Z7U Camera Test - Uh Oh, Need Help!

[Jason Aumont]

I went by Abel Cine here in NYC to test out the Sony HVR-Z7U to see how it felt and to get sample footage for a workflow test. First off, I have to say, Charlie down there was very cool and accommodating. They just lent out their display model, and he opened a brand new box so I could test it out.

I loved the build and ergonomics, Sony sure knows where to put things. I came away quite impressed with the Z7U.

When I got home and loaded up the footage, and the picture looked great. Until I noticed that on quick pans, tilts and zooms it seems like the image shows a lot of compression (digital blocks). I did these tests to see if the rolling shutter was a big issue. I can live with some slight skewing (which I didn't notice too much), but these compression artifacts are a deal-breaker for me.

The below link shows 4 frame grabs:
1) A locked-off shot on the tripod.
2) The same object during a quick pan which shows these digital artifacts.
3) This was from a test where I recorded HDV to the tape and DV to the CF card. This is from the DV.
4)This is from my Canon XH-A1. It's a fast move taken right from the .M2T file from the HDV capture.

http://www.theputtydivision.com/z7utest.html


I shot HDV-24P-24f shutter-5.6f. Since this was my first time ever touching the Z7U, I hope I just didn't set something properly. Please look at these shots and let me know if you think it's a setting, is it a CMOS thing, is it the nature of HDV/long GOP (if so, why have I never seen this in my A1 footage)?

This is pretty important because I was leaning towards this camera for a documentary I'll begin shooting in May and I wont be able to shoot some of the action without quick camera movements.

Thanks,
Jason

[Paul Therrien]

What's really strange is the CF card fast motion is much more clear then what was recorded to tape. How is that posible if the same 1"s and 0's are being recorded?

The A1 looks good, it must be the way Sony designe the codec to handle motion and the way it is sent through the process of recording with the heads.

Very odd, but is it visible it real time? Would this get in your way while really using it since you would never see the movement so well with your eyes?

Either way, make your choice based on what you need the camera to do for you. If this is a problem check out another cam like the EX or wait to see if Canon etc comes out with something.

Good Luck,

Paul

[Jason Aumont]

You absolutely see this at normal speed. But I have to imagine that something wonky is going on. I can't believe that Sony would release a camera with this type of quality. That's why I asking if any Z7U users have seen this in their HDV footage with camera movement.

[Juan Martinez]

Hi I work for Sony. I agree that Abel Cinetech's staff is very professional. You did not give enough detail on your description. Were you shooting in progressive or interlaced? It also appears that you set the CF card to record DVCAM. Is this so? IN any case, the streaks you see in the image are the result of motion blur caused by a long exposure time. Shorten the shutter time/ angle and try again. You will see the difference.

Juan Martinez
Senior Manager Sony Electronics.

[Jason Aumont]

Juan,
Thanks for responding. I actually spoke with a Sony tech earlier in the week, I sent him sample clips and those stills...I haven't heard back from him yet.

I was shooting 1080i, 24P, 1/24 shutter, f/5.6.

I am used to normal motion blur. I expected to see it. What I don't like to see are those digital artifacts in shot #2 from the link above. Which was very noticeable playing back in real-time. #2 was a frame grab taken full-res from the HDV/M2T clip from the timeline.

I'm actually fine with the motion blur in shot #3 which was recorded to DV on the CF card.

Please look at the artifacts in shot #2. http://www.theputtydivision.com/z7utest.html

Thank you,
JA

[Greg Hartzell]

You normaly move shots this quick? I don't see how this is a relevant test, I would have to see the clips in full motion but I don't see the artifacting you are talking about (unless #2 and #3 were the same shot and then ya, we're in for a big problem).

To me, it looks like #2 was shot with a lower shutter angle than #3 and #4. Again, I'd need to see the motion clips.

All three pans look as the pans were moved at different rates, and faster than most camera ops would do and most editors would pick shots from. I don't see this as a huge problem for most users.

Please please though. Post complete settings for each shot and try to aproximate pan speed in the next test so we can all be in the know. Thanks for posting, hope this helps.

[Don Greening]

Originally Posted by Jason Aumont

I was shooting 1080i, 24P, 1/24 shutter, f/5.6.

It's generally considered that the minimum shutter speed for smooth motion should be twice the frame rate. In your case you need to have the shutter set to 1/48th sec.

- Don

[Jason Aumont]

"You normaly move shots this quick? I don't see how this is a relevant test, I would have to see the clips in full motion but I don't see the artifacting you are talking about (unless #2 and #3 were the same shot and then ya, we're in for a big problem."

I am prepping for a documentary shoot. Because we'll be shooting in a verite style, this is a very relevant test. Moving or zooming the camera quickly to catch something on tape is a regular occurrence. I would not want to have to cut around it or slowly pan the camera over to the subject to avoid these artifacts.

Here is a link to a frame grab from a quick zoom. The digital artifacting is ridiculous:
http://img33.picoodle.com/img/img33/...im_4b2aef5.png

This is what I see with any movement that isn't slow. I don't see how this is related to the shutter-speed. I have been shooting video for 15 years, and the only time I have seen this is on a camera-phone or web-video.

I normally shoot at 1/48th, but for this test we shot at a number of shutter speeds. Unfortunately, I didn't have enough camera movement samples at those speeds to show a good example...but it is there.

If anyone wants to download one of the actual M2T files and see for yourself, here is the link: http://www.box.net/shared/gi243x3sws

Once again, I know what normal motion blur looks like, and I know what normal 24P stutter looks like, this is not that. This seems like the camera was unable to write/compress the image to the tape properly.

Thanks,
JA

[Zack Birlew]

Does it look as bad when recorded to CF?

If it's strictly related to HDV tape, then I suggest that you stock up on some 32gb CF cards if you're going to use it.

Yeesh, that quick zoom artifacting was embarassingly bad! I hope it's just something wrong with the camera you were shown.

[Jason Aumont]

It looks the same for tape and CF...the more I think about it, the more I think it was a bad unit. I can't imagine under any menu settings where it would look like this.

JA

[Zach Love]

any updates?

where you able to go back to Abel Cine & test out their display camera? I'm hoping that it is a bad apple, but if that is the norm...

[Ethan Cooper]

I've seen #2 (hehe) in low light situations with my FX7 and just chalked it up to HDV and lots of gain. But your test was well lit so I don't know.
If you're worried about artifacts in the style of shooting you'll be doing maybe HDV isn't for you. It's a known fact that at some point, HDV will give artifacts no matter who the manufacturer is. That's the trade off for high resolution at low bit rates and low prices.

Maybe renting a higher priced non-HDV system for around the same cost as buying the Z7u might be better for your situation depending on how long you'll be shooting.

**EDIT**
Wait, you already own an A1? Why do you feel the need to get a Z7? Is it for the better low light capability?

[Jason Aumont]

I need to have 2 cameras for the shoot. So the Z7U would be the "A" cam because of low-light and the CF card workflow. We will be dealing with hundreds of hours of footage, so the CF will save us days and days in post.