Digital vs Film (Page 2): :sm24: :sm24: :sm24:...

Tronjo Loc: Canada, BC

burkphoto wrote:

I helped guide a pro portrait lab through the tran... (show quote)

TriX Loc: Raleigh, NC

Strodav wrote:

How about hearing from a semi-professional photogr... (show quote)

I would only add that I don’t think there’s any 35mm DSLR that approaches 16 stops of DR - 12 max is what I would expect. ( http://www.photonstophotos.net/Charts/PDR.htm ) There’s no question (I think) that DSLRs have higher ISO capability for reasonable noise/grain. As far as resolution, I’d refer you to Scotty’s (selsmie) recent in-depth analysis of resolution of both media.

therwol Loc: USA

Sshlitz wrote:

I am (was) a film photographer. I've used Nikon F4... (show quote)

I was going to suggest borrowing or renting what you're thinking of buying to come to your own conclusions, but you'll also need some software on your computer to do anything meaningful with the images.

I'm not a professional, but in my opinion, you can do so much more with digital than film that it's no contest. There is also a quantum leap in IQ between a typical point and shoot with a tiny sensor and a full frame DSLR. (I have a Nikon D810 and can see it plainly.)

Strodav Loc: Houston, Tx

TriX wrote:

I would only add that I don’t think there’s any 35mm DSLR that approaches 16 stops of DR - 12 max is what I would expect. ( http://www.photonstophotos.net/Charts/PDR.htm ) There’s no question (I think) that DSLRs have higher ISO capability for reasonable noise/grain. As far as resolution, I’d refer you to Scotty’s (selsmie) recent in-depth analysis of resolution of both media.

https://petapixel.com/2015/05/26/film-vs-digital-a-comparison-of-the-advantages-and-disadvantages/
On dynamic range: "A release by Kodak showcased that most film has around 13 stops of dynamic range. Today’s modern digital cameras all average around 14 stops of dynamic range, with high-end units such as the Nikon D810 reaching almost 15 stops. Film continuous to deliver incredible dynamic range, but today’s digital technology can easy match it."

On Resolution: Just as different sensors produce different resolutions, different types of film will also produce different resolutions. Roger N. Clark’s analysis of standard 35mm film showcased that depending on the type of film used, the resolution fell between 4 and 16 million pixels. For example, Clark’s study noted that Fujifilm’s Provia 100 film produced a resolution around 7 MP while Fujifilm’s Velvia 50 produced a resolution around 16 MP. Considering that entry cameras such as Nikon’s D3330 produce around 24 MP, 35mm film doesn’t have much of an advantage in this scenario."

Here is another useful reference although I believe it's a bit dated it does show that low ISO fine grain films can have a higher equivilent resolution compared to older DSLRs. http://www.clarkvision.com/articles/film.vs.digital.summary1/index.html

burkphoto Loc: High Point, NC

rpavich wrote:

I think when people ask about the "comparable... (show quote)

Very well stated! You make entirely valid points. Choice of medium is personal. We’re fortunate that folks like you are plentiful enough to support the existence of the vendors who supply materials and equipment.

rpavich Loc: West Virginia

burkphoto wrote:

img src="https://static.uglyhedgehog.com/images/s... (show quote)

Why thank you!

TriX Loc: Raleigh, NC

Bh

Strodav wrote:

https://petapixel.com/2015/05/26/film-vs-digital-a... (show quote)

Different sites measure DR differently, but I have found the Photons to photos site to be very reliable and accurate. here’s the DR chart for one of the best 35mm DSLRs in that respect - the Nikon D850 - just under 12 stops at base ISO. and here’s the link to Scotty’s discussion on the subject of film and DSLR resolution: https://www.uglyhedgehog.com/t-545214-1.html.

Here are a couple of other references, but remember this: the best case DR, assuming no noise and the MSB (most significant bit) of the A/D just set on the brightest pixel, is determined and limited by the A/D, and the A/D of the best DSLRs is 14 bits (stops), so if everything were perfect and there was no noise limiting the low end, the DR would be 14 bits. In reality, noise (and digitization error which is typically + or - 1 LSB) limits the low end, so the total DR will always be less than the max resolution of the A/D.

https://www.camerastuffreview.com/en/camera-guide/review-dynamic-range-of-60-camera-s
https://www.cambridgeincolour.com/tutorials/dynamic-range.htm

Strodav Loc: Houston, Tx

TriX wrote:

Bh ... Different sites measure DR differently, but I have found the Photons to photos

The Dynamic Range, i.e., the range of light that can be seen by the camera is a property of the sensor and not the the A/D converter. The A/D should be adjusted for all bits to 1 at photon saturation, i.e. max volts out, to all 0 bits at no photons hitting the sensor. Yes I agree, there is some noise that may randomly flip the last bit. So an 8 bit A/D will have more course steps (256 shades) than a 14 bit A/D (16384 shades), but all 1s and all 0s will mean the exact same thing whether its 8 bits or 14 bits. So, when I see DxOMark rating the sensor of the D850 at 14.8EV, I see a dynamic range of 28,500 to 1 with fairly fine steps at 14 bits, where Kodak Vision3 film is 8000 to 1.

On resolution, it is more limited by the lens than the camera, but fine grain film may be capable of 150 line pairs / mm, but an FX camera may be capable of about 75 line pairs / mm, but even good lenses start to loose contrast at 30 lines / mm.

Apaflo Loc: Anchorage, Alaska

Strodav wrote:

The Dynamic Range, i.e., the range of light that c... (show quote)

In fact the A/D converter sets the upper limit of Dynamic Range.

When the maximum analog sensor output voltage is set to all 1's you have no option as to what the voltage is that produces all 0's. That voltage is purely a function of how many bits of resolution the A/D converter has. It will be a voltage half again lower for each added bit.

Hence a 9 bit A/D will resolve as all 0's a voltage that is 1/2 the voltage where all bits are 0 with an 8 bit A/D. A 14 bit A/D will not be all 0's at the same analog voltage as an 8 bit A/D, but will be all 0's only when the voltage has dropped to 1/64 that voltage (and in 6 steps).

It needs to also be pointed out that values for maximum Dynamic Range can only be compared if they are defined in exactly the same way. The figures quoted from the Photos to Photons site are not the "engineering" dynamic range used by some sites, but rather are "Photographic" dynamic range, and will be slightly lower values for the same camera.

Apaflo Loc: Anchorage, Alaska

Strodav wrote:

... On resolution, it is more limited by the lens than the camera, but fine grain film may be capable of 150 line pairs / mm, but an FX camera may be capable of about 75 line pairs / mm, but even good lenses start to loose contrast at 30 lines / mm.

Most lenses, even older ones, will have greater resolution than you suggest.

Current FX sensors are most often greater than 75 lp/mm, currently running about 102 lp/mm. Current APS-C sensors are about 128 lp/mm.

The system resolution is a combination of the lens and the sensor and it is incorrect to say either is limiting the total resolution. The system resolution using the same lens will always be higher with a higher resolution sensor.

selmslie Loc: Fernandina Beach, FL, USA

Apaflo wrote:

Most lenses, even older ones, will have greater resolution than you suggest.

No, they won't. The very best FX lens tested barely reaches 3200 lines per picture height at the center. That woks out to about (3200/24)/2=66.67 lp/mm. Most FX lenses have a good deal less resolution, especially as you move away from the center of the image.

Apaflo wrote:

Current FX sensors are most often greater than 75 lp/mm, currently running about 102 lp/mm. Current APS-C sensors are about 128 lp/mm.

Even a Leica Monochrom sensor cannot record a separate target line for each row of pixels. That would happen only if the target's pattern conveniently lined up with the rows of sensels. Since that can't happen, even with a Leica Monochrom, the actual resolution is somewhat less.

You are also overlooking the Bayer array. No matter how much resolution the sensor might produce without the Bayer array, the physical resolution is always about 1/2 what it would be without it.

The fact that the resulting RGB image has nearly the same number of pixels as in the original raw data is due to the way the image is assembled from the raw data. It takes at least four raw values to interpolate each RGB value. The value recorded at each raw pixel is used to create at least four adjacent RGB pixels.

Apaflo wrote:

The system resolution is a combination of the lens and the sensor and it is incorrect to say either is limiting the total resolution. ...

When you work out the numbers you will see that both the sensor and the lens define an upper limit on resolution:

1/Rt^2 = 1/Rs^2 = 1/Rl^2 where Rt=total resolution, Rs=sensor resolution and Rl=lens resolution.

There is no way that Rt can be greater than either Rl or Rs.

The lower resolution sets the limit of resolution, even if the other resolution could be infinite. And if both are the same, for example 50 lp/mm, the combined resolution will be [Rl or Rs]/SQRT(2) or 50/1.414=35.36 lp/mm.

Apaflo Loc: Anchorage, Alaska

selmslie wrote:

... [noise deleted] ...

So the first thing you do is demonstrate that what I said about lens resolution is true, and yet you then deny that 66 lp/mm is clearly much greater than 30 lp/mm.

And the rest of your article is equally confused between accurate numbers and formulas that are then not used correctly.

kernowp

FF digital surpasses 35mm film but larger film formats may still be superior at optimal iso. Newer digital models like Nikon D810 and D850 outperform any film at high iso. Processing, storage etc all better with digital.

cameraf4 Loc: Delaware

Sshlitz wrote:

I am (was) a film photographer. I've used Nikon F4... (show quote)

S, I was in your shoes a few years back. I grieved big time when the writing on the wall about digital became clear to me. I loved my F4 and my Nikon FM3. I only shot fine grain, low RMS slide film and used the best photo techniques to make the sharpest, highest detailed images possible. The 6MP DSLRs at that time couldn't hold a candle to Fuji Velvia or Kodachrome. And the Nikons were a joke. Yes, I could use my trusty Nikkor lenses on them, but with an APS-C size sensor, my landscape images really suffered.

That was then. This is now. The Nikon D800-series cameras are capable of much more image detail than even Kodachrome 25 was. The Dynamic Range at ISO 100 (or now, ISO 64) is just amazing. Yes, prices are certainly higher than our old F4S used to sell for new. But if you take the plunge, you may (like me) rediscover a love for the craft. And if you managed to save any of your older Nikkor glass, these new Nikons really make it shine.

selmslie Loc: Fernandina Beach, FL, USA

Apaflo wrote:

So the first thing you do is demonstrate that what I said about lens resolution is true, and yet you then deny that 66 lp/mm is clearly much greater than 30 lp/mm.

And the rest of your article is equally confused between accurate numbers and formulas that are then not used correctly.

Most FX lenses, especially your zoom lenses, have resolutions between 30 and 50 lp/mm throughout but only peak near the center.

And the rest of my post went over your head because you don’t understand basic algebra - or much else, for that matter.