Sony RX1 - Full Spectrum Conversion

I was unable to find anyone else who had tried converting this camera to infrared (to test its fixed 35mm f/2 lens)... So I decided to have it converted to full spectrum anyway. What could possibly go wrong?

DSC09483_combo (1).jpg

Various filters from the Sony RX1 - Schwende Disctict, Switzerland

THE 35mm IR DREAM

Ever since I started using the Sony A7 for infrared photography I have wanted a compact autofocus 35mm lens that was good for IR. Unfortunately I could not find one (check my hotspot lens test here) from Sony, Sigma, Samyang, Zeiss or Tamron. It seems like almost all modern lens coating are built for visible light at the expense of infrared.

 

I had loved using the Fuji X100 for IR (just useable hand-held without conversion) in the past, but the Fuji X100's 23mm f/2 lens isn't great for IR either. I was curious if the Sony RX1 (with its real 35mm f/2 lens) was any good for IR, but unfortunately there was no way to tell. I could not find anyone else who had it converted. Only one person who tried an IR filter on a non-converted camera with a long exposure. He thought he was seeing a hotspot, but it was actually vignetting from the thick hot mirror (which Alan hypothesized).

THE RISK

Considering how expensive the RX1 was originally I couldn't justify the risk of converting it for another £300 in case its fixed lens was not good for IR. Recently I saw one come up for a reasonable price on Tradera (A Swedish auction site) so I bit the bullet. It was a big risk, but I sent it off to Alan Burch (of IRCC UK) almost immediately. He did a fantastic job, discovering that it needed a glass layer to replace the hot mirror to focus correctly (unlike most of the cameras he converts. He bought a batch of the glass so now offers the conversion on his website here.

DSC00092 copy.jpg

 

INFRARED PERFORMANCE

This RX1's lens looks pretty great for IR, at least for a 35mm. Only at f/22 does a very faint hotspot start to show in some situations, but this is a non-issue as diffraction means most photographers will avoid that setting anyway. A score of 8.8/10 @f/8 is very impressive and it doesn't get much worse than that stopping down to f/16. Blooming and ghosting of bright objects are a bigger issue however (see below)

 

If you're not familiar with my IR testing, a perfect score is a straight line along the top. A straight line would be a consistent score of 10/10 at every aperture. This is what you would see in colour. To see more details about the IR performance of this lens click here.

RX1.jpg

BLOOM

The red line ("worst") is a bloom / glow effect from the objects in the scene that reflect infrared light. This can be seen in the below image as a large yellow tint to the sky in the middle of the frame. Ideally the sky (clear or clouds) should not be yellow in a false colour 590nm scenario. This is disappointing as it's impossible to fix and doesn't really change at any aperture. I think it is happening here more due to overcast conditions, but I need to do much more testing to confirm that.

GHOSTING

The RX1 also occasionally suffers from ghosting, at least that's what I'm going to call it. This is where bright objects get repeated in the frame as if you overlayed another frame slightly zoomed in. Although boosted to show the effect a bit better here (below) it's unfortunately still noticeable with normal processing, so it can be annoying. This is caused by very high contrast lighting and unfortunately doesn't differ much throughout the aperture range. 

 

I noticed a zooming effect similar to this when using Astonomik's filters (with their mirror like surfaces) on the Sony A7, so this could be IR light bouncing off the new glass layer, but that's purely speculative. So far this doesn't seem to happen too often.

 

The Zeiss FE 55mm f/1.8 lens also does something like this, but not as extreme and in a slightly different way (mirrored rather than zoomed). In my experience with that lens - this is probably not related to IR light bouncing off of the external filter. Not that we have much choice to try alternatives here, unless you want to try a straight IR conversion.

Just how common these infrared image quality issues are I'm not willing to speculate on just yet. I haven't had enough testing with enough weather types to say how this happens yet, but I will come back to update this section then.

THE CAMERA

Being a precursor to the original Sony A7, the RX1 uses the same 24mp full frame sensor. This is still competitive on sharpness, dynamic range and ISO noise today. Unlike the mkII it doesn't have any PDAF striping issues to adversely affect its IR quality, so its only weak point for IR is its lens (see above). It has a nice common filter size (49mm) and a metal screw thread (rare these days). Despite being 8 years old now this is still the smallest and lightest full frame camera and lens available. The Zeiss Sonnar lens is very good in general and pretty much worth the price you pay for these cameras second hand these days.

The down sides here are pretty obvious. Battery life, grip, buttons & menus are pretty poor. Much of this is simply down to its age however and some of those are fixed on the mkII, although for IR use this is likely not a good idea.

DSC08940.jpg

WISH LIST

Much of what I wish for here is in the next version (Sony RX1R II). The tilting rear screen, built-in EVF & newer sensor. Unfortunately the latter brings with it a PDAF array that likely introduces a striping issue that will adversely affect infrared images. This and the reason the camera is considerably more expensive put me off trying that one for conversion.

I would also like a better battery and grip, but at the same time I'm not sure I'd like to live with the extra size and weight, so this is a tough one. Perhaps something like the grip in the Fuji X-T1 would be nice here. That would still be pretty small compared to the lens, so it wouldn't take up much more space in a bag.

 

Nicer buttons and menus would be a huge bonus here too. What the RX1 currently has is pretty disappointing for such an expensive camera. This should get naturally fixed if Sony were to make a new camera as the Sony A7SIII is pretty nice in this regard. Although the Sony A7C was a clear step backwards, so maybe I should make it clear that Sony needs to keep pushing forward with this.

 

Speaking of the lens, it would be nice if they could get it smaller (like the Fuji X100). Now I know this is a proper 35mm f/2, sothere's a reason it's this big. However, I can't help but wonder how much smaller they could make it if Sony could develop a curved sensor. In theory that would drastically reduce the need for such a complex lens construction. To my knowledge curved sensors have never been done before, but it could be considerably more viable in a fixed prime lens system like the RX1, even if it would likely cost a small fortune. While you're at it please give us a global shutter. I don't ask for much :P. Come on Sony, you know you wanna :).

CONCLUSION

Battery life may not be great, the grip is non-existant, there's no built in EVF, the rear screen doesn't tilt, the buttons and menus are just terrible... but for small and light camera combined with a full frame sensor (that's still competitive today) and a sharp 35mm lens that's pretty good for IR - this is an impressive all-in-one option for full spectrum.

 

Hotspot may be a non-issue with the RX1 in my opinion, but considering its other infrared issues (ghosting and blooming) I wouldn't recommend this camera for a straight IR conversion. In the few cases where you run into these problem, a full spectrum conversion can avoid the issues by simply switching filters to a different wavelength. IR Chrome being the most friendly, in most cases. 

  • White Facebook Icon
  • White Flickr Icon
  • White YouTube Icon
  • White Instagram Icon