Teleconverter or Gigapixel?

tomhongkong said:

.........
Is 1 aperture stop loss worth the image degradation from upsizing.?...

........

Click to expand...

tomhongkong said:

OK, I am not necessarily advocating Gigapixel but it is pretty popular upsizing software. There are others.

However this is an interesting question. Assuming you had a 300f4 for example, and the bird image from 300mm is too small (low res) after cropping, are you better off using a TC or to upsize? Let's assume a 1.4xTC, the most commonly used size. (Bigger magnifications run into problems with both approaches)

The TC loses 1 stop of light. Gigapixel introduces artifacts as a result of upsizing (but not much with just a 1.4x upsize) Often I am balancing iso and shutter speed and 1 stop makes quite a difference.

Click to expand...

tomhongkong said:

Is 1 aperture stop loss worth the image degradation from upsizing.?... that's iso 1600 to 3200, or 3200 to 6400. DXO is getting pretty near the limit! How much does the 1.4x TC also introduce degradation? perceived wisdom is that the 2x TC does do so quite a lot.

What is your experience? Any samples?

Full disclosure: I am trying to decide if an MC14 is really an advantage on my 300, (or even an MC20!) or if I am better off doing it with software

tom

Click to expand...

drj3 said:

tomhongkong said:

OK, I am not necessarily advocating Gigapixel but it is pretty popular upsizing software. There are others.

However this is an interesting question. Assuming you had a 300f4 for example, and the bird image from 300mm is too small (low res) after cropping, are you better off using a TC or to upsize? Let's assume a 1.4xTC, the most commonly used size. (Bigger magnifications run into problems with both approaches)

The TC loses 1 stop of light. Gigapixel introduces artifacts as a result of upsizing (but not much with just a 1.4x upsize) Often I am balancing iso and shutter speed and 1 stop makes quite a difference.

Click to expand...

You do not lose 1 stop of light. You have exactly the same amount of light on the target with the TC as you have with the lens without TC shot at one stop lower ISO setting. You have the same amount of light on the TC spread over more pixels. The cropped image without the TC loses that extra stop of light when you crop it to the same target size. You cropped away that stop of light.

Click to expand...

drj3 said:

You lose one stop of light, only if you can move closer without the TC so the target occupies the same amount area of the frame as in the TC image.

A good TC will always give a better image at the same distance than a bare lens.

A good 2X TC should give a better image than a 1.4X TC which should give a better image than the lens without a TC.

The TC images will always be poorer than the lens without a TC when shot at different distances so that each gives the same target size in the frame where you do lose stops of light with the TCs compared to lens without a TC.

drj3 said:

Is 1 aperture stop loss worth the image degradation from upsizing.?... that's iso 1600 to 3200, or 3200 to 6400. DXO is getting pretty near the limit! How much does the 1.4x TC also introduce degradation? perceived wisdom is that the 2x TC does do so quite a lot.

What is your experience? Any samples?

Full disclosure: I am trying to decide if an MC14 is really an advantage on my 300, (or even an MC20!) or if I am better off doing it with software

tom

Click to expand...

Click to expand...

cba_melbourne said:

drj3 said:

tomhongkong said:

OK, I am not necessarily advocating Gigapixel but it is pretty popular upsizing software. There are others.

However this is an interesting question. Assuming you had a 300f4 for example, and the bird image from 300mm is too small (low res) after cropping, are you better off using a TC or to upsize? Let's assume a 1.4xTC, the most commonly used size. (Bigger magnifications run into problems with both approaches)

The TC loses 1 stop of light. Gigapixel introduces artifacts as a result of upsizing (but not much with just a 1.4x upsize) Often I am balancing iso and shutter speed and 1 stop makes quite a difference.

Click to expand...

You do not lose 1 stop of light. You have exactly the same amount of light on the target with the TC as you have with the lens without TC shot at one stop lower ISO setting. You have the same amount of light on the TC spread over more pixels. The cropped image without the TC loses that extra stop of light when you crop it to the same target size. You cropped away that stop of light.

Click to expand...

This is not correct. With a TC, each sensor pixel covers a smaller area of the FOV and hence collects less light.

Click to expand...

cba_melbourne said:

The TC trades light for a higher resolution. A digital zoom trades resolution for keeping the light constant.

Click to expand...

cba_melbourne said:

cba_melbourne said:

You lose one stop of light, only if you can move closer without the TC so the target occupies the same amount area of the frame as in the TC image.

A good TC will always give a better image at the same distance than a bare lens.

A good 2X TC should give a better image than a 1.4X TC which should give a better image than the lens without a TC.

The TC images will always be poorer than the lens without a TC when shot at different distances so that each gives the same target size in the frame where you do lose stops of light with the TCs compared to lens without a TC.

cba_melbourne said:

Is 1 aperture stop loss worth the image degradation from upsizing.?... that's iso 1600 to 3200, or 3200 to 6400. DXO is getting pretty near the limit! How much does the 1.4x TC also introduce degradation? perceived wisdom is that the 2x TC does do so quite a lot.

What is your experience? Any samples?

Full disclosure: I am trying to decide if an MC14 is really an advantage on my 300, (or even an MC20!) or if I am better off doing it with software

tom

Click to expand...

Click to expand...

Click to expand...

tomhongkong said:

OK, I am not necessarily advocating Gigapixel but it is pretty popular upsizing software. There are others.

However this is an interesting question. Assuming you had a 300f4 for example, and the bird image from 300mm is too small (low res) after cropping, are you better off using a TC or to upsize? Let's assume a 1.4xTC, the most commonly used size. (Bigger magnifications run into problems with both approaches)

The TC loses 1 stop of light. Gigapixel introduces artifacts as a result of upsizing (but not much with just a 1.4x upsize) Often I am balancing iso and shutter speed and 1 stop makes quite a difference.

Is 1 aperture stop loss worth the image degradation from upsizing.?... that's iso 1600 to 3200, or 3200 to 6400. DXO is getting pretty near the limit! How much does the 1.4x TC also introduce degradation? perceived wisdom is that the 2x TC does do so quite a lot.

What is your experience? Any samples?

Full disclosure: I am trying to decide if an MC14 is really an advantage on my 300, (or even an MC20!) or if I am better off doing it with software

tom

Click to expand...

Skeeterbytes said:

Have yet to see a 300 Pro uprez vs. MC14 or MC20 comparison where the teleconverter was not better. Can't, however, cite all the software used in these tests.

The 300+MC20 combo I find very challenging to handhold, while the MC14 is still a reasonable magnification for me to wrangle, so that's perhaps a side consideration. It's gratifying that both work so well with the lens (see Marko Finland and others here for excellent samples).

If you already own the software, then experiment before spending more on the TC. Maybe it will be good enough for your purposes. F:8 I find pretty slow for light less than full daylight. 5.6 is still okay into the evening.

Cheers,

Rick

Click to expand...

lokatz said:

Have yet to see a 300 Pro uprez vs. MC14 or MC20 comparison where the teleconverter had an edge. Most of the time, it is a dead race.

This assumes that the upscaling is done properly. Adobe used to do a mediocre job here, though Photoshop recently learned how to do it properly. Gigapixel is hard to beat.

The problem is that most people tend to over-sharpen along the way, which makes the upscaled image look unnatural and the one with the TC better. Doing it properly, however, renders the TC nothing more than a convenience feature.

I've tested this on Nikon with a 1.4x and 2x, on Canon with a 1.4x, and on Oly/OMDS with a 1.4x. I keep reading that the Oly 2x is far worse than the excellent 1.4x, but did not test this myself.

There is a limit to what I am saying, though it is hard to quantify: if you crop A LOT, the TC starts having an edge. The explanation for that is simple: the fewer pixels there are, the more upscaling becomes a guessing game.

Click to expand...

tomhongkong said:

That's what I suspect, as a guess looking at the upscaled output, without any hard evidence. I did limit my option to 1.4x,although I guess 2.0 times might be very close.

I hoped I could get an answer without forking out for a TC (which might not give me anything extra)

tom

Click to expand...

tomhongkong said:

Skeeterbytes said:

Have yet to see a 300 Pro uprez vs. MC14 or MC20 comparison where the teleconverter was not better. Can't, however, cite all the software used in these tests.

The 300+MC20 combo I find very challenging to handhold, while the MC14 is still a reasonable magnification for me to wrangle, so that's perhaps a side consideration. It's gratifying that both work so well with the lens (see Marko Finland and others here for excellent samples).

If you already own the software, then experiment before spending more on the TC. Maybe it will be good enough for your purposes. F:8 I find pretty slow for light less than full daylight. 5.6 is still okay into the evening.

Cheers,

Rick

Click to expand...

Thanks for the reply

I have Gigapixel and have tried.....it's not bad, but is a TC better? (always looking for better IQ)

Hopefully someone with both can post actual examples

Click to expand...

lokatz said:

Have yet to see a 300 Pro uprez vs. MC14 or MC20 comparison where the teleconverter had an edge. Most of the time, it is a dead race.

This assumes that the upscaling is done properly. Adobe used to do a mediocre job here, though Photoshop recently learned how to do it properly. Gigapixel is hard to beat.

The problem is that most people tend to over-sharpen along the way, which makes the upscaled image look unnatural and the one with the TC better. Doing it properly, however, renders the TC nothing more than a convenience feature.

I've tested this on Nikon with a 1.4x and 2x, on Canon with a 1.4x, and on Oly/OMDS with a 1.4x. I keep reading that the Oly 2x is far worse than the excellent 1.4x, but did not test this myself.

There is a limit to what I am saying, though it is hard to quantify: if you crop A LOT, the TC starts having an edge. The explanation for that is simple: the fewer pixels there are, the more upscaling becomes a guessing game.

Click to expand...

rreichar said:

I am using an OM-1 with the PL 100-400 and I have been getting great results with Topaz Photo AI which sort of combines their standalone apps into one. Here is an example of a White-crowned Sparrow in terrible light just after sunset last week. ISO 8000. The second pic is after running it through Topaz Photo AI.


Original photo


2X upscale using Topaz Photo AI

Click to expand...

drj3 said:

lokatz said:

Have yet to see a 300 Pro uprez vs. MC14 or MC20 comparison where the teleconverter had an edge. Most of the time, it is a dead race.

This assumes that the upscaling is done properly. Adobe used to do a mediocre job here, though Photoshop recently learned how to do it properly. Gigapixel is hard to beat.

The problem is that most people tend to over-sharpen along the way, which makes the upscaled image look unnatural and the one with the TC better. Doing it properly, however, renders the TC nothing more than a convenience feature.

I've tested this on Nikon with a 1.4x and 2x, on Canon with a 1.4x, and on Oly/OMDS with a 1.4x. I keep reading that the Oly 2x is far worse than the excellent 1.4x, but did not test this myself.

There is a limit to what I am saying, though it is hard to quantify: if you crop A LOT, the TC starts having an edge. The explanation for that is simple: the fewer pixels there are, the more upscaling becomes a guessing game.

Click to expand...

It is actually very easy to show. If you have the TCs, just shoot a very detailed image with color, with both at the same distance with the TC at 1 stop (1.4XTC) or 2 stops (2XTC) higher ISO. Then downsize the TC image target to the same size as the lens without the TC. The TC image should show better detail and much better color than the image without a TC.

Below is one example of a shot of my monitor where you can see the detail and color of the monitor pixels. I shot these at ISO of 200-1600 with the 300mm f4 and at ISO of 800-6400 with the MC20 and the TC images were always better. Four times the number of pixels with 1/4 the light produce a better image than the four times the amount of light on 1/4 the number of pixels.



Below is an example of shooting both at ISO 200, so that the MC20 target image on the left has 4 times the amount of light as the 300mm image.

Click to expand...

lokatz said:

Your examples, as well as the other one a poster shared, demonstrate perfectly that one cannot prove anything with examples.

Specifically, you don't even say which software you used for upscaling, the non-TC examples are in dire need of sharpening, which was obviously not done or not done correctly, whereas the MC20 shots beg for CA removal. Which looks better after doing the job right? Impossible to tell from these shots.

On top of that, shooting your monitor is useless for where you really want to use enlargement, namely when the bird is far away. Lens performance at close distances vs long distances isn't the same thing.

Again, I've tested this numerous times, with various lenses, TCs and software, but always with the same (distant) target and shooting position. Before Photoshop learned how to get it right, I followed the proper upscaling approach I learned in a course many years ago (reduce noise, upscale by no more than 25%, sharpen, then repeat the process until you're done). The results remained about even.

I still sometimes use a TC, but for convenience, not for better results.

Click to expand...

Skeeterbytes said:

This is where you can show your work to support your assertion.

Click to expand...

lokatz said:

Skeeterbytes said:

This is where you can show your work to support your assertion.

Click to expand...

Nah. I'm not storing these kinds of test shots.

And why should I? To prove what I know to someone who believes examples can prove anything? Science recognized centuries ago that they can't and won't.

Click to expand...