Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

a bit disappointed with the 4K coming from the Scarlet

The possibility of cropping/reframing is a creative bonus for if/when required. But logically not a best practice if you want the best images out of your Scarlet.

As to 4k out of Scarlet not being good... I honestly can't say. For two reasons.

1) I don't own a 4K monitor or projector. So I either have to trust in the professional opinions of those who do. test as Mark has, or concede the point.

2) What other 4k am I supposed to be comparing it to? If there someone doing better 4K than RED yet? The closest I see to a possible comparison is where again someone like Mark has compared Epic to Scarlet footage - which I have heard the from others as well compares very favourably. I guess there's also the R1, or we'd need someone with an F65 I guess, if we want to have a discussion on 4k, and just how good Scarlets "4k" is. Personally I'd love to see, R1 4.5, Epic's 5, and Scarlet's 4, and see what is to be seen or not, on a RED ray.

Anyhow, those times are soon going to be upon us. And I'm not particularly looking forward to the inevitable, my 4k is better than yours wars, and reading endless necessary posts on the relevance of debayering, compression, codec, is it sharpened in camera, power of RAW... and on it goes.
 
Not saying that Scarlet footage isn't the most absolutely jaw-dropping slice of heaven and thanks-to-my-Bomb-Squad-rep never seems to get a good response.
 
I'm not sure as I don't have Photoshop but that comparison doesn't work, or isn't showing what you are looking for. The reason the difference seems minimal is you are viewing it on a sub 4k screen the 1080 you have taken from 4k but not fully scaled yet as it's still in the ap (export the stills then compare would be better but I still think this method is unusual... to me) plus it benefits from oversampling so it will be a much better looking 1080 than normal so when compared to it's 4k version on a screen that can hardly resolve it yes they will look similar. Plus if it's a good scale I'd think on the low res screens the up res could bring it just about back to what it looked like in the first place, though I have been drinking and it's been a long family night so who knows I could be way off.

thinking more about it, export one at 4k and one at 1080, zoom 2x or more and compare from there just not in the application... though could be the same thing can't test it here sorry.

and remember that DSLR stills are a whole different ball game no or little motion blur heavy sharpening... now they do tend to produce sharper images more easily, though I have seen scarlets and epics shoot very comparable images to DSLR's and considering it's in motion that is amazing. Heck look at the amount of fashion shoots that have used them they can't be that bad plus the whole 5k aspect makes them even that much better when using epic.

have fun but trust the wealth of knowledge on this site... not me, cross check anything I ever say... or just ignore me! Have a good one!
 
Which lens(es) ?

Did you check back focus ?
 
Andy Corleone said:
I think anyone can do this Now:
grab a 4K RAW frame from the scarlet and load it to photoshop:
1) in Photoshop with the frame open Go to Layer tap /"duplicate Layer" and select "document" new
2) this new document select Image/Image Size/ and change it to 1920x960 pixels
3) Go to the Layer tap and select again "Duplicate Layer" but his time you wil sellect the original 4096x2048 pixel Image as the destination
4) you will end up with two images one 4K and the other one 1920x960 pixels in a different layer
5) Now select the Small 1080p one and scale up to match the 4096x2048 size

from now you can switch off and on the layers to compare the 4K Scarlet to the 1080p "Zoom in" Scarlet. you will notice some extra minimal resolution on the 4k but nothing great
...

Hope this make sense.
Click to expand...

So you are taking a 4k frame and resizing it to 1080, and then taking that same 1080 frame and blowing it up to 4k... then you are comparing it to the original? Then you are compare it to the native output of 1080 cameras? Based upon these comparisons you declare that the 4k Scarlet output is disappointing?

All of your comparisons are apples to oranges, viewed through a sieve. I'm sorry but your methodology makes absolutely no sense whatsoever.

When you resize the 4k down to 1080, you are throwing away all of the resolution that makes it 4k. When you then blow it back up you are no longer looking at a Scarlet frame. What you are looking at is a Photoshop frame because Photoshop has used it's own algorithm to duplicate pixels and artificially create it's own resolution. To make matters worse you are looking at this "4k" material on a 1080 monitor and drawing conclusions based upon that.

Andy Corleone said:
For Comparison you can do the same with a 16MP to 8MP DSLR or with a 1080p to 720p and you will see how different the resolution is. while with the scarlet 4K o 2K is not that great.
Click to expand...

Sorry, I am trying to understand what that is meant to illustrate, but can't. What is it supposed to mean? Red 4k 16:9 is what 11.5 MP? Are you now comparing that with a 16MP image? ... So a 16MP image scaled down to 1080 looks better than a 11.5MP image scaled to 1080 therefore the 4k is disappointing?

Is this 16MP DSLR generating 16MP moving pictures? The answer is No. So why make that comparison, are you planning to use the Scarlet as a still camera? I suppose you are not. Otherwise there would be no reason to scale it down to 1080.
 
Last edited:
I'm sorry but you didn't understand. for this example that anyone can do at home: I'm comparing to Native 4K. I'm taking this 4K and Scaling down to 1080p and them Scale Up to 4K again and comparing it with the unaltered 4K.


Scott Crawley said:
So you are taking a 4k frame and resizing it to 1080, and then taking that same 1080 frame and blowing it up to 4k... then you are comparing it to the original? Then you are compare it to the native output of 1080 cameras? Based upon these comparisons you declare that the 4k Scarlet output is disappointing?

All of your comparisons are apples to oranges, viewed through a sieve. I'm sorry but your methodology makes absolutely no sense whatsoever.

When you resize the 4k down to 1080, you are throwing away all of the resolution that makes it 4k. When you then blow it back up you are no longer looking at a Scarlet frame. What you are looking at is a Photoshop frame because Photoshop has used it's own algorithm to duplicate pixels and artificially create it's own resolution. To make matters worse you are looking at this "4k" material on a 1080 monitor and drawing conclusions based upon that.



Sorry, I am trying to understand what that is meant to illustrate, but can't. What is it supposed to mean? Red 4k 16:9 is what 11.5 MP? Are you now comparing that with a 16MP image? ... So a 16MP image scaled down to 1080 looks better than a 11.5MP image scaled to 1080 therefore the 4k is disappointing?

Is this 16MP DSLR generating 16MP moving pictures? The answer is No. So why make that comparison, are you planning to use the Scarlet as a still camera? I suppose you are not. Otherwise there would be no reason to scale it down to 1080.
Click to expand...
 
Andy Corleone said:
I'm sorry but you didn't understand. for this example that anyone can do at home: I'm comparing to Native 4K. I'm taking this 4K and Scaling down to 1080p and them Scale Up to 4K again and comparing it with the unaltered 4K.
Click to expand...

I'm sorry to say and not trying to put you down but that is an extremely moronic test.
 
Andy Corleone said:
I'm sorry but you didn't understand. for this example that anyone can do at home: I'm comparing to Native 4K. I'm taking this 4K and Scaling down to 1080p and them Scale Up to 4K again and comparing it with the unaltered 4K.
Click to expand...

Yes, I understood your method but can't make any sense of it. Why would you do that? What is that meant to show? As I suggested before, your method could be used to evaluate Photoshop's scaling abilities, but it says nothing about the quality of Scarlet's output.
 
Okay, I'll play along...

There is no CC, sharpening, contrast adjustments etc done to any of the below frames. Shot with a Zeiss ZE 35mm @ f2.8. For best results click on the image and view the original size to see the differences as intended.

Here is the full frame in 1080p
Full Frame

Here is a 1:1 crop from the R3D

4K_100

Here is the 1080p scaled to 200%

1080p_200

And just for fun...

Here is the R3D scaled up to 200%

4k_200

And the 1080p with the equivalent zoom

1080p_400

I am on a retina MBP and I can see the difference between a 1080p shot and a 4k shot. It is quite noticeable too...

One more thing. The stills are one thing, but if you do the same thing with a moving image the difference is HUGE!
 
Hey Drew in your example the difernt between 4K and 1080p scale up is very noticiable I will upload one in which is not
 
Andy Corleone said:
Hey Drew in your example the difernt between 4K and 1080p scale up is very noticiable I will upload one in which is not
Click to expand...

My buddy shot the ocean in 720 HDV and they used it in Pirates of the Caribbean 3... The subject matters a lot. So does focus. Sometimes you can mix in lower rez footage, or crop 4k 1:1, sometimes you can't.
 
looking at resolution

looking at resolution

If I understood correctly, he is using a simple test for resolution: Take the image, lets say it's a possible 4K image, and resize it down to 2K. Now resize it back up to the 4K. If you put these side by side and inspect ( the original 4K and the one that was sized down and up ) you should see a sharpness difference. If you don't , then the 4K was really only resolving 2K.
It's a good 'poor mans' way to test the real sharpness of an image.

Try it on my footage that I shot in a cemetary at ISO 1300 on the Scarlet on a monopod ;) ;)

In the spirit of posting actual 4K or 5K material, here is a grab from work in progress.

( People really should start posting some FULL RESOLUTION stuff here, on the famed and glorious 4K Reduser site. Enough talk. Show me the money... )

link to 4K image grab is : https://dl.dropbox.com/u/61349012/mi...orever0004.jpg

1K thumbnail :
1K_iso1280_hollywood_forever0004.jpg
 
Les Dittert said:
If I understood correctly, he is using a simple test for resolution: Take the image, lets say it's a possible 4K image, and resize it down to 2K. Now resize it back up to the 4K. If you put these side by side and inspect ( the original 4K and the one that was sized down and up ) you should see a sharpness difference. If you don't , then the 4K was really only resolving 2K.
It's a good 'poor mans' way to test the real sharpness of an image.
Click to expand...

I don't see how that can be true, but maybe I'm a dullard. The software is responsible for throwing away a majority of the image information, then multiplying the remaining pixels to re-create the lost resolution. This causes edges to alias and soften. The vast majority of the resulting image is the creation of the software, in this case Photoshop, not the camera system. How can you look at that image and attribute any of it's sharpness to the original capture when most of it was created by Photoshop?

Other factors such as contrast contribute to perceived sharpness, but if you are using your eyes to make a determination as opposed to a measurement, then you are dealing strictly in the realm of perception and empirical conclusions like "the 4K was really only resolving 2K" can not be made with any accuracy.

Andy Corleone said:
As I said in the begining some extra resolution but not a huge different: only really noticiable at 200% and up

http://www.flickr.com/photos/89764934@N08/8158502867/in/photostream

http://www.flickr.com/photos/89764934@N08/8158532118/in/photostream/
Click to expand...

Again, I don't know where you get the idea that you have extra resolution. There is no "extra" resolution. It's a trick, a fabrication. The upscaled image is being built from a fraction of the original source. Once you throw pixels away, they are gone. An upscaled image may have the same pixel density as the original, but it can only be a fraction it's actual resolution. It is in no way representative of the quality of the original image IMO.
 
Last edited:
Scott, you realize you're proving him correct, right?

In your explaination, you're saying software is responsible for deleting (reduction) and rebuilding (upscaling) image/sharpness using it's own software/processing and that it has nothing to do with the quality of the capture device. What he's saying is that the reduced/upscaled image (regardless of the software) looks very close to as good as the source 4k... Basically that the software is upscaling to an image that is as good as the original AFTER having thrown away data. The 4k source isn't going to get any better, so if the upscale looks as good, there is a bit of a problem.

That said, if the lenses used are poor resolvers, you won't see the advantages of higher resolution capture anyway (which is what I think is happening.)

Scott Crawley said:
I don't see how that can be true, but maybe I'm a dullard. The software is responsible for throwing away a majority of the image information, then multiplying the remaining pixels to re-create the lost resolution. This causes edges to alias and soften. The vast majority of the resulting image is the creation of the software, in this case Photoshop, not the camera system. How can you look at that image and attribute any of it's sharpness to the original capture when most of it was created by Photoshop?

Other factors such as contrast contribute to perceived sharpness, but if you are using your eyes to make a determination as opposed to a measurement, then you are dealing strictly in the realm of perception and empirical conclusions like "the 4K was really only resolving 2K" can not be made with any accuracy.



Again, I don't know where you get the idea that you have extra resolution. There is no "extra" resolution. It's a trick, a fabrication. The upscaled image is being built from a fraction of the original source. Once you throw pixels away, they are gone. An upscaled image may have the same pixel density as the original, but it can only be a fraction it's actual resolution. It is in no way representative of the quality of the original image IMO.
Click to expand...
 
Get a bigger 4K screen to view, I dont think the retina display is big enough to judge the image.
 
Dave Blackham said:
Get a bigger 4K screen to view, I dont think the retina display is big enough to judge the image.
Click to expand...

That logic doesn't work because you can view the files 1:1, which should actually make the quality differences appear even more obvious... Since he's not seeing much of a difference "zoomed in" to a 2880x1800 crop of the image, he's less likely to see differences when the entire image is displayed at 1:1 on a 4k monitor.
 
You must log in or register to reply here.
Back
Top