RD-1 resolution question

[georgef]

Hi everyone,

I have in p-shop set up a command tree I like to use on almost all RD-1 images before cataloguing; upon close inspection, zooming to a high magnification in the image, I find I see the out-of-focus before I see the pixelation.

Let me explain: I used to have the NIKON D100, which has the same sensor as the RD1. When I zoom into those images I see the pixelation before the out-of-focus. By that, I mean if I keep on zooming in, I start seeing the individual pixels but the image seems still in focus. This is not the case with the RD1. It starts to look soft before I get to the individual pixels. Its as if I cannot get the same degree of focus with it that I did with the D100.
I thought it may be that the RD1 needs adjustment for focus, which I will attempt tomorow.
I was just wondering if anyone else noticed this.

 

[RF-Addict]

Could eaily be the lens you are using, or you are not sharpening the same amount you did with the Nikon. What are you shooting - JPEGs or RAW?

 

[jlw]

I used to have a D100 also, and still have an R-D 1. I can't prove it, but it wouldn't surprise me to learn that the D100 has a higher level of default sharpening than the R-D 1 does.

The amount of default sharpening does vary from camera model to camera model, and yes, to some extent this does apply to raw images as well as JPEG images. Despite what we read and hear, there is no such thing as a completely unprocessed digital image -- the camera has to do some processing to convert the sensor's analog charge-level readouts of one-color pixels into a digital record of three-color pixels, and there are always some algorithms involved in that.

I believe that Epson deliberately kept the default sharpening level fairly low, leaving it to the user to apply more sharpening in software if desired. I noticed that my low-light photos with the R-D 1 usually looked "smoother" than those with the D100, and in fact they look somewhat smoother than those from my current D80 even when I take the D80's greater pixel count (10 mp vs. 6 mp) into account.

 

[Jim Watts]

As jlw says just because the cameras have the same sensors it does not necessarily mean that images from the camera will look the same even in Raw format.

The choices for the camera maker - firmware writer are though more complex than just one of whether or not to add some amount of additional sharpening in the conversion. The optical to electronic conversion function controls the tonal reproduction and also has a bearing on the apparent sharpness, amount of noise and artifacts present in the image. These are a balancing act for the manufacturer. If he chooses a very inverse-S-shaped function strategy this boosts the mid-range brightness areas contrast and enhances the perceptual sharpness of the images (which also has the bonus of giving the camera high scores in most reviewers resolution tests) this will also give softish shadows and highlights and a greater dynamic range. The disadvantages are that it produces inaccurate tone reproduction, higher artefacts (jaggies, ghost lines) at high magnification, more apparent noise at higher iso's and increased corner shading (vignetting). The use of different anti-alaising moire filtering and microlenses in front of the sensor will also have a bearing on how the maker carries out the conversion.

Epson have taken the strategy of a more linear conversion from the sensor in the R-D1 partly I believe to avoid the greater vignetting problems with a sensor in a rangefinder camera and partly to keep noise low. The added benefit is the good colour\tonal reproduction for which the R-D1 is rightly praised and the downside is even raw file comparisons show less apparent sharpness and 'snap' than some other cameras.

Nikon seem to have gone for higher mid-range contrast with the D100 conversion, which gives a greater preceptual sharpness (but not necessarily a higher true resolution), but also more artefacts (and the pixels become more apparent, especially on diagonals).

One strategy is not necessarily better than the other (within limits), but will of course result in subjective preferences.

 

[georgef]

Thanks for your thoughs,
RF-Addict, I use Jpeg /Raw most of the time.

I also agree about the different processors between the Nikon and Epson.
JLW, I agree also thet eh RD-1 gives me quite a beter set of low light results than the Nikon, hands down. I am still under the impresion that my RD-1 may be off allignment, so I will try re-calibating the rangefinder.

I hopeI was successful in uploading an image as an example. The door shot is shrunk to fit the size, but the centre lock crop is as shot. I find it soft even without sharpening. It was shot with the CV 28 1.9 ultron at f.5.6, 1/125/sec.

 

[georgef]

ooops, here are the images. Downscaled to fit the upload criteria.

 

[Leica Geek]

I noticed the same thing. How do the prints look. I have the D100 and the RD-1 and I noticed the same thing. I thought something was wrong with the RD-1.

 

[Philip Whiteman]

I too use a D100 alongside an R-D1 (which I chose instead of upgrading the D100 to a D200). I'm very pleased with the prints from both (up to A3, where quality does become critical). Depending how much space I have on cards and the assigment, I shoot RAW/RAW+jpeg or just plain jpeg (I find jpeg losses are not much of an issue, provided you treat the original as a digital negative and make copies, never saving changes)

I can't say I have noticed a systematic difference in sharpness between the two, but I did find the Nikon's default sharpening on jpegs was too soft for my liking, and long ago tweaked it up to maximum – haven't really got to grips with this on the Epson, but nor do the pre-processed jpegs look unduly soft.

Interestingly, on the R-D1s default jpeg setting shots using a 50mm Summicron often do not need sharpening, at least for modest enlargement, wheras all exposures with the 15mm Super-wide Heliar most certainly do. The Heliar is sharp when used with film – so I don't really know why!

I'm tempted by the D300, but the D100 and R-D1 are still magnificent cameras. Just got an Epson Stylus Photo 1400 printer and the quality in prints you can produce amazes this one-time B & W darkroom fanatic

 

[nemjo]

I'm sure, there are better solutions, but do you find it better?

nemjo

 

[georgef]

I noticed the same thing. How do the prints look. I have the D100 and the RD-1 and I noticed the same thing. I thought something was wrong with the RD-1.

The prints seem to show the same results; I hear people rave about how sharp certain lenses I have and used, but I am not getting the results.
I just put the camera on a tripod, put the heliar 15mm and the ultron 28mm on, then shot both at f.8, 1/30th sec. at ISO 200. The images are nice, wonderfully rendered, but not "wow" sharp! At no point anywhere in the image, is there a spot that impressed me!

Has anyone blown up prints to, say 11"X17"? Mine look great on 5X7, good on 8.5X11, but only ok-ish at 11X17. I understand what I should expect from 6mps, but comparing the D100, this is sub-par in sharpness.
I will calibrate the rangefinder again and try again.😡 😕

 

[georgef]

I'm sure, there are better solutions, but do you find it better?

nemjo

Hi nemjo,
I have also sharpened the image to print, but to the point where it becomes artificial.
Its just seeing the image comming out of the camera and not being.."3D" if may use that description...hard to explain, just overall soft enough to be noticable.

Maybe I am being too picky? I was getting much sharper images straight out of my CANON 1D at 4mp!

 

[Tuolumne]

This jpeg photo was taken with a 28mm M-Hexanon wide open (f2.8) on an R-D1s. The final image (below) is a 50% crop from the original. In an A3 print it is a knockout with the area from the blanket under the chin to the top of the nose particularly sharp and detailed. The apparent sharpness is enhanced by the out-of-focus foreground.

/T

 

[louisb]

I can only add some anecdotal experience. I recently submitted A3 sized prints from my R-D1 for a public exhibition which were accepted. One was taken with my Elmarit 90/2.8 on a tripod and the second was with my lux35asph hand held. In my opinion neither was fantastically sharp when I inspected the prints (even though I can normally blow up the lux35 images on my screen to 100% and they are still amazingly sharp). However, everyone who saw the prints oohed and aaahed at the images and two out of five were chosen by the selectors.

Here's the point. How many people are going to stand with their nose to the glass looking at your pictures? Most are going to stand 3-5 feet away at which distance everything looks sharp!

I'd worry more about colours and tones.

Really guys, move on and take some pictures!

LouisB

 

[georgef]

...Here's the point. How many people are going to stand with their nose to the glass looking at your pictures? Most are going to stand 3-5 feet away at which distance everything looks sharp!

I'd worry more about colours and tones.

Really guys, move on and take some pictures!

LouisB

Louis, for one, I am going to stand with my nose to my images🙂

Every year, there is a show of the year's best photojournalism images here in Toronto, in the BCE place. I have been awed by almost everyone of them, and wish I had taken them. 8 out of 10 are technically sub-par, but the subject is so powerful, all else is secondary.
Neither that example nor yours, however, helps me in figuing out if there is somethiing wrong with my RD1 or if this is something everyone sees.

This was the point of my original post. There is a wealth of experience and enthousiasm here I am hoping to learn from.

 

[Jim Watts]

ooops, here are the images. Downscaled to fit the upload criteria.

George,
Sorry I'm not clear exactly what you have posted here so its hard to tell.

Is the lock a crop at 100% pixels from the larger image? Has the image had any added sharpening or is it straight from the camera?

If not can you post a full size image (downsized) and a 100% crop at screen resolution (before downsizing) that are a conversion from Raw to Jpeg without further procssing at a high quality jpeg setting.

 

[Johnmcd]

Hi George,

I have posted the image below in another thread. http://www.rangefinderforum.com/forums/showthread.php?t=47932

But I have included a 100% crop with no sharpening as produced by Capture 4 Beta from raw. The resized image does have some sharpening applied as required and would print very sharp indeed.

My RD1 with the 40mm Nokton is as sharp as anything I can produce with my 20D and an 'L' lens.

John

 

[Johnmcd]

George,

I had another look at your examples and definitely think that something is not right. I am sure that the RD1 can do better than that. Though I don't have a D100 to compare, my RD1 compares very favourably with my Canon 20D.

Here is another example I took today. The 100% crop has no sharpening at all and was converted from raw with Capture 1 (4) beta. The resized image has sharpening, saturation etc.

Hope this helps or if nothing else shows that the RD1 can be as sharp as any other camera.

BTW, both images have been compressed with a jpeg setting of '8' in Photoshop.

Cheers,

John

 

[georgef]

Thanks John,
That is the kind of input I was hoping for. I do not seem to get these results with my R-D1, whether I am using my 40 nokton, 28 ultra or 15 heliar!?! so I think I will endeavour to adjust the rangefinder this weekend.
Very nice and sharp images by the way.
cheers, George

 

[JNewell]

I had another look at your examples and definitely think that something is not right. I am sure that the RD1 can do better than that. Though I don't have a D100 to compare, my RD1 compares very favourably with my Canon 20D.

I do have a D100, and I think the two compare very well (that is, the results are generally similar, given some allowance for different lenses). Check Rich Cutler's site for the best instructions on the RF and *be patient* with yourself and the camera! 😀