IQ 100 for B&W

[Pemihan]

Not that I'm in a position right now to get an IQ 100, but nevertheless (in case I should win the lottery) I'm very interested in learning how it behaves when it comes to B&W conversions especially when using tech cam with Rodie wides. My biggest issues when doing B&W with the CCD files are issues like tiling, micro lens ripples and noise. As far as I understand it the CMOS backs doesn't suffer from tiling but still have issues with weird lines different from micro lens ripples which is greatly exacerbated when doing B&W and there might be other issues I don't know about.
My main thing is landscapes and the issues I have with CCD files are predominantly visible in the sky.

Real time experience comments and images are most welcome. Would also appreciate any RAW files + LCC I can play with.

Thanks
Peter

 

[Paul2660]

Hi Peter,

I don't think you will see any tiling. The straight hard micro lens ripple is also not present. The 100 and 50 however due have a ripple issue, much less visible (only shows on a solid like blue sky). The latest LCC correction in C1 vr 10.x seems to do a very good job on this now. Enough that I don't see it anymore.

Noise, that's a bit more difficult. You can push the base ISO of 50 and ISO 100 quite a bit almost 3 full stops. ISO 200 and up are a lot less pushable (if that's a word) IMO. If exposed within limits of a push of 1.2 to 1.5, the upper ISO's also seem to do very well.

For tech camera work, lenses that prefer a CF still need to use on on the CMOS backs due to the extra exposure fall off.

Paul Caldwell

 

[dchew]

Peter,
I'll search for some images that might be relevant and send you a link. The only Rodi wide I have is the 40. Some may be the 60xl since I don't shoot the 40 all that much, but I'll look!

In the mean time I'll gather some thoughts too.

Dave

 

[Pemihan]

Thanks Paul, I have one file from IQ 100 with blue sky and the ripple issue is still present with C1 10 although to a much much lesser degree than in C1 9. Regarding tiling as far as I know the CMOS chips only have one read out channel thus eliminating tiling.

Peter

Hi Peter,

I don't think you will see any tiling. The straight hard micro lens ripple is also not present. The 100 and 50 however due have a ripple issue, much less visible (only shows on a solid like blue sky). The latest LCC correction in C1 vr 10.x seems to do a very good job on this now. Enough that I don't see it anymore.

Noise, that's a bit more difficult. You can push the base ISO of 50 and ISO 100 quite a bit almost 3 full stops. ISO 200 and up are a lot less pushable (if that's a word) IMO. If exposed within limits of a push of 1.2 to 1.5, the upper ISO's also seem to do very well.

For tech camera work, lenses that prefer a CF still need to use on on the CMOS backs due to the extra exposure fall off.

Paul Caldwell

 

[Pemihan]

That would be great Dave, I think I got one from you before with extreme shift. Maybe I should point out that I don't shift that much, seldom more that 10mm left/right/rise/fall,
I have both thr 40HR and the 60XL so both are relevant.

Peter

Peter,
I'll search for some images that might be relevant and send you a link. The only Rodi wide I have is the 40. Some may be the 60xl since I don't shoot the 40 all that much, but I'll look!

In the mean time I'll gather some thoughts too.

Dave

 

[voidshatter]

Thanks Paul, I have one file from IQ 100 with blue sky and the ripple issue is still present with C1 10 although to a much much lesser degree than in C1 9. Regarding tiling as far as I know the CMOS chips only have one read out channel thus eliminating tiling.

Peter

There's one possible tiling in the horizontal direction (along the longer edge) if a lens is heavily shifted vertically (along the shorter edge). This is hardly visible under normal usage and a lot less problematic than the CCDs.

There're also the vertical banding artefact there when you shift the lens along the shorter edge of the sensor and capture the sky. This is the most likely reason why Digital Transitions chose not to publish the RAW files of the vertically shifted 32HR, 40HR, 50HR etc for their outdoor tests where you'd be able to see the artefact in the sky. I read from the others that this can be fixed with Topaz noise reduction.

 

[Paul2660]

The LCC for the IQ100 in C1 Ver 10, IMO takes 90 to 95% of the vertical banding out, unlike the previous versions where nothing was done. And as VS mentioned, the de-banding feature of Topaz Denoise will take out the rest, however remember to create a duplicate layer when you do it as Topaz will not leave the results as a separate layer.

Paul Caldwell

 

[dougpeterson]

There're also the vertical banding artefact there when you shift the lens along the shorter edge of the sensor and capture the sky. This is the most likely reason why Digital Transitions chose not to publish the RAW files of the vertically shifted 32HR, 40HR, 50HR etc for their outdoor tests where you'd be able to see the artefact in the sky. I read from the others that this can be fixed with Topaz noise reduction.

Please, stop spouting conspiracy theories.

We didn't "chose not to publish" severe rise-tests from the test "X41_IQ3 100mp Tech Cam Testing (Shift)" - we didn't capture any. It was a boring overcast day with a gray sky so we decided to focus on shift (hint: it was in the name of the test folder). Test X42 shot around the same time includes 25mm rise on a 32HR, 30mm on a 50HR, and 30mm on a 90HR.

If you'd like to donate a raw file or two with significant rise into a blue sky that shows artifacts I'll be very glad to add them to our publicly available raw catalog. We're not afraid of clients finding out where the limits of usable image circle are on various lens+sensor combinations - if anything we're afraid of a client NOT finding out before they purchase; we want happy customers! That's why DT has done more to define the performance limit of technical cameras and various sensors than any other person, company, or entity; I've personally invested dozens of hours into such tests. The tech camera test we put out contained numerous examples of failures (artifacts or excessive noise) that help define where a sensor+lens+movement combination will not produce high quality.

Also, please kindly link to the IQ3 100mp Raw File Download Page or the specific tests not the direct/naked link to the Google Drive.

 

[voidshatter]

Please, stop spouting conspiracy theories.

We didn't "chose not to publish" severe rise-tests from the test "X41_IQ3 100mp Tech Cam Testing (Shift)" - we didn't capture any. It was a boring overcast day with a gray sky so we decided to focus on shift (hint: it was in the name of the test folder). Test X42 shot around the same time includes 25mm rise on a 32HR, 30mm on a 50HR, and 30mm on a 90HR.

If you'd like to donate a raw file or two with significant rise into a blue sky that shows artifacts I'll be very glad to add them to our publicly available raw catalog. We're not afraid of clients finding out where the limits of usable image circle are on various lens+sensor combinations - if anything we're afraid of a client NOT finding out before they purchase; we want happy customers! That's why DT has done more to define the performance limit of technical cameras and various sensors than any other person, company, or entity; I've personally invested dozens of hours into such tests. The tech camera test we put out contained numerous examples of failures (artifacts or excessive noise) that help define where a sensor+lens+movement combination will not produce high quality.

Also, please kindly link to the IQ3 100mp Raw File Download Page or the specific tests not the direct/naked link to the Google Drive.

Thank you for your reply.

It is a bit odd that you captured such test shots with some of the less-sold lenses e.g. 23HR, 35XL, but selectively did not capture such test shots with the revenue generating (flagship) lenses e.g. 32HR, 40HR etc. Maybe it's just coincidence!

As for Test X42 it was indoors where the artefact would not be shown. For Morgan Library the IQ250 was not a flagship anyway.

Anyway, thanks for your tests!

 

[Pemihan]

What I'm interested in is real time experience - what do people experience in real time usage especially when doing B&W conversions. I'm not particular interested in tests pushing the files to it's far limits thus breaking them apart. Doing that will show issues on ALL cameras.

Also please do not use/pollute this thread to create arguments about what dealers may or may not have done, that is NOT the purpose here!

Peter

 

[voidshatter]

You should kindly ask your dealer to provide such test shots to you then, given that DT is unlikely to provide such.Or you could download DT's 35XL rise outdoors and play with it. I have provided you the worst case scenario so you know what to look for.

Such test shots reflect problems and affect potential sell/resell values of gear and pride of ownership so I doubt anyone would be generous to provide. (See what happened in the IQ3100 long exposure issue thread)

 

[Pemihan]

The worst case scenario is nice to know in order to know what to look for, I appreciate that as long as the thread doesn't turn into theoretical discussions.

You should kindly ask your dealer to provide such test shots to you then, given that DT is unlikely to provide such. I have provided you the worst case scenario so you know what to look for.

 

[dchew]

Peter,
A few thoughts. I too came from a 180 to the 3100. I did a test to demo the 3100 and compared it directly in the same scene to the 180. That post is floating around here somewhere (?). From what I remember, the LCC on the 60xl looked very different between the two backs. Neither was worse, just different. Note everything in my experience is limited to 18mm shift or rise/fall.

I can dig up those files if you like. That may have been what I sent you before. It was a lake with blue sky. If you don't have that one let me know; I will send it along with the others I just sent you. That was one of the original dealer demo units.

After 6 months I can't think of a photo that had issues other than one shifted image from the 40hr of a white house with blue sky. It had a few lines in the sky. I rarely shift the 40 that much, and the few lines were easily cleaned up. I bet the 180 would have been either the same or worse.

So tiling is better, color cast is either the same or better. I have not had any issues with the 60xl, which is frankly surprising because the LCC has significant cast before conversion.

For some reason I have not done a lot of B&W conversions lately, but feel free to experiment with the files I sent you.

Dave

 

[Pemihan]

Thanks Dave,

I don't have those files, would be great to be able to compare them with my normal B&W processing.
Got the other files, thanks so much.

Peter

Peter,
A few thoughts. I too came from a 180 to the 3100. I did a test to demo the 3100 and compared it directly in the same scene to the 180. That post is floating around here somewhere (?). From what I remember, the LCC on the 60xl looked very different between the two backs. Neither was worse, just different. Note everything in my experience is limited to 18mm shift or rise/fall.

I can dig up those files if you like. That may have been what I sent you before. It was a lake with blue sky. If you don't have that one let me know; I will send it along with the others I just sent you. That was one of the original dealer demo units.

 

[drewharty]

Not that I'm in a position right now to get an IQ 100, but nevertheless (in case I should win the lottery) I'm very interested in learning how it behaves when it comes to B&W conversions especially when using tech cam with Rodie wides. My biggest issues when doing B&W with the CCD files are issues like tiling, micro lens ripples and noise. As far as I understand it the CMOS backs doesn't suffer from tiling but still have issues with weird lines different from micro lens ripples which is greatly exacerbated when doing B&W and there might be other issues I don't know about.
My main thing is landscapes and the issues I have with CCD files are predominantly visible in the sky.

Hello Peter,

I work with a Credo 50 doing B&W Prints. In the past I worked with an Aptus 65 and then Aptus 56 CCD backs. I make 40x60 BW prints using matte papers and Cone K7 inks. My images have large areas of smooth middle and light grey values, like your skies, and most are stitched together from multiple exposures. (www.drewharty.com). I shoot with Rodenstock digitar lens (only one HR lens) in lengths from 45mm to 135mm. Short of some initial issues with the Aptus 65, I have never had problems with any types of micro banding of values or micro lens ripples in files. The digitar lens have much large image circle sizes than the newer wide Rrodenstocks, so I am rarely shifting to the limit of the lens. I do tend to title the front lens significantly at times.

The Credo 50 noise performance is much better than the Aptus CCD backs but, from my experience printing B&W, not as great as many claim. With the Aptus 65 backs, I could not use anything above the base ISO without getting significant noise in smooth grey values--particularly in zones 4 - 6. With the Credo 50, I can work up to 400 ISO. I try not to work above 200 ISO, but 400 ISO is acceptable. The Credo 50 is also much cleaner with 5s and longer exposures.

I also fine working with a live back to offer significant advantages in focusing and composing images.

Hopes the helps, and I very nice images on your site.

Drew Harty

 

[Pemihan]

Hi Drew,

Thank you for your kind words and thanks you for your input. I have been looking into Piezography my self but haven't yet gone there. From what I can read it sounds great!

I mostly shoot at base ISO, but if I at some point upgrade to a CMOS back who knows, it might change. I enjoy shooting around dusk so it could certainly be useful.

I took look at your website, great work!

Peter