With Z7 you will get horizontal bands in dark areas pretty easily. I have not tested Z7 ii but it is supposed to be better.
When you are in infrared zone, you lose details anyway (because all light except ir is blocked). For that reason I think Z6 and Z6II are better choices.
Of course, it you have full spectrum converted camera, and use it within visible spectrum, you still get the full benefit of extra resolution.
indusphoto wrote:
With Z7 you will get horizontal bags in dark areas pretty easily. I have not tested Z7 ii but it is supposed to be better.
When you are in infrared zone, you lose details anyway (because all light except ir is blocked). For that reason I think Z6 and Z6II are better choices.
What are "horizontal bags"? I have a Z7 IR converted to 590nm by LifePixel, and have been happy with it.
As stated above the Z7 has a tendency to produce PDAF banding when shot in IR. The workaround is to use the PDAF filter in RawTherapee when it shows up. I found it annoying enough to code up my own banding removal for DNGs that I integrated into LR instead. Anyway, it is annoying and fairly frequent.
I believe the whole PDAF banding mess was addressed in the Z6II/Z7II which should apply equally as well to IR converted bodies, but you'll probably want to double check that.
I have plenty of IR shots with plenty of per pixel sharpness from the Z7, but I agree in general you are going to be getting less optical performance from most lenses than in the visual. So the 45MP vs 24MP advantage is diminished. That said, there is certainly no advantage to the 24MP sensor, so if the price difference for the body you are going to convert is small I'd tend to go with the 45MP option. But if I could get a great deal on a 24MP body then I'd do that with no regrets as well.
Be aware you have to choose Z leases very carefully for IR use. Many have bad hot spots. Specifically the otherwise excellent 24-120/4S is not good for IR. The 24-70/4S is marginal. The 24-200 is not good. The 24-70/2.8S is poor. Hard to find a good "walk around" zoom for IR that is native Z mount. I believe in general Tamron lenses often do well for IR, so you might see check on the 28-70/2.8 options.
Here's an example of Z7 PDAF banding in IR along with how well banding removal can work (this was my own code, but the RawTherapee filter works just as well):
While these are present in even non converted pictures, in IR where sky goes dark, these are more readily visible. At 590mm there is still a bit of blue light left. Use a 720nm+ filter and you will see strong banding in dark areas.
kwilliam8 wrote:
What are "horizontal bags"? I have a Z7 IR converted to 590nm by LifePixel, and have been happy with it.
Choosing nikon lenses for IR is a struggle. My current choices are 14-30mm, 40mm f2, and 28-400mm. I have also used tamron 28-200mm with megadap adapter and it works well.
Most other lenses, as you listed, have significant issues. The 24-200mm was a big disappointment as I was hoping to use it as the walk around lens.
A feature that Z6/Z6II have which is very useful is that you can add audio notes to the images. I find it convenient to record these details, when working with different filters. This is not available on Z7/Z7ii.
kwalsh wrote:
As stated above the Z7 has a tendency to produce PDAF banding when shot in IR. The workaround is to use the PDAF filter in RawTherapee when it shows up. I found it annoying enough to code up my own banding removal for DNGs that I integrated into LR instead. Anyway, it is annoying and fairly frequent.
I believe the whole PDAF banding mess was addressed in the Z6II/Z7II which should apply equally as well to IR converted bodies, but you'll probably want to double check that.
I have plenty of IR shots with plenty of per pixel sharpness from the Z7, but I agree in general you are going to be getting less optical performance from most lenses than in the visual. So the 45MP vs 24MP advantage is diminished. That said, there is certainly no advantage to the 24MP sensor, so if the price difference for the body you are going to convert is small I'd tend to go with the 45MP option. But if I could get a great deal on a 24MP body then I'd do that with no regrets as well.
Be aware you have to choose Z leases very carefully for IR use. Many have bad hot spots. Specifically the otherwise excellent 24-120/4S is not good for IR. The 24-70/4S is marginal. The 24-200 is not good. The 24-70/2.8S is poor. Hard to find a good "walk around" zoom for IR that is native Z mount. I believe in general Tamron lenses often do well for IR, so you might see check on the 28-70/2.8 options.
Here's an example of Z7 PDAF banding in IR along with how well banding removal can work (this was my own code, but the RawTherapee filter works just as well):
Yep, the 14-30/4S is pretty good if not quite perfect - very workable. The 40/2 is a delight and easy to always have with you. The 50/1.8S is very slightly better, but I rarely have it with me.
The 28-400 is quite serviceable, a bit worse for mild hot spot on the short end than the long end. On the short end it is very similar to the 24-70/4S as far as mild hot spot goes. Both usually need only a little work in post to deal with a very diffuse hot spot in certain conditions.
The Tamron 70-300 is excellent for IR. It is sort of the reason I didn’t end up keeping the 28-400, the 28-400 wasn’t adding much I didn’t already have between the 24-70/4S and the 70-300. But honestly, the fact the 28-400 works decently for both visual and IR makes it very compelling - especially paired with the 14-30/4S. Pretty amazing lens!
I’m thinking I’m going to try the Tamron 28-75/2.8 G2. From some E-mount tests it looks really good for IR. I don’t need F/2.8, but it’s a bit lighter than my 24-120/4S, optically better for visual, and possibly the best IR normal zoom. So I could have the 14-30/4S, 28-75/2.8 G2, and 70-300 both fully usable for visual and IR. Only weakness is the 70-300 has field curvature bad enough to still affect corners at F/8 below about 100mm.
Just reporting back so the information is at least someplace! (I hate it when someone posts they are going to do something and then never return with results).
I did acquire a Tamron 28-75/2.8 G2 and tested it on a 590 nm camera. In my experience it is a lot easier to see hotspots in such "color IR" shooting. I did a quick side-by-side test with the 24-70/4S by shooting the blue sky in the center of the frame with bright foliage along the bottom. The Tamron was actually slightly worse than the 24-70/4S.
Primarily I tested at F/8 (practical aperture) and at F/16 (to intentionally make hot spots easier to see). In both cases for both lenses the mild hot spots did not have sharp transitions or structure. They were quite diffuse which means they'd be easier to deal with in post processing than something like the 24-120/4S or 24-200 which have nastier hot spots.
Anyway, the Tamron 28-75/2.8 G2 is not a miracle IR lens unfortunately. It is pretty good for a modern zoom, but no better than the 24-70/4S.
As an aside, the 28-75/2.8 G2 that I received from Amazon actually has a fairly sizable gold colored hunk of metal loose in the front end of the lens!!! Well, I have a good excuse to return it then...
I think you should have tried the nikon version or the G1. The G2 has new optical glass which makes it sharp but also less suitable for IR. My guess is that new optics with advanced coatings are not good for infrared.
kwalsh wrote:
Just reporting back so the information is at least someplace! (I hate it when someone posts they are going to do something and then never return with results).
I did acquire a Tamron 28-75/2.8 G2 and tested it on a 590 nm camera. In my experience it is a lot easier to see hotspots in such "color IR" shooting. I did a quick side-by-side test with the 24-70/4S by shooting the blue sky in the center of the frame with bright foliage along the bottom. The Tamron was actually slightly worse than the 24-70/4S.
Primarily I tested at F/8 (practical aperture) and at F/16 (to intentionally make hot spots easier to see). In both cases for both lenses the mild hot spots did not have sharp transitions or structure. They were quite diffuse which means they'd be easier to deal with in post processing than something like the 24-120/4S or 24-200 which have nastier hot spots.
Anyway, the Tamron 28-75/2.8 G2 is not a miracle IR lens unfortunately. It is pretty good for a modern zoom, but no better than the 24-70/4S.
As an aside, the 28-75/2.8 G2 that I received from Amazon actually has a fairly sizable gold colored hunk of metal loose in the front end of the lens!!! Well, I have a good excuse to return it then......Show more →
indusphoto wrote:
I think you should have tried the nikon version or the G1. The G2 has new optical glass which makes it sharp but also less suitable for IR. My guess is that new optics with advanced coatings are not good for infrared.
The Tamron G1 for E-mount was indeed slightly better than the G2 as reported here:
But based on what I saw on the G2 and what is reported above, it seems unlikely the G1 is going to work any miracles. Also not 100% sure the Nikon Z version uses the same coatings as the Tamron branded versions. Probably does?
Anyway, doesn’t appear any of the 28-75/2.8 versions are going to be much different than the smaller, lighter 24-70/4S that I already own. But that still puts them on the short list of halfway decent Z mount standard zooms for IR! If someone wants F/2.8 they are a decent choice.
At this point I’m going to instead write some software to automatically remove diffuse hot spots from RAW files instead of chasing the never to be found perfect standard zoom for IR. Both the 24-70/4S and the 14-30/4S are good enough with diffuse enough hot spots that an adaptive steepest descent nuller provided two somewhat offset images of the same scene should be able to form an excellent estimate of the hotspot and subtract it from the image. Fun project for the winter I guess…
Interesting… I’ve found that list only semi-reliable as in certain cases “Very Good” and “Excellent” are reversed from my measurements. Probably expected as it started as two reviewers’ results combined and appears to have added other third parties since then. Still, I’ve found it a super useful starting point and it has saved me time and money by eliminating the obviously poor choices.
Searching more broadly there seem to be a handful of reports regarding the Nikon Z 28-75/2.8 specifically being excellent for IR. One person comparing it to the 50/1.8 saying it is better than the 50 - which would be impressive. Quite possible it is using different lens coatings than the Tamron E-mount version then. Hmmm…
Unfortunately it isn’t any better, or even barely worse, for landscape in the visual than the 24-70/4S. But I’m genuinely curious now if it is significantly better in IR.
Thanks for posting! I’d been assuming the Nikon Z 28-75/2.8 performance would be the same as the Tamron G1 E-mount, but maybe that’s not the case!
