Monochrome converted A7 series - I just discovered this today

Have a look for yourself how good the quality is:

https://www.monochromeimaging.com/technical/resolution/

I've believed for a long while now that digital is better for monochrome, while film is better for colour.

In what way is film better for colour?

You've not been reading that Petapixel BS have you?

There is no trickery here

Except that the regular A7 has an AA filter and most likely the modified A7 does not (This camera is a full spectrum converted camera; meaning that there are no filters over the sensor.). How much of the "resolution gain" is due to the AA filter being removed and how much due to the "bayer filter" being removed?

Joe, I don't read Petapixel. IMO, film colours look better.

Dieter, it is certain that both have an effect. Debayering takes off about 20% of resolution. I don't know about the low-pass filter.

"Dieter, it is certain that both have an effect. Debayering takes off about 20% of resolution. I don't know about the low-pass filter."

- Hmmm. I'm not convinced that removing the RGGB filtering is a good idea. The microlenses would have to go too, since as far as I can tell the filters and microlenses are a single unit.

Also, this wouldn't help resolution unless the entire image processing algorithm was re-programmed. The stepping needed for a Bayer array samples 4 adjacent photosites and combines them into one pixel. Simply removing the colour filtration wouldn't alter that processing and would effectively average the Y component over 4 photosites.

In order to take complete advantage of the Bayer array removal (if that's even possible - which I doubt) the image processing would have to be re-written to remove the RGGB combination algorithm and replace it with individual photosite-by-photosite sampling.

I doubt that the individual that 'converts' these cameras has the necessary skill to completely re-program the camera. Therefore the camera is just made colour-blind without the full advantage of a true monochrome sensor.

My guess is that any improvement in resolving power is simply due to removal of the AA filter, and it would be interesting to see before and after sample images from an already AA-free sensor.

I know little but I am recalling: When the Leica Monochrom came out results from it weren't as much significantly better than Bayer-M9 shots as some people expected them to be according to their theories. - Monochrome conversions seem nice to have for their extra speed.

What are micro lenses doing on a sensor tailored towards retrofocus constructions, like the A7 family's (besides maybe making sure that the elements of a Bayer pattern get hit properly)?

"Have a look for yourself how good ...bless his white cotton socks .Karim.

Wow, incredible, unbelievable....beyond believe.

Just the same as any, did I say any,,,any digital camera.

If anyone is still breathing, in the dusty closets of PN...... they might find the two videos of interest.

Maybe not....but I did hear a creaking chair and a cough...probably just my imagination.

- Hmmm. I'm not convinced that removing the RGGB filtering is a good idea. The microlenses would have to go too, since as far as I can tell the filters and microlenses are a single unit.

You are right that this process is not as easy as we might think. Have a look:

Monochrome2DNG - File Converter for «Monochrome-Converted» Cameras | FastRawViewer

Also, this wouldn't help resolution unless the entire image processing algorithm was re-programmed.

It's not the camera that needs reprogramming - but you should use software that eliminates the debayering step. Of course you can use normal RAW converters, but you will lose some resolution. Applications like Monochrome 2DNG deliberately eliminate debayering when converting a monochrome file to DNG.

The metering will automatically adjust itself (it's a mirrorless camera). However, the ISO value in the metadata will be way off.

My guess is that any improvement in resolving power is simply due to removal of the AA filter, and it would be interesting to see before and after sample images from an already AA-free sensor.

That's not correct. Debayering costs some resolution. The Bayer filter is, as smarter people than I have pointed out, a form of data compression. It is efficient but it does come with a cost. Combine debayering with an AA filter and you lose even more resolution.

The RED Epic (a now superceded cinema camera) is 5K for a reason. A 5K debayer, with an AA filter in the image chain, produces 4K worth of resolution. Those who shoot for a 4K output image shoot a 5K RAW image, then they downsample.

I know little but I am recalling: When the Leica Monochrom came out results from it weren't as much significantly better than Bayer-M9 shots as some people expected them to be according to their theories. - Monochrome conversions seem nice to have for their extra speed.

I cannot comment on the first statement, but the extra speed is a genuine feature. It's up to 3 stops more sensitive, depending on how you judge image quality. If you're conservative, it's a 2 stop advantage. If you like or prefer b&w, you'll take it. The original MM was compared to a Nikon D800E and resolving power was almost the same. Not bad!

What are micro lenses doing on a sensor tailored towards retrofocus constructions, like the A7 family's (besides maybe making sure that the elements of a Bayer pattern get hit properly)?

The flange focal distance is very short. That's all I can think of. They might have added microlenses for the sake of RF lenses. But RF lenses, especially classic ones, did not work well with the first generation of A7 bodies.

Just the same as any, did I say any,,,any digital camera.

Computer says...

no.

- Why she shoots film.

Apparently to needlessly rub her fingers all over it and produce boring pictures of a friend wearing sunglasses.

"What are micro lenses doing on a sensor tailored towards retrofocus constructions, like the A7 family's.."

- The microlenses are doing the same as they do on a Leica sensor; making the photosites more efficient.

"Applications like Monochrome 2DNG deliberately eliminate debayering when converting a monochrome file to DNG."

- There's no mention of any RAW processing except LR in the originally posted article.

I'll have to look into how much RAW pre-processing is done in-camera, but I'm pretty sure that most 'raw' files don't deliver the image data down to individual photosite level.

The only difference I see in the sample images is a small amount of 'smudging' in the roof tiles reflected in the car window. Does this represent a major improvement in resolution? Not in my view. Certainly not worth ruining a perfectly good camera for.

Apparently to needlessly rub her fingers all over it and produce boring pictures of a friend wearing sunglasses.

1. She is probably removing the Remjet. 2. The photos are... ok. Better than a lot of photos on pic-of-the-week threads on PN.

- There's no mention of any RAW processing except LR in the originally posted article.

Link from above (hyphen added because PN doesn't always make it obvious where the URLs are:

-https://www.fastrawviewer.com/Monochrome2DNG

I'll have to look into how much RAW pre-processing is done in-camera, but I'm pretty sure that most 'raw' files don't deliver the image data down to individual photosite level.

They do.

The only difference I see in the sample images is a small amount of 'smudging' in the roof tiles reflected in the car window. Does this represent a major improvement in resolution? Not in my view. Certainly not worth ruining a perfectly good camera for.

The difference is noticeable to me.

What is the photosite : gap ratio on a sensor?

The fill factor (or fill area) is unknown, but the microlenses help compensate for that, apparently. This article is sort of interesting:

-https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5375906/

This second link from Allen is much more interesting. As are the pictures.

The point made is that much more manipulation can be done when converting to monochrome from colour, than is possible with a B&W original. Even carrying a whole swathe of filters around, it's not possible to selectively filter, say, a blue sky and not affect green foliage.

Therefore the pictorial possibilities of working with a colour original far outweigh the tiny resolution gain. Resolution means nothing if nobody gives the pictures a second glance.

"1. She is probably removing the Remjet."

- You only get Remjet on colour negative movie film. And rubbing doesn't do anything to reduce remnant anti-halation dye, which makes no difference to the negative anyway.

The point made is that much more manipulation can be done when converting to monochrome from colour, than is possible with a B&W original. Even carrying a whole swathe of filters around, it's not possible to selectively filter, say, a blue sky and not affect green foliage.

That is true. But it depends which compromises you want. Some people don't even use filters for b&w work. In fact... I don't either!

The girl in the video bulk loaded cinema film (50D, presumably 5203) into recycled cassettes, BTW.

"The girl in the video bulk loaded cinema film (50D, presumably5203) into recycled cassettes, BTW."

- If so, why develop it as black and white? The thicker set of colour layers loses you definition, and the film costs more than B&W. Only reason I can think of for using it, is if you get it for nothing.

And why shoot film only to scan it and end up with a digital image? Scanned, I might add, with a scanner that costs more than a very good digital camera.

A lot of people have said that Portra 400 is their favourite b&w film, but they don't use it because it's so expensive. I would imagine that cinema films might have those qualities, too. Buying it in bulk - especially short ends - makes it cheap. Personally I wouldn't self-develop movie film. Too much trouble.

As for expensive scanners, they're the only choice if you want to bring out as much as you can from the emulsion.