[ElJamoquio]

Kodak Technical Pan 320 lppmm


The highest resolution film is therefore 320 line pairs per mm, that's phenomenal resolution

Where can I buy some? How do I develop it? How's the ISO 100 version stack up?

---

Last edited by ElJamoquio; 06-04-2014 at 02:11 PM.

[stevebrot]

The 28-70 isn't a modern lens, it's 21 years old and it resolved ~130 lp/ph on old sensors.

Modern lenses on modern sensors are resolving 225 lp/ph with color, see e.g. Sony E 10-18mm f/4 OSS (SEL-1018) - Lab Test / Review - Analysis

That's much higher than any color film resolution I'm aware of, but I don't know much about film.

Did you mean 225 lp/mm? 225 lw/ph would be dog poor.

As for films, Modern photography tested the best lenses of the time (early 1980s) against the highest resolution films available at the time.* The best performance was 102 lp/mm with the Leitz Summicron 50/2 on Micro Ektachrome. The Pentax-M 50/1.7, along with the other five lenses in the test, was just behind at 100 lp/mm. Evaluation of resolution was done by direct microscopy of the negative. 100 lp/mm = 4800 lw/ph = 2400 lp/ph

I may be wrong, but 4800 lw/ph is quite competitive by current standards.

The Summicron-M 50/2 had the same optical design from 1979 until 2013 and for that period was considered to be the performance benchmark for the 50mm focal length. I don't know what lens currently holds the title (Zeiss Otus 50/1.4 or maybe Apo-Summicron-M 50/2?). It would be my expectation that the Pentax-M 50/1.7 if tested directly against the Summicron on, say, the Sony a7r would probably still perform in the same ballpark.

In other words, the best of the vintage lenses should be well up to the task of supporting current generation digital sensors.


Steve

* Nester's scans of the original Modern Photography article:

https://www.flickr.com/photos/nesster/4424744296
https://www.flickr.com/photos/nesster/4424744224

[ElJamoquio]

Did you mean 225 lp/mm? 225 lw/ph would be dog poor.

Thanks, edited.

As for films, Modern photography tested the best lenses of the time (early 1980s) against the highest resolution films available at the time. The best performance was 102 lp/mm with the Leitz Summicron 50/2 on Micro Ektachrome. The Pentax-M 50/1.7, along with the other five lenses in the test, was just behind at 100 lp/mm. Evaluation of resolution was done by direct microscopy of the negative. 100 lp/mm = 4800 lw/ph = 2400 lp/ph

I may be wrong, but 4800 lw/ph is quite competitive by current standards.

Very.

In other words, the best of the vintage lenses should be well up to the task of supporting current generation digital sensors.

I agree to an extent. It's not like the 31mm or 77mm are garbage, they're certainly more than adequate resolution wise. Problems seem to be more along the lines of purple fringing IMO. Fortunately the more recent (4&5) versions of lightroom have been able to fix that.

[stevebrot]

Where can I buy some? How do I develop it? How's the ISO 100 version stack up?

You can't, Technical Pan has been out of production for some time. Comparable emulsions are available:

Rollei ATP1.1 Advanced Technical Pan Film 35mm x 36 exp. - Single Roll Unboxed | Freestyle Photographic Supplies

ATP is best when developed in Rollei's developer and is nominally rated to ISO 32. You can shoot it at whatever you wish, of course, but your mileage may vary. Or you can simply shoot any of the available tabular-grain films. Resolution is less than Tech Pan or ATP, but still somewhat better than most of the conventional films of the past. In other words, any quality film produced today should have adequate resolution to support prints to about 16"x20" as an optical enlargement with no special technique or 12"x18" with a 4000 dpi scan.*


Steve

* Assuming 300 dpi at the printer

---------- Post added 06-04-14 at 02:54 PM ----------

Very.

Did you mean that 4800 is very competitive or I am very wrong?

I did a quick survey of reviews at photozone.de before I posted my previous comment and did not find any with higher than 4800 lw/ph. For example, the best the highly-rated Sigma 35/1.4 Art could manage on the Nikon D3x was 3960 lw/ph. The above mentioned benchmark Summicron-M 50/2 on the Leica M9 had a maximum of 3265 lw/ph. The best I could find was the Zeiss Sonnar T* FE 35/2.8 on the Sony a7r at 4454 lw/ph.

I definitely stand by my statement. 4800 lw/ph (Pentax-M 50/1.7 on film) is quite competitive by current standards. It is tempting to speculate that the same lens would do somewhat less well on the best of today's FF sensors, but it would not be the fault of the lens.


Steve

---------- Post added 06-04-14 at 03:05 PM ----------

Thanks, edited.

BTW...I would like to see the test results where somewhat got 225 lp/mm on any medium. That translates to 10800 lw/ph on a FF sensor (225 lp * 24mm * 2). Woo! Hoo!


Steve

---

Last edited by stevebrot; 06-04-2014 at 03:17 PM.

[ElJamoquio]

I did a quick survey of reviews at photozone.de before I posted my previous comment and did not find any with higher than 4800 lw/ph. Perhaps you could provide an example of a few lens/camera combinations that are significantly higher?

I was agreeing with you.

BTW...I would like to see the test results where somewhat got 225 lp/mm on any medium. That translates to 10800 lw/ph on a FF sensor (225 lp * 24mm * 2). Woo! Hoo!

Sorry, meant 225 lw/mm.

If you're looking for the best of today (at Photozone) to compare to the best of yesterday, I'd look at the Sony E-mount page. It's only APS-C but it's a 24MP APS-C, which is among the best you're going to get right now.

---------- Post added 06-04-14 at 03:11 PM ----------

Interesting (if irrelevant) factoid - photozone has the Q + kit 8.5mm spec'd to 403 lw/mm.

[stevebrot]

I was agreeing with you.

That is good. I like it when people agree with me.

Seriously though, there are a lot of generalities that get thrown around regarding vintage vs. current lenses. Usually the discussion winds down to thoughts on the superiority of modern coatings and tendency to display purple fringing. The coatings are undoubtedly better on the newer lenses, though whether there is field significance is debatable. As for PF, most claims are not backed up by actual example photos. I own quite a few vintage lenses and I don't know that any of them are particularly prone to PF. I also own the FA 35/2 and FA 77/1.8*, both of which are supposed to be PF demons. I have specifically gone out on occasion with the intent of taking PF photos and have been unable to do so with either lens. I guess my region just simply lacks the light intensity or spectral balance to cause PF.


Steve

* I know that some folk consider these two models to be "vintage", but I am hesitant to assign that label to any product currently in production.

---------- Post added 06-04-14 at 03:37 PM ----------

I was thinking about the lenses on my shelf and trying to determine if any would be eligible for a short list of "not recommended" for the K-3. Here is the short list based on my collection:

• Auto Rikenon 55/1.4: This is a nice lens, but it has a very wide flat rear element that actually reflects back off the sensor creating a low-contrast hot spot in some light and blue circle artifact in others. This is sad.

Ummm...that is indeed a very short list and I have a lot of lenses. There are a few that I might want to add simply because they were not that great on film or my K10D, but since I haven't actually used them on the K-3, I thought it best to not subject them to humiliation. (Hint: both are in my signature and both are 200mm focal length)

Steve

[ElJamoquio]

I guess my region just simply lacks the light intensity or spectral balance to cause PF.

Maybe.

[monochrome]

To the OP:

In very good outdoor light (where it performs best) my DA55~300 is better on my K-3 than on my K-01 (K-30, Jr.) and different than on my K10D. Anyone else can tell me the 55~300 is an average quality, consumer zoom with merely average apertures and construction quality, and they'd be quite correct. But at higher f/ stops and higher shutter speeds (mid-day sunlight sports action @ 200mm) additional resolution is overkill for my needs. Frame rate, AF improvements, buffer speed, dual cards - all those benefits of the K-3 are worth the investment for that specific use and the lens does the job well enough that it needn't be replaced.

If OTOH I really wanted the resolution for cropping birds and I was going to add the 1.4x TC - yeah, that's not worth it. In that case you'd be better off with a DA*60~250 or DA*300 and if necessary the additional TC (normhead here has lots of shots to demonstrate the 1.4x and 60~250).

Your decision should be based on your intended uses, your preferences and your needs.

[Imageman]

Steve, id like to know what you think we disagree on because so far I agree with everything you say, I think we are of the same opinion with regards to lenses and recording medium whether that be film or sensor.


What I do have a problem with is that test you quote performed by Modern Photography. They tested lens resolution on colour film stock.


The information I have shows colour film to be limited to around 100 to 110 lppmm. these lpmm resolution readings were always an interpretation and one tester would consistently assess resolutions differently to another. I have done my share of lens testing and I know the issues and difficulties faced.


You have to set up a bench rig, set up a chart and choose the film carefully (this was film days) take half a dozen separate images focussed as accurately as you can, and then under the microscope you examine the negatives directly, and you take the best result from those 6 images.


I have seen the same lenses tested by different testers and the results have varied between testers by as much as 30%


It is revealing that in their test results, all the 5 lenses if I am reading what you say correctly, vary only by 2% and are between 100 to 102 lppmm this is highly unusual. A selection of lenses can be expected to vary by more than 10% yet these results didn't vary as expected, the difference between 100 and 102 in testing terms is trivial and essentially the same result.


One of the lenses you quote, the Leitz Summicron 50, that they achieved 102 lppmm I have seen other testers publish resolution figures for this very lens that vary anywhere between 50 lppmm and 160 lppmm Whats going on here, why such a monumental variance. Do these testers not know how to test lenses or is something else at work.


My interpretation is that the difference can only be the recording medium they are using. The 102 lppmm that Modern Photography see from this lens is so close to colour film stock ultimate resolution that it is I believe the other thing at work and the limiting factor in these tests.


How can a lens deliver more resolution than the film stock is capable of. This is the very point I have been making all along.


If they wanted to test the resolution of a lens beyond 100 lppmm they should have used film stock that can resolve well beyond 100 lppmm and that means Monochrome film not colour and not just any monochrome film high resolution monochrome film Is called for. They chose colour film and I regard their results therefore as dubious.


I would not use colour film in my testing Fujicolor Superia 100 color negative film resolves around 63 lppmm and is therefore useless for testing. I routinely used monochrome film that would resolve beyond 110 lppmm as the lenses I tested all resolved below this figure had any or the lenses resolved close to 110 lppmm I would have had to move to higher quality film.


I am reminded of a set of published results by a respected source similar to Modern Photography but in the digital age. They tested a large number of lenses, and none of them was able to resolve beyond 58 lppmm. all lenses maxed out at 58 lppmm, something fishy I thought, considering my own testing in the 1970s revealed standard cheap kit lenses delivering 80 lppmm.


Why were all the lenses tested limited to 58 lppmm, I was expcting to see some lens reach 100 lppmm and yet none in these tests could get beyond 58 lppmm it was a puzzle.


I looked closer, and it turned out that the digital cameras they had used for the tests had sensors that resolved 60 lppmm. This was the answer and the limiting factor.


The limiting factor in the resolution of a lens is always going to be the recording medium.


Take 2 lenses, a Nikon Ultradoodah that can resolve 120 lppmm and a Japanese something or other cheapo with Nikon mount that is only capable of 50 lppmm, put them both on a Nikon camera that has a sensor that can only resolve 40 lppmm, test them as many times as you like, but both lenses will deliver identical results. No awards for predicting what thy are. Just 40 lppmm for both lenses A lens cannot resolve beyond the resolving power of the recording medium.


Back to k3. The sensor on this resolves 120 lppmm. I would expect therefore that any lens generally available will do well on this camera and reach its full performance. And since this resolution exceeds colour film resolution. I expect lenses to deliver images on k3 which are identical to those they delivered on colour film stock. Exactly what you are seeing.

[ElJamoquio]

Anybody have an ISO chart shot done with color film, scanned at high resolution?

[stevebrot]

Maybe.

Yep, that would be classic PF.


Steve

---------- Post added 06-04-14 at 08:39 PM ----------

They tested lens resolution on colour film stock.

Huh? According to the article,* films tested:

• Panatomic-X (classic low ASA, fine grain, B&W)
• Technical Pan (Microfilm stock confectioned for camera use, B&W)
• Hi Contrast Copy (What it says, also B&W)
• Kodachrome II (The historic standard for fine grain, high detail color reversal)
• Micro-Ektachrome (Color reversal film used for photomicrography)

At the time the article was written any of the above would have been on the short list of go-to films for high detail work. At least that was the stated rational in the article regarding their choice of films. I have personally worked with three of the five and would rank them in the order Tech Pan > KII > Pan-X for resolving power, though Pan-X would be my choice for general pictorial use. The five materials listed were among the best available at the time and probably as good or better than any available today. (Ektar 100 is widely regarded to have resolution on the same order as Kodachrome.)

As for the point of disagreement...I have no interest in a point-by-point rebuttal. Mostly my objection would be the matter of the claimed incredible resolving power of photographic film. 100 lp/mm for lens/film combo measured 30 years ago was and is impressive and is better than current digital technology. Current tabular grain materials are good and can probably come close to 100 lp/mm. None of these are even one order of magnitude better resolution than current digital sensors. They are playing in the same ballpark in terms of that aspect of recording quality when paired with any lens currently available.

In the end, we came to same general conclusion and that is worth something, eh?


Steve

* I was fortunate enough to have read it in print when it was published. It caused quite a stir, primarily because of the tight grouping of the lenses tested. Nobody could believe that brand-X was essentially equivalent to brand-Y and that the Leitz was only a little better than the rest.

[stevebrot]

Anybody have an ISO chart shot done with color film, scanned at high resolution?

Google is our friend. There are a number of tests that have been done over the years, including one that tested a number of films scanned at 8000 dpi as well as the Sony A900. Unfortunately the test was done against a low resolution target.

Technique - Kodak Ektar 100 test - Resolution

Although flawed, the test is still an interesting read in that it readily shows the difference in evaluating film vs. digital captures of a resolution target. The last photo is particularly telling.


Steve

[ElJamoquio]

Google is our friend.

Google might be your friend, but Google searching killed my father.

There are a number of tests that have been done over the years, including one that tested a number of films scanned at 8000 dpi as well as the Sony A900. Unfortunately the test was done against a low resolution target.

Thanks; but if we're comparing to digital sensors, in my mind the preferred method should be slant edge to slant edge. Those converging lines are OK telling the difference between "X" resolution and "2X" resolution, but e.g. some people on the web (imaging resource) have variations of 20-30% in reading their charts compared to slant edges, which makes their conclusions a bit nonsensical sometimes.

On top of that subjective rating, we're then comparing it to a slant edge technique when we use photozone numbers, and an unknown MTF% at that.

That said, subjective-to-subjective, that A900 looks 2x as good as the film examples in that chart to me, and state of the art digital should be ~40% better with higher pixel count and no AA filter.

[stevebrot]

That said, subjective-to-subjective, that A900 looks 2x as good as the film examples in that chart to me

Absolutely. Did you see the last image, the one of the high res scan of the chart itself The A900 image is clean where there is no clean.


Steve

[ElJamoquio]

Absolutely. Did you see the last image, the one of the high res scan of the chart itself The A900 image is clean where there is no clean.


Steve

Yup, I saw it. Actually considered calculating the test chart area (larger would help quite a bit) and then I decided I'd have another beer instead.

It'd be interesting to see a test with today's no-aa, 36MP sensors at ISO 100 and ISO 800.