Digital zoom test

[Noa Put]

I did a quick test how the digital zoom performed on the stocklens and my Samyang 85mm, I had to shoot inside through a window as it was bad weather but I thought the view was perfect with all those details.

First of all, you will see on the more extreme zooms that the image is moving a bit and you can see some white smoke appear, this was because there was a chimney just on the left of me that was releasing some heat causing some heatwaves in the image. It was the best I could do for this test but I think it's clear enough.

To explain a bit: In the first part I zoomed in with the stocklens optically and then tried to find the same frame by zooming out optically and then zooming in digitally so I would get the same frame as with optically alone. The result is pretty obvious, just check the roofs where suddenly aliasing appears out of nowhere and also check the softness of tge image and loss of fine detail. Only thing I can conclude is don't use the digital zoom with the stocklens ever.

The second part I filmed with the 85mm Samyang lens and then zoomed in digitally, then I added the stocklens and zoomed in optically to get the same frame to compare, the stocklens and the 85mm lens are optically almost the same, the 85mm is just a tiny bit sharper but from a distance you need to look hard. Here you also see the image softening up with loss of fine detail.

I mirrored the images from both camera's so it's easier to compare.

The results when filming people up close with a prime lens might not look so bad but for high detail objects that are located further away I would not touch the digital zoom.

Sony nex ea50 digital zoom test on Vimeo

[Robert Moran]

Through experience, I've learned that DZ on the zoom will work if carefully done and if the subject matter is appropriate. Seen below is a short clip titled Undercurrent, shot with the kit lens with many shots using the dz to the max as the subject matter was too far away to get anything decent. For me, it has been a godsend as it allows me to do shots not possible by any other means save for buying really expensive telephoto lenses to get the look I want.

Undercurrent on Vimeo

[Noa Put]

I noticed on my quick test with the 85mm that it doesn't look "that" bad compared to the stocklens but I will test more with a 24mm next week so I have a bit wider field of view that I can compare with the stock lens. I do have a 14mm as well but the stock lens starts at 18 so I can't put them exactly side by side.
With the stocklens you normally should never use the digital zoom as it has an optical one but even if you are thinking of using the digital vs the optical zoom with the stocklens I wouldn't recommend doing so based on the test I did.

[Jerome Cloninger]

I think the digital zoom would be fine IF Sony didn't perform the "line skipping" from the sensor and sampled the whole image and downscaled like you would in photoshop or the like. Of course, we wouldn't have the moire nor aliasing issues then either.

I don't understand why Canon & Sony does the line skipping routine. I'm sure there's processing power that is demanded by sampling the entire image, but to the best of my knowledge, Panasonic & Nikon doesn't do the line skipping.

Think about this too... 1/4, 1/3 and 1/2" sensors doesn't do line skipping so why can't they do the same with the APS-C?

[Noa Put]

I don't understand why Canon & Sony does the line skipping routine.

I think they cripple the camera on purpose to protect the sales on their higher end cameras like the fs100

[Jerome Cloninger]

Originally Posted by Noa Put

I think they cripple the camera on purpose to protect the sales on their higher end cameras like the fs100

Probably so!

[Noa Put]

You see that with more manufacturers, like canon with their 5dIII and 6D, the 6d is cheaper, has the same low light performance but suffers a lot more from moire and aliasing, so those that don't want that have to pay extra, just like the ea50, if you want a better image you pay a premium price.

[Ray Lee]

I found a test video on the EA50, at 3:07 they do some digital zoom testing with a couple Canon FD primes

SONY NEX - VG20 VS NEX - EA50 - YouTube

Loss of some image quality for sure but for my eye far less objectionable than the image bobbing around from manual zoom.

[Noa Put]

Thx for that link Ray, too bad we can't download a higher rez file from youtube to compare, the tests he did where from up close and then it doesn't seem that bad, I also experienced this a bit with my 85mm but will be doing some separate tests next weeks but only with some wider primes and might compare sharpness with my ea50 lens using the same frame. You also need to consider there is hardly any fine detail in those digital zoom tests because it is then that the image starts falling apart, if you look close at the guys eyebrows at 04:00 you"ll see artifacts starting to appear, important to know is how that looks on higher detail surfaces or backgrounds and especially in it's native form because a youtube compression will give a wrong impression.

The vg20 looks like a good match to the ea50, the ea50 only needs a more comparable picture profile. If I can spare the cash I might sell one of my dslr's and get a vg30 instead to get 50p on my steadicam, that would be sweet. :)

[Ray Lee]

Agreed about not being able to download better files.... your sample confirmed digital zoom is far from lossless but the sample I found gives me hope than there might be some instances where I can use it to add a bit of interest or movement to some shots when using fast primes.

I am not a techie and just cant wrap my head around why digital zoom would look so bad (like your test and a few other samples I found) the regular 1080 footage is down sampled from a 16.6 megapixel sensor seems like you could get the same IQ from a much smaller piece of that sensor.

So much great information on this site, pretty sure you all helped me find my next (first real) video camera and have a good idea of what to expect out of it.

[Noa Put]

nd have a good idea of what to expect out of it.

When doing these kind of tests I do get reactions that you can't expect the camera to be flawless at it's price point, but that's not the reason why I do this, first I"m performing these tests for myself, to get to know what the limitations of this camera are so I can safely decide how far I can push it before it gets ugly. And while I"m doing that I like to share as well. Just for the record, I"m not techie at all either as my tests are pretty straightforward. The ea50 is not a bad camera at all and once you find what it's limitations are, you can work around it to get better results, that's what I"m trying to do. :)
Tommorow I"m doing a wedding and will be taking the ea50 along for the first time, will be very interested to see how it deals with low light in the real world. If I"m allowed by the couple I"ll post some footage.

[Dan Eskelson]

More power to you, Noa, for being the premiere scientific mind on this forum. I realize that the multitude of tests you perform are largely for your own improvement, but your willingness to share your observations and results is truly extraordinary and most beneficial to us all.

Thank you, and keep up the very good work!

Dan