Sony EX1 vs Z7
By Guy Barwood - Glass Eye Vie
w (May 08)

There has been a lot of discussions of late based around the decision to purchase either the EX1 or the Z7. Sony have definitely released two models into the market with a lot of pro's and con's each. Deciding which model (if either) is for you seems to be causing some interesting dialogue (particularly in the AVPA Members Email Forum).

During the May 2008 AVPA meeting, as it was held at out sponsor's premises (VideoCraft) I was able to shoot with the EX1 and the Z7 at the same time. Putting both cameras next to each other also allowed for an almost identical frame for comparison. This test s not meant to be perfect, it was done in about 10 minutes quite late in the evening. The EX1 was captured on SxS, the Z7 was captured onto one of my 8GB Lexar 300x UDMA CF cards I pulled out of the D3 (Nikon DSLR) using the Z7's CF card recorder.

My first task was to drop the light levels and open each camera up. Which camera is more sensitive at its natural open (0dB gain) setting? I have seen one comparison on DVInfo (dvMISSinfo perhaps) which showed images suggesting the Z7 would be close to at least a full stop slower than an EX1. As you will find this is not the case.

The following are frame grabs from Liquid Edition 7.2 on a 1920x1080 uncompressed timeline (sequence). As noted in the table, the EX1 is on the left, the Z7 is on the right. Both cameras, being showroom floor demo models, were still set at all factory defaults. I had little time, or inclination, to change the stock standard settings of the cameras. Both can be changed of course. Both were manually white balanced on the wall, lens full open and gain set at 0dB.

The frame was take with the stock lens on the Z7 and fixed lens on the EX1 with the centre markers set right on the Arri Fresnel.

EX1 1920x1080 50i 35Mbps Z7 1440x1080 50i 25Mbps

The first thing you will notice is the EX1 is more sensitive, but not by much. You can also see the EX1 is slightly wider than the Z7. The Z7 frame does look softer, you can even see this in the 25% reduced frame above, but when this frame was taken the EX1 was in auto focus mode (so was reasonably focused) while it wasn't until later I discovered the Z7 was in manual focus and was a little of focus.

The above are100% crops of the colour chart and the grey section of the wall just to the left of the Arri Fresnel light. Now you can clearly see the Z7 was out of focus in the colour chart. The noise in the grey appears to me to be fairly equal. I don't think you could look at either and know which camera it came from without being told.

Here you can see on a waveform monitor the EX1 is achieving higher levels from the same scene. Depending on the section of the scope you look at, the difference ranges from about 5-10 IRE higher in levels for the EX1.
The frame above now represents the same scene zoomed into the cameras full zoom ability. This was where I noticed the Z7 being out of focus and was manually focused on the Arri using the LCD screen. To my eye, the EX1 is a clear winner here. The image is much sharper and the reach of the lens is better too. This makes the EX1's lens wider and tighter than the stock lens on the Z7 (although you could have worked that out by the math on the lens mm ratings anyway).
The above is a 100% crop from the full zoom frames above. The difference is striking, there is simply no comparison. The Z7 is as soft as a kitten while the EX1 is very sharp. The Z7 is most likely hit badly by the quality of the stock lens. This is still an important test however as most Z7's will more than likely continue to use the stock lens throughout their whole life. Even though it is removable, the cost of an upgraded lens for the Z7 will make the EX1 a cheaper option even with the high cost of the SxS cards.

Zoomed fully in there is very little difference in exposure levels between the two cameras in the mid range. The EX1 is higher but only about 5 IRE in the upper and lower thirds, and about 2 IRE in the mid range.

Black Crop

Black Crop
The next test was to close the lens hoods, add the ND and close the iris (everything I could do to stop any light getting in) and shoot some straight black at 0dB. As you can see from the 100% crops above, black looks like black, I can't see any visual noise in either.


Back Noise HIGHLY Exaggerated


Back Noise HIGHLY Exaggerated

What I have done here reveals and highly exaggerates the noise in the image. First I used Photoshop to produce a "Difference" with the original frame and a pure black artifical frame. This produced a VERY light grey frame. I then adjusted the levels raising the black level to 230 (white is 255, black level is usually 0). This in effect amplifies the noise a huge amount. As you can see above, to my eye the Z7 actually has less noise in the black. Not by much when you consider how exaggerated the noise levels above are but this result goes against other reports I have heard stating the Z7 was very noisy at 0dB.

Conclusion

These results are not all that surprising. The EX1 and Z7 share similar sensor technologies. The EX1 has 1920x1080 pixels in a 1/2" type sensor while the Z7 has 960x1080 pixels in a 1/3" type sensor. This means the Z7 has only half the number of pixels in a sensor about 16% smaller. Its pixels should be much larger then the EX1's so really the Z7 should be more sensitive. The EX1 however has more advanced processing making the results from the EX1 stunning across the board. The Z7 appears to be either heavily let down by the stock lens, or really shows the limitations of the lower resolution sensor. Most likely it is a combination of both with the lens being more at fault than the pixel offset sensor.

So where does that leave the Z7. Certainly with the Z7 appears to be quite capable when it comes to sensitivity and noise compared to the EX1 which has set the standard in this regard in the handycam form factor since its release. It does appear to be dramatically let down by the stock lens however (I would love to test it with a decent 1/3" HD lens such as the Fujinon 13x HD lens originally released for the HD100 series). Adding such a lens, or any other half decent 1/3" HD lens to the Z7 dramatically increases its price, so much so the EX1 even with a swag of SxS cards may present a decent cost saving (the Fujinon 13x lens is well over AU$10K itself). If you are only going to add the new lens and never change again then you might as well go with the EX1 to start with as the fixed lens is better than most other entry level HD glass. If you are really going to be spending a lot on various lens and swapping them a lot then the lens cost will far outweigh the difference in camera costs so perhaps you should just pay the extra and get the EX3 (being an EX1 with removable lens) to get the best of both worlds.

The Z7 offers cheap CF card recording but this is not completely comparable to the SxS option. A single card limits record time (EX1 is practically unlimited), the CF recorder is much more quirky than the EX1 is with SxS. It can take some time to stop recording if you shoot a small clip for example. There have also been reports of corrupted frames. The cards fault? or the recorder? Does it matter? It has not happened with the SxS (or P2) professionally designed Solid State recording options. There is definitely something to be said for paying extra for a professionally engineered product rather than an add on which may be very cheap and handy but just not as streamline. As long as you also record to tape as well, and start/stop like you expect with a tape transport the CF option will probably work just fine.

Copyright (c) 2008 Guy Barwood
Images or text may not be reproduced in part or whole without permission (email glasseye@glasseye.com.au)