DVX+Andromeda, a budget HVX for the rest of us?

Hey Juan, good to see you here and thanks for participating.

Regarding the dynamic range: our complaint is that the DVX shot is clearly not exposed properly -- none of the comparison shots have shown the DVX shot exposed properly, it's always way blown out. Can't you just do a simple test where you shoot with the andromeda at its proper exposure, and then shoot a second shot with the DVX at its proper exposure level, and let us compare them?

Use a waveform to set the Andromeda to its proper exposure. Then use a waveform on the DVX to change the exposure so that the DVX is at its proper exposure. Then import the DVX frame from tape and let us compare.

Thanks!
 
..the sensitivity of the ccd dosent necissarily have to change.. the dsp and gain and such applied to the signal from the CCD in the DVX pipeline to record DV ..and all that other proccessing technical gibble gabble that im sure you understand more than i do could possibly be completely responsible for the induced overexposure of the DV image.. but again.. just because one is overexposed and one is underexposed doesnt mean that the lattitude in one image exceeds the other.. i could put two dvxs side by side and over expose one and underexpose the other.. it doesnt really matter why its happening or how.. the reality is that he cams have the same lattitude and somehow one is overexposed and the other is underexposed. in order to tell which one has more lattitude you need to properly expose them both the same. thats why we have a problem with the comparision images. not the product.. if it does what its suposed to why would we have an issue? wed all love it to death.. most of us do and just have issues with these promotional materials. were just saying that your not exhibiting the extended latitude properly, and the comparision images are unfortunately misleading.. because they are not a proper comparision. as far as the varicam images. whats the point.. i already pointed out that my DVX grabs are better.. almost all the dvx grabs ive ever seen are better. we all know beyond a shadow of doubt that the varicam bows the psnts off the DVX. im sorry, but the reality is that theres really no need for me to proove that the varicam is capeable of better imagery.

Juan .. in all honesty.. im not trying to knock your product at all .. i like it.. im a business person myself, and im consdering buying it, i just want some cold hard real testing to be done before i consider buying one further.. i research the noddles out of everything i buy.. so if you could please conduct the experiment we are all asking for above it would be much appreciated..
 
I completely agree with the camparison shots, but I also see where Juan is coming from. Obviously the DVX frame came from the Andromeda frame after processing, and if you bring the andromeda frame in, and color correct it, to expose the trees and deck the same as the DVX, you could easily keep the sky detail in there as well. I shoot with a lot of Digital SLR's, and I have to tell you, that no matter the brand make or model, when shooting raw, I can almost always pull at least an extra stop through 1.5 stops out of the image than from the jpeg or tiff file. Sure it requires more post processing, levels, curves mainly, but it can be done, and having that latitude is great. Looking at the andromeda, it seems to be the same type of thing. All of the raw CCD info is there which allows you make drastic adjustments to shadows, brining up under exposed levels to normal while maintaining all the highlight detail.
 
yeah ..i know what you mean about correcting the andromeda shot.. but objectively , you could do the same with a properly exposed DVX (or underexposed same as andromeda grab)shot as well. i completely agree that raw data is capeable of yeilding some nice advantages in terms of lattitude and other areas and the andromeda should be able to provide those benefits. all I want is to see a real comparison to be able to visualize the realistic extent of the actual benefit.
 
Thanks for your comments, we will carry out a test as Barry suggested.

I still have to point out, the grand mayority of the people who have commented negatively about the Varicam frame all act incredulous and have never touched a varicam. Many even ask me why we 'reduced' the image to 960x720!

I feel their confusion when I see a setup that costs $100k produce those images, but this is no reason to simply conclude that there is no way they could look a certain way or not. Again, please send some Varicam frames that really do represent what it supposedly can do. The argument that "we all know it can do better because it costs $60k++" is not an argument at all. There are many things that can cause the cost to be that high from market placement, to 'professional' features, to actual cost of manufacturing of aspects that don't necessarily affect image quality. We've all seen the tricked-out hummer that costs more than a Ferrari, this doesn't mean the hummer goes 0-60 in 4 seconds. :)

Cheers,
Juan
 
Hey Juan, maybe de-anamorphizing the image correctly would be a good start. Afterall, since you do electronic manipulation to get that huge frame of the andromeda, why not give the end-result 16x9 720p image of the varicam?

What is really srtiking from that Varicam image, besides the fact it is not very well shot (I've seen better from that poster), is the color fringin on the lens that makes me wonder what glass was he using with that Varicam! That is really odd. If you want to see some good varicam images, just rent a copy of colateral or catch a rerun of arrested development. They are nothing like the shot you posted...

I too can't wait to see some fair comparisson, not because I don't trust you but to measure exactly how much improvement can the mod bring me, you know? Then people will judge better if they should spend their 3000 with you or not.
 
Actually, even though the tester put the two image comparison together, i would still vote against stretching the image horizontally. The Varicam records 960x720, stretching it horizontally to 1280x720 is an up-rez which is completely dependent on what algorithm is used to do so. For the Andromeda image we did not take the standard definition output and up-rez it, but we rather built the large image using detail that is there from the pixel shift, just like an HD mosaic sensor.

The Varicam does it's own internal 'electronic manipulation', we do ours. That's the raw result. If we somehow modified the Varicam image in post, then we'd have people complaining that we did something to make it look bad, or that we didn't up-rez it horizontally with a good enough algorithm. We decided to leave both images as they come out of the systems. Ours is the output straight from our exporting software. That image is actually out of our very first algorithm so it doesn't look as good as the latest build.

I don't know either what lens he used, i'm sure he'll tell you if you ask him.

Cheers,
Juan
 
Alright, it's your call. I'm certainly not relying on that image for a quality comparisson, but it does work very well for comparing both systems sizes.

I hope you get to your final release model soon. With more and more modded DVXs around it will be easier to get any image we need for our evaluation. Besides, I can't wait to see this site's new HD 4-way shootout (HVX, Andromeda, Z1 and HD100)!

Good Luck!
 
iamloser said:
yeah ..i know what you mean about correcting the andromeda shot.. but objectively , you could do the same with a properly exposed DVX (or underexposed same as andromeda grab)shot as well. i completely agree that raw data is capeable of yeilding some nice advantages in terms of lattitude and other areas and the andromeda should be able to provide those benefits. all I want is to see a real comparison to be able to visualize the realistic extent of the actual benefit.
Click to expand...

Actually Niel, I would have to dissagree with that, as correcting dvx footage, to the same extent that you can correct the andromeda footage (Yes I have tried for myself) is no where in the same boat. DV footage is extremely compressed, so much so, that any exposure adjustments, above a slight lightening, will show extreme artifacting and blocked up shadow detail. That is non negotiable with 4:1:1 compression. Now the Andromeda footage does have the ability to be taken much much further, therefore, if done correctly, the foreground of the deck shot, could look like the DVX footage while the sky could still be in the latitude range. I didn't believe it at first when I first downloaded the 4:4:4 image and tried the adjusments, but it is really amazing. Now don't take this as I'm against you, I completely agree with you in that I want to see some actual properly shot test footage. I want to see a measured latitude scene (like kodak has on their film page), and I want to see a resolution chart ( I don't know why no one will just download a simple resolution chart and shoot it with the andromeda). If I could see these things, and see the technical details in action with this mod, then I would be ready to buy right now.
 
I know that proper evaluation benchmarks are important. But keep in mind that Juan and his colleagues are doing what no others have done. 4:4:4 is the Holy Grail for us, and I'm sure that Juan does not have a staff of 50 engineers, and a 100,000 sq/foot manufacturing plant. So lets give him some latitude (pun intended). I would be more interested in him completing the development of his product, then getting the exact figures and full marketing hype on a work in progress at this point in time.
 
WEll, I was really directing my coments more towards, say, Obin :) You out there big man, how about helpin us out some :)
 
..yeah .you spelled my name wrong so obviously you hate me..lol

actually stefan, your right, the 4:4:4 image can be pushed much more than a DVX image in terms of keepng the image intact without introducing artifacts and such.. your 100% right on that. I only meant to comment on the ability to retain lattitude info.. i suppose regardless of how bad the compression screwws it up.. cause were not really comparing compression schemes for post processing.. were comparing given latitude within the original image. so i guess my way of looking at it was that if you took a dvx image the same as the andromeda image given. which i believe the dvx is capeable of. and uncompressed it so that you were using the same exact process to push the image.. you could perform the same levels adjustments albiet any artifacts from the original DV compression. at any rate, were on the same side.. we both lke the product,, and just want to see some real comparisions done.


juan, as far as the varicam image.. here are some of the requested sample varicam frame grabs:

frame-grab-05.JPG


image_page3-1.jpg


dsc_0125_800x532.jpg




heres some footage from the varicam:

click here. ..go to "click here" under adlib over on the right



http://www.lucykeyes.com/lucy/trailer/
 
You must log in or register to reply here.
Back
Top