The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Cooke Speed Panchro 50mm/2.0/T2.3 on the G1


View attachment 13649

Cymbidium Orchid
G1 - Cooke Speed Panchro Ser 2 50mm/f2.0/T2.3 on Spacer Ring - ISO=100 1/800 @ f/2.0 - Raw Therapee + Gimp


View attachment 13650

100% crop

Technical:

This 35mm cine lens had previously been converted to Leica M mount
Lens covers whole sensor : Yes
Vignetting : No
Smearing : No
Adapter : Novoflex M-> m4/3
 
R

Ranger 9

Guest
Holy cannoli!

I'd expect excellent results from high-end professional cinematography lenses, which I assume this is... but it's still impressive to see it "in the pixels."
 
man you have some exotic lenses...
(Taylor, Hobson) Cooke Speed Panchros can be often found on ebay.

If you add up what some people here are spending on c-mount lenses that do not cover the whole sensor, and/or smear and/or vignette, they might even be considered cheap.:).

If you want one, make sure it is Series 2 or 3, they are much better, and of course check they are in good condition. Speed Panchros are workhorses of the motion picture industry. Often they have not been handled with gloves but might have perfect glass (or vice-versa)
 
Last edited:
M

milapse

Guest
If you add up what some people here are spending on c-mount lenses that do not cover the whole sensor, and/or smear and/or vignette, they might even be considered cheap.:).
$1000+ vs under $200... Hmmm strange math you have there!
 

Jonas

Active member
Cymbidium Orchid
G1 - Cooke Speed Panchro Ser 2 50mm/f2.0/T2.3 on Spacer Ring - ISO=100 1/800 @ f/2.0 - Raw Therapee + Gimp
To my eyes everything look just fine, details are crisp (I see the image is sharpened) at the same time as the background details are rendered ina pleasing way. There is something that may be an out of focus highlight point in the background, or it is just something bright, that looks good.

But hey, where does all this noise come from? Not only the crop but the first downsized image suffers from blotchy ugly noise. Did you raise the shadows considerably? regards, /Jonas
 
But hey, where does all this noise come from? Not only the crop but the first downsized image suffers from blotchy ugly noise. Did you raise the shadows considerably?
Jonas, thank you for your comments.

I have analyzed the image, both the original RW2 in both Silkypix and Raw Therapee, as well as the ex-camera JPEG. The ex-camera JPEG was underexposed. The flower was lit by direct sunlight, the background was in the shade. The exposure the camera set in "A" mode and spot metering was 1/800, not enough even for the white flower

Raw Therapee automatically increased brightness by 0.91, in Silkypix I had to manually increase brightness and contrast. Explicit sharpening of the posted image was minimal, only slightly on top of the defaults of the raw converters, applied after downsizing for the web.

My conclusion is that this camera is much more critical about underexposure than both the DMR and the M8, my usual cameras. With these Leicas you can gain 2-3 EVs in raw conversion without generating objectionable noise. It is even a known technique in high contrast situations to explicitly underexpose by 2-3 EVs and gain them in raw conversion in order not to burn highlights. For this picture it was not of my doing, I let the camera set the exposure.

What do you think, have you had any similar situation? It might be interesting to start a thread about this noise problem.

Regards
Peter
 
$1000+ vs under $200... Hmmm strange math you have there!
Some of the best c-mount lenses fetch similar prices, a Steinheil Quinon1.5/25mm is currently offered for $999, a Navitar 0.95/50mm for $1,299. An ugly looking Kinoptik Macro Apochromat 2/25mm recently sold for $700 within a few minutes with buy it now. Compared to that, a Schneider Xenon 25mm F:1.4 that fetched $499 seems a steal.

At the time I bought the Speed Panchro, 3 years ago, lens prices were more reasonable, I think the G1 craze lately raised prices significantly, crisis or no crisis.
 

Jonas

Active member
I have analyzed the image, both the original RW2 in both Silkypix and Raw Therapee, as well as the ex-camera JPEG. The ex-camera JPEG was underexposed. (...)
The flower was lit by direct sunlight, the background was in the shade. The exposure the camera set in "A" mode and spot metering was 1/800, not enough even for the white flower.
What do you think, have you had any similar situation? It might be interesting to start a thread about this noise problem.
Hmm, a white flower and spot metering? As a flower doesn't run away I would have increased the exposure and dialed in the exposure by looking for the white blinkies when reviewing the image. ;)

Yes, I have run into similar problems with my 4/3 cameras, first with the E-510 and now with the G1. Before the E-510 and between the E-510 and the G1 I shot mainly with the Canon 5D and there is no comparison. I think that's why I am a bit (not much but still...) frustrated by the noise in the G1 images, also at ISO100.

When the images are correctly exposed the problems is small and I see noise only when cropping ("printing big"). If the image is a little underexposed, or if I need to raise the shadows to improve the image the noise is there, immediately.

The 12 megapixels from the g1 are certainly more vulnerable than the 12 megapixels from the 5D. I don't know why that is but guess somebody with more in depth technical knowledge than me can explain it.

All I know is that my small sensor images kind of fall apart far more easily when I'm heavyhanded with the post processing. If you had to compensate the exposure one stop that explains it to me.

My "workaround" is to check the exposure whenever I have the time. I don't apply any sharpening during the raw conversion. In an image like your orchid image I would have applied some local sharpening only in Photoshop. If I use a tripod I often take several images using bracketing and then use the best parts from them in Photoshop using layers.

The problems are the limited dynamic range, noise at base ISO and the vulnerable images. Heh! That sounds worse than it really is... but the precautions mentioned are sometimes really needed.

regards, /Jonas
 
The 12 megapixels from the g1 are certainly more vulnerable than the 12 megapixels from the 5D. I don't know why that is but guess somebody with more in depth technical knowledge than me can explain it.
Jonas,

Thank you for your explanations and workaround. My technical knowledge is not very deep, but I repeat what I read, for instance in DPReview tests of small sensor cameras.

The more pixel you pack into a given surface, the smaller each pixel gets. Being smaller, it will need more signal amplification by the electronics and that is a source of noise. It is a bit like if you tune a radio to a distant station with a low signal level and turn up the volume. The 5D has a bigger sensor surface and thus each one of the 12 megapixel is bigger and needs less amplification. The result is less noise.

Sometimes, less megapixel is better than more, but the market seems to be focused mainly on the megapixel battle. The Digilux 2 is still capable of delivering excellent results in spite (or because) its pixel count is less than in current cameras.

regards,
Peter
 

Jonas

Active member
The more pixel you pack into a given surface, the smaller each pixel gets. (...) (5D) and thus each one of the 12 megapixel is bigger and needs less amplification. The result is less noise.(...)
I guess it is as simple as that. While I prefer the handling and features coming with the G1 I prefer the image quality from cameras with bigger sensors.

It's a good thing that the smaller sensors seem to develop faster. OTOH I also see the new generation of FF cameras having lower noise (and more) than ever and the gap seem to still be about three stops. Physics is physics. Luckily I don't always need super performance.

kindest, /Jonas
 
R

Ranger 9

Guest
My conclusion is that this camera is much more critical about underexposure than both the DMR and the M8, my usual cameras.
I've noticed the same thing in comparing G1 photos to those from my Epson R-D 1, and remarked on it in another thread.

My observation is that basically, the G1 has good noise performance even at high ISOs as long as you expose correctly, but you dare not underexpose with it at all.

By comparison, it's usually no problem to rescue one or sometimes even two stops of underexposure in raw files shot with the R-D 1. I assume that the difference is mostly because the R-D 1's sensor, even though an older design, is only trying to pack six megapixels onto an APS-C-size sensor. The G1, with twice as many pixels on a smaller sensor, poses a greater challenge in electrical isolation, as Jonas noted. Your M8, with 10 megapixels on a larger-than-C-size sensor, also represents an easier-to-solve noise problem.

Another thing that occurred to me: In comparing the G1's noise results to those of other cameras, you have to keep in mind that the G1's ISO-sensitivity settings are on the "hot" side, as you can see by the graphs on the DxOMark website. I had been comparing noise in photos made with the G1 set to ISO 1600 against photos made with my Nikon D300 at the same ISO setting, and noted that the G1 photos seemed noisier.

When I compared the graphs on DxOMark, though, I realized that's partially because the G1 is almost one full stop more sensitive at any given ISO setting than the D300 is. In other words, if I had compared actual exposures rather than ISO settings, I would have found that the Nikon has to be set at almost EI 3200 to deliver the same exposure level that the G1 delivers at 1600! When I compared the noise performance of ISO-3200 shots with the Nikon against ISO-1600 shots with the G1, I found their results to be more similar.

I don't know if this will work for you, but I've been experimenting with the following procedure for getting the best exposures from the G1:

When I start photographing in a given lighting situation, I choose an ISO setting that gives me roughly the range of f/stop and shutter speed I need, and use the AEL to lock in a representative midrange exposure.

Then, as I compose pictures, I adjust the exposure compensation dial while watching the in-finder histogram. I adjust the dial to move the histogram toward the right until the lightest highlights just begin to clip. This assures that I'm giving the most exposure the scene can handle​

It sounds more complicated than it is; once the exposure is locked, it's very easy to check the histogram and turn the front dial a bit if needed. Incidentally, this exposes a bit of clever design in the G1: turning the front dial to the right (from the photographer's viewpoint) moves the histogram to the right, and turning the dial to the left moves the histogram to the left.

Sure, I'd rather have the G1's 12 megapixels and small size AND the tonal range and noise performance of a larger, lower-pixel-count sensor. But this is a new world and, as is the case with any new category of camera, it takes some experimenting to figure out how to get the best out of it.

Incidentally, one of the claims made for the GH1 (that's what they're calling the newly-announced video version, right?) was an improved sensor with better low-light performance. It'll be interesting to see how true that is and whether it has significant still-photography benefits. For now, though, I'm managing with the current version. Just remember: don't underexpose!
 
B

butterdada

Guest
I have cooke panchro 40mm F2
My friend modify it to Leica M mount with rangfinding.

The perfomance is pretty good both on Epson R-D1s and Panasonic DMC-G1.

I think better than Kinoptik 40/2.
 
I've been experimenting with the following procedure for getting the best exposures from the G1:
When I start photographing in a given lighting situation, I choose an ISO setting that gives me roughly the range of f/stop and shutter speed I need, and use the AEL to lock in a representative midrange exposure.

Then, as I compose pictures, I adjust the exposure compensation dial while watching the in-finder histogram. I adjust the dial to move the histogram toward the right until the lightest highlights just begin to clip. This assures that I'm giving the most exposure the scene can handle​
It sounds more complicated than it is; once the exposure is locked, it's very easy to check the histogram and turn the front dial a bit if needed. Incidentally, this exposes a bit of clever design in the G1: turning the front dial to the right (from the photographer's viewpoint) moves the histogram to the right, and turning the dial to the left moves the histogram to the left.
Thank you for your tips, I shall try it out. It certainly SOUNDS complicated when you read it, but every camera has a learning curve if you want to get the best results.
 
I guess it is as simple as that. While I prefer the handling and features coming with the G1 I prefer the image quality from cameras with bigger sensors.

It's a good thing that the smaller sensors seem to develop faster. OTOH I also see the new generation of FF cameras having lower noise (and more) than ever and the gap seem to still be about three stops. Physics is physics. Luckily I don't always need super performance.
Jonas,

One could dream of a top range Leica(?) with the handling and features of the G1 and a bigger (full size?) sensor. They could leave out some features like film mode, though.

kindest,
Peter
 
R

Ranger 9

Guest
Thank you for your tips, I shall try it out. It certainly SOUNDS complicated when you read it, but every camera has a learning curve if you want to get the best results.
I suppose a simpler method to accomplish the same thing would be to use manual exposure mode and set your exposures by referencing the histogram rather than the little pointer at the bottom of the viewfinder.

I like using the AEL/exposure compensation method because it keeps the finder brightness more constant.
 
Top