16x9 Rez Test

[braw]

I have been toggling between the XL2 and DVX for sometime like I'm sure many. I have been running some tests on footage to squelch my internal debate.

The two things that had me torn were the higher rez 16x9, and the image. I have some footage I shot with a friends DVX and cropped it to 16x9, uprezzed it with Photozoom, and put it on a DVD to compare with the untouched footage. Not knowing what settings were best, I have gathered from this forum that changing the size to 720 or 1080 was silly since it would be down converted again. Seeing Barry's experiment with the FX1, I knew I should uprez at the same resolution but a higher dpi. It seemed that would be accurate since I am utilizing the exact same CCD area as the XL2, just less pixels. So after researching what the pixel count of 35mm and 16mm were, I arbitrarily settled on 150 dpi since I wasn't seeing much improvement above that to my eye on a 21" monitor.

I can notice the difference on a 30" TV however slight, but it seemed that I just transformed the footage into having the same problems as I thought the XL2 had. It seemed way too sharp and had that "videoish" look and the latitude seemed slightly different. The best way to describe it, would be to look at the tunnel grabs posted for the XL2, DVX, FX1 comparison. Uprezzed footage looked more along the lines of that tunnel image.

My personal, and I stress personal conclusion was that I would probably only uprez wide or establishing shots. I have heard others say this before, and I now fully agree, the DVX hit on something whether intentional or coincidental with this machine. It has a very sharp picture, but is somehow balanced it's colors and softness/grain that makes it look more like film than the XL2 to me.

I was also torn with the whole post processing thing. Sure almost everything gets post processed now a days, but the DVX has pumped out images that need little if any correction, and I feel (until someone proves me wrong) that almost all XL2 footage would have to be fiddled with to give me a satisfying film like quality. So no matter which one I would get, I would have to do some sort of alteration to the footage to get what I want.

I just thought I would share my point of view to anyone else in my position, please keep in mind this is only an opinion. I do not yet have a camera so I *AM* unbiased.

[icicle22]

I have discovered the same thing. When I uprezzed using Photozoom some of my DVX footage it took on a "Xl2" sharpness. Not sure how to handle this. However when I used vegas to convert from letterbox to anamorphic 16x9 it softened the footage.

If the DVX just had real 16x9 it would rock! But with the price of chips it would still have to be a 1/3 chipper so a smaller area would be used, thus smaller pixels leaving us with lower light sensitivity.

I guess you can't have it all! At least not all in one camera anyway.

[Rick_Rock]

"I guess you can't have it all! At least not all in one camera anyway."


More people need to realize this, especially on the whole DVX vs. XL2 vs. ETC ETC ETC deal.

[braw]

Interesting. I cropped the images to 16x9 first and then uprezzed them. Opened in Vegas and rendered to MPEG. I can see a slight jump in sharpness but so close to just "sharpening", the time didn't seem worth it. I can see it's use in up converting for HD and maybe to help sharpen wide shots since the DVX is a little soft in that area, but it almost destroys close shots and turns it a little more video like.

I wish they had a native 16x9 DVX, it would end the debate. Native 16x9 DVX would be king. For now, it's a sea of compromises.

[Shaw]

I'm not sure that classifying film as soft and video as sharp is a good enough classification. Sure video can be too sharp but I think what most people percieve as "too sharp" is "too sharp over a large area" - in essence the lack of shallow DOF. Film is sharp. Quite sharp. What makes the difference I think is that most of the frame is out of focus enough to not make it overly noticable. Just my theory though.

If the additional sharpness is a problem try adjusting the settings in photozoom. It adds a minor sharpening to the footage inherently. This can, however, be turned off.

[braw]

I'm not sure that classifying film as soft and video as sharp is a good enough classification. *Sure video can be too sharp but I think what most people percieve as "too sharp" is "too sharp over a large area" - in essence the lack of shallow DOF. *Film is sharp. *Quite sharp. *What makes the difference I think is that most of the frame is out of focus enough to not make it overly noticable. *Just my theory though.

If the additional sharpness is a problem try adjusting the settings in photozoom. *It adds a minor sharpening to the footage inherently. *This can, however, be turned off.

I agree. It's hard to discribe and I don't know if my termonology is the best, but since 1/3" cams have a pretty deep DOF, the XL2, and the DVX uprezzed the way I did it seems to be digitally sharp in the sense that things are too defined and crisp. Film is indeed very crisp but the colors tend to blend slightly into each other causing a softer image. To me it looks like the DVX blends in a a similar fashion possibly in the DSP or something.

It's not that the DVX is low rez so much as it is the way the image looks. Something about the color curves, resolution, lens, or firmware seem to make it a little more filmic.

Granted this isn't the most fair comparison, but the end result looked very similar to me.

So Shaw, what settings are you changing in Photozoom?