My thoughts after 48 hours with an IQ260

[Craig Stocks]

I decided to trade my P65+ for the new IQ260, which I just received from Capture Integration on Friday. I've been playing with it on my 645 DF and thought I'd share some of my observations so far.

I don't really have any experience with the IQ1 series, so many of the comments may apply to those backs as well.

I did most of the processing in Lightroom 5 since that's my normal workflow. Also, I've been unable to coax Capture One 7 into outputting a jpeg. It outputs tiffs just fine though - even when I've selected jpeg settings in the process recipe.

Overall image quality: It's hard to tell any difference between the P65+ and the IQ260 at ISO 50. The IQ260 seems to yield a slightly brighter file, so its base ISO may be just a touch higher than the P65+.

The LCD screen is a lot brighter than the P65+. I've turned it down to 20% brightness, but it's still rather bright. I'll have to recalibrate my eyes and/or pay more attention to the histogram for a while.

The WiFi seems to work OK with a 1st generation iPad, though screen refreshes are sometimes pretty slow.

Battery life seems terribly short compared the P65+, even with WiFi turned off. I only have two batteries, so I may be in the market for one or two more.

The power button is very sensitive, and a quick tap will turn the back off - even if it's doing a dark frame or still writing to the card. Obviously, you loose that frame if you accidently bump the power button. I prefer the P65+ arrangement where you have to momentarily hold the power button to turn off the back. I hope this gets addressed in a firmware upgrade.

Long exposure mode: Phase One suggest using LE mode for anything longer than 10 seconds. In my quick tests, I'm seeing some image degradation when using ISO 140 LE compared to ISO 50. My most recent test was a one minute exposure at each ISO, adjusting the aperture to compensate for the ISO. The ISO 50 shot was much crisper, but showed a fair amount of single pixel noise (though perhaps slightly less so than the P65+). The ISO 140 LE frame didn't show any single pixel noise, but was softer overall.

My usual approach with single pixel noise is to run Photoshop's Dust and Scratches filter at 1 or 2 pixels. I get the best result using a 1 pixel radius and manually healing the remaining spots. Comparing the results, to the ISO 140 LE frame, 1 pixel radius was still slightly crisper, but left a few spots, a 2 pixel radius eliminated all of the single pixel noise, but the result was slightly softer than the ISO 140 LE frame.

My conclusion for now - I'll probably stick to ISO 50 for less than 60 seconds and plan to do some extra processing to clean up the file.

I also noticed some color shift when going from ISO 50 to ISO 140 LE, especially in the shadows.

Lastly - there seems to be a bug in the communication between the camera and back. I'm using the default 60 seconds for the LCD to shut off. When using ISO 140 LE, if the back goes to sleep during the dark frame, the camera will allow you to start another exposure before the back's ready for it. The result is corrupted colors in the 1st frame, and you never get the 2nd frame. I hope that can be fixed in the back's firmware since Phase One stated that they've abandoned firmware updates for the DF body. For now, I'll need to make sure the dark frame is fully complete before starting another exposure.

 

[PeterA]

Interesting observations Craig- thanks for your thoughts so far.

 

[ondebanks]

I'm seeing some image degradation when using ISO 140 LE compared to ISO 50. My most recent test was a one minute exposure at each ISO, adjusting the aperture to compensate for the ISO. The ISO 50 shot was much crisper, but showed a fair amount of single pixel noise (though perhaps slightly less so than the P65+). The ISO 140 LE frame didn't show any single pixel noise, but was softer overall.

That's very worrying. If the only thing happening in LE mode is dark frame subtraction, there is NO WAY that there could be "image degradation" or "softer overall" image quality.

It appears that Phase One are cheating here - filtering the RAW files to eliminate single pixel noise that remains, even after dark frame subtraction.

That is a big no-no in my book - if a manufacturer correlates pixel values by filtering prior to writing out the RAW file, they have irretrievably compromised the IQ.

It would be ok only if there's a setting in the back giving the user the option to turn off that behaviour? (Allowing a straight dark frame subtraction and no filtering)

Ray

 

[kdphotography]

....
I did most of the processing in Lightroom 5 since that's my normal workflow. ....

I really think you need to use Capture One Pro 7 to process all your Phase MFDB raw files. Imho, that can make all the difference in the world.

There's probably some minor quirk on your C1Pro install or a small check box skipped---once ironed out, the current iteration of C1 Pro 7 is really good.

ken

 

[dougpeterson]

I did most of the processing in Lightroom 5 since that's my normal workflow. Also, I've been unable to coax Capture One 7 into outputting a jpeg. It outputs tiffs just fine though - even when I've selected jpeg settings in the process recipe.

Right now you are highlighting the recipe you want to use for output rather than checking the checkbox for the recipe you wish to use.

Just check the box. :chug:

If it's not that then it's some other small thing; I'd suggest a Capture One class to get you up the learning curve without hassle.

Overall image quality: It's hard to tell any difference between the P65+ and the IQ260 at ISO 50. The IQ260 seems to yield a slightly brighter file, so its base ISO may be just a touch higher than the P65+.

a) use Capture One rather than LR so you're getting the most out of the information recorded
b) Unadjusted you won't notice much difference; LR and C1 will both ignore the data at the extremities of the dynamic range (otherwise the image at default settings would look very flat considering the range of tone it captures). But shoot something with strong color and a huge dynamic range and add some shadow and highlight recovery and you'll be able to examine the differences. I've also seen some difference in color rendering, but have not done any formal study of it.

The LCD screen is a lot brighter than the P65+. I've turned it down to 20% brightness, but it's still rather bright. I'll have to recalibrate my eyes and/or pay more attention to the histogram for a while.

The histogram and user-adjustable highlight-warning are the only way to go for exposure judgement. Otherwise you'll make very different decisions on how to expose when you're facing the sun or not (or in cloudy vs bright vs nighttime shooting). Fortunately the wide-aspect screen lets you see the histogram without obscuring the overall image.

Quick tip: if you tap and hold on the histogram and bump the histogram to #3 of the 5 tools then it's always visible - if you're scrolled to the top-most tools it will be the bottom slot and if you're scrolled to the bottom-most tools it will be the top spot.

[other tool]
[other tool]
[histogram]
[other tool]
[other tool]

The WiFi seems to work OK with a 1st generation iPad, though screen refreshes are sometimes pretty slow.

Make sure bluetooth is off on the iOS device.

Also, if you're commenting on "On" vs "Adhoc" mode I'd encourage you to try both (using a high quality wireless router like an Airport Extreme with no other traffic on it).

Since you haven't given any numerical times it's hard to tell if you're getting normal results but just not impressed or if you're getting sub-par results. Compared to the speed most shooters are used to shooting tethered the direct-wireless time-to-screen and zoom-to-100% is pretty impressive (especially if they've used any other wireless solutions). Compared to the LCD on the back of the camera it's probably not.

Battery life seems terribly short compared the P65+, even with WiFi turned off. I only have two batteries, so I may be in the market for one or two more.

Make sure you're using the higher capacity 2900 mAh batteries, preferably less than a year old (they do wear down, as any battery does).

The power button is very sensitive, and a quick tap will turn the back off - even if it's doing a dark frame or still writing to the card. Obviously, you loose that frame if you accidentally bump the power button. I prefer the P65+ arrangement where you have to momentarily hold the power button to turn off the back. I hope this gets addressed in a firmware upgrade.

I agree!

Long exposure mode: Phase One suggest using LE mode for anything longer than 10 seconds. In my quick tests, I'm seeing some image degradation when using ISO 140 LE compared to ISO 50. My most recent test was a one minute exposure at each ISO, adjusting the aperture to compensate for the ISO. The ISO 50 shot was much crisper, but showed a fair amount of single pixel noise (though perhaps slightly less so than the P65+). The ISO 140 LE frame didn't show any single pixel noise, but was softer overall.
[...]
I also noticed some color shift when going from ISO 50 to ISO 140 LE, especially in the shadows.

You shouldn't make any assessments in this area until you switch to Capture One.

When you do. Make sure to play with noise reduction settings on both ISO50 and ISO140 to your taste/aesthetic. ISO50 has greater DR but gains noise more quickly. IQ140 has very good DR, but not as much as ISO50 but gains noise very very slowly (i.e. 8 min looks a lot like 1 min).

My usual approach with single pixel noise is to run Photoshop's Dust and Scratches filter at 1 or 2 pixels. I get the best result using a 1 pixel radius and manually healing the remaining spots. Comparing the results, to the ISO 140 LE frame, 1 pixel radius was still slightly crisper, but left a few spots, a 2 pixel radius eliminated all of the single pixel noise, but the result was slightly softer than the ISO 140 LE frame.

C1's dedicated single-pixel noise reduction algorithm will do MUCH better at this than your current PS workflow. You can't switch fast enough.

Lastly - there seems to be a bug in the communication between the camera and back. I'm using the default 60 seconds for the LCD to shut off. When using ISO 140 LE, if the back goes to sleep during the dark frame, the camera will allow you to start another exposure before the back's ready for it. The result is corrupted colors in the 1st frame, and you never get the 2nd frame. I hope that can be fixed in the back's firmware since Phase One stated that they've abandoned firmware updates for the DF body. For now, I'll need to make sure the dark frame is fully complete before starting another exposure.

Bummer. I haven't seen this, but haven't done any long exposure work with a 260 and a DF non plus.

 

[dougpeterson]

That's very worrying. If the only thing happening in LE mode is dark frame subtraction, there is NO WAY that there could be "image degradation" or "softer overall" image quality.

It appears that Phase One are cheating here - filtering the RAW files to eliminate single pixel noise that remains, even after dark frame subtraction.

LOL, talk about jumping the gun.

Expose the images being especially careful to avoid any ambient or user-induced vibration. Then open in C1 (not LR) and turn noise reduction off. Then talk to me about the results .

As I wrote about in my IQ260 a Sensor Story article there are some differences in the way the sensor itself gathers and stores light/electrons and I've seen a difference in the rendering of fine details (not really better or worse, just different enough to notice). But I haven't come across any black-magic or softening beyond those caused by camera shake or software. Would gladly add a more definitive test for that to the list of tests I'm behind on getting to though .

 

[Craig Stocks]

Here's a link to a screen shot from Capture One zoomed to 200%. The ISO 140 frame is on the left. I copied and applied the adjustments so the two frames are being processes the same.

http://www.craigstocksarts.com/C1-ISO-Comparison.jpg

 

[jagsiva]

Certainly a noticeable difference there. Could it be that you are loosing to diffraction at f22 vs. f12 in the ISO140 shot?

I find things starting to go a little south around f11-14 on my IQ180.

 

[mmbma]

That looks like the reason. I've been testing out mine and it certain doesn't show any IQ degradation at ISO 140 vs my older P65+. I don't have the P65+ anymore but I tried to shoot a similar shot after reading this article.

 

[dougpeterson]

+1 diffraction

F/22 will lose significant amounts of detail when viewed at 100%.

 

[ondebanks]

+1 diffraction

That's a relief!

Sorry for jumping the gun. I couldn't think of any other reason why detail would be softer when the ISO was changed. :facesmack:

Ray

 

[dougpeterson]

They don't call them "confounding variables" or "lurking variables" because they are immediately obvious!

 

[Craig Stocks]

Well, don't I feel silly. I was so focused on keeping the exposure time constant that I didn't think about diffraction. I just re-did the test keeping 1 minute exposure at f/11 by moving the light closer or further away. As Doug suggested, I don't see any deterioration in sharpness, and ISO 140 LE is clearly better on noise.

Sorry for jumping to the wrong conclusion. Thanks for helping me sort it out.

 

[ondebanks]

Here's a link to a screen shot from Capture One zoomed to 200%. The ISO 140 frame is on the left. I copied and applied the adjustments so the two frames are being processes the same.

http://www.craigstocksarts.com/C1-ISO-Comparison.jpg

I know that we've solved the detail/diffraction issue, but looking at that screengrab where you've zeroed all the noise reduction settings, something else strikes me. I wonder about those dark pixels ("pepper noise" I guess you'd call it) scattered here and there around the images? They are not at consistent positions between the two frames. Maybe Doug can explain what's going on with them.

Ray