Sean Reid's observations on M mount lenses on the G1

[monza]

The issues with the adapters haven't been with regard to coplanarity, rather, they've set the lens either too close or too far from the sensor. I suppose it's possible they could add a bit of tilt, but they'd have to work hard to err in that regard.

 

[Jonas]

Here are some boring and quick wallpaper shots using the 18, 25, 35, and 50 mm Zeiss lenses. (...)

Thank you, speak volumes, at least if we are talking f/5.6 only.

I don't mind boring targets. They can tell us a lot.

 

[johnastovall]

Oh, where is Roland Barthes when we need him? In the end it's the studium and the punctum, which truly make the photograph not the sensor or the lens.

 

[scho]

Thank you, speak volumes, at least if we are talking f/5.6 only.

I don't mind boring targets. They can tell us a lot.

You are welcome Jonas. I also shot wide open (lens dependent f/2, f/2.8, f/4) and all at f/4. Same conclusion.

 

[Jonas]

G1 and CV28/2 Ultron results

I mentioned the f/5.6 comparison because I got another result today. I wonder what you can do out of this:

Earlier today I received a Cosina Voigtländer M-mount lens, the CV28/2 Ultron. I had high hopes for this lens as I have seen a lot great images taken with it (at flickr that is).

Pretty early when playing with it I noticed a (quite severe) focus shift problem. Then I aimed the camera towards a book shelf to see what the borders/corners would be like. I shot two series; one without refocusing between the exposures and then another one where I did refocus thinking of the mentioned focusing shift.

The images below are from the second series - refocusing was, simply put, necessary with this lens.

First an overview:

Then, in order, f/2 - f/5.6 100% crops. There is no sharpening applied to the crops:

I can't believe the lens is supposed to behave this way and the images I have seen (Bessa and M8 images) didn't prepare me for this. There is no difference between the left and right side or any of the corners.

Now the lens will go back anyway due to the focus shift. I knew about the focus shift but it was worse than anything I had read at the rangefinder forum.

What do you think about all this? Should I buy Zeiss lenses instead?

 

[Cindy Flood]

Re: G1 and CV28/2 Ultron results

I mentioned the f/5.6 comparison because I got another result today. I wonder what you can do out of this:

Earlier today I received a Cosina Voigtländer M-mount lens, the CV28/2 Ultron. I had high hopes for this lens as I have seen a lot great images taken with it (at flickr that is).

Pretty early when playing with it I noticed a (quite severe) focus shift problem. Then I aimed the camera towards a book shelf to see what the borders/corners would be like. I shot two series; one without refocusing between the exposures and then another one where I did refocus thinking of the mentioned focusing shift.

The images below are from the second series - refocusing was, simply put, necessary with this lens.

First an overview:

Then, in order, f/2 - f/5.6 100% crops. There is no sharpening applied to the crops:

I can't believe the lens is supposed to behave this way and the images I have seen (Bessa and M8 images) didn't prepare me for this. There is no difference between the left and right side or any of the corners.

Now the lens will go back anyway due to the focus shift. I knew about the focus shift but it was worse than anything I had read at the rangefinder forum.





What do you think about all this? Should I buy Zeiss lenses instead?

I do not understand what this is all about. Focus shift????
I have / have had lenses with focus shift on my M8. These are fast lenses in my case. The Noctilux was the worst. I had to focus it (when wide open) and then give it a little twist to the right. Stopped down, there was no shift. The problem was that the focus patch indicated that the lens was in perfect focus when it was at f/1, but it wasn't.

With the G1, focus shift is not a problem. Just focus the lens for the shot and it is always perfect. This is the one of the reasons that I like the G1 so much. When you focus, you can be sure that you are in focus.

Smearing corners is a different thing, but I haven't seen it in my photos yet. Thanks Carl for doing testing on this. I am interested in taking photos, not testing, so I'll wait to hear what info others gather.

If I'm missing something here, please enlighten me.

 

[Guy Mancuso]

Well if you focus with the lens wide open than stop down after you make the focus , than focus shift can occur. On the G1 you should be really focusing at the working aperture than it should not happen. Alas the G1 has gain with it so it does make it finder friendly to focus at the working aperture .

The question I have here and maybe the case is if you focused wide open than left that alone and just stopped down and shot WITHOUT refocusing at the working aperture than the focus shift could occur. In this case you really need to refocus each time with the G1 at the working aperture

 

[scho]

Re: G1 and CV28/2 Ultron results

I mentioned the f/5.6 comparison because I got another result today. I wonder what you can do out of this:

Earlier today I received a Cosina Voigtländer M-mount lens, the CV28/2 Ultron. I had high hopes for this lens as I have seen a lot great images taken with it (at flickr that is).

Pretty early when playing with it I noticed a (quite severe) focus shift problem. Then I aimed the camera towards a book shelf to see what the borders/corners would be like. I shot two series; one without refocusing between the exposures and then another one where I did refocus thinking of the mentioned focusing shift.
I can't believe the lens is supposed to behave this way and the images I have seen (Bessa and M8 images) didn't prepare me for this. There is no difference between the left and right side or any of the corners.

Now the lens will go back anyway due to the focus shift. I knew about the focus shift but it was worse than anything I had read at the rangefinder forum.

What do you think about all this? Should I buy Zeiss lenses instead?

I have this lens also and yes, it does focus shift. I get around it by either shooting wide open or stopping down to about f/8 for landscapes on the M8. However, focus shift, as Cindy noted, is really a non-issue on the G1. What you are seeing with your series of bookshelf shots is exactly the same as what I found with my wallpaper shots and the Zeiss lenses. This peripheral distortion/smearing seems to be most prominent in the 20 to 35 mm focal lengths that I have tested on the G1 with the M adapter and it is worse at wider apertures. It has nothing to do with focus shift.

 

[Cindy Flood]

Re: G1 and CV28/2 Ultron results

I have this lens also and yes, it does focus shift. I get around it by either shooting wide open or stopping down to about f/8 for landscapes on the M8. However, focus shift, as Cindy noted, is really a non-issue on the G1. What you are seeing with your series of bookshelf shots is exactly the same as what I found with my wallpaper shots and the Zeiss lenses. This peripheral distortion/smearing seems to be most prominent in the 20 to 35 mm focal lengths that I have tested on the G1 with the M adapter and it is worse at wider apertures. It has nothing to do with focus shift.

Carl and Guy, thanks for the input and clarification.

 

[Jonas]

Re: G1 and CV28/2 Ultron results

I do not understand what this is all about. Focus shift????

It is known the CV28/2 suffer from some focus shift. I mentioned it most of all to make sure anyone reading the post should know I knew about it, and hence did refocus between the shots.

I have / have had lenses with focus shift on my M8. These are fast lenses in my case. The Noctilux was the worst. I had to focus it (when wide open) and then give it a little twist to the right. Stopped down, there was no shift. The problem was that the focus patch indicated that the lens was in perfect focus when it was at f/1, but it wasn't.

With the G1, focus shift is not a problem. Just focus the lens for the shot and it is always perfect. This is the one of the reasons that I like the G1 so much. When you focus, you can be sure that you are in focus.

If I focus the CV28/2 with the G1 (with the lens opened up to f/2) and then stop down the focus plane has moved. That is focus shift. With the CV28/2 stopping down to f/8 didn't help; the focus plane has moved away so far that the DOF doesn't cover the are first focused at. That was way more severe than I expected.

Smearing corners is a different thing, but I haven't seen it in my photos yet. Thanks Carl for doing testing on this. I am interested in taking photos, not testing, so I'll wait to hear what info others gather.

If I'm missing something here, please enlighten me.

My post was about smearing corners and I hope we are talking about the same thing. Maybe i shouldn't have mentioned about the focus shift at all.

 

[jlm]

i'm confused...isn't the crop from the center, not the corners?

 

[Jonas]

Re: G1 and CV28/2 Ultron results

I have this lens also and yes, it does focus shift. I get around it by either shooting wide open or stopping down to about f/8 for landscapes on the M8. However, focus shift, as Cindy noted, is really a non-issue on the G1. What you are seeing with your series of bookshelf shots is exactly the same as what I found with my wallpaper shots and the Zeiss lenses. This peripheral distortion/smearing seems to be most prominent in the 20 to 35 mm focal lengths that I have tested on the G1 with the M adapter and it is worse at wider apertures. It has nothing to do with focus shift.

To anyone:

My post was about the smeared borders. I shouldn't have mentioned "focus shift" as there is nothing making people more confused. And I did mention I took two series and the crops are from the second one where I did re-focus between the images (meaning I refocused for every step I moved the aperture ring). OK now?

==

@ Carl:

Thank you.

To me it is clear there is something with the G1 and the short focal length lenses that doesn't work very well. The left border crop in my images are approximately 8mm from the center when looking at the sensor. I can't imagine the lens behaves like that on film where it is about 21mm from the center to the corner.

The problem is probably also lens dependent, some lenses may be more telecentric than other ones.

 

[Jonas]

(...)
The question I have here and maybe the case is if you focused wide open than left that alone and just stopped down and shot WITHOUT refocusing at the working aperture than the focus shift could occur. In this case you really need to refocus each time with the G1 at the working aperture

Hi,

As mentioned in my post i refocused between the shots. I started with f/2, focused, triggered the shutter, stopped down to f/2.8, refocused (10x magnification activated), took the shot, stopped down to f/4... and so on.

I think the only mistake I did was to mention the focus shift problem i noticed with the lens...

 

[Cindy Flood]

Jonas, Thank you for your explanation of your method. Did you use a tripod?

 

[Jonas]

Jonas, Thank you for your explanation of your method. Did you use a tripod?

Hi Cindy,
Yes, I indeed used a tripod. The images are ISO100, f/2 1/8 sec and so on, taken with the camera on a tripod and with the 2 sec timer turned on.

 

[barjohn]

If you looked at Sean's test of the CV28/2 you would see that even on the M8 it is soft in the corners at f2. It is softer focused even in the center at f2 compared to the f1.9 nockton and the 28 cron. It isn't until f5.6 that it is getting close and at f8 it looks pretty good.

 

[Brian Mosley]

Jonas, I think you're just confirming for yourself what we already know from Sean's review.

Maybe this is why the 20mm f1.7 won't arrive for some months... The 25mm f2.8 pancake for 4/3rds was quite an achievement. And that's for Olympus.

At this short FL you can have slow or big, pick one

For street shooting, maybe the lumix 14-45mm kit lens is the optimum choice for now?

Kind Regards

Brian

 

[peterv]

Hi Jonas, sorry to see so many got confused. Of course your test is not about focus shift. What you get from your new CV lens is, from what I've seen from my own tests, not a problem specific for this lens. I've seen it with the 28 Cron, and with the 50 Lux and the 90 Cron, too. I would like to understand what it is that makes the G1 produce so much distortion around the edges of the frames with my (your) M-mount lenses.

 

[Brian Mosley]

Interesting that there's a problem with your 90 Cron... maybe it's the registration distance of the M-mount lenses being too close, hence greater angle of incident light rays at the edges?

With the Hexanon lenses and the 4/3rds adapter the lens is at normal 4/3rds distance - I'll have to check whether I'm getting the same problem - I can check with a 21mm f4 I have.

Kind Regards

Brian

 

[jklotz]

pixel peepers delight!

I've always believed that if you want the scoop on your own lenses, one must test them for themselves. Pixel peepers delight! Heres the test shot, the wall in my studio. Processing in LR, no processing, sharpening or manipulation. I took this shot, wide open on the following lenses. I then cropped the BEST corner at around 100%, results to follow.