1/3 of a stop

http://www.j4p.com/overview.jpg
This is with a fuji 125mm LF lens 1/60sec F8 Iso 50
The digital back is set to use Flash Sync

If we look at the detail
http://www.j4p.com/fuji_125_fsync_zoom.jpg
There is a slight magenta cast to the image. Also very high luminance noise ( even at ISO 50 !! )

In comparison if we look at the same scene photographed with a 200 series camera with a 100CF lens at 1/60sec F8 Iso 50
The digital back is set to use 200 Series Sync
http://www.j4p.com/Has_cf100_200series_zoom.jpg
The image has no observable luminance noise and no colour cast.

The complete files are here for comparison
http://www.j4p.com/fuji_125_fsync_0165.jpg
http://www.j4p.com/Has_cf100_200series_0177.jpg

I have tried with several large format lens and after some time I can produce the same dud image with each lens.

The factors to make this happen seem to be:-
Very high dynamic range scene, large areas of sunshine and shadows in the scene.
Use of a sync cable between lens and digital back

I looked in the Hasselblad manual and it says there is a shutter delay menu setting which is a corrective feature which adjusts the time a little to ensure the image is read from the sensor. There is a picture of the menu and setting in the manual. Unfortunately on the actual digital back the setting is not there in the menu system. ( firmware – 256 – the latest version)

Any advice on how I can narrow down or solve this issue appreciated.

 

[Stuart Richardson]

Are you sure that the back is only exposing for a short period of time? My guess would be that while your lens is only doing a short exposure (1/250th etc), your back is firing for 1 second or some other longer time. This would lead to proper exposure in the image, but increased noise in the actual file due to longer sensor activity. I know that on my sinar back you could set the exposure time of the back itself for when you were making images without a dedicated connection between the back and a body made to be used with it.

 

[neil]

Stuart, thanks for the advice. I just took a look at the back. It is set for exposure time < 1/8 sec. The manual seems to say this is the correct setting. It says if I go lower for example 1/4 then I have to set 1/4 on the back and 1/4 on the lens. But for settings above 1/8 it should be good....

This is such an odd problem. If I point the camera away from the bright sunshine and take a picture for example all in shade. The picture is correct and shows no noise.

 

[Jack]

Exposure: LF lens shutters are mechanical, and just because they're marked 1/60th, doesn't mean they actually are firing at 1/60th -- and in fact, most will start to run slow as they age. Most every serious LF shooter purchased a shutter speed tester for the very reason to know what the actual shutter speeds of their lenses were -- most of us used a piece of tape on each with the corrections for that lens. So here, being off by 1/3 stop is not surprising, and actually expected... (Apertures can vary too, so your marked f8 could easily be off by 1/6th stop or more. Combine a shutter that is 1/6th off with an aperture that is 1/6th off, and you have a 1/3rd stop error which of course is enough to blow the highlights on a DB.) Thus, my advice for tech camera shooters is to do a planned -1/3 or -1/2 exposure comp since most DB's will easily handle a 1/3 to 2/3 push in RAW without any serious negative effects, especially when shot as base ISO to begin with. And of course let's not forget you have the histo on your back to immediately confirm exposure

Magenta cast: Purple fringing along high-contrast edges and in specular highlights is fairly common with digital, and especially Kodak sensors -- I would suspect that Hassy's converter has a setting to attenuate it? (I know C1 does for Phase backs.)

Magenta shift: Can be caused, again especially with Kodak sensors, when the lens is off-axis at all, as in a shift or tilt -- another problem with using digi backs on view or tech cameras, and the reason for the LCC or "white frame" corrections.

Magenta 3 edit: I just realized that older Fuji LF lenses had coatings that ran toward yellow or magenta depending on the lens and focal. This was the present in the series prior to the ones marked "EBC". If your 125 is pre EBC, then that could explain the magenta. It also probably would explain the slower shutter -- many older Fuji LF lenses used Seiko shutters which were not as consistent as Copal or Compur shutters from the same period.

Noise: I have to agree with Stuart, in that it sounds like sensor "on" time (heat) is excessive, as it shouldn't happen with a less than a 1 second exposure at ISO 50 on any DB. The fact you don't see it when you point away from the Sun is puzzling for sure...

 

[neil]

Jack, Thanks for the advice. Your right on with the explanation of Exposure and Magenta cast. For the exposure I will be not so trusting of the LF lens settings. Some of the lens are quite old and never had any maintenance. For the purple fringing will deal with it using software. Seems to be a given using MF sensors.

The luminance noise at iso 50 is worrying as this seems to be unusual and should not be happening. I had a chance today to do some more testing to see if I could reproduce the issue again. I simplified the setup for easy comparison. Am trying to discover if my digital back has a problem or User error is occurring. Any advice on how I might narrow this down appreciated.

If there are any Hasse CVF39 users reading this could you give it a try using the sync cable and high dynamic range scenes to see if you can reproduce the same noise issue.

I used a 553ELX camera with a cf180 lens set to 1/60 at F8, ISO 50. The first 4 shots are taken with a sync cable attached between lens and digital back. The digital back is set to flash sync in the menu. The 5th shot is taken with no sync cable and the digital back is set via the menu to sync internally with the 553ELX camera. All shots use the shutter in the CF180 lens. A series of pictures was taken which include and then exclude a strong band of highlight. The pictures are taken within a few minutes of each other.

The picture below is an overview of the scene. It is a picture of a brick wall with strong sun light hitting the scene from the right side. On the left hand side there is a band of strong highlight running down the side of the next building.
http://www.j4p.com/overview2.jpg

1.In the first picture the camera is pointed slightly to the right and does not include the bright band of strong highlight. A sync cable is attached between lens and camera back.
http://www.j4p.com/01shade12.34PM_flashsync_0692_zoom.jpg
If we look at the detail. The image has no observable luminance noise. ( Look at the surface texture on the gas boiler. Quite amazing that you can read the letters on something the size of a small book which is 30 meters a way)

2.In the second picture the camera is move back to the left and the bright band of strong highlight is included in the picture. A sync cable is attached between lens and camera back.
http://www.j4p.com/02brightline12.37PM_flashsync_0695_zoom.jpg
The image has high luminance noise

3.In the third picture the camera is again move to the right to not include the bright band of strong highlight. A sync cable is attached between lens and camera back.
http://www.j4p.com/03shade12.38PM_flashsync_0696_zoom.jpg
The image has no observable luminance noise.

4.In the fourth picture the camera is moved back and the highlight is again included in the scene. A sync cable is attached between lens and camera back.
http://www.j4p.com/04brightline12.39PM_flashsync_0698_zoom.jpg
Again the image has very high luminance noise

5.The fifth picture is exactly the same position as the fourth picture including the highlight band but the sync cable is removed from the lens and the digital back is set to sync using the internal 553elx sync.
http://www.j4p.com/05brightline12.41PM_elxsync_0699_zoom.jpg
The image has no observable luminance noise. ( surface texture on the gas boiler is smooth )

In conclusion:-
Pictures 1 and 3 have no noise. They use a sync cable and do not have strong highlights in the scene.

Pictures 2 and 4 have high lumiance noise ( iso 50 ). They use a sync cable and have strong highlights in the scene.

Picture 5. This is the most interesting as it has strong highlights in the scene the same as picture 4 but there is no lumiance noise. All that has changed between pic 4 and pic 5 is that the sync cable was removed and the camera was set to internally sync between lens and back.

The question is does this point to there being something wrong with the sync cable ability of the digital back. If so why does it only occur if there is a strong highlight band in the scene ? or is there something else going on here which I have missed.

The complete files are here for comparison

http://www.j4p.com/01shade12.34PM_flashsync_0692.jpg
http://www.j4p.com/02brightline12.37PM_flashsync_0695.jpg
http://www.j4p.com/03shade12.38PM_flashsync_0696.jpg
http://www.j4p.com/04brightline12.39PM_flashsync_0698.jpg
http://www.j4p.com/05brightline12.41PM_elxsync_0699.jpg

 

[Stuart Richardson]

Neil -- I think you might be best off talking to Hasselblad about this. To be honest, I don't see a major difference between the files...definitely 2 and 4 have more noise than the others, but it does not seem to be ruinous to file quality. But still, they are probably the only ones who can tell you what exactly is going on. I wouldn't really think that sync cable would do anything to the noise performance of an image unless it is holding the exposure open for a longer time.

 

[neil]

Stuart, Jack, thanks for the comments. This is helping me to understand what is an issue and what is not an issue. I just went and looked back over the images and noticed that the jpg files are actually smoothing out the noise in the original files. Below I have added in some links to small tif files. The numbers 01 to 05 correspond to the numbers above. Anyone that is interested can download those files and compare. I will experiment a bit more with a large format camera and get this in a form which I can report it to Hasselblad Tech support. Will first wait and see if there is any feedback from other Hasselblad users. Also in a couple of weeks time I will have a phase one back on my large format camera so I can compare with how that renders the same scenes. If they are close in performance then I can conclude its a non issue.

http://www.j4p.com/01shade12.34PM_flashsync_0692_zoom.tif

http://www.j4p.com/02brightline12.37PM_flashsync_0695_zoom.tif

http://www.j4p.com/03shade12.38PM_flashsync_0696_zoom.tif

http://www.j4p.com/04brightline12.39PM_flashsync_0698_zoom.tif

http://www.j4p.com/05brightline12.41PM_elxsync_0699_zoom.tif

 

[Stuart Richardson]

Perhaps I am a bit more forgiving than you, but I am not sure I would have even noticed the difference, let alone figured it was a problem. Are you sure this is worth the time and effort to worry about? I am not trying to be dismissive about it -- there is clearly a difference between the files, but at least from my perspective I think my time would be better spent just writing off an email to hasselblad tech support to see if they have an answer, otherwise just chalking it up to the strange vagaries of modern technology. The only other thing I could suggest would be to open up the files in a different RAW converter to see if the differences are caused by some automatic correction that is applied -- perhaps the camera or software applies different corrections based on different shooting configurations. You might try something like Raw Photo Processor which applies nearly no corrections at all by default.

 

[Jack]

I am going to stick with sensor heat build up for the noise theory. Perhaps the synch cable triggers the sensor "on" sooner than the direct connection, and thus heat (and noise) have an opportunity to build up to the slightly higher levels we see in the shadows.

 

[jotloob]

In conclusion:-
Pictures 1 and 3 have no noise. They use a sync cable and do not have strong highlights in the scene.

Pictures 2 and 4 have high lumiance noise ( iso 50 ). They use a sync cable and have strong highlights in the scene.

Picture 5. This is the most interesting as it has strong highlights in the scene the same as picture 4 but there is no lumiance noise. All that has changed between pic 4 and pic 5 is that the sync cable was removed and the camera was set to internally sync between lens and back.

The question is does this point to there being something wrong with the sync cable ability of the digital back. If so why does it only occur if there is a strong highlight band in the scene ? or is there something else going on here which I have missed.

Neil

Images 1 to 4 all have the following in common .

aaa) They are all shot using the sync cable .
Please give your exact settings you entered in the back .

bbb) All 4 images have a magenta cast , which is obviously the result of a
timely wrong syncronization .

Image 5 has obviously no magenta cast .

The CFV-39 BACK needs a kind of "wakeup" before your shot can be captured 100% properly synchronized .
The same is valid for PHASE ONE BACKS .
I have learned this from the ALPA people and could experience what they say with my CFV-16 and CFV-39 using my HASSELBLAD 905SWC and also an ALPA 12 SWA camera .
The issue is more obvious , when using short exposure times , like 1/125sec and shorter .

Just to put one thing right , I do not want to say , that your images suffer from this "wakeup" issue , but it smells very much like it .

Let me give you an example :

Setup is HASSELBLAD 905SWC + CFV-16 BACK , NO SYNC CABLE .
Exposure time is 1/125 . From what I found , Iso setting does not matter .
Back set to SWC and <1/8 sec .
The result is in 90% of the exposures , aimages with a magenta cast .

Setup is HASSELBLAD 905SWC + CFV-16 BACK , SYNC CABLE . Exposure time 1/125 , back set to FLASH SYNC and < 1/8 sec . Results are the same as above .

Setup is HASSELBLAD 905SWC + CFV-16 BACK , SYNC CABLE . Exposure
time 1/125 , back set to FLASH SYNC and <2 sec .
NO magenta cast and a proper histogram displayed .

Please note , that the correct exposure time is given by the shutter and not by the digital back . The 2 sec setting just means , that the back is "armed" for 2 sec . But the back received a kind of "wakeup" signal , before the actual shot was taken .

I experienced the very same using an ALPA 12 SWA and a CFV-39 back , operating in the very same sequence , as described above .

Please give your settings on the back for the two situation (image 1 to 4 and image 5) . Perhaps you would like to do some testing , with what I have tried to point .
When using the HASSELBLAD "internal" sync (501CM 553ELX or others which do not require the flash sync cable) the "wakeup" seems to work fine but i still experience a very slight magenta cast when using <1/125 sec shutter times .

Jürgen

 

[neil]

Jurgen,

Thanks very much, I will give this a try at the weekend. This sounds promising.

Exposure time was set to < 1/8 of a second for all the shots. For image 1 to 4 and for image 5.

Back was set to Flash sync for images 1 to 4. For image 5 the back was set to ELX.

In the menu on my back under the camera menu when you set the body. Under that I only have one other setting which is Exposure time.

If you look at the manual on page 64. The picture shows there should be another menu called Shutter Delay. This does not appear on my back menu.

This is sad because the manual says:-

"There can be differences in the amount of time required in preparation (raising the mirror, opening/closing the shutter etc) for digital capture between various camera models. This corrective feature adjusts the time a little to ensure the image is read out from the sensor. The normal setting is Default and should only be changed if problems are encountered."

Since my back does not have this option in the menu I cannot change it away from the default setting. Does your back have the Shutter Delay option in the menu under flash sync or elx or 200 or 500 or swc camera setting ?

At the weekend I am going to reshoot the scene and change the back exposure time to 2 seconds as you suggest.

I am still a bit confused about this wake up signal on Hasselblad backs. I read the manual for phase one backs so can understand the setup for a phase one back. For Hasselblad I cannot understand how the back gets a wake up signal before the actual shot is fired. Or do I need to take two shots one after the other the first being the wake up ?

Your right about the histogram in some of the images. It does not seem to be properly formed.

Neil

 

[jotloob]

In the menu on my back under the camera menu when you set the body. Under that I only have one other setting which is Exposure time.

If you look at the manual on page 64. The picture shows there should be another menu called Shutter Delay. This does not appear on my back menu.

This is sad because the manual says:-

"There can be differences in the amount of time required in preparation (raising the mirror, opening/closing the shutter etc) for digital capture between various camera models. This corrective feature adjusts the time a little to ensure the image is read out from the sensor. The normal setting is Default and should only be changed if problems are encountered."

Since my back does not have this option in the menu I cannot change it away from the default setting. Does your back have the Shutter Delay option in the menu under flash sync or elx or 200 or 500 or swc camera setting ?

At the weekend I am going to reshoot the scene and change the back exposure time to 2 seconds as you suggest.

I am still a bit confused about this wake up signal on Hasselblad backs. I read the manual for phase one backs so can understand the setup for a phase one back. For Hasselblad I cannot understand how the back gets a wake up signal before the actual shot is fired. Or do I need to take two shots one after the other the first being the wake up ?

Your right about the histogram in some of the images. It does not seem to be properly formed.

Neil

Neil

The shutter delay option is not shown on my CFV-16 menue and also not shown on my CFV-39 menue .
But I do believe , that this setting , which is at default , has got nothing to do with the issue you describe .

Let me explain a bit the term "sleeping back" and why the back must get a "wakeup" .
LEAF backs do not have this "problem" , at least that is what my dealer said .
PHASEONE and HASSELBLAD backs need a kind of wakeup signal .
They are in "sleeping mode" to safe power from the battery and to avoid a "heat up" of the sensor , which in turn would produce noise .

When you use a HASSELBLAD 500 , 501 , 503 , 203FE modified etc . , where you do not need a sync cable , the wakeup is done by the small rectangular plunger
(camera rear side , right bottom) when you press the shutter release button . That is why your exposure was correct , when the back was set to 553ELX .
Have a look to page 4 of the CFV digital data sheet , which is the connectivity chart .

When you use any HASSELBLAD SWC , the shutter release behaves different and therefore I always use the sync cable . Still you can encounter a faulty sync situation , when using exposure times of 1/125 sec or shorter .
For view cameras , using HASSELBLAD digital backs , the Flash input sync cable is always required .
But here , with lenses having most likely COPAL shutters or PRONTOR PROFESSIONAL , what ever your lenses shutter is , we encounter the problem , that the back has to wake up , before the shot is done .
Again the short exposure times are affected .

A possible solution you could try out is , like I did with my 905SWC , the actual exposure time , set at the shutter ring was a 1/125 sec . (I did not try shorter time settings) and the setting for the "exposure" at the back was 2sec . (or even 3sec.) That worked , but might vary with the shutter you use , as the times are not that precise .

http://www.kapturegroup.com/ have solutions , but I have no experience with any of their special cables . If you are interested , you should contact them . The cables are rather expensive . (approx.400$) . Tripod use only .

ALPA , which must , in this case , be regarded as a view camera , has produced their own solution , which they integrated into the cameras hand grip . The difference in price between a hand grip with or without wakeup function is also high .
I don't know what kind of electronic is inside the grip , but it works as follows . You do not need a flash sync cable from the lens to back .
The handgrip has a little integrated pushbutton and a cable goes from there to the sync plug in the CFV back .
You set the back to FLASH SYNC MODE . Set the time in the menue to 2 or 3 sec. Now , to expose , you press the little button in the grip and then within your chosen time (2 or 3 ec. for example) you release the shutter .
This solution can also be used handheld .

Please come back with your findings from the weekend .
Might be , that some ALPA users in this forum can add some more information .

Regards Jürgen

 

[neil]

Jurgen,

Thanks for the explanation and the suggestions. I got a chance today to do a couple of hours testing. It was a cloudy day so I do need to do the tests again on a bright day but my initial tests show that this does sort out the issue I was having.

First of all I tried the Alpa solution:-

I bought a Fujifilm GX645AF remote cable release. ( found one in a second hand shop - Luck was with me this week ). It is an electronic cable release with an electronic push button on the end. It plugs into the same socket that the sync cable plugs into.

Set the digital back to Flash sync mode and set the time in the menu to 3 seconds.
Do not use a sync cable instead plug the Fuji cable release into the digital back sync socket.
Push the Fuji cable release which wakes up the back.
Push the manual cable release which triggers the large format lens.
The back hits the 3 second point and writes out the image.

All is good with the image. There is no magenta cast and no extra noise in the image.

One downside of this method is that I suspect the battery in the digital back will run down a lot quicker. I also wonder about having the back active for 3 seconds may add any extra noise to the images. There must be a good reason why the back is only usually activated for 1/8 of a second for each image.

Next I tried using a 903SWC camera.

I set the back to SWC sync mode and put the time in the menu to 1/8 second. No sync cable attached. No Fuji Cable attached.

The shutter release on the SWC has a long push down to release the shutter. Also the push of the shutter is not distinct. You push several millimeters before it releases. What I found is that when I start to push after a millimeter or so the Digital back will wake up and after another millimeter or so the shutter will fire.

If you just push the shutter release a little the digital back will wake up and write out a blank image. The shutter does not fire.

Try again, If you push slowly the back wakes up keep pushing slowly the shutter fires and the digital back will write out a lovely heavy blur bright magenta image.

Try again. Give the shutter release a short sharp stab all the way down and release quickly. The digital back writes out an image and it looks ok.

Ran out of time here. I need to repeat this test with the back set to 2 seconds to see what happens.

Do you think that both these issues look like timing issues. The large format case is a timing issue at the start of the image capture ? Back does not wake up in time and shutter already opened before recording starts. The SWC case is a timing issue at the end of the image capture ? Back wakes up but shuts down before the shutter has finished firing ?

Regards


Neil