Megapixel Equivalence

EASmithV · Jan 28, 2009

Hey a friend was joking and said that his new DSLR could beat mine. I then told him that I could pull out my 4x5 and tell him to beat 600+ megapixels...

Which raises the question- Was my random estimate accurate? Is there a formula for calculating a rough megapixel equivalent? I know that the nature of film is such that it is impossible to have a "Megapixel", but how do you compare film to digital in layman's terms?

I'm specifically curious as to TXP320, TMX100 and Velvia 100 as those are my favorite films.

domaz · Jan 28, 2009

600 megapixels is probably a bit optimistic even for 4x5, a quick calculation yields that scanned at 3000 dpi is 171 megapixels. But still there is no way a DSLR can beat 4x5 quality. At least not any DSLR he can afford.

keithwms · Jan 28, 2009

Here's a very rough rule of thumb. Take the lens resolution in lines/mm, square it, and multiply by the area of your film. E.g., 35mm film is ~24x36 mm = 863 mm^2, assuming ~100 lp/mm resolution, this gives you roughly 17 mp. (Don't get bent out of shape about factors of two, this is just a rough estimate.) For 4x5, you're looking at a film area of ~12903 mm^2, and assuming a lens resolution of ~80 lp/mm across the frame, this corresponds to 165 mp.

Technically, you'd have to multiply by film MTF and take into account diffraction etc., but the above approximation gets you pretty close.

And yes, there is grain beyond what the lens resolves, and that is very important to some people. (Do I have enough disclaimers in place now? We'll see...)

Photo Engineer · Jan 28, 2009

There is a way to calculate it by using about 300 mg of silver per square foot of film. Translate that into moles by divinding by 108,000 (approximate mg /mole) then to atoms by multiplying by 6.023 x 10^23 and then to sensitivity centers by dividing by 3. This will give you the maximum number of centers per square foot of film.

Now, of course, it is much lower than this due to turbidity, the grain itself, and a number of other factors including the development ampification factor, but you can guesstimate that by dividing this huge number by 10 and 100 respectively to get the possible range for 10 atoms / silver clump to 100 grains / silver clump. That should put you in the ball park. Even if you estimate by dividing by 1000 grains / clump then you will still have a huge number in a sheet of film.

PE

sanking · Jan 28, 2009

I don't think they like us to compare film to digital on this forum, except maybe in the Lounge or Soapbox.

However, I will answer your question as best I can. 600 mp is way more than what you can get from scanned 4X5 film. I think that would require film resolution on the order of 240 l/mm, which is basically impossible. A more realistic figure would be 100-200 mp. 150 mp from 4X5 would represent extremely refined technique with very high quality camera and lens and a very fine grain film.

The best approach is never to try to compare mp in terms of film and digital. These arguments always end unhappily for all concerned. You would have a better chance of success serving as a intermediary between Hamas and Israel.

Sandy King

EASmithV said:
Hey a friend was joking and said that his new DSLR could beat mine. I then told him that I could pull out my 4x5 and tell him to beat 600+ megapixels...

Which raises the question- Was my random estimate accurate? Is there a formula for calculating a rough megapixel equivalent? I know that the nature of film is such that it is impossible to have a "Megapixel", but how do you compare film to digital in layman's terms?

I'm specifically curious as to TXP320, TMX100 and Velvia 100 as those are my favorite films.

keithwms · Jan 28, 2009

Haha Ron

P.S. Haha Sandy

Oh dear, even on the hybrid forum, I do recall that this topic took us on quite a lengthy tour.

Poisson Du Jour · Jan 28, 2009

I agree they don't like digital vs film comparisons on this forum and frankly I am tired of digital fanatics.

So, just a few quick words. Don't lose sleep over digital. A recent sub-convert to 4x5 and 3-year digital user myself, the quality of the former is far, far superior to digital.

People enamoured with digital often overlook one critical thing: the lens on the front of the camera. What a pity so many are so carried away with camera bodies.

Anyway, here is the formula for 35mm that I've kept in my Little Red Book. As you can imagine, the results would be significantly higher for 4x5.

36 * 24 * (85 * 2)2 = 24.96 million (for 85lp/mm)

Note that the obscenely hyped and over-priced Nikon D3X has a 24.5mpx sensor, so compared to 35mm, the gap is indeed closing but the price will be a profound disincentive to the masses. A camera costing many thousands is unlikely to stir the blood when the same mpx equivalent can be had much cheaper in a film camera! C'est la vie.

JBrunner · Jan 28, 2009

Also, with all due respect to Sandy and the others and their calculations, it is a bit silly. The technologies don't work the same way, and the variables approach the infinite. People like nice quantifiable things, and so the tendency to try and equate one thing into another. There are so many ways to measure a picture, and for my money Mp is the most worthless, simply because it is the most pedestrian of arguments, clung to by magic bullet chasers, gear heads and other non hackers. I'm gonna laugh myself silly when the affordable 45Mp cameras are out and the people with the bulging shorts who are using them still make crappy images. The supreme measure of a photograph isn't expressed in numbers.

Photo Engineer · Jan 28, 2009

Keith, I thought you might like that! It actually is proper and better than using resolution as in the other posts. After all, resolution is not similar, or related to the number of megapixels, but the figure I used is. Even so, it is a trivial answer.

My point is that silver halide imaging is on the atomic level with several Angstroms as the "yardstick" and Digital is on the Micron level with about 8 - 10 microns as the "yardstick". If you include color, muliply the "yardstick" by 3 in each case.

PE

Poisson Du Jour · Jan 28, 2009

JBrunner said:
Also, with all due respect to Sandy and the others and their calculations, it is a bit silly. The technologies don't work the same way. People like nice quantifiable things, and so the tendency to try and equate one thing into another. There are so many ways to measure a picture, and for my money Mp is the most worthless, simply because it is the most pedestrian of arguments, clung to by magic bullet chasers, gear heads and other non hackers. The supreme measure of a photograph isn't expressed in numbers.

Amen.

nickandre · Jan 28, 2009

Scroll up to the top of the page and look at some quote by flotsam.

keithwms · Jan 28, 2009

Jason, of course, I agree with you in spirit. Yet I do think we sometimes need rough translations, and this is useful info when thinking about the future of that whole market.

Anyway, I am still chuckling over Ron's answer

sanking · Jan 28, 2009

I agree with you. For many people image quality is directly tied to income and they want/need to understand the relationship of digital mp to film dp. Obviously it is an extremely complicated subject but that does not mean that it can not and should not be addressed seriously. At the same time, APUG is definitely not the place where it can be addressed with any degree of seriousness, which is the reason for my flippant response.

As you note, even on the hybrid forum the very hint of digital versus film stirs up misunderstandings.

Sandy King

keithwms said:
Jason, of course, I agree with you in spirit. Yet I do think we sometimes need rough translations, and this is useful info when thinking about the future of that whole market.

Anyway, I am still chuckling over Ron's answer

Aurum · Jan 28, 2009

Poisson Du Jour said:
People enamoured with digital often overlook one critical thing: the lens on the front of the camera. What a pity so many are so carried away with camera bodies.

I would agree. The results for compact digitals I have used in the past bear this up.

I have an old Olympus Digital that I picked up from fleabay for a song. 1.6MP IIRC. The pictures it produces are much higher quality than a 5MP cheap brand version the wife bought from Aldi

Unfortunately it is only recently that digital manufacturers in the compact end of the market have realised that the size of the sensor, is only one part of the equation and that if you have a cheap disposible camera lens, thats the results you will get

Photo Engineer · Jan 28, 2009

Keith;

That is my way of being 100% right and 100% wrong at the same time. What do you say about that?

PE

sanking · Jan 28, 2009

Choice of lens is equally important for film photography. I still see many people with 35mm Nikon and Caon SLRs shooting exclusively with zoom lenses. With rare exceptions zoom lenses for 35mm can not touch high quality prime lenses.

Sandy King

Aurum said:
I would agree. The results for compact digitals I have used in the past bear this up.

I have an old Olympus Digital that I picked up from fleabay for a song. 1.6MP IIRC. The pictures it produces are much higher quality than a 5MP cheap brand version the wife bought from Aldi

Unfortunately it is only recently that digital manufacturers in the compact end of the market have realised that the size of the sensor, is only one part of the equation and that if you have a cheap disposible camera lens, thats the results you will get

keithwms · Jan 28, 2009

Photo Engineer said:
Keith;

That is my way of being 100% right and 100% wrong at the same time. What do you say about that?

PE

There is something appealing to that logic, I must admit!

Michael W · Jan 28, 2009

To the OP - why don't you & your friend go to some place & photograph the same scenes or objects with your respective cameras. Keep ISOs the same. Then print & compare. Seems a more practical & simple way to answer your question.

haryanto · Jan 29, 2009

even thousand gigabyte camera digital cannot make tone that i love from silverprint

Tom Kershaw · Jan 29, 2009

What is a thousand gigabyte camera?

Tom.

haryanto · Jan 29, 2009

i dunno Tom, for me it's not about pixels it's about ink compared to silver

Thomas Bertilsson · Jan 29, 2009

If you love your tools and know how to use them, you can make outstanding photography with either method, or even a hybrid workflow. It is not about the stupid camera.

I too am tired of the pissing contest.

Andy K · Jan 29, 2009

EASmithV said:
Hey a friend was joking and said that his new DSLR could beat mine.

Then he obviously is insecure in his choice of medium. Anyone who has to say their 'thing' is better than your 'thing', is commenting from a position of insecurity.

jim appleyard · Jan 29, 2009

Andy K said:
Then he obviously is insecure in his choice of medium. Anyone who has to say their 'thing' is better than your 'thing', is commenting from a position of insecurity.

OH YEAH! A dear friend has an old K-1000, an old AE-1 and an FM-2 that got old before its time. these 35's all have problems. She also has a fairly new Bronica SQ-A. She has no desire to make her 35 systems compatable or to fix them, no desire to upgrade from her Gossen Pilot meter, and no desire to add lenses. However, she takes great photos! Her composition is so simple it is startling. I don't know how she does it, but she gets many a good photo. She is proof that it is not what you use, but how you use it.

Anon Ymous · Jan 29, 2009

Megapixels vs gigagrains? Who cares? Well, some obviously do, but I'm not one of them. That's especially weird when those who do only print 10x15cm, if they even print. And then there's aesthetics... Will a photo ever be nice just because it's taken with an ultra high resolution sensor/film? Ever seen a nice photo of a test chart? Not me, but I've seen beautiful photographs taken with mobile phones. Yep, I'm not ashamed to admit that

Lastly, if anybody wants to make a resolution estimation of a film photograph, contrast must be taken into account. Reala for example can reach 125 lines/mm, but with 1000:1 chart contrast. Resolution drops to 63l/mm with 6:1 chart contrast. And this leads me to the questions:

1) What's a "normal" (or average) contrast value and how much would the "resolution" be in that case?
2) Can we convert these contrast values to contrast index values? How?

Megapixel Equivalence

Moderator

Privacy & Transparency

Privacy & Transparency